The authors have no conflicts of interest to declare.

Medical PhysicsVolume 44, Issue 1 p. 3-6 Point/CounterpointFree Access Advocating for use of the ALARA principle in the context of medical imaging fails to recognize that the risk is hypothetical and so serves to reinforce patients' fears of radiation Jeffry A. Siegel Ph.D., Jeffry A. Siegel Ph.D. nukephysics@comcast.net 856-899-9767 Nuclear Physics Enterprises, Marlton, NJ, 08053 USASearch for more papers by this authorCynthia H. McCollough Ph.D., Cynthia H. McCollough Ph.D. mccollough.cynthia@mayo.edu 507-284-4104 Department of Radiology, Mayo Clinic, Rochester, MN, 55905 USASearch for more papers by this authorColin G. Orton Ph.D., Colin G. Orton Ph.D. ModeratorSearch for more papers by this author Jeffry A. Siegel Ph.D., Jeffry A. Siegel Ph.D. nukephysics@comcast.net 856-899-9767 Nuclear Physics Enterprises, Marlton, NJ, 08053 USASearch for more papers by this authorCynthia H. McCollough Ph.D., Cynthia H. McCollough Ph.D. mccollough.cynthia@mayo.edu 507-284-4104 Department of Radiology, Mayo Clinic, Rochester, MN, 55905 USASearch for more papers by this authorColin G. Orton Ph.D., Colin G. Orton Ph.D. ModeratorSearch for more papers by this author First published: 22 November 2016 https://doi.org/10.1002/mp.12012Citations: 18 Suggestions for topics suitable for these Point/Counterpoint debates should be addressed to Colin G. Orton, Professor Emeritus, Wayne State University, Detroit: ortonc@comcast.net. Persons participating in Point/Counterpoint discussions are selected for their knowledge and communicative skill. Their positions for or against a proposition may or may not reflect their personal opinions or the positions of their employers. AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Overview The ALARA (As Low As Reasonably Achievable) principle is based upon the assumption that low doses of radiation might be harmful and, therefore, should be minimized for medical imaging procedures. Some consider, however, that such low doses are not only harmless but might also even be beneficial, and that advocating for use of the ALARA principle in the context of medical imaging fails to recognize that the risk is hypothetical and so serves to reinforce patients' fears of radiation. This is the claim debated in this month's Point/Counterpoint. Arguing for the Proposition is Jeffry A. Siegel, Ph.D. Dr. Siegel obtained his M.S. degrees in Chemistry and Radiological Physics from the University of Cincinnati and his Ph.D. degree in Medical Physics from the University of California Los Angeles. After working for over 15 years as a medical physicist and Associate Professor, Diagnostic Imaging, at Temple University School of Medicine, as Director, Section of Nuclear Medicine Physics and Physics Research and Development at Cooper Hospital/University Medical Center, and as Clinical Professor of Radiology at the University of Medicine and Dentistry New Jersey – Robert Wood Johnson Medical School, Dr. Siegel assumed his current position as President and CEO, Nuclear Physics Enterprises, Marlton, NJ, USA. This is an international consulting firm specializing, among other things, in evaluation of new radioactive drug therapies and clinical trial design, translational research, biokinetic modeling, quantitative nuclear medicine/radiological imaging, internal and external dosimetry, radionuclide therapy patient release, radiation protection, and FDA and NRC regulatory issues, topics on which Dr. Siegel has published extensively. He has authored more than 350 publications, including two books providing guidance for compliance with NRC regulation of nuclear medicine. Arguing against the Proposition is Cynthia H. McCollough, Ph.D. Dr. McCollough obtained her M.S. and Ph.D. degrees in Medical Physics from the University of Wisconsin, Madison. Upon graduation she began working in the Radiology Department at Mayo Clinic, Rochester, Minnesota, where she is currently Professor of Medical Physics and Biomedical Engineering. As Director of Mayo Clinic's CT Clinical Innovation Center, Dr. McCollough leads a multidisciplinary team of physicians, scientists, research fellows, and graduate students on projects seeking to detect and quantify disease using CT imaging. She has particular expertise in the use of CT for quantitative assessment of material composition, disease progression or regressions, and organ function, as well as methods to quantify and reduce radiation dose. Dr. McCollough is internationally recognized for her contributions to the fields of CT imaging physics and technology, and radiation dosimetry and protection. She has served in numerous capacities in the AAPM including on the Board of Directors and the Editorial Board, and has been elected Fellow of the AAPM, the American College of Radiology, and the American Institute for Medical and Biological Engineering. For the proposition: Jeffry A. Siegel, Ph.D. Opening Statement Medical imaging, particularly CT, is said to produce iatrogenic cancer risk from radiation exposure. Yet, credible evidence of imaging-related low-dose (< 100 mGy) carcinogenic risk is nonexistent; it is a hypothetical prediction derived from the demonstrably false linear no-threshold hypothesis (LNTH). On the contrary, low-dose radiation does not cause, but more likely helps prevent, cancer. Countless experimental and observational studies show this benefit.1, 2 Epidemiological studies purporting to establish low-dose radiogenic risks fail to consider basic scientific research and employ circular reasoning, rendering their conclusions false and indefensible.3 The LNTH and its offspring ALARA are fatally flawed, focusing only on molecular damage, while ignoring protective, organismal biological responses. DNA double-strand break repair and other adaptive protections more than eliminate the low-dose radiogenic damage, repairing or removing even the far greater damage from endogenous processes.4, 5 Many radiologists and medical physicists grant that imaging's radiation-associated risks are minute, and may be nonexistent, with benefits far outweighing these putative risks, yet nevertheless, advocate the “prudence” of dose “optimization” (i.e., using doses that are ALARA); but this is a radiophobia-centered approach. For example, the goal of the Image Gently Alliance is to lower the potential risk of CT-caused cancer in children by providing information on dose management and “optimization” (based on notional LNTH-predicted risks) creating the false perception that some risk exists. There is nothing prudent about ALARA dosing: radiophobia's far greater actual risks arise from patients' fear-driven imaging avoidance and physician-recommended use of alternative procedures, such as long-duration MRIs in children requiring anesthesia. True iatrogenic risk arises not only from such alternative procedures but also from misdiagnoses that are secondary either to patient refusal of medically indicated imaging or to nondiagnostic scans resulting from insufficient exposure.6 All medical procedures require the justification of medical indication, but such justification does not involve imaging's radiation levels. The problem is radiophobia, not radiation. Dose “optimization” efforts only multiply illnesses, injuries, and deaths without justification. Therefore, the ICRP-recommended fundamental principles of radiation protection – justification and optimization – are mutually contradictory and without merit for radiological imaging. Moreover, imaging's dual benefits remain hidden: first, the valuable diagnostic information it provides, which either strengthens confidence in suspected diagnoses or leads to more accurate diagnoses and better treatments;7 and second, the far more likely low-dose health benefits of reduced lifetime cancer risk and all-cause mortality.8 Medical imaging achieves a diagnostic purpose and should be governed by the highest science-based principles and policies (use of proper procedures, appropriately calibrated equipment, etc.). The LNTH is an invalidated anti-scientific hypothesis, spawning the ALARA policy: neither errs on the side of caution. Rather, LNTH and ALARA are responsible for misguided concerns and uninformed policies promoting radiophobia that leads to actual risks far greater than the hypothetical carcinogenic risk purportedly avoided, all while ignoring imaging's benefits. Therefore, these policies have no place in managing imaging's usage. Radiophobia can no longer be ignored: medical imaging's low-dose radiation exposure has no documented pathway to harm, while LNTH/ALARA most assuredly do. Against the proposition: Cynthia H. McCollough, Ph.D. Opening Statement The fundamental principles of radiation protection in medicine require that two criteria are met.9-11The first is justification – any exposures to ionizing radiation must be justified by an anticipated medical benefit. The second is optimization – justified exposures should be applied using the lowest dose necessary to accomplish the required task. This latter principle is referred to as ALARA – As Low As Reasonably Achievable. The premise of Dr. Siegel is that this admonishment to keep doses as low as possible implicitly teaches that radiation is something dangerous, the obvious question being “why aim for low doses unless radiation is a bad thing?” To address this question, I could discuss the topic of whether or not low doses of radiation are in fact dangerous. However, this is irrelevant to the need for the ALARA principle. Large doses of ionizing radiation are a known carcinogen. The evidence for this is unassailable and, because current biological and epidemiological evidence cannot definitively prove that low doses of radiation are safe, the precautionary principle of risk management must be invoked.9-12The precautionary principle is the precept that an action should be undertaken with great care if the consequences are uncertain and potentially dangerous.4 Under the precautionary principle, it is the responsibility of a proponent (e.g., CT provider) to establish that the proposed activity (e.g., receiving a CT) will not result in significant harm (e.g., cancer induction).12 Advising people to take the lowest effective dosage of a medicine (or receive the lowest appropriate dosage of radiation) is always the right thing to do if we know that at high doses, significant harm can occur. Medical imaging providers should not stop aiming to use the lowest radiation dosage that accomplishes the diagnostic task just because Dr. Siegel is concerned about the public perception of radiation. Those who accept existing evidence that the low doses of radiation delivered by medical imaging are associated with risks too small to be definitively demonstrated, including the AAPM,13 IOMP,14 HPS,15 and BEIR VII committee,16 acknowledge that the linear nonthreshold (LNT) hypothesis is a reasonable model for radiation protection. This is absolutely not the same thing as endorsing the hypothesis that risk actually exists from low doses of radiation. Neither does it mean that patient care should ever be compromised in the name of ALARA. ALARA simply means that we should treat radiation as the carcinogen that we know it is (at higher doses) and avoid unnecessary exposures. We should absolutely not abandon such common sense. Rebuttal: Jeffry A. Siegel, Ph.D. Dr. McCollough asserts that whether or not low-dose radiation is dangerous is irrelevant to ALARA dosing. But that is precisely the key relevant point. She bases the possibility of low-dose harm, even if undetectable, on the undisputable fact that high doses are harmful, and advocates erring on the side of caution. However, she ignores voluminous scientific research demonstrating that the body repairs/eliminates low-dose radiation damage, and at the same time is stimulated to repair the much greater endogenous metabolic damage, resulting in a net benefit, through a variety of protective adaptive mechanisms. At high doses, repair is overwhelmed if not inhibited, indicating a different mechanism of action, thereby invalidating the LNTH. Without this evidence, Dr. McCollough's advocacy of ALARA would indeed, as she says, derive from common sense. Therefore, the question of low-dose danger could not be more relevant. It is precisely the proven benefit of low-dose radiation that renders the ALARA principle a source of radiophobia. Furthermore, her invocation of the precautionary principle one-sidedly ignores the harms of radiophobia, including patient refusals of radiation-associated medical imaging and numerous deaths caused by unnecessary forced relocations of mass populations in the aftermath of the Fukushima nuclear accident. She refers the reader to the ICRP and other organizations/committees that adhere to the LNTH. They all concede, with her concurrence, that low-dose medical imaging “risks [are] too small to be definitively demonstrated.” Like Dr. McCollough, they too ignore/dismiss the mountains of evidence that the LNTH-derived cancer risk is a fiction, and that benefit has been proven as presented in my Opening Statement. ALARA-dosing fosters radiophobia because denials that low-dose radiation confers a net benefit, and averrals that it confers risk, are demonstrable falsehoods that neglect the sciences of biology, chemistry, and physics that demonstrate the falsity of the LNTH and the reality of the hazards caused by any policy based on the ALARA principle. Rebuttal: Cynthia H. McCollough, Ph.D. Dr. Siegel and I agree that “credible evidence of imaging-related low-dose (< 100 mGy) carcinogenic risk is nonexistent.” However, I challenge his claims that, instead, “low-dose radiation does not cause, but more likely helps prevent, cancer.” The effect of low doses of radiation – if they exist – are simply too small to demonstrate.17 That holds true for hormetic effects just as it does for harmful effects. As much as biology adds to our understanding of radiation effects, epidemiological studies are the only way to take into account a whole organism's biological response to a radiation exposure. The protective and other adaptive responses to which Dr. Siegel refers can only be shown in the context of the whole organism (i.e., epidemiology), and there it is just as difficult to prove hormesis as it is to prove carcinogenic risk. I further disagree with Dr. Siegel's assertion that Image Gently, and other professional efforts that promote ALARA, are “creating the false perception that some risk exists.” The general public, and many medical professionals, already have the strong bias that radiation is bad. Images from Hiroshima and Nagaskaki, Chernobyl, and Fukushima, and hair loss from CT overexposures are what most people think of when radiation is mentioned. ALARA did not create a bias against radiation. Frightening events associated with high doses of radiation did. To disregard this public perception would be to ignore the beliefs and concerns of our patients. Finally, there is irrefutable evidence that before the advent of Image Gently and other efforts like it, which seek to promote ALARA in medical imaging, children were being irradiated with adult doses. There was not universal attention to optimization of the exams, such as has evolved since these ALARA-focused campaigns. Without a focus on optimization, a cavalier approach to imaging – one that aims for the best pictures and not the best balance of overall care – would ensue. Such disregard of the actual dosage applied would erode the public's faith in imaging providers because of people's underlying belief that radiation is dangerous. Failure to acknowledge potential risks would ignore these beliefs and undermine trust, which is at the core of the patient–doctor relationship. Clear recognition of potential risks and demonstration of technical expertise to minimize risk and maximize benefit is essential in maintaining the trust of our patients. Conflicts of interest The authors declare no conflicts of interest. References 1Doss M. Counterpoint: should radiation dose from CT scans be a factor in patient care? No. Chest. 2015; 147: 874– 877. 2Aurengo A, Averbeck D, Bonnin A, et al. Dose Effect Relationships and Estimation of the Carcinogenic Effects of Low Doses of Ionizing Radiation. Paris: Academy of Sciences – National Academy of Medicine; 2005. 3Sacks B, Meyerson G, Siegel JA. Epidemiology without biology: false paradigms, unfounded assumptions, and specious statistics in radiation science (with commentaries by Inge Schmitz-Feuerhake and Christopher Busby and a reply by the authors). Biol Theory. 2016; 11: 69– 101. http://link.springer.com/article/10.1007/s13752-016-0244-4. 4Siegel JA, Welsh JS. Does imaging technology cause cancer? Debunking the linear no-threshold model of radiation carcinogenesis. Technol Cancer Res Treat. 2016; 15: 249– 256. 5Feinendegen LE, Pollycove M, Neumann RD. Low-dose cancer risk modeling must recognize up-regulation of protection. Dose-Response. 2010; 8: 227– 252. 6Cohen MD. CT radiation dose reduction: can we do harm by doing good? Pediatr Radiol. 2012; 42: 397– 398. 7Pandharipande PV, Reisner AT, Binder WD, et al. CT in the emergency department: a real-time study of changes in physician decision making. Radiol. 2016; 278: 812– 821. 8Scott BR, Sanders CL, Mitchel REJ, Boreham DR. CT scans may reduce rather than increase the risk of cancer. J Am Phys Surg. 2008; 13: 8– 11. 9 ICRP. The 2007 recommendations of the international commission on radiological protection. Ann ICRP. 2007; 103: 1– 332. 10 ICRP. The 1990 recommendations of the international commission on radiological protection. Ann ICRP. 1990; 60: 1– 201. 11 ICRP. Protection of the patient in diagnostic radiology. Ann ICRP. 1982; 34: 1– 82. 12 Australian Radiation Protection and Nuclear Safety Agency, Radiation Health and Safety Advisory Council. Council advice on precautionary approaches in radiation protection. 2002, 1– 4. http://www.arpansa.gov.au/pubs/rhsac/prec.pdf 13 American Association of Physicists in Medicine. AAPM position statement on radiation risks from medical imaging procedures. 2011 http://aapm.org/org/policies/details.asp?id=318&type=PP¤t=true. 14Hendee WR. Policy statement of the International Organization for Medical Physics. Radiol. 2013; 267: 326– 327. 15 Health Physics Society. Radiation risk in perspective: position statement of the health physics society. 2016 http://hps.org/documents/risk_ps010-3.pdf 16 The National Academies of Sciences. Health Risks from Exposure to Low Levels of Ionizing Radiation: BEIR VII Phase 2. Washington, DC: The National Academies Press; 2005. 17Land CE. Estimating cancer risks from low doses of ionizing radiation. Science. 1980; 209: 1197– 1203. Citing Literature Volume44, Issue1January 2017Pages 3-6 ReferencesRelatedInformation

The ultimate goal of radiotherapy treatment planning is to find a treatment that will yield a high tumor control probability (TCP) with an acceptable normal tissue complication probability (NTCP). Yet most treatment planning today is not based upon optimization of TCPs and NTCPs, but rather upon meeting physical dose and volume constraints defined by the planner. It has been suggested that treatment planning evaluation and optimization would be more effective if they were biologically and not dose/volume based, and this is the claim debated in this month's Point/Counterpoint. After a brief overview of biologically and DVH based treatment planning by the Moderator Colin Orton, Joseph Deasy (for biological planning) and Charles Mayo (against biological planning) will begin the debate. Some of the arguments in support of biological planning include: • this will result in more effective dose distributions for many patients • DVH-based measures of plan quality are known to have little predictive value • there is little evidence that either D95 or D98 of the PTV is a good predictor of tumor control • sufficient validated outcome prediction models are now becoming available and should be used to drive planning and optimization • Some of the arguments against biological planning include: • several decades of experience with DVH-based planning should not be discarded • we do not know enough about the reliability and errors associated with biological models • the radiotherapy community in general has little direct experience with side by side comparisons of DVH vs biological metrics and outcomes • it is unlikely that a clinician would accept extremely cold regions in a CTV or hot regions in a PTV, despite having acceptable TCP values Learning Objectives: 1. To understand dose/volume based treatment planning and its potential limitations 2. To understand biological metrics such as EUD, TCP, and NTCP 3. To understand biologically based treatment planning and its potential limitations

The ultimate goal of radiotherapy treatment planning is to find a treatment that will yield a high tumor control probability (TCP) with an acceptable normal tissue complication probability (NTCP). Yet most treatment planning today is not based upon optimization of TCPs and NTCPs, but rather upon meeting physical dose and volume constraints defined by the planner. It has been suggested that treatment planning evaluation and optimization would be more effective if they were biologically and not dose/volume based, and this is the claim debated in this month's Point/Counterpoint. After a brief overview of biologically and DVH based treatment planning by the Moderator Colin Orton, Joseph Deasy (for biological planning) and Charles Mayo (against biological planning) will begin the debate. Some of the arguments in support of biological planning include: • this will result in more effective dose distributions for many patients • DVH-based measures of plan quality are known to have little predictive value • there is little evidence that either D95 or D98 of the PTV is a good predictor of tumor control • sufficient validated outcome prediction models are now becoming available and should be used to drive planning and optimization • Some of the arguments against biological planning include: • several decades of experience with DVH-based planning should not be discarded • we do not know enough about the reliability and errors associated with biological models • the radiotherapy community in general has little direct experience with side by side comparisons of DVH vs biological metrics and outcomes • it is unlikely that a clinician would accept extremely cold regions in a CTV or hot regions in a PTV, despite having acceptable TCP values Learning Objectives: 1. To understand dose/volume based treatment planning and its potential limitations 2. To understand biological metrics such as EUD, TCP, and NTCP 3. To understand biologically based treatment planning and its potential limitations

Construction of new proton therapy facilities is expanding rapidly, despite the order-of-magnitude higher expense of building them compared to conventional radiotherapy. Despite this, many have argued that proton therapy is cost effective, at least for some types of cancer and some populations of patients. In this month's Point/Counterpoint, we debate the claim that proton therapy is the most cost-effective modality for partial breast irradiation (PBI). Proton beam irradiation has become the modality of choice for tumors difficult to treat with other modalities. Among those clearly supported by the literature are the treatment of intraocular melanoma, tumors near or at the base of skull, and those requiring treatment with craniospinal irradiation.1 Proton therapy value—meaning benefit as a function of cost—in the treatment of other diseases is more difficult to show. The case of prostate cancer has become the center of a growing discussion on the cost-effectiveness of protons in terms of clinically meaningful gains of expensive new technologies. In this context, it is not intuitive that using protons for accelerated partial breast irradiation (APBI) would be a good value. Yet the data show that proton APBI compares favorably to all the external beam and brachytherapy alternatives.2–5 The dose is very homogeneous within the target region and delivers negligible dose to nontarget breast, heart, and lung. The initial clinical outcomes of proton APBI have been reported from three institutions, and additional Phase II studies are ongoing. Two of these experiences have been updated recently with up to five years of follow-up showing excellent local control rates and high patient satisfaction.6,7 The group at Loma Linda University has the largest published experience with proton beam APBI showing good to excellent cosmetic outcomes in 90% of patients. Problems seen with other types of APBI—high infection rates, declining cosmesis with time, and fat necrosis—have not been seen with proton therapy. In particular, the poor cosmesis seen in the TARGIT-A and RAPID trials suggests that Linac-based APBI may not be a favorable option.8,9 Because of this promising clinical role for protons in the treatment of early stage breast cancer, cost becomes relevant. Only one cost analysis comparing proton APBI to other breast irradiation techniques has been published to date.5 In 2006, Taghian et al. found protons to be more costly than 3D-conformal partial breast irradiation with photons and classic whole breast irradiation (WBI) with a boost. At PTCOG-NA, Ovalle et al. presented their analysis of current costs of proton APBI compared to seven other partial- and whole breast irradiation techniques.10 Using 2014 Medicare reimbursement rates, proton APBI costs were similar to six weeks of whole breast irradiation including a boost, and less costly than APBI with interstitial brachytherapy using a multilumen device or IMRT whole breast treatments. The key factors in these unexpected results are that (1) the small number of treatments required by proton APBI offsets the higher costs per fraction and that (2) reimbursement rates for the various options have significantly changed in the last decade. With lower costs than multilumen brachytherapy APBI or whole breast IMRT, proton APBI appears to be an appealing alternative for the treatment of early stage breast cancer and deserves further investigation. In general, one would not dispute the proposition that with protons one should be able to design a dosimetrically superior treatment plan compared to external beam Linac-based treatments. What may be surprising is the notion that a marvelous physics-rich modality as proton therapy can be considered a less expensive alternative to conventional photon-based treatments for PBI. In 2014, the Alberta Health Services11 issued a report on referral of patients for proton beam therapy (PBT). In this report, they constructed a framework for determining which situations were most likely to benefit from a referral for treatment with protons, recognizing the premium placed on this expensive resource. They estimated the average cost of each referral to be $200,000. Additionally, the report summarized national guidelines for England, Denmark, and the Netherlands, none of which lists breast irradiation as a standard indication for protons. Similarly in 2014, ASTRO released its model policy for PBT12 detailing the indications for insurance coverage based upon medical necessity and adequate clinical data. They state that (1) it is necessary to understand and document the associated clinical benefits of PBT and (2) PBT should not be considered in lieu of a photon-based schema that delivers quality clinical care with low normal tissue toxicities. PBI did not meet their criteria. Recently, Loeffler13 laid out the current and future landscape for particle beam therapy and reviewed sites of proven clinical benefit and sites of ongoing investigation such as breast. He reminds us that dosimetric advantage does not necessarily correlate with a clinical advantage and that a definitive advantage needs to be established for PBT to become common treatment. He lists sites with the highest priority, acknowledging that the data are not yet available to warrant inclusion for breast treatment. There are a few published clinical trials with protons6,7 supplemented with dosimetric studies.2 An early Phase I trial testing the feasibility of proton PBI reported more late skin toxicities compared to photon-based PBI,6 a consequence, they postulated, of limiting daily delivery to a single proton treatment field due to machine time availability. Five-year results for a proton PBI phase 2 trial were reported by Bush et al.7 demonstrating very good results with 5-yr disease-free survival and overall survival rates of 94% and 95% and excellent cosmesis using 40 Gy over two weeks. In comparison, Formenti et al.14 reported 5-yr results on 100 PBI patients treated with photons with one recurrence out of the 100 patients (1%) with 95% disease-free survival. Updated results (private communication) with 397 patients and median follow-up of 40 months indicate a predicted 5-yr recurrence rate of 0.4% and overall survival rate of 98.2% using daily fractions of 6 Gy over five consecutive days. Of course, cost-effectiveness is not based solely on either the expense of a treatment or even a Medicare-based reimbursement analysis but also upon nonmedical costs, weighting factors for normal tissue toxicities, outcomes, and costs of salvage treatments. Shah et al.15 reviewed the cost-effectiveness of accelerated PBI compared with WBI using 2011 Medicare schedules with costs ranging from $8500 (3D planning), $12,500 (IMRT), up to $18,400 for multilumen brachytherapy. Their analysis included estimates of incremental cost-effectiveness ratios and costs per quality adjusted life year. For protons to be considered cost-effective, we need to demonstrate low costs or improved outcomes. Indeed, for protons to be considered cost-effective, they must be proven to be of lower cost than the alternatives, or to have improved outcomes, or both. In the absence of cost-effectiveness analysis, one can analyze these separately. Costs. Costs of therapy can be classified as medical and indirect costs. As previously mentioned, there are two available reports on APBI costs in the USA. The first one used the 2006 Standard Medicare Payments Schedule for professional and technical charges to estimate costs of proton APBI and compare them to a mixed-modality 3D-conformal external APBI schedule and a six-week WBI technique. Protons had higher medical costs than the alternatives but the lowest patient related costs. Today, many other treatment options are utilized for patients with early stage breast cancer. These include APBI with brachytherapy devices and partial- or whole breast irradiation with IMRT. Using current Medicare reimbursement rates, Ovalle et al. compared these treatments (amongst others) to proton APBI.10 The cost of protons was lower than APBI with a brachytherapy device and whole breast IMRT, and very similar to a 6-week WBI schedule with a boost. Effectiveness. One of the most common and probably the most relevant way of comparing treatment effects is survival. Other outcomes specifically significant to radiotherapy are local and local-regional recurrence. Published data on proton APBI on all these fronts are promising, with a 5-yr overall survival of up to 95% and an ipsilateral breast tumor recurrence-free survival of 97%. Assessments of toxicity have been erratic and will be a vital factor when determining the role of proton APBI. Ongoing trials will aid in answering these questions. Finally, according to ASTRO's proton beam therapy model policy (2014),12 coverage with evidence development is suggested for disease sites such as breast as long as the patient is enrolled in a clinical trial. Hopefully, this will facilitate generating more clinical data and cost-effectiveness analysis will likely follow. A driving principle for initiating APBI protocols and certainly one of the major justifications for exploring APBI was an expansion of women's access to breast conserving therapy (BCT). It was noted that access to BCT was particularly an issue in areas underserved by radiation oncology, where traveling long distances for up to 30 daily visits pushed women toward mastectomies and away from BCT. The number of proton facilities, although increasing, remains small and they are largely situated in high population areas, a situation that may not address this original motivation and may increase the costs for those women who feel a pressure to seek out proton therapy but who live very far from a proton center. Thus, although the use of protons provides an alternative for BCT, it may not expand access for the population pools seeking an alternative to mastectomy. As noted in my opening statement, acute and late effects following proton treatment are dependent upon regimen and technique similar to photon-based treatments, suggesting that protons are not inherently superior. They demand attention to dose-fractionation regimens and planning and treatment specifics just as photon-based APBI.6 The APBI toolbox is already crowded and confusing, at least for the patient. Perhaps, the addition of protons will trigger a review of available protocols to determine which techniques are competitive on both a cost and outcome basis since both of these vary substantially. Protons and multicatheter brachytherapy represent the high end of costs, and five-fraction external beam the simpler, low-cost alternative.14 Dr. Ovalle wishes to acknowledge the assistance of her mentor Eric A. Strom, M.D.

Medical PhysicsVolume 42, Issue 4 p. 1474-1476 Point/counterpointFree Access GPU technology is the hope for near real-time Monte Carlo dose calculations Xun Jia Ph.D., Xun Jia Ph.D. Department of Radiation Oncology, The University of Texas Southwestern Medical Center, Dallas, Texas 75390 (Tel: 214-648-3224; E-mail: [email protected])Search for more papers by this authorX. George Xu Ph.D., X. George Xu Ph.D. Nuclear Engineering Program, Rensselaer Polytechnic Institute, Troy, New York 12180 (Tel: 518-276-4014; E-mail: [email protected])Search for more papers by this authorColin G. Orton Ph.D., Colin G. Orton Ph.D. ModeratorSearch for more papers by this author Xun Jia Ph.D., Xun Jia Ph.D. Department of Radiation Oncology, The University of Texas Southwestern Medical Center, Dallas, Texas 75390 (Tel: 214-648-3224; E-mail: [email protected])Search for more papers by this authorX. George Xu Ph.D., X. George Xu Ph.D. Nuclear Engineering Program, Rensselaer Polytechnic Institute, Troy, New York 12180 (Tel: 518-276-4014; E-mail: [email protected])Search for more papers by this authorColin G. Orton Ph.D., Colin G. Orton Ph.D. ModeratorSearch for more papers by this author First published: 11 March 2015 https://doi.org/10.1118/1.4903901Citations: 14AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL OVERVIEW Monte Carlo (MC) dose calculations are recognized as being the most accurate modality for radiotherapy treatment planning but, because of the excessive computational time required, they cannot presently be used for near real-time dose calculations. Currently, the most common way to accelerate MC dose calculations is to use clusters of central processing units (CPUs), but some believe that the future of near real-time MC dose calculations lies not with clusters of CPUs but with the use of graphics processing unit (GPU) technology. This is the claim debated in this month's Point/Counterpoint. Arguing for the Proposition is Xun Jia, Ph.D. Dr. Jia received his Masters degree in Applied Mathematics and Ph.D. degree in Physics, both from UCLA. He is currently an Assistant Professor in the Department of Radiation Oncology, University of Texas Southwestern Medical Center. Dr. Jia's research focuses on GPU-based high-performance computing for medical physics and medical imaging. He has developed several Monte Carlo packages to improve efficiency for photon, electron, and proton transport. Dr. Jia's research has been supported by government and industrial grants and he has published 60 peer-reviewed papers. He is currently a section editor of the Journal of Applied Clinical Medical Physics. Arguing against the Proposition is X. George Xu, Ph.D. Dr. Xu obtained his Ph.D. in Nuclear Engineering from Texas A&M University, College Station, TX and, for the past 20 years, he has been on the faculty of Rensselaer Polytechnic Institute, Troy, NY, where he currently holds the Edward E. Hood Endowed Chair of Engineering. Dr. Xu's research has centered around applications of Monte Carlo methods to problems in radiation protection, imaging, and radiation therapy. He has been continuously funded by the NIH over the past ten years, including an R01 grant to develop a new Monte Carlo code, archer, for heterogeneous computing involving GPUs and coprocessors. He is the author of more than 150 journal papers and book chapters, and 270 conference abstracts. Dr. Xu is a Fellow of the American Association of Physicists in Medicine, the Health Physics Society, and the American Nuclear Society. In 2014, he was re-elected to a 6-yr term as a council member of the National Council on Radiation Protection and Measurements. FOR THE PROPOSITION: Xun Jia, Ph.D. Opening Statement Clinical applications of MC dose calculations have been limited by the long computation time to achieve a sufficient precision level. Over the years, great efforts have been devoted to accelerating MC simulations. Recently, with the success of GPU-based high-performance computing,1,2 particularly for MC simulations, near real-time (e.g., seconds or subseconds) dose calculation is becoming feasible. Achieving this will not only facilitate its routine utilization, but also realize novel applications to advance radiotherapy practice, such as MC-based inverse treatment planning. To date, the computation time for a typical photon plan has been reduced to less than a minute with ∼1% uncertainty using only one GPU, and the speed can be further boosted with multiple GPUs by a factor proportional to the number of GPUs. Also reported are computation times as low as seconds to tens of seconds for different applications.3,4 Notably, the group at UT Southwestern5 has developed a GPU application to visualize an MC-reconstructed dose delivery process in almost real-time during beam delivery, with a refresh frequency of >10 Hz. These achievements have clearly demonstrated the potential of near real-time MC dose calculations. Besides advantages in speed, GPUs also hold other favorable features for clinical applications. First, GPUs are orders of magnitude lower in cost than a conventional high-performance-computing structure with a similar processing power. Second, GPUs are locally hosted and managed. This is particularly important for problems aiming at near real-time applications, since data-transfer and job-scheduling times cannot be neglected if the computation facility is remotely placed and shared by many users. Patient privacy may also be a concern when transferring medical data to a remote facility. Of course we cannot neglect disadvantages of using GPUs for MC. As a new platform, redevelopment of codes is necessary. However, burdens of initial code development have been overcome to a large extent, and several packages have been successfully built. Efforts have also been initiated to write MC packages in OpenCL to increase portability.6 While there are also technical issues hindering computational efficiency, e.g., thread divergence and memory writing conflicts, many solutions exist to remove or alleviate them.4,7 I would also like to mention a strong competitor of the GPU, the Intel many integrated core (MIC) processor. What makes this particularly attractive is its x86 compatibility, which can run existing CPU codes with minor modification. However, just like for GPUs, substantial effort is needed to achieve optimal performance.8 Simply running an existing code may not achieve high acceleration, because parallel-computing specific issues such as memory access and vectorization were not considered sufficiently in the conventional CPU code. As of today, there has been only limited study regarding MC dose calculations on MIC processors. While it holds the potential to improve efficiency, a lot of research is needed. In conclusion, GPU technology has the capability of substantially accelerating MC simulations. Its advantages and extensive research efforts demonstrate the hope for near real-time dose calculations. AGAINST THE PROPOSITION: X. George Xu, Ph.D. Opening Statement Since the invention of computers in the 1940s, MC codes have been developed for nuclear engineering, high-energy physics, and, recently, medical physics applications. However, most radiation treatment planning is done currently using dosimetry algorithms that are extremely fast, but only "approximately" correct.9 Given the lasting interest in accelerating MC methods, the recent hype related to the GPU is not surprising. Originally marketed by NVidia as household devices, GPU-based game consoles offered amazingly fast graphics at an affordable price. It did not take long, however, for the scientific community to realize that these desktop toys were actually parallel computers. As summarized in two review papers,1,2 GPU adopters from the medical physics community wasted no time in reporting overwhelmingly positive experiences, including a dozen studies that focused specifically on MC dosimetry. Impressive, but inconsistent, "speedup factors" ranging from single digits to several hundreds were reported within months, sometimes by the same group. It has become a cliché to highlight how fast an MC-based dose calculation can be done with a GPU. Such results indeed attracted a lot of attention from medical physicists who are notoriously busy and seeking expediency. There are two strong indications that GPU technology is only hype and not the hope for near real-time, fully MC dose calculations. First, we have not seen any convincing evidence that the GPU is indeed better than traditional solutions for running MC dose calculations. Both of the above review papers1,2 enjoyed referencing the rapidly increasing number of GPU-related journal articles—which only reinforces the concept of a "hype cycle." Furthermore, the authors of the GPU-accelerated MC studies obscure the issue by omitting details on how they compared GPU performance with traditional CPUs. CPU-based clusters are currently so cheap that one can assemble a desk-side 32-core cluster for about $3000US—the cost of a high-end CPU/GPU system. Using software optimization schemes and hyperthreading, such a CPU cluster may achieve a speedup similar to the best reported for GPUs, without the painful process of rewriting the MC code for the GPU/compute unified device architecture (GPU/CUDA) environment. But few of the GPU enthusiasts optimized the CPU code in order to make fair performance comparisons. It has been observed that a lack of "fair comparison" measures is responsible for exaggerated GPU performance.10 Second, competing technologies are mostly ignored by GPU adopters. Intel's Xeon Phi coprocessor, for example, which comes with 60 embedded Pentium cores, is capable of achieving a similar level of parallelism as GPUs.11–13 Adopting the coprocessor is relatively easy and a large number of them are, in fact, used in Tianhe-2—the world's number-1 supercomputer. The "heterogeneous computing" era has just begun and it is uncertain which hardware (and software) technology will dominate the market.14 The excitement brought by the GPU has reignited our interest in achieving real-time MC dose calculations and one should take full advantage of the research opportunities.15 However, an inflated expectation can be counterproductive, especially when investing in a single technology that may be obsolete in ten years. Rebuttal: Xun Jia, Ph.D. I agree that variations in reported GPU-acceleration factors exist due to different degrees of software/hardware utilization and optimization. However, it is quite difficult, if not impossible, to conduct an absolutely fair comparison. For example, I would like to mention the software aspect that unfairly treats GPUs: Software optimization schemes, such as variance reduction techniques widely employed in CPU-based MC packages, have been barely explored for GPUs. The deterministic nature of such algorithms is expected to be particularly favorable for GPU's single-instruction-multiple-thread structure. Yet it is absolute computational efficiency, rather than performance relative to CPUs, that determines the feasibility of near real-time MC calculations. The fact that a single GPU can already compute dose in seconds strongly supports this feasibility. Practicality should also be considered. While a low-end cluster with 4–8 computers may offer high speed, it is more advantageous in a clinical environment to use GPU-enabled computers in terms of energy efficiency, ease of management, etc. The utilization of GPUs in scientific computing is absolutely more than hype. Among the world's top 500 supercomputers, 46 of them use GPU-based coprocessors compared to only 17 systems with MIC coprocessors. A few major vendors in radiotherapy, e.g., RaySearch and Elekta, already employ GPUs in their products. I agree that multiple options are available to substantially accelerate MC in this era of booming technology. Intel MIC is a great example. Nonetheless, it too may be hype which only emphasizes the ease of programmability based on existing CPU codes but hides the required efforts of performance tuning. There is probably no single technology that is undoubtedly better than others. However, based on the overall consideration of GPU's advantages and developments so far, I believe that GPU technology is the hope for near real-time MC dose calculations. Rebuttal: X. George Xu, Ph.D. I agree with Dr. Jia that the capability of real-time MC dose calculations is within reach owing largely to the innovative technology and marketing strategies by Nvidia. The greatest roadblock to GPU is the fact that the effort to translate legacy MC codes to the new CUDA programming environment is prohibitively expensive. GPU also faces tough technological challenges, including limited memory and data bandwidth.14 Given the steep investment and market risk, for everyone to jump onto the GPU wagon is costly and unwise. To CPU enthusiasts, multithreading techniques such as OpenMP and Pthreads are readily available for parallel computing. Intel CPUs come with hyperthreading for concurrent execution, and various compiler options can be used for optimization. As a competing architecture, Intel's MIC is much easier to adopt. To avoid "unfair comparison" between GPU and CPU,11 one should consider the above-mentioned software optimization techniques and pick a "multicore" CPU (instead of a "single-core") at a similar price to the GPU implementation. Comparative studies should also consider software related labor expenses. When we recently compared the performances of ARCHER—an MC dosimetry code developed from scratch by my Ph.D. students11–13—in the CPU, GPU, and MIC platforms, we found that GPU's advantages as a dose engine are less dramatic than some of those reported in the literature. All things considered, traditional CPU clusters and MIC remain serious competitors to GPUs when energy efficiency is not the priority. In the next five years, all these technologies are expected to evolve rapidly. The potential waste of capital and human resources due to hype and misleading information should be avoided. To this end, peer-reviewed journal publication and grant application processes should emphasize balanced GPU studies that offer the best methodologies and practices to the medical physics community. REFERENCES 1X. Jia, P. Ziegenhein, and S. B. Jiang, "GPU-based high-performance computing for radiation therapy," Phys. Med. Biol. 59, R151– R182 (2014).10.1088/0031-9155/59/4/R151 2G. Pratx and L. Xing, "GPU computing in medical physics: A review," Med. Phys. 38, 2685– 2697 (2011).10.1118/1.3578605 3S. Hissoiny, M. D'Amours, B. Ozell, P. Despres, and L. Beaulieu, "Sub-second high dose rate brachytherapy Monte Carlo dose calculations with bGPUMCD," Med. Phys. 39, 4559– 4567 (2012).10.1118/1.4730500 4X. Jia, J. Schuemann, H. Paganetti, and S. B. Jiang, "GPU-based fast Monte Carlo dose calculation for proton therapy," Phys. Med. Biol. 57, 7783– 7797 (2012).10.1088/0031-9155/57/23/7783 5F. Shi, X. Gu, Y. Graves, S. Jiang, and X. Jia, "A real-time virtual delivery system for photon radiotherapy delivery monitoring," Med. Phys. 41(6), 432 (2014).10.1118/1.4889184 6Khronos OpenCL Working Group, "The open standard for parallel programming of heterogeneous systems" (2013), available at: https://www.khronos.org/opencl/.others. 7S. Hissoiny, B. Ozell, H. Bouchard, and P. Despres, "GPUMCD: A new GPU-oriented Monte Carlo dose calculation platform," Med. Phys. 38, 754– 764 (2011).10.1118/1.3539725 8D. Mackay, "Optimization and performance tuning for Intel®Xeon Phi™ coprocessors–Part 1: Optimization essentials" (2012), available at: https://software.intel.com/en-us/articles/optimization-and-performance-tuning-for-intel-xeon-phi- coprocessors-part-1-optimization.others. 9D. W. O. Rogers, "Fifty years of Monte Carlo simulations for medical physics," Phys. Med. Biol. 51, R287– R301 (2006).10.1088/0031-9155/51/13/R17 10V. W. Lee, C. Kim, J. Chhugani, M. Deisher, D. Kim, A. D. Nguyen, N. Satish, M. Smelyanskiy, S. Chennupaty, P. Hammarlund, R. Singhal, and P. Dubey, "Debunking the 100X GPU vs. CPU myth: An evaluation of throughput computing on CPU and GPU," in Proceedings of the 37th Annual International Symposium on Computer Architecture (ACM, New York, NY, 2010), Vol. 38(3), pp. 451– 460. 11T. Liu, X. G. Xu, and C. D. Carothers, "Comparison of two accelerators for Monte Carlo radiation transport calculations, NVIDIA Tesla M2090 GPU and Intel Xeon Phi 3120 coprocessor: A case study for x-ray CT imaging dose calculation," in Joint International Conference on Supercomputing in Nuclear Applications and Monte Carlo (SNA + MC 2013), Paris, France, 27–31 October (EDP Sciences, Les Ulis, France, 2014). 12L. Su, Y. M. Yang, B. Bednarz, E. Sterpin, X. Du, T. Liu, W. Ji, and X. G. Xu, "ARCHERRT—A photon-electron coupled Monte Carlo dose computing engine for GPU: Software development and application to helical tomotherapy," Med. Phys. 41, 071709 (13pp.) (2014).10.1118/1.4884229 13X. G. Xu, T. Liu, L. Su, X. Du, M. J. Riblett, W. Ji, D. Gu, C. D. Carothers, M. S. Shephard, F. B. Brown, M. K. Kalra, and B. Liu, "archer, a new Monte Carlo software tool for emerging heterogeneous computing environments," in Joint International Conference on Supercomputing in Nuclear Applications and Monte Carlo (SNA + MC 2013), Paris, France, 27–31 October (EDP Sciences, Les Ulis, France, 2014). 14B. R. Gaster, L. Howes, D. R. Kaeli, P. Mistry, and D. Schaa, Heterogeneous Computing with OpenCL, 2nd ed. (Elsevier, Inc., Waltham, MA, 2013). 15T. Friedman, Do believe the hype, New York times, 2 November, 2010, available at: http://www.nytimes.com/2010/11/03/opinion/03friedman.html?_r=0.others. Citing Literature Volume42, Issue4April 2015Pages 1474-1476 ReferencesRelatedInformation

HDR brachytherapy uses radionuclides such as iridium-192 at dose rates of 20 or more cGy/min to a designated target point or volume.1 The US Nuclear Regulatory Commission (NRC) regulation requires an authorized user and an authorized medical physicist to be physically present during the initiation of all patient treatments involving the unit.2 The rules and regulations of most of the states in the US require that the licensee shall comply with the provisions and the requirements of NRC regulations (Subpart H of 10 CFR Part 35). For example, “A licensee shall ensure that operators, authorized medical physicists and authorized users participate in drills of the emergency procedures, initially and at least annually thereafter.” However, it does not specify “who” would be the “HDR operators.” Many physicists interpret the regulation as being fulfilled when physicists are operating the unit but with the physician pressing the treatment-delivery button on the HDR console. On the other hand, some states do specify that either an authorized user or a licensed radiation therapist must operate the HDR unit during the administration of radiation to cancer patients.3–5 Based on my recent survey among physics colleagues, the operation of an HDR unit during treatment is still mainly done by physicists in many institutions. In my opinion, analogous to the regulations governing linac-based radiation treatments, the HDR unit, being a radiation-delivery device, should be operated by a licensed radiation therapist for patient treatments. One of the clinical benefits of having therapists operate the HDR unit is to improve the communication in the radiation oncology team and also have extra sets of safeguards to crosscheck treatment plans before delivery to the patient. Plus, most therapists are enthusiastic and willing to learn, and are motivated to operate the HDR unit, which promotes and increases their communication and technical skills in the field. It also shows to administrators that HDR is a joint effort including radiation oncologists, nurses, physicists, and therapists. Furthermore, in some clinics, therapists also perform the HDR daily QAs under a physicist's supervision; this practice is supported by numerous official documents.6–8 For example, in AAPM TG-59, it is clearly stated that the “radiation therapist executes daily QA protocol the morning of the procedure”;6 similarly, in IAEA-TECDOC-1257, it is also stated that “daily tests can be performed by a technician.”7 The physicist's role is to define the organization and responsibilities of the treatment-delivery team members and to provide for their training.6 In the past, since older models of HDR units were not directly interfaced and able to communicate with a Record and Verify (RV) system, physicists were heavily relied upon to ensure that treatment plans were being correctly transferred from the treatment planning system to the HDR unit. Currently, with many HDR units integrated into centralized information RV servers (i.e., ARIA or MOSAIQ), the chances of an HDR treatment plan being incorrectly transferred to a delivery unit is greatly minimized, and thus allows therapists to play more active roles in the operation of the HDR unit during treatments. From both the regulatory and treatment safety points of view, a therapist should not be the only person to operate an HDR unit for patient treatments. The U.S. Nuclear Regulatory Commission (USNRC) regulates the use of all reactor-produced materials (byproduct materials) including their medical use. The Code of Federal Regulations 10 CFR Part 35 requires, for HDR remote afterloader units, (i) an authorized user (AU) and an authorized medical physicist (AMP) to be physically present during the initiation of all patient treatments involving the unit; and (ii) an AMP and either an AU or a physician, under the supervision of an AU, who has been trained in the operation and emergency response for the unit, to be physically present during continuation of all patient treatments involving the unit.2 Currently, 37 states have entered into agreements with USNRC and assumed regulatory responsibility over most activities involving radioactive material within their states.9 Based on a survey of physicists working in different states, the current HDR treatment team can be a combination of the following professionals: AU, AMP, radiation therapist (RT), registered nurse (RN), and radiation safety officer (RSO). It varies across hospitals and from state to state. For instance, it can be AU + AMP + RN in Mississippi, AU + AMP + RT or AU + AMP + RN in California, AU + AMP + RN + RSO in Georgia, and AU + AMP + RT in Ohio. Who operates the console to administer HDR treatment also varies by state. Some states, such as Florida and New York, specify that only an AU or RT should administer an HDR treatment.3,4 Some states do not specify who should operate an HDR unit for treatment, e.g., in Mississippi, where an AU or AMP or RT can push the treatment button.10 In both agreement states and nonagreement states, the AU can execute an HDR treatment to meet the regulatory requirements. Both AUs and AMPs play essential roles in HDR treatments as recommended by different professional associations such as the American College of Radiology (ACR), the American Society for Radiation Oncology (ASTRO), and the American Association of Physicists in Medicine (AAPM).1,6 Although comprehensive pretreatment quality assurance is performed before an HDR treatment, the treatment process involves potentially high safety risks because it is given with a high activity source over a short duration. The AMP, who best understands the brachytherapy sources, risks, safety issues, machine delivery mechanism, and treatment planning, should take ownership of the treatment rather than simply act as a participant. When emergent situations appear, the AMP must take immediate action to minimize any potential risk. The potential roles of the RT during an HDR treatment can be fulfilled by existing team members, e.g., AU, AMP, and RN. The AU operates the console to execute the treatment; the AMP monitors the machine performance; the RN monitors the patient; the AU oversees the process. In conclusion, it is not necessary to have a therapist involved in an HDR patient treatment process, as it adds cost (of staff) and does not improve treatment quality or safety. I appreciate my colleague having listed all the different combinations of HDR treatment team members across the country, and I agree with his definitions of the essential roles of AUs and AMPs in the HDR process. However, I disagree with him that having a therapist in an HDR process adds extra cost and does not improve safety of HDR delivery. First, Dr. Yang stated, “The potential roles of the RT during an HDR treatment can be fulfilled by existing team members.” I disagree with this statement. Currently, most HDR brachytherapy treatments require pretreatment imaging for verification purposes.11 The RT plays very important roles in the imaging process, and is there to operate an imager—either by C-arm (standalone HDR suite), O-arm (intraoperative brachytherapy), or OBI kV imager (linac room). This imaging role cannot be fulfilled by any existing team member because none of those members have the proper licensure to operate the imagers. Since RTs are already involved in the process, their operating the HDR unit really adds no additional cost. Second, Dr. Yang stated that adding an RT does not improve safety. My view is that adding an RT would be definitely adding an additional layer of safety to the HDR process. On the other hand, having only a physicist operating the HDR unit may present a higher error probability due to lack of a self-checking mechanism. For HDR error prevention, the AAPM TG-100 has performed an FMEA for HDR brachytherapy, constructing fault trees and failure modes as a reminder of what could go wrong.12 Also, we should provide ongoing training, including equipment usage, for all parties involved, in addition to mandatory annual training. With all the safety procedures in place, now is the time to make it a requirement that therapists should operate HDR units for patient treatments. When using the analogy between an HDR unit and a linac, one cannot ignore the fact that federal or state regulations do not require the physical presence of both an AU and the AMP for linac-based radiation therapy, but for HDR treatments, they do. For linac treatments, the AAPM only recommends the physicist's physical presence for special treatment procedures involving large fraction doses, such as for stereotactic body radiation therapy.13 It is clear that during the initiation of all patient treatments involving HDR units, an AU can initiate treatment delivery while satisfying the regulations at both federal and state levels. The therapists do not have to be the only machine operators in radiation therapy. Since treatment quality and patient safety are assured by the existing HDR team members of AU, AMP, and RN, adding unnecessary personnel (a therapist) could be a distraction and may potentially lead to confusion in an emergency situation. Having the therapist as the exclusive operator of the HDR unit improves neither quality nor safety and is unlikely to improve communication between the HDR team members. In addition, the physicist has to make sure that the therapist understands the treatment procedure and treatment plan, which will need extra effort and time. Furthermore, it is important to be able to demonstrate to administrators that the HDR program is cost-effective. Adding a therapist to the HDR team potentially increases personnel costs. In summary, an HDR team of AU, AMP, and RN is effective in both cost and communication while satisfying regulations and maintaining optimal patient safety.

Arguing against the Proposition is Kevin L. Moore, Ph.D. Dr. Moore obtained his Ph.D. in Physics from the University of California, Berkeley, subsequently training and working at Washington University in St. Louis before moving back west to the University of California, San Diego. He is certified in Therapeutic Radiological Physics by the American Board of Radiology and is currently Associate Physics Director and Medical Physics Residency Director in the UC San Diego Department of Radiation Medicine and Applied Sciences. Dr. Moore's major research interests lie in knowledge-based treatment planning, treatment plan quality control, and informatics applications in clinical radiotherapy. He has published nearly 30 peer-reviewed papers and one book chapter, and is lead inventor on a patent regarding knowledge-based dosimetric prediction. Within a decade, radiation treatment planning will become fully automated without the need for human intervention because (i) we will exploit pertinent trends in the manufacturing and informatics industries, (ii) the precedent is already established, and (iii) it is imperative to improving quality and continuing advancements in care. Impressive technological advances have occurred in radiation oncology over the past two decades. Image-based planning, optimization, and guidance progressed rapidly from compelling concepts to routine tools because outside influences like high-performance computing, networking, and robotics became widespread and affordable. IMRT and IGRT have become ubiquitous tools and have altered the paradigm of treatment. But we wish to do more for our patients. The "adaptive" concept was also described more than 15 years ago. The concept has been developed extensively and now includes biologically motivated adaptation.1 However, more effort is needed to overcome the complexities and impact on workflow to realize adaptation as it was conceived. Future efforts will benefit surely from the "third wave of computing" from which image processing and information technologies will produce insights from large quantities of unstructured treatment planning data. The need to estimate dose distributions made automation central to the earliest developments of computerized treatment planning. Today, even more advanced functions are automated, such as image registration, organ delineation, and dose optimization. Using commercial tools, it is now possible to control workflow so as to fully create, evaluate and document a plan with minimal intervention.2–4 Interestingly, applications involving tangential breast irradiation remain controversial: In spite of the obvious improvements in personalization and efficiency afforded by IMRT and related automation techniques, modern innovation is discouraged because entrenched reimbursement guidelines confuse the technologies and the "modality" they enable.5 Providing healthcare is one of the most complex and demanding of human endeavors. Radiation oncology treatment relies on distributed decisions and tasks that are shared across highly skilled medical and technical staff. We strive to assess and respond to each patient's personal needs; but our tools, skills, and processes are stretched to the limit. Procedures can become error prone, sometimes with tragic and very public consequences.6 Consequently, practice guidance is limited to the structures and inspections required to achieve safety today.7 The dynamic nature of patients and their response to treatment were recognized long ago as a control problem. Adaptive control provides a means to account for anatomical and physiological variations and supports highly personalized treatment.1 Adaptive radiation therapy must become "more than safe." It must embrace a broader definition of quality to ensure that clinical decisions and technical procedures are evidence based, effective, equitable, timely, and highly tailored to each patient.8 A higher level of robust quality is required and we must do more than embrace automation. Robust quality is achieved by design rather than through organic innovation followed by inspection for quality control.9 Adaptation requires a framework to achieve robust quality that is safe, consistent, and highly customized. Within such a framework, care will become more complex unless automation is used to make it "merely complicated." As someone who intends to spend the next decade working to advance the proposition, I nonetheless contend that the odds of fully automated treatment planning being the norm in ten years' time must be rated as extremely unlikely. A close read of the proposition could make my task relatively easy, i.e., interpreting "fully automated" treatment planning to imply end-to-end automation, whereby all steps between radiotherapy simulation and first treatment are performed without human intervention. Impressive though the last decade has been for the field of autosegmentation, it strains credulity that a decade's time would be enough to herald a universal autosegmentation platform that not only identifies all normal anatomical structures across all imaging modalities but also flawlessly incorporates every patient's unique clinical circumstances into fully automated tumor volume contouring. Making my task somewhat more difficult, we could interpret the proposition to "merely" imply full automation from segmentation to treatment. Both my opponent3 and I10,11 have clinically implemented automated treatment planning using present-day technologies, and, undoubtedly, research and commercial offerings in this space will advance in the next ten years. However, we should appreciate the enormity of the challenge in effecting universal automation for all clinical scenarios. Using the impressive work of my opponent as an example, tangents in early-stage breast cancer can clearly be automated to a great effect, but I am skeptical that this algorithm can be easily extended to all breast cancer treatments, e.g., bilateral postimplant chest wall irradiation with internal mammary chain and axillary lymph node involvement, including electron scar boosts, for a patient with cardiac comorbidities. Such a case is both complicated and outside of normative experience, making the work of algorithmic development simultaneously more difficult, more time consuming, and less beneficial (in a utilitarian sense). As automated treatment planning advances, by necessity it will expand from common and standardized treatment sites to infrequent and nonstandardized cases. To automate everything we treat in radiotherapy will take time, and ten years is simply not enough of it. In fairness to the spirit of the proposition, I feel I should stake my own claim for 2024. Semiautomated (i.e., computer-assisted) treatment planning will be used in the large majority cases, with some form of knowledge-based and/or computer-aided multicriterial optimization removing most of the present-day human variability from the optimization process.12,13 The clinical expertise of humans will still be regularly employed to evaluate and adjust plans for patients whose circumstances fall outside of normative treatments. Automated software systems will be commonly available for online plan adaptation. Some reductions in treatment planning staff will occur, although job descriptions might expand to encompass increased needs in clinical informatics and adaptive plan management. Ironically, human-driven planning will probably retain the largest foothold in 3D-conformal/palliative treatments, where patient anatomical variations can be very large and nonstandard clinical considerations are a frequent occurrence. These changes will be breathtaking and practice altering, but will fall short of delivering fully automated treatment planning by the year 2024. As for 2034… I appreciate Dr. Moore's flexible viewpoint and the challenges he presents. Indeed, image segmentation is a major hurdle; contours are vital for communicating decisions and intent. We are poised to exploit vast stores of images and manually delineated organs,14 but current clinical practices may not provide what is required for algorithm training. We do build on "shifting sands" to some degree as technologies and practice standards evolve. But, the proposition does not "strain credulity" if manual contouring is approached with consensus and consistency. Dr. Moore believes clinical variation limits our capacity to automate planning. I agree to the extent that the "Pareto Rule" governs progress; i.e., 20% of our efforts will succeed for 80% of the cases. Clearly outliers require significant human effort, but reducing arbitrary variation and building anatomically related evidence to support continuing development could help. Dr. Moore also speculates that automation will reduce staffing, but concedes it could free skilled staff to add value to challenging cases and to advance appropriately personalized adaptation. As automation is introduced, it influences the tasks remaining, i.e., what staff are asked to do. Will we continue in familiar territory, or will new tasks differ qualitatively or become disconnected? We must also assure that it is possible to monitor and compensate for system deficiencies. If these deficiencies are ignored, there is a risk of new types of errors and system failures. In short, automation does not solve all problems.15 In his opening statement, Dr. Moore contends that the "odds" of fully automated treatment planning being the norm in ten years' time must be rated as extremely unlikely. In my opinion, the future should not be left to chance. We must move to achieve robust quality by design. I conclude by quoting Dr. Dennis Gabor, the Nobel Laureate who invented holography: "The future cannot be predicted, but futures can be invented."16 I absolutely align myself with the large portion of Dr. Sharpe's statement dedicated to how automation could improve radiotherapy, and thus will focus on the narrow portion of my opponent's argument that attempts to explain how automated treatment planning might come to pass. Relying on a third wave of computing to make this happen perhaps confuses more data with more knowledge. I would respond that analyzing prior information is a necessary but not sufficient condition, and while we must exploit prior information, we cannot expect that the mere possession of large quantities of data will herald miraculous gains. The recent advance of knowledge-based planning yielded useful predictions only when new techniques were brought to widely available data.2,11,13,17,18 Extending automation will rely on further research and development and, as argued in my opening statement, this will take time to expand to all clinical scenarios. As for the precedents that Dr. Sharpe introduced—image registration, organ delineation, and dose optimization—I would argue that none of these are yet fully automated. Image registration comes close, but in my experience automatic registrations are ultimately adjusted more often than not. I have contended (with great respect) that autosegmentation is not fully automated even after more than a decade's development. As for optimization, this is unfortunately the least automated of all, demanding further technological development to eliminate human-caused variability.12 Examples of full automation in radiotherapy are actually very few. One example is beam aperture definition, now automated by programmed multileaf collimators. Arguably the elimination of human block cutters did occur on a decade's timescale, but mere citation of this does not inform predictions for other technologies. In closing, I am not at all pessimistic about the future of automation in treatment planning; deployed in tandem with the clinical expertise of highly trained human beings, automation will improve patient care and make radiotherapy more efficient. We should, however, work toward this future with eyes open about the significant effort that remains to achieve fully automated treatment planning.

Arguing against the Proposition is Richard L. Maughan, Ph.D. Dr. Maughan obtained his Ph.D. in Nuclear Physics in 1974 from the University of Birmingham, England. From 1974 to 1983, he worked as a member of the scientific staff of the Cancer Research Campaign Gray Laboratory at Mount Vernon Hospital in England, where he was involved in basic radiation physics, chemistry, and radiation biology research. He moved to the USA in 1983 when he took a position as a medical physicist and a member of the faculty in the Radiation Oncology Department of Wayne State University, Detroit, where he played a major role in development and application of a superconducting cyclotron for neutron radiation therapy. In 2000, Dr. Maughan moved to the University of Pennsylvania as Professor and Director of the Medical Physics Division, where he is currently Department Vice-Chair. His research interests are particle therapy with neutrons, heavy ions, and especially protons. In conventional therapy, multileaf collimators are used for 3D conformal and intensity modulated dose delivery. This debate on proton therapy mostly revolves around the use of MLCs for simple final collimation. Gottschalk1 describes the potential detriment of employing a universal MLC (a single device for all patients) for this purpose in passively scattered proton fields. The bulky device necessitates a large air gap which, in combination with a large double-scattering source size, degrades the lateral penumbra. The effect is aggravated by the range compensator, an additional source of scatter placed far away from the patient surface. Decreased distal conformality is an additional consequence of a large air gap. Depending on leaf size, the scalloping effect, even though mitigated at treatment depth by multiple Coulomb scattering, worsens lateral conformality and falloff. Additional issues arise from thick collimators, ranging from even further degradation of lateral penumbra to added complexity in dose modeling. For a significant fraction of proton indications, the dosimetric effects listed above will result in lower target doses due to the targetˈs proximity to critical structures. The optimum dose distribution is achieved by a thin collimator as close to the patient surface as possible. This holds true for uniform (US) and pencil beam scanning (PBS) deliveries, albeit to a lesser extent given the smaller source sizes. Efficiency in todayˈs proton therapy is still not on a par with conventional photon radiotherapy. The use of MLCs could help but, unlike many other improvements the proton community is presently tackling, it comes at the cost of compromised dose distributions. Pencil beam scanning is generally regarded as the future of proton therapy delivery, replacing passive scattering as the primary mode for large proton field production. PBS systems promise great flexibility in dose application, comparatively simple system QA, and much more efficient treatment planning and delivery. PBS permits intensity modulated proton therapy. This can also be achieved with a multileaf collimator, analogous to conventional therapy. The general feasibility has been shown,2 but its performance depends on implementation and the properties of the delivery system. Treatment time may prove problematic, and maximum field size continues to pose a challenge. In summary, the use of MLCs for final collimation in passive scattering and scanned beam delivery comes with degradation of the dose distribution. Use for intensity modulation in passive scattering may improve the quality of dose distributions, but is linked to increased effort in quality assurance and will most likely be outperformed by pencil beam scanning systems. The use of a multileaf collimator for proton therapy has not been widely investigated. The HIMAC facility in Hyogo, Japan, uses an MLC with a 12C beam in the treatment of extracranial lesions.3 Daartz et al.4 investigated the use of a mini-multileaf collimator (MMLC) for use in passively scattered proton beam therapy for intracranial lesions. At the University of Pennsylvania Roberts Proton Therapy Center, four MLCs are used on gantry mounted nozzles for shaping passively scattered and uniformly scanned beams for all treatment sites. These MLCs were built in a joint collaboration between the University of Pennsylvania, Ion Beam Applications, SA (Louvain-La-Neuve, Belgium) and Varian Medical Systems (Palo Alto, CA). The primary reason for installing these devices was to improve beam delivery efficiency. The MLC has a leaf width giving a 5 mm projection at the isocenter and beams of up to 25 × 18 cm2 are delivered in double scattering and uniform scanning mode. Data show that the 20%–80% penumbra achieved with a beam of range 22 cm, modulation 10 cm, measured using film at the isocenter, at depth of 17 cm in solid water, with a collimator-to-surface distance (CSD) of 16 cm, and a 10 × 10 cm2 field size is 9.6 ± 0.6 mm. This penumbra can be compared with data from Fig. 3 of Safai et al.5 where the penumbra for an unmodulated beam of 22 cm range, with a CSD of 10 cm and a brass aperture, of unspecified size, is 7.9 ± 0.2 mm. Oozeer et al.6 showed that penumbral width as a function of depth is practically independent of modulation and that variations with field size are also small. The differences in these penumbras can most likely be attributed to the differences in the CSD, which introduces more air scattering in the case of the MLC measurements. CSD may be an issue with an MLC, since their large dimensions compared to brass apertures when used for treating small and intermediate sized fields, dictate that the MLC be positioned at a greater CSD than for an aperture. The collimator housing, therefore, is designed to have a D-shape allowing access over the patientˈs shoulder, minimizing the air gap for brain and base of skull treatments. For head-and-neck patients many fields are large, extending below the shoulder, which require larger air gaps even with brass apertures. The collimator has tungsten leafs; concerns about leaf activation and excessive neutron dose7,8 have proved to be unsupported.4,9,10 The University of Pennsylvania experience in using an MLC with double scattered and uniform scanning beams has shown it to be convenient in use and well suited to a department where a large number of complex treatments are delivered with two or more fields. It increases efficiency and eliminates the need for the storage of brass apertures. Our clinical experience confirms the conclusions of Daartz et al.4 that there are “only small differences in the dose distributions obtained with brass apertures and the MMLC.” We are focusing the debate on the dosimetry of multileaf collimators as used for final collimation. Once more we end up in a stalemate between “penumbra is always worse” and “but not significantly.” It seems that we agree: it is worse. How much worse? That depends on the device. It should be obvious that the results of our 2009 study performed using a mini-MLC with 2.5 mm leaf size and a maximum field of 8 × 6 cm2 cannot be transcribed to a large universal MLC, necessitating larger CSD, with 5 mm leaf thickness.2 The penumbra in a collimated beam is a function of depth, energy, CSD, SAD, source size, and range compensator thickness. Unless obtained under the same conditions, since the varying parameter is the mode of collimation, penumbral widths should not be compared. Dr. Maughan mentions penumbral widths for a single beam energy, without information on source size or SAD, CSD varying by 6 cm and a depth of measurement of 17 cm for the MLC and 21 cm for the aperture. But assuming this comparison is valid—even with a D-shaped MLC housing, minimizing CSD, there is little doubt that air gap is still larger than for custom-milled apertures. Considering the cranial component of a head-and-neck treatment plan, where the most critical organs are to be spared, the air gap with apertures is 2–3 cm. Widening of the penumbra by 2 mm will result in notably decreased target coverage. In addition, these numbers neglect the effect of a range compensator—amplifying the impact of increased CSD. This reasoning disqualifies the use of a large, thick-leaf universal MLC from application to cranial sites. Does the argument of gained efficiency still hold if one has to employ a different mode of collimation for a rather large subset of our typical proton patients? Other than qualitative statements, there are no data quantifying the effect of universal MLCs on efficiency. Commissioning and routine quality assurance for an MLC are a substantial effort. In addition, it is one more device that potentially fails during operation and causes downtime. The limitation in field size necessitates splitting fields into abutting areas more frequently. Despite the promise, MLCs have not been widely adapted in the field. Perhaps, the practical advantages are not that convincing after all. As pointed out by Dr. Daartz, the future of proton therapy lies in the development of more efficient pencil beam scanning systems; this requires faster scanning times and, more importantly, faster layer switching times. PBS has multiple advantages over passive scattering; better dose conformality (especially on the proximal field edge), the ability to treat large fields (up to 30 × 40 cm2) without field matching or patching, and the possibility of intensity modulated proton therapy. However, presently the treatment of moving targets with PBS remains problematic, since gating, breath-hold, and over scanning all reduce efficiency significantly. The University of Pennsylvaniaˈs five treatment room system was originally configured with two gantries with passive scattering, US, and PBS, two gantries with passive scattering and US, and a fixed horizontal beamline with PBS only. Experience treating with PBS in the fixed beam and in one gantry room has convinced us that our final room configuration will comprise three PBS rooms (two gantries, one fixed beam) and two gantries with passive scattering. With this combination, it will be possible to select the patients best suited for treatment with MLC based passive scattering and those requiring the improved dose conformality of PBS beams, which have beam-spot diameter σ-values of 3 and 4 mm in air at 230 MeV, for the fixed beam and gantry rooms, respectively. Targets with appreciable motion will be treated with passive scattering until robust solutions for motion management of PBS delivery are established. The use of an MLC with PBS will be challenging as the MLC required to treat large fields would be impractically large. Developing beams with a smaller sigma may be a better solution for improving penumbra. In conclusion, we find that using an MLC for proton beam shaping does not compromise our treatment plans, but leads to greater efficiency in handling the patient throughput. This work was supported in part by the US Army Medical Research and Materiel Command under Contract Agreement No. DAMD17-W81XWH-04-2-022. Opinions, interpretations, conclusions and recommendations are those of the author and are not necessarily endorsed by the US Army.

Arguing against the Proposition is Clive Baldock, Ph.D. Dr. Baldock is the Executive Dean of the Faculty of Science at Macquarie University, Sydney, Australia which he joined in 2012 from the University of Sydney where he was previously Head of the School of Physics. He graduated from the University of Sussex, Brighton, United Kingdom with a B.Sc. (Hons) in Physics and was subsequently employed as a trainee medical physicist at Guyˈs Hospital, London while studying for his M.Sc. in Radiation Physics at St Bartholomewˈs Medical College, University of London. He subsequently worked in a number of UK hospitals providing scientific support to clinical nuclear medicine and MRI services. His main research interests were in the field of the MRI of radiation sensitive gels for improved three-dimensional radiotherapy dosimetry for which he received his Ph.D. from Kings College, University of London. Dr. Baldock moved to Queensland University of Technology, Brisbane, Australia in 1997 and subsequently worked at the University of Sydney from 2003 to 2012. In 2010, he completed a Master of Tertiary Education Management at the University of Melbourne, Australia. His current research interests continue to be in the fields of radiation therapy, dosimetry, and medical imaging in which he has published over 140 research papers. Ranking organizations conduct an extraordinary amount of data collection, statistical analysis, and number crunching on such a wide variety of topics to boil down questions such as “Which institution has the best graduate program in a given area, or is the ‘best’ overall?” into a single numerical result. University level ranking systems frequently generalize across a large number of responses, rarely take into account posteducation performance, and inherently suffer from conflicts of interest.1 Subject-specific rankings can help ameliorate some of these issues, but such an index does not presently exist for medical physics educational programs. Ideally, career decisions should involve a broad range of metrics focusing on the field, the location, and the desires of the individuals involved. Rankings are not an absolute indicator of performance or quality and can be disregarded safely. The European University Association (EUA) has published two reports, one in 2011 and another in 2013, analyzing international university rankings and describing the benefits and pitfalls of each.1,2 They conclude that rankings most accurately reflect research produced overall and not education quality. Metrics used to evaluate education quality vary drastically between the systems analyzed by the EUA, some metrics bearing little relevance to education. Rankings are typically done via surveys of the higher education deans and presidents of the universities being ranked, oftentimes asking only about the reputation of a university. We must also consider that none of these metrics evaluate such a small subspecialty as medical physics in any statistically relevant manner. A more appropriate metric in North America for an initial evaluation of candidates or programs would be CAMPEP accreditation status, since applications for board certification are now contingent upon program accreditation. The accreditation process is a stringent and well-defined system to help standardize medical physics graduate education.4 Accreditors consist of board-certified physicists reviewing the work of other physicists. Accredited programs ubiquitously adhere to the standards proposed by TG-197 defining the requirements for graduate education in medical physics.4,5 Students coming from accredited programs have a proven higher passing rate for the ABR certification exams compared with students from nonaccredited programs. For the sake of argument, let us assume that university rankings do reflect the relative quality of a medical physics program. If this were true, then we would expect to see a correlation between university ranking and CAMPEP accreditation.6 As of July 2013, there were 44 CAMPEP accredited graduate programs. We binned these with U.S. News and World Report College Rankings3 for undergraduate education, research achievements, and medical school and found that there is no distinct correlation with ranking quality. About half of schools with CAMPEP programs are listed as unranked or unpublished. Those that are ranked are relatively uniformly distributed across all subgroups. The overall goal of any career or job search is to find the best match between employer and employee. While institutional rankings and educational acclaim are useful tools in evaluating the options, they should not be used as a method to filter out potential candidates. A number of metrics are used to assess the success and productivity of universities and their researchers therein.7 Internationally, much emphasis is given to the rankings of universities and the production of associated league tables,8,9 with much anticipation each year among university administrators, funding agencies, and students when a number of international ranking agencies publish their latest ranking lists.10 Such rankings, now a standard feature, are playing a significant role in a changing higher education landscape internationally with implications for many, whether realized or not. The practice of university rankings dates back to the beginning of the 20th Century with the publication of Where We Get Our Best Men. The backgrounds of “Englandˈs most prominent and successful men of the time” were evaluated with particular reference to where each studied. This resulted in a listing of universities ranked by the number of distinguished alumni that the universities could claim.11 Subsequently, graduate programs in United States universities were ranked on the basis of peer reputation.12 More comprehensive rankings of universities began being published from 1983 when the US News and World Report initiated ranking college undergraduate education programs with this ranking being published annually from 1987. Since 2003 numerous university rankings have been published with some now becoming particularly popular. Some of the most well known include the Academic Ranking of World Universities (ARWU) from Shanghai Jiao Tong University in China, the QS World University Rankings, the Times Higher Education World University Rankings, and more recently, the Leiden University Rankings. Despite ongoing debates about the use and validity of university rankings, they enable students as consumers to compare institutions within a country and around the world as they make decisions regarding which university to potentially attend. Further, for many university presidents and administrators rankings influence organizational missions, strategies, personnel, recruitment, and public relations.8,9 Furthermore, rankings often drive the allocation of resources with decision makers and administrators sensitive to the resulting prestige that may be associated with ranking performance.13 Internationally, governments and funding agencies are also increasingly using rankings as policy instruments to assess the performance of higher education institutions.14 Students will potentially make future choices of what and where to study, whether it is a graduate program in medical physics or biomedical engineering, based on where a university lies in a particular ranking. Such choices will not necessarily be based on which graduate programs are of higher “quality.” It would be unfortunate for prospective students to reference only rankings as an indication of program quality. As indicated in “The Role and Relevance of Rankings in Higher Education Policymaking,”14 rankings should be used “only as part of overall system assessment efforts and not as a standalone evaluation of colleges.” We should urge up-and-coming students to view rankings as a metric that does not necessarily guarantee program quality, student success, or eligibility for future career goals. Arguably, at least in North America, the best indicator to guide program decisions in medical physics is CAMPEP accreditation. Our experience at the University of Toledo supports the idea that rankings can be safely ignored. The radiation oncology medical physics program was CAMPEP accredited in 2009, but has been in existence since 1979, producing highly successful medical physicists who are leaders in our community today. Since 2001 alone, we have produced 52 graduates with masters and 12 with doctoral degrees in medical physics. One hundred percent of these graduates have found employment in the field of medical physics. Our program has been involved in testing several emerging systems, such as the first MLC programs, compensator-based IMRT, and retrofitting a micro-MLC and IORT onto a linear accelerator. Our graduates have been directly involved in these efforts, exposing them to a broad range of clinical and research experience, more than the majority of graduate programs. However, the University of Toledo is one of many schools whose rank is not published in any of the three categories (undergraduate, research, or medical school). Our program, with a throughput of five students per year, and our specialty in general, are too small to be accurately sampled by a large rankings program. Present experience and numerous reports from experts within higher education all point toward the ineptitude of large-scale ranking systems to adequately capture the true quality of a program dedicated to medical physics. Ignoring the overall rank of an institution can be done without any added peril to the student. Over the past decade, the Council of Graduate Schools (CGS) has conducted a multiyear examination of international graduate application, admission, and enrollment trends from overseas students seeking masters and doctoral degrees from US colleges and universities.15 With international students comprising 15% of all graduate students in the US,16 the US has for many years been the destination of choice for students from overseas with the significant majority enrolling from China. Recently however, this trend has started to change, with the number of Indian students entering US graduate schools increasing remarkably while the share of new graduate students from China has increased only modestly. In 2013, graduate enrollments from India increased by 40% with, interestingly, those from Brazil rising significantly by 17% as a result of the Brazilian government funding large-scale scholarship programs to send students abroad, particularly in the sciences.15 After seven consecutive years of double-digit growth, however, the number of Chinese students enrolling in US graduate programs in 2013 increased by only 5%. This should be cause for concern for US graduate programs that have in recent years relied on student growth from China to offset weak domestic enrollments, particularly in the sciences and engineering. Without the significant number of new students from India, the second-largest source of overseas students, overall international enrollments would have increased only slightly. University and clinic-based medical physicists often rely on graduate students to assist them as part of the normal practice and culture of undertaking research. As the number and profile of overseas students enrolling in graduate programs changes, there may be fewer students over time choosing to enroll in graduate programs in medical physics to pump-prime the pipeline of future graduate students undertaking research in this discipline. Overseas students interested in enrolling in North American medical physics graduate programs will undoubtedly seek to be well informed as they consider and make their decisions. Inevitably university rankings will be a consideration for some.17 Universities that are proactive in recruiting international students are potentially able to overcome, to some extent, a perceived low ranking in internationally recognized university league tables. To this end, it is valuable for medical physicists aspiring to have a research career to proactively avail themselves of university rankings as they develop their careers and research teams into the future. If the trend of students using university rankings to inform and assist in the making of their choices, all who ignore this issue potentially do so at their peril. Dr. Parsai wishes to thank Sean Tanny and Nicholas Sperling, Ph.D. candidates in UT Medical Physics program, for helpful discussions.

Arguing against the Proposition is Alan Wassyng, Ph.D. Dr. Wassyng earned his Ph.D. in Applied Mathematics from the University of the Witwatersrand, Johannesburg, South Africa in 1979. After spending 14 years as an academic, first at the University of Witwatersrand and then at the University of Minnesota, he incorporated a computer consulting company in Toronto, Canada. He returned to academia in 2002, joining the Department of Computing and Software at McMaster University, Hamilton, ON, Canada, where he is currently Associate Professor and Director of the Centre for Software Certification. Dr. Wassyng has published widely on software certification and the development of dependable embedded systems. He is cofounder of the Software Certification Consortium, and is Co-PI on the highly funded “Certification of Safety-Critical Software Intensive Systems” project led by McMaster University. Presentations at the Canadian Organization of Medical Physicists (COMP) Winter School1 showed that medical physicists are deeply imbued in a safety culture. They react instinctively within this culture, pay attention to human-technology interaction, and exhibit due process in the light of safety concerns. I compare this to the environment of a software engineering colleague who specializes in testing: she lives in a volatile, market-driven, and cost-minimizing environment. Even though she has years of experience in testing different products, her instincts and her quality goals are different from those of a medical physicist. The software engineering literature does not acknowledge the need for the conjunction of computational software design processes with a deep safety culture, which is required for deployment of software used to support clinical decision making. Instead, such software is confused with either control software, which directly operates a medical device, or commercial products where patient wellbeing is not directly affected by the correctness of the output. As a result, there are no guidelines in the software engineering literature that address the specific characteristics and needs of clinical software. When advising on software quality guidelines, a typical software engineer takes a broad-spectrum approach. This approach suffers from two serious flaws. First, it encourages the perception that software is correct unless proven otherwise. This dangerous assumption has been a contributor to several fatal accidents in the safety-critical world.2 A recent article3 talks about problems “when a computer lulls us into a false sense of security”. Second, this broad-spectrum approach does not use the knowledge of the people associated with the software, and does not acknowledge the specifics of the operational environment that these people understand. Vessey and Glass4 criticize the software engineering community for their broad, generic solutions, calling them weak solutions. They contend that the most effective and strong solutions are those that target the specific environment or situation. Medical physicists, with their knowledge, can provide this strong solution. In a 2012 survey of software development and maintenance practices sent to medical physicists across Canada, the medical physics community demonstrated that it already has a level of understanding necessary to provide meaningful software quality guidance to its own community. The devil is in the details and the medical physicists who responded to the survey understood the details of their environment and the implications of problems. This understanding is far beyond what a software engineer outside the clinic can bring to the table. Kendall and Post, after studying decades of development of nuclear arsenal simulation software at the Los Alamos National Laboratory, concluded that the best people to draw up a list of “best practices” for software development and maintenance are the members of the code project teams themselves, and that these practices are those “… that the teams have judged useful for improving the way they do business”.5 It is the same for medical physicists. They know the best ways to assess their software in order to safely do their business. Software safety – under defined conditions, software should not contribute to unsafe behavior or generate results that can lead directly to harm; Software security – protection afforded the software to keep it from harm, and from causing harm through users maliciously bypassing the softwareˈs designed-in safety and dependability; Software dependability – in its intended environment, the software can be trusted to produce the outputs for which it was designed, with no adverse effects. Lack of continuity8 (Moderator: in software engineering, continuity refers to a continuous function for which, intuitively, small changes in the input result in small changes in the output) – If a bridge is built to withstand a force of 100 tonnes, because of the mostly continuous behavior of physical objects and our mathematical models, the bridge is likely to be safe for loads less than or equal to 100 tonnes. We do not expect it to collapse if a bicycle goes across it. In contrast, a simple error in software that finds a number within a list of n numbers could easily result in the software working only when n is even. This happens if the designer separates the behavior into two cases (n is even and n is odd) and forgets to deal with the one case (n is odd). Thus, in this case, the software will work when n = 1000, but not for all n < 1000 – it fails when n = 19, for example. Information hiding9 (Moderator: information hiding is a software development technique in which each moduleˈs interfaces reveal as little as possible about the moduleˈs inner workings and other modules are prevented from using information about the module that is not in the moduleˈs interface specification)7 – This is a specific way of performing modularization so that the resulting design significantly improves safety and dependability when changes are made. SQA monitors the development process of the software so that the resulting product is, and remains, safe, secure, and dependable. This should include ways of evaluating the information hiding aspects of the design. Similar principles are software testing,10,11 hazard analysis,12 the absolute need for requirements traceability,13 and semantics for module interface specifications,14 and numerous others. How can medical physicists take these principles into account when they do not possess this fundamental software knowledge? The person who does have this knowledge is a software engineer (or perhaps, a computer scientist). The Professional Engineers Act of Ontario [“Professional Engineers Act”, Professional Engineers Ontario, http://www.e-laws.gov.on.ca/html/statutes/english/elaws_statutes_90p28_e.htm/] states that the mandate of an engineer is “to ensure that the public interest may be served and protected.” A software engineer is tasked with developing software-dependent systems and protecting the public from harm caused by those systems. There are definite differences in SQA requirements in different domains – which is why some people think it is appropriate to have an MP establish SQA guidelines. However, there are more commonalities across different domains than there are differences, so software knowledge is of primary importance. SQA is a team effort, with domain expertise coming from the MP. However, the core knowledge and guidance must come from a software engineer. My colleague built his argument around a software engineering list of software qualities: safety, security, dependability. Why not a list from scientists: accuracy, trustworthiness, readability, consistency with the physics, and simplicity? Apparently, scientists focus on simplicity far more than software engineers.15 For any set of qualities, we still need to achieve them. My colleague suggests, for example, information hiding and requirements traceability as fundamental principles for anyone. David Parnas, who first published the information hiding principle, complained in an invited talk16 that most software engineers do not know how to properly apply the principle. Several Standish Group surveys reported that only 9%–16% of software projects are delivered successfully, largely because software engineers do not understand user requirements.17 But medical physicists do understand their user requirements because they are the users. How do we add quality to a piece of software? It is well understood18 that there is a disconnect between the desire for the high-level quality and what low-level activities actually achieve it. There are no common activities or “silver bullets” such as “information hiding” to achieve, for example, “dependability”. There is no research that demonstrates that particular code-level activities guarantee any high-level quality. The best we can do is to understand the particular software in front of us. In the case of clinical software, medical physicists have the best understanding of what is in front of them in terms of use and the physics embedded in it. By keeping the software code very simple (which scientists have a tendency to do15), medical physicists are in the best position to decide what further keeps them out of trouble. My colleague suggests that software engineers have a mandate to ensure that the public is served and protected. Software engineers do not always live in that culture. Medical physicists live in a safety culture and they have the capacity to fully understand what will best achieve quality software for their own work. I have been involved in the COMP Winter School1 since its inception (I missed one year), and give a talk each year on medical device software. I have been told that the medical physicists find this talk disturbing because they are surprised to find out how much they do not understand about software! Their safety culture is, indeed, admirable, but this does not make up for a lack of technical (software and system safety) knowledge. SQA needs a team of people, including domain experts and software experts. The SQA lead must have both the software expertise and the requisite safety knowledge. The fact that some software engineers are not immersed in the system safety world should not lead us astray. We see more and more software engineers who work in the medical domain, understand software, and have developed expertise in what is required in a regulated domain, in which safety is a primary concern. If there are not enough software engineers with this safety focus, we should be championing changes to the software engineering curriculum. There are conferences19,20 targeted at software engineering in the medical domain. These conferences focus on medical devices and reporting/planning software. IEC standard 62304 focuses on safety of software used in medical devices.21 Just because the standard applies to medical devices does not mean it is irrelevant for other clinical software. Software that can impact the health and safety of patients is deemed to be a medical device in most regulatory regimes. There is a (growing) body of software engineers who do have the specific software expertise required, as well as familiarity with basic safety concepts and regulatory guidelines. They are in a much better position to establish quality assurance guidelines for medical software than are medical physicists.

Medical PhysicsVolume 41, Issue 7 070601 Point/CounterpointFree Access Point/Counterpoint: Low-dose radiation is beneficial, not harmful Mohan Doss Ph.D., Mohan Doss Ph.D. Diagnostic Imaging, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111-2497 (Tel: 215-214-1707; E-mail: mohan.doss@fccc.edu)Search for more papers by this authorMark P. Little Ph.D., Mark P. Little Ph.D. D.Phil. Radiation Epidemiology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892-9778 (Tel: 240-276-7375; E-mail: mark.little@nih.gov)Search for more papers by this authorColin G. Orton, Colin G. Orton Moderator Diagnostic Imaging, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111-2497 (Tel: 215-214-1707; E-mail: mohan.doss@fccc.edu)Search for more papers by this author Mohan Doss Ph.D., Mohan Doss Ph.D. Diagnostic Imaging, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111-2497 (Tel: 215-214-1707; E-mail: mohan.doss@fccc.edu)Search for more papers by this authorMark P. Little Ph.D., Mark P. Little Ph.D. D.Phil. Radiation Epidemiology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892-9778 (Tel: 240-276-7375; E-mail: mark.little@nih.gov)Search for more papers by this authorColin G. Orton, Colin G. Orton Moderator Diagnostic Imaging, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111-2497 (Tel: 215-214-1707; E-mail: mohan.doss@fccc.edu)Search for more papers by this author First published: 12 June 2014 https://doi.org/10.1118/1.4881095Citations: 26AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat OVERVIEW The recent rush to embrace the concept that diagnostic x-ray procedures are being overused, or that doses are too high and need to be reduced, is based upon the assumption that low doses of radiation are harmful and should be avoided as much as possible. On the other hand, some believe that such low doses of radiation are not harmful and might even be beneficial. This is the premise debated in this month's Point/Counterpoint. Arguing for the Proposition is Mohan Doss, Ph.D. Dr. Doss obtained his Ph.D. in Physics in 1980 from Carnegie-Mellon University, Pittsburgh, PA and then spent the next ten years in research positions at the University of Washington, Seattle, Lawrence Berkeley Laboratory, Berkeley, CA, and the Saskatchewan Accelerator Laboratory, Saskatoon, Canada. He then began his career as a Diagnostic Physicist at Regina General Hospital in Regina, Canada. In 2001 he joined Fox Chase Cancer Center Philadelphia, where he is now Associate Professor. He is certified in Nuclear Medicine Physics by the Canadian College of Physicists in Medicine. Dr. Doss's major research interests include biodistribution and dosimetry of new PET imaging agents, small animal PET imaging, and health effects of low dose radiation, and he has published over 50 papers. He is the recipient of the 2014 Outstanding Leadership Award in the field of dose-response by the International Dose-Response Society. Arguing against the Proposition is Mark P. Little, D.Phil. Dr. Little obtained his D.Phil. in Mathematics from New College, Oxford in 1985. He then worked for the next six years at British Coal, Harrow, London, and Berkeley Nuclear Laboratories, Nuclear Electric, Berkeley, UK. He then continued with his career in epidemiology first as Principal Scientific Officer, Epidemiology Group, NRPB, Chilton, UK, and then in the Department of Epidemiology and Biostatistics, Imperial College Faculty of Medicine, London, UK. In 2010 he moved to the USA as Senior Investigator at the Radiation Epidemiology Branch, National Cancer Institute, Rockville, MD. Dr. Little's major research interests have included models and epidemiological studies of cancer induction by radiation, risks associated with mobile phones, cancer risks of radiation exposure of children, and deleterious effects of occupational radiation exposures, on which he has published over 150 papers and supervised the work of 20 researchers and graduate students. FOR THE PROPOSITION: Mohan Doss, Ph.D. Opening Statement The process of oxidative metabolism in living beings sometimes results in the production of free radicals which can cause oxidative damage. Our body has an elaborate system of antioxidants to neutralize these free radicals. This system is not perfect, and a small amount of damage does persist. There is evidence that accumulation of such damage contributes to causing many of the aging-related diseases. When free radical production is increased, e.g., from low-dose radiation (LDR) exposure (or increased physical/mental activity), our body responds with increased defenses consisting of increased antioxidants, DNA repair enzymes, immune system response, etc. referred to as adaptive protection.1 With enhanced protection, there would be reduced cumulative damage in the long term and reduced diseases. The disease-preventive effects of increased physical/mental activities are well known. There is considerable evidence from animal studies supporting the hypothesis that LDR reduces the likelihood of cancer as well as nonmalignant diseases.2 For humans, (i) epidemiological studies of irradiated populations exhibit reduced risk of cancer from LDR,3–5 (ii) interspersed adjuvant LDR treatment has resulted in better tumor control and reduced metastases in radiation therapy of non-Hodgkin's lymphoma patients,4 and (iii) tissues subjected to LDR have shown reduced second cancers per kg in radiation therapy patients.4 For noncancer diseases in humans, LDR has been shown to control many such diseases.2,6 Thus LDR is indeed beneficial, as it results in reducing cancer and noncancer diseases. The present concerns over the carcinogenic potential of LDR are based on the concepts that LDR causes DNA damage resulting in increased mutations, and that the accumulation of mutations can transform a normal cell into an uncontrollably dividing cell, causing cancer.7 This argument unjustifiably ignores LDR adaptive protective responses.1 If the effect of LDR adaptive protection is included, there would be reduced DNA damage following LDR,1 reducing the likelihood of transformation of normal cells into those with malignant phenotypes. Also, the above mutation model of cancer cannot explain the more than 100% increase in cancers in organ transplant patients (and in AIDS patients), in whom the immune system is suppressed.8 Hence there is little credibility in the prediction of a small percentage increase in cancer from LDR based on this model. On the other hand, using immune system deficiency as the cause of clinical cancer, many of the characteristics of cancer incidence can be explained.2 Since LDR boosts the immune system, LDR would be expected to reduce rather than increase the risk of cancer.1,2 For both cancer and noncancer diseases, there is a threshold dose below which no increased risk of disease has been observed. The atomic bomb survivor data, considered to be the most important data for estimating radiation effects in humans, have traditionally been used to justify LDR carcinogenic concerns. Recent reanalysis has shown the data are more consistent with a threshold, or radiation hormesis, model than the linear nonthreshold (LNT) model.4,5 In view of the above, we can conclude confidently that low-dose radiation is beneficial, not harmful, from both mechanistic and epidemiological considerations. AGAINST THE PROPOSITION: Mark P. Little, D.Phil. Opening Statement The detrimental tissue-reaction (deterministic) and stochastic effects associated with moderate and high dose ionizing radiation exposure are well known.9 In contrast to tissue-reaction effects, for stochastic effects scientific committees generally assume that at sufficiently low doses there is a positive linear component to the dose response, i.e., that there is no threshold, or beneficial effect.9 Moreover, there is accumulating direct evidence of excess risk of cancer and various other health endpoints in a large number of populations exposed at moderate and low doses. I review some of this evidence below. There is evidence of excess cancer incidence of most types associated with radiation exposures of the order of 10–20 mGy from diagnostic x-ray exposure in the Oxford Survey of Childhood Cancers and in various other groups exposed in utero.10 These data remain somewhat controversial, but as Wakeford and Little note “the consistency of the childhood cancer risk coefficients derived from the Oxford Survey and from the Japanese cohort irradiated in utero supports a causal explanation of the association between childhood cancer and an antenatal x-ray examination found in case-control studies. This implies that doses to the foetus in utero of the order of 10 mSv discernibly increase the risk of childhood cancer.”10 There is also evidence of excess risk of childhood leukemia associated with natural background radiation exposure, at doses above 5 mGy, in a large UK population-based case-control study.11 At slightly higher doses, increased risks of leukemia and brain cancer have been observed in patients who were exposed as children to multiple computerized tomography examinations resulting in doses of about 60 mGy to the respective tissues (red bone marrow, brain).12 The excess risks in all of these studies are consistent with those in the Japanese atomic bomb survivor data.10–12 The health risks of low-level exposure to ionizing radiation have been assumed to be related primarily to cancer.9 Evidence has recently emerged of an association between lower doses (<0.5 Gy) and late circulatory disease. In particular, a recent systematic review and meta-analysis suggested an excess radiation-associated risk at occupational and environmental dose levels (<0.5 Gy).13 However, the presence and magnitude of the excess circulatory disease risk at low doses is still relatively controversial, and much remains unknown as to the shape of the dose-response curve.13 There is also accumulating evidence from the Japanese atomic bomb survivors and various other moderate- and low-dose exposed groups of excess risk of cataracts.14 There are data, reviewed in Ref. 15 suggesting an increase in stable chromosome aberrations and other markers of biological damage in the peripheral blood lymphocytes of nuclear workers and other groups with protracted radiation exposures. Chromosome changes play a major role in carcinogenesis and there is increasing evidence that the presence of increased frequencies of chromosome aberrations in peripheral blood lymphocytes in healthy individuals could be a surrogate for the specific changes associated with carcinogenesis and therefore indicative of risk.15 Much other in vitro and in vivo radiobiological data suggest small adverse effects of moderate dose exposure—in particular there is little data to suggest a threshold in dose, or possible hormetic (beneficial) effects of low-dose radiation exposure.9,15,16 In summary, excess cancer risks have been seen in a number of (largely pediatrically- or in utero-exposed) groups. Excess risks of circulatory disease and cataracts have also been observed in a number of groups exposed to low or moderate doses. The available data on biological mechanisms do not provide general support for the idea of a low-dose threshold or hormesis for any of these endpoints. This large body of evidence does not suggest, indeed is not statistically compatible with, any large threshold in dose (>10 mGy), or with possible beneficial effects. Rebuttal: Mohan Doss, Ph.D. Dr. Little quotes the consistency of childhood cancer risk factors from Oxford and Japanese studies as evidence for carcinogenicity of in utero LDR.10 However, for the Japanese cohort, leukemias were observed only following high dose radiation (HDR), and the risk coefficients were calculated using an assumed LNT model, creating the illusion of increased risk of leukemias from LDR whereas none was observed.10 Also, cohort studies, which are superior to case-control studies, have not shown increased leukemia risk.17 The study of childhood leukemias correlated with background radiation11 does not consider confounding factors such as breastfeeding. Small changes in the results from consideration of such factors could make the increased leukemias statistically insignificant. The study of childhood cancers following CT scans12 has methodological issues including the lack of a control group, raising major doubts about its conclusion.18 With regard to heart disease, the meta-analysis13 combined LDR and HDR data, effectively transferring HDR risk to LDR as described in a detailed critique.19 Regarding cataracts, Chernobyl and atomic bomb survivor data do show a threshold dose for cataracts requiring surgery.14 Although Dr. Little expressed concerns regarding LDR-induced chromosome changes, mutation is not the primary determinant of clinical cancer, whereas deficiency in immune system is an important factor.2 Since LDR increases immune system response,20 it would reduce the cancer risk.2 Finally, Dr. Little quoted the UNSCEAR 1993 Report16 as lack of evidence for the beneficial effects of LDR. However, Annex B of the UNSCEAR 1994 Report did discuss the beneficial effects of LDR. Also, many publications in recent years have demonstrated the disease-preventive effect of LDR for cancer and noncancer diseases.2,4,21 In conclusion, since the opposing arguments presented by Dr. Little are explainable as discussed above, considering the arguments and evidence presented in my Opening Statement, we can indeed conclude confidently that LDR is beneficial, not harmful. Rebuttal: Mark P. Little, D.Phil. Dr. Doss discusses the well-known involvement of the immune system in cancer, and more generally the role of adaptive response. The critical issue is whether the up-regulation of the immune system or other forms of adaptive response that may result from a radiation dose offsets the undoubted carcinogenic damage that is caused. The available evidence, summarized in my Opening Statement, is that it does not, and that, given the similarities in risks per unit dose following exposures to very low doses of radiation and with those after moderate dose radiation exposure,10–12,15 the nonlinearities induced by any adaptive response cannot be substantial. While adaptive response modulating the effect of relatively high challenge doses of radiation (of several Gy) following a smaller priming dose (of usually at least several tens of mGy) is well known experimentally (mostly in vitro), it is not universally observed in all experimental systems, nor does it last more than a few days, and there is little or no evidence for its involvement at low priming and challenge doses.22,23 Responding to the points relating to existence of a possible dose threshold, or hormetic effect, there is no evidence for these either for cancer24,25 or for noncancer disease26 in the Japanese atomic-bomb survivors. Naturally, thresholds below a certain size cannot be ruled out by the Japanese data, but the evidence suggests that thresholds cannot be larger than about 60 mSv for cancer24,25 or larger than about 0.9 Sv for noncancer disease.26 Taken together with the other data discussed above,10–12 thresholds or hormetic effects much above 10 mGy can be largely discounted for cancer. REFERENCES 1L. E. Feinendegen, M. Pollycove, and R. D. Neumann, “Hormesis by low dose radiation effects: Low-dose cancer risk modeling must recognize up-regulation of protection,” in Therapeutic Nuclear Medicine, edited by R. P. Baum ( Springer, Berlin, 2013). Google Scholar 2M. Doss, “Shifting the paradigm in radiation safety,” Dose-Response 10, 562– 583 (2012).10.2203/dose-response.11-056.DossCrossrefPubMedWeb of Science®Google Scholar 3B. Cohen, “The cancer risk from low-level radiation,” in Radiation Dose from Adult and Pediatric Multidetector Computed Tomography, edited by D. Tack and P. Gevenois ( Springer-Verlag, Berlin, 2007), pp. 33– 49. CrossrefGoogle Scholar 4M. Doss, “Linear no-threshold model vs. radiation hormesis,” Dose Response 11, 480– 497 (2013).10.2203/dose-response.13-005.DossCrossrefPubMedWeb of Science®Google Scholar 5M. S. Sasaki, A. Tachibana, and S. Takeda, “Cancer risk at low doses of ionizing radiation: artificial neural networks inference from atomic bomb survivors,” J. Radiat. Res. 55(3), 391– 406 (2014).10.1093/jrr/rrt133CrossrefCASPubMedWeb of Science®Google Scholar 6E. J. Calabrese and L. A. Baldwin, “Radiation hormesis: its historical foundations as a biological hypothesis,” Hum. Exp. Toxicol. 19, 41– 75 (2000).10.1191/096032700678815602CrossrefCASPubMedWeb of Science®Google Scholar 7E. J. Hall and A. J. Giaccia, Radiobiology for the Radiologist, 6th ed. ( Lippincott Williams & Wilkins, Philadelphia, 2006). Google Scholar 8C. M. Vajdic and M. T. van Leeuwen, “Cancer incidence and risk factors after solid organ transplantation,” Int. J. Cancer 125, 1747– 1754 (2009).10.1002/ijc.24439Wiley Online LibraryCASPubMedWeb of Science®Google Scholar 9 United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), “ UNSCEAR 2006 Report, Annex A: Epidemiological studies of radiation and cancer,” (United Nations, New York, 2008), pp. 13– 322. Google Scholar 10R. Wakeford and M. P. Little, “Risk coefficients for childhood cancer after intrauterine irradiation: a review,” Int. J. Radiat. Biol. 79(5), 293– 309 (2003).10.1080/0955300031000114729CrossrefCASPubMedWeb of Science®Google Scholar 11G. M. Kendall et al., “A record-based case-control study of natural background radiation and the incidence of childhood leukaemia and other cancers in Great Britain during 1980–2006,” Leukemia 27(1), 3– 9 (2013).10.1038/leu.2012.151CrossrefPubMedWeb of Science®Google Scholar 12M. S. Pearce et al., “Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study,” Lancet 380, 499– 505 (2012).10.1016/S0140-6736(12)60815-0CrossrefPubMedWeb of Science®Google Scholar 13M. P. Little et al., “Systematic review and meta-analysis of circulatory disease from exposure to low-level ionizing radiation and estimates of potential population mortality risks,” Environ. Health Perspect. 120, 1503– 1511 (2012).10.1289/ehp.1204982CrossrefPubMedWeb of Science®Google Scholar 14M. P. Little, “A review of non-cancer effects, especially circulatory and ocular diseases,” Radiat. Environ. Biophys. 52, 435– 449 (2013).10.1007/s00411-013-0484-7CrossrefPubMedWeb of Science®Google Scholar 15M. P. Little et al., “Risks associated with low doses and low dose rates of ionizing radiation: why linearity may be (almost) the best we can do,” Radiology 251, 6– 12 (2009).10.1148/radiol.2511081686CrossrefPubMedWeb of Science®Google Scholar 16 United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). “ Sources and effects of ionizing radiation. UNSCEAR 1993 Report to the General Assembly, with scientific annexes” (United Nations, New York, 1993), pp. 1– 922. Google Scholar 17R. L. Brent, “Carcinogenic risks of prenatal ionizing radiation,” Semin. Fetal Neonatal Med. 19(3), 203– 213 (2014).10.1016/j.siny.2013.11.009CrossrefPubMedWeb of Science®Google Scholar 18M. Cohen, “Cancer risks from CT radiation: Is there a dose threshold?,” J. Am. Coll. Radiol. 10, 817– 819 (2013).10.1016/j.jacr.2013.03.029CrossrefPubMedWeb of Science®Google Scholar 19S. Akiba, “Circulatory disease risk after low-level ionizing radiation exposure,” Radiat. Emerg. Med. 2, 13– 22 (2013) (available URL: http://www.hs.hirosaki-u.ac.jp/~hibaku-pro/rem/file_pdf/2013_vol2-2/rem_vol2_2_03_suminori_akiba.pdf). Google Scholar 20F. Zakeri, T. Hirobe, and K. Akbari Noghabi, “Biological effects of low-dose ionizing radiation exposure on interventional cardiologists,” Occup. Med. (London) 60, 464– 469 (2010).10.1093/occmed/kqq062CrossrefCASPubMedWeb of Science®Google Scholar 21M. Doss, “Low dose radiation adaptive protection to control neurodegenerative diseases,” Dose-Response 12(2), 277– 287 (2014).10.2203/dose-response.13-030.DossCrossrefPubMedWeb of Science®Google Scholar 22 United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), “ Sources and effects of ionizing radiation. UNSCEAR 1994 Report to the General Assembly, with scientific annexes” (United Nations, New York, 1994), pp. 1– 272. Google Scholar 23 International Commission on Radiological Protection (ICRP), “Low-dose extrapolation of radiation-related cancer risk,” Ann. ICRP 35(4), 1– 142 (2005). Google Scholar 24M. P. Little and C. R. Muirhead, “Curvature in the cancer mortality dose response in Japanese atomic bomb survivors: absence of evidence of threshold,” Int. J. Radiat. Biol. 74(4), 471– 480 (1998).10.1080/095530098141348CrossrefCASPubMedWeb of Science®Google Scholar 25D. A. Pierce and D. L. Preston, “Radiation-related cancer risks at low doses among atomic bomb survivors,” Radiat. Res. 154(2), 178– 186 (2000).10.1667/0033-7587(2000)154[0178:RRCRAL]2.0.CO;2CrossrefCASPubMedWeb of Science®Google Scholar 26M. P. Little, “Threshold and other departures from linear-quadratic curvature in the non-cancer mortality dose-response curve in the Japanese atomic bomb survivors,” Radiat. Environ. Biophys. 43(2), 67– 75 (2004).10.1007/s00411-004-0244-9CrossrefPubMedWeb of Science®Google Scholar Citing Literature Volume41, Issue7July 2014070601 ReferencesRelatedInformation

In linear-accelerator based radiotherapy for prostate cancer, both target and normal tissues are known to change position and shape considerably during a course of therapy. The need to track and compensate for these motions is well established and adaptive planning is widely accepted. Some claim that this should be accomplished online while the patient is lying on the couch waiting for treatment and this is the premise debated in this month's Point/Counterpoint. Arguing for the Proposition is X. Allen Li, Ph.D. Dr. Li is Professor and Chief of Medical Physics in the Department of Radiation Oncology, Medical College of Wisconsin. He is certified in Radiation Oncology Physics by both the Canadian College of Physicists in Medicine and the American Board of Medical Physics. Dr. Li has served for the AAPM in the capacity of chair or member of various subcommittees and task groups including the Biological Effects Subcommittee, TG-74, TG-106, TG-166, and the Editorial Board of Medical Physics, and is a Fellow of the AAPM. He has been a peer reviewer for 14 scientific journals and seven public and private research funding agencies. He has edited a book entitled “Adaptive Radiation Therapy” and has authored over 125 peer-reviewed papers. Dr. Li's research interests range from adaptive radiation therapy, outcome modeling, and recently to MRI guided radiation treatment planning and delivery. Arguing against the Proposition is Qiuwen Wu, Ph.D. Dr. Wu obtained his Ph.D. in Physics from Columbia University, New York, and subsequently worked in the Department of Medical Physics, Memorial Sloan-Kettering Cancer Center, New York, NY, the Department of Radiation Oncology, Virginia Commonwealth Univer-sity, Richmond, VA, William Beaumont Hospital, Royal Oak, MI, and Wayne State University, Detroit, MI, before moving to Duke University Medical Center where he is currently Professor in the Department of Radiation Oncology. He is certified in Therapeutic Radiological Physics by the American Board of Radiology, is a Fellow of the AAPM and is a member of the Board of Editors of the Journal of Applied Clinical Medical Physics. Dr. Wu's major research interests include adaptive radiation therapy, online image guidance, and applications of flattening filter free linacs, for which he has published about 70 peer-reviewed papers. It has been widely reported that interfraction variations in both targets and organs at risk (OARs) can be substantial during radiation therapy (RT) for prostate cancer.1 These variations, including translational and rotation shifts, deformation (e.g., volume and shape changes) and independent organ motions, can be systematic and random in nature. To account for these variations, large CTV-to-PTV margins are necessary to ensure adequate target coverage. These large margins inevitably result in increased doses to the OARs, leading to increased treatment-related toxicities that may prevent safe delivery of more effective and/or socially/economically favorable treatments such as dose escalated RT, hypofractionated RT, or stereotactic body RT (SBRT). To address the negative impacts listed above, a large amount of effort has been expended in recent decades on the development of technologies and strategies to correct for interfraction variations.2 Image-guided RT, where images acquired immediately prior to a treatment are used to guide patient repositioning, is currently the standard practice used to correct translational shifts and rotational errors, if the machine is properly equipped, but does not correct for anatomy deformations and independent organ motions. Adaptive RT (ART) that may be performed online or offline has been introduced to correct for this problem.3 While offline ART may be used to account for systematic variations,4 online ART that generates and delivers a new dosimetric plan optimized based on the anatomy of the day (fraction) can fully account for interfraction variations including both systematic and random (unpredictable) deformations and independent organ motions.5,6 Online replanning needs to be fast so that it can be completed within a few minutes while the patient is lying on the table waiting for treatment. Although such fast planning is generally challenging using conventional planning technologies, adaptive replanning does not need to start completely from scratch. For example, it can start with an initial plan fully optimized from the planning images for the same patient. Technologies on the quality of inroom imaging, image registration and segmentation, plan optimization algorithm, and computing hardware are advancing significantly and rapidly. For example, improved quality of onboard cone-beam CT, addition or integration of diagnostic-quality CT or MRI in the treatment room, graphic-processing unit accelerated autosegmentation and dose calculation,7 rapid plan modification with aperture morphing algorithms,6 and plan adaptation based on previous knowledge or a previously created plan library, are among the technological advances that can speed up adaptive planning significantly. Commercial treatment planning systems including some of these advances are becoming available. Extensive literature indicates that the interfraction deformations of prostate, bladder, and rectum and independent motions between prostate, pelvic nodes, rectum, and bladder during RT for prostate cancer, are generally unpredictable and can be substantial. For example, based on the MRIs acquired during the course of prostate RT, Nichol et al.8 reported that prostate volume could decrease by 20% and its shape could deform by 13 mm. Patients with a transurethral resection of prostate were prone to prostate deformation.8 It is necessary and possible to fully account for these variations by using online adaptive replanning for a treatment strategy requiring a small CTV-to-PTV margin. With online ART, a margin can reach as low as 3 mm, depending mainly on intrafraction variations. Such a small margin would be highly desirable to reduce treatment-related toxicities. In conclusion, online adaptive planning for prostate cancer radiotherapy is necessary and ready now, particularly for cases with large deformations or for dose escalated RT, hypofractionated RT, or SBRT. With advanced onboard imaging systems, standard online image guidance (IG) techniques currently used in many clinics is efficient and can correct setup errors and rigid organ motions. The residues, including uncorrected rotations, execution errors, deformations, and intrafraction motions, are handled by the margins in treatment planning, which are significantly reduced from those used for conventional treatment planning without IG.9 In principle, online planning, either adapted geometrically to the target or dosimetrically to include dose from previous fractions, can compensate additionally for interfractional deformations of both targets and critical organs in real-time while the patient is still on the table, potentially allowing the use of minimal margins.10 However, there are steep challenges ahead, both technically and practically. The crucial process in treatment planning is the delineation of regions of interest (ROIs). Unfortunately, current state-of-the-art online CBCT images are not adequate for this task for prostate cancer treatments.11 Radiation oncologists need significantly longer time and yet with less confidence to manually contour on CBCT than conventional CT. Also, the accuracy of computerized segmentation and deformable image registration algorithms depends heavily on image quality and, therefore, margins for the uncertainty of ROI definition must be added. In addition, online adaptive planning will undoubtedly increase the time between imaging and treatment and therefore needs increased margin for intrafractional motion and deformation, which increase with time.12 The potential savings in margins for online planning can quickly diminish. The prolonged treatment session is not economical either. Many practical factors also limit the use of online planning. The physician is required to approve the ROIs on online CT as well as the treatment plan, which the planner needs to generate quickly. In addition, the physicist needs to validate the plan and perform quality assurance creatively with the patient on the table. In the meantime, the therapists need to check all new plan parameters. Ideally, all these tasks need to be done at the treatment console for efficient communication and minimization of errors. However, this is unrealistic for scheduling in many clinics since it requires additional personnel at the linac at the same time for each prostate patient. There are competing alternatives to handle nonrigid organ motions such as the hybrid strategy combining online IG and offline adaptive planning.13 The advantage is that time-consuming treatment planning is handled offline, the same way as traditional treatment planning. With no added burden to the online process, the online IG is the same efficient process that patients and therapists are used to. Many studies have shown that organ motion and deformation in prostate cancer are patient-specific in nature.14,15 Offline or hybrid adaptive radiotherapy can take advantage of this by analyzing repeated images during early treatments and incorporating them into the treatment planning. Dose guided adaptive radiotherapy and equivalent margin reductions can be achieved.16 In summary, online adaptive planning faces many technical and practical challenges and is not yet ready for clinical implementation. Increased margins for target definition uncertainty and intrafraction motion offset the savings in margin for deformation. In the meantime, clinically proven alternative offline and hybrid adaptive strategies have been shown to be equally as effective. The offline and hybrid adaptive strategies cannot adequately account for random (unpredictable) organ deformations and independent organ motions and hence they are certainly not as effective as the online adaptive approach. They would require larger margins for cases with moderate and large deformations. Also, I have to disagree with Dr. Wu on his point that the additional time for online adaptive planning required in order to increase margins to account for the increased intrafraction variations may diminish the benefits of online replanning. For cases with moderate and large deformations, the interfraction variation is dominant. It has been documented that the dosimetric loss due to the intrafraction changes during the application of online replanning is relatively small compared to the gain from the online replanning. Furthermore, the time required for online replanning can be reduced substantially (to a few minutes) with the recent advances in online imaging, image registration, replanning algorithms, and/or computing technology. Undoubtedly, large efforts and resources are needed for online replanning, similar to any other emerging techniques. However, the benefit of the reduced margins from online replanning is clear. These reduced margins would lead to reduced radiation injury to normal tissues and/or offer opportunities for more effective treatments (e.g., dose escalation, hypofractionation). For prostate radiotherapy, the increased effort can be socioeconomically justified in cases with large deformations or when using hypofractionated RT or SBRT. Both Dr. Li and I agree that there are a variety of substantial interfractional organ motions in prostate cancer patients. We disagree on the effectiveness and efficiency of online adaptive planning to handle them as it stands now. Its advantage over alternatives is too little, and it is too soon for clinical implementation. Many pelvic organ motions, including deformations and intrafraction motions, are not totally random but rather patient-specific; they can be characterized based on a few measurements from previous fractions, and compensated through either offline or hybrid adaptive planning. In principle, in-room MRI could improve the accuracy and shorten the time of ROI delineation. However, either MRI-only based treatment planning must be clinically proven, or CT-MRI registration be shown to be equally as accurate as CT-CT registration. Pretreatment verification of “a new dosimetric plan optimized based on the anatomy of the day” is a practical concern. With the standard of care of prostate radiotherapy moving toward IMRT and VMAT, the online adaptive plan must be verified before treatment as this is required by regulations. This poses considerable challenges while the patient is on the treatment table. Current workflow in many radiation oncology clinics is sequential in nature, in which each task is performed by a specially trained staff member and completion of a task triggers the next one in line. It is not a trivial task to bring physicians, planners, physicists, and therapists together to the treatment console at the same time for each fraction. All this being said, I believe that online planning represents the holy grail of prostate cancer IGRT, and I am all for the research toward its successful application. Pointing out its current deficiencies should provide us better directions to work on in the future.

From Radium Teletherapy to Cobalt‐60 Teletherapy ‐ Peter R. AlmondThe history of radioisotope teletherapy machines will be traced from the beginning of the twentieth century, when radium was used, to the widespread use of cobalt‐60 machines in the 1970s and 1980s. Some early radium teletherapy machines will be discussed showing their development up to around 1940. By then their limitations were well understood and as early as 1937 a suggestion was made that an artificial radioisotope might be found to replace the radium. Although cobalt‐60 was recognized fairly early as a possible candidate, it was not until World War II and the introduction of the nuclear reactor that a suitable source of cobalt‐60 was identified. Following the war cobalt machines were developed in Canada and the United States resulting in commercial units being available primarily from Atomic Energy of Canada Limited and in the U.S. the Picker X‐ray Company. However cobalt‐60 machines were not the first artificial radioactive teletherapy treatment units. The first to treat a patient was a radioactive iridium machine built in the United Kingdom. Because radioactive cesium‐137 was readily available it was also used in treatment machines. Both the iridium and the cesium units were short SSD machines with various problems associated with such machines. They could be considered the logical development of the radium units, with the radium being replaced with the radioactive isotope. The cobalt units were a complete departure from the radium machines. The source size and the activity available allowed the cobalt units to be developed as a replacement for orthovoltage X‐ray units. The clinical implications of this will be presented. Cobalt‐60 units reached their peak use around 1980, but there were inherent problems with them and they have now been mainly replaced with linear accelerators.Early X‐Ray Therapy Machines ‐ Colin G. OrtonIn January, 1896, just months after Roentgen discovered X rays, the 1st cancer patient treatments began. Unfortunately, the majority of radiation treatments were not for cancer but for many other purposes, such as removal of hair and the treatment of acne, eczema, migraines, etc., many with dire consequences. Even when used to treat cancer, there were often tragic effects due to the use of very crude equipment during the 1st 20 years of X‐ray therapy. The machines were often unshielded, with several patients and staff in the room being exposed to scattered radiation. Hundreds of the early pioneers (the “Martyrs of Radiology”) died of radiation‐induced cancers. Also, since X‐ray energies were relatively low (80–120 kVp) and outputs extremely low and highly unreliable, most treatment schedules lasted months and had to be individualized based on skin reactions. Untoward reactions were commonplace. It was not until the advent of the Coolidge hot‐cathode X‐ray tube in 1913 that higher outputs could be achieved and controlled, and energies could be increased such that standardized schedules could be prescribed and deeper cancers treated. Prior to this, the low‐energy machines (later called “superficial X‐ray units”) were really only suitable for the treatment of skin lesions. It was not until the 1920s that “deep” therapy machines of energies 120 – 300 kVp (initially often referred to as “cannons” but later to be called “orthovoltage X‐ray units”) began to be developed for the treatment of deeper lesions. These usually used bearings and cranks to adjust tube position and angulation so that patients could be treated from different directions. By 1930 the quest for higher‐and‐higher energy machines began. The 1 st of these were “supervoltage” X‐ray machines built for radiotherapy between 1930 and 1934. These included machines of 500 kVp installed at Harper Hospital, Detroit, 600 kVp at Caltech, 750 kVp at Memorial Hospital, New York, 800 kVp at Mercy Hospital, Chicago, and 1,000 kVp at Swedish Hospital, Seattle and Caltech. These were enormous machines. The 1,000 kVp unit at Caltech, for example, occupied a vault 42 m long, by 20 m wide, by 15 m high, with the treatment room, where up to four patients could be treated simultaneously, situated above the ceiling.In the pursuit of more compact machines with even higher energies, John Trump and colleagues in Boston developed the 1st Van de Graaff machines for therapy. A 1 MeV unit was installed at Huntington Hospital in 1937 and soon ceiling‐mounted machines of energies up to 2 MeV were developed. These could be rotated about a horizontal axis aimed at a patient sitting in a chair that could be rotated about a vertical axis. Multidirectional, and even rotational, megavoltage X‐ray therapy was born. By 1965, about 40 Van de Graaff radiotherapy machines were in use worldwide, but their energies were restricted to just a few MeV. The betatron, on the other hand, was capable of much higher energies. The 1st betatron used for radiotherapy was a 6 MeV machine built by Konrad Gund at Siemens, Erlangen, Germany and reportedly used to treat a patient in 1942, but there is little documentation of this. Elsewhere, the 1st well‐documented treatments were with the 24 MeV betatron built by Donald Kerst at the University of Chicago and developed for therapy in 1948 by Lester Skaggs, John Laughlin, Gail Adams, Larry Lanzl, and colleagues. Although over 100 betatrons were being used for radiotherapy up until about 1970, because they were bulky machines, not amenable for isocentric mounting, and with a low X‐ray output, they were soon replaced by linear accelerators for high energy X‐ray therapy.Linacs grew out of the development of klystrons in the USA and magnetrons in the UK in the 1930s, but much of this development was kept secret during the 2nd World War due to their use in radar. It was not until late 1946 that the 1st linac (0.5 MeV) was developed by Fry and colleagues at the Atomic Energy Research Establishment in Harwell, UK, and in early 1947 by Bill Hanson, Stanford University, California and Varian Associates (1.7 MeV), which was soon increased to 4 MeV by late 1947. The 1st therapy linac was an 8 MeV stationary machine built by Metropolitan Vickers and installed at Hammersmith Hospital, London, with the 1st patient treated on 19 August, 1953. Later that year, a 4 MeV isocentric machine was installed by Mullard Research Laboratories at the Newcastle General Hospital. The 1st linac for radiotherapy in the USA was a 6 MeV machine built by Varian, which began patient treatments at Stanford in 1956. This was the predecessor of the Clinac 6, a fully rotational linac, introduced by Varian in 1960, and later the Clinac 4, which used standing‐wave technology to reduce size, complexity and cost, which became the workhorse of radiotherapy in the USA in the 1970s. Modern, linear‐accelerator based high‐energy X‐ray therapy had been born.Learning Objectives:1. Understand the early development of teletherapy treatment machines2. Understand the benefits of the evolution from radium to cobalt sources for teletherapy3. Understand the advantages and disadvantages of early x‐ray treatment machines

Being a referee for a journal article is an important yet daunting task. It is especially worrisome for someone who has little reviewing experience. It is important to remember is that it is an honor and perhaps a duty, but not a burden. Someone considers you to be an expert in the subject and it is with that mindset that you should approach the task. Each part of the review process has its own difficulties:• Grammar and Language: How much of this do you need to do? Does poor grasp of the English language warrant rejection. Is it effective or excessively verbose?• Science: Is this project novel? How do you determine that? How similar can it be to other previously published manuscripts?• References: Are the proper studies referenced? How much does this affect the review?• Methods: How do you determine if the methods are sound? How much can you suggest without trying to be a collaborator?• Results: Are graphs and tables being utilized wisely? Are there an excessive number of them?All of these issues are important part of the review process and need to be considered. But how do you weight them in your final recommendation and communicate them to the authors and editors? This panel of senor medical physics, which has extensive experience in refereeing, will help guide the audience through the entire review process. These experts will share their experiences and methods to help new referees best prepare for the opportunity of reviewing manuscripts.Learning Objectives:1. Learn the role of the referee and what is expected of them.2. Learn how to perform a review.3. Learn how to communicate a review with authors and editors.