Methods: Patients with DFA measurements on postoperative days 1 and 3 (DFA1/DFA3) following PD or distal pancreatectomy (DP) between 10/2016-4/2018 were identified from a prospective database.Preoperatively determined risk-stratified pathways were labeled "low-risk PD," "high-risk PD," and "distal pancreatectomy" (DP).Receiver operator characteristic (ROC) analysis was performed to determine clinically relevant sensitivity thresholds for ruling out Grade B/C POPF.Results: 174 pancreatectomy patients were stratified as low-risk PD (n = 78, 45%), high-risk PD (n = 51, 29%), and DP (n = 45, 26%).B/C POPF developed in 3% (n = 2) lowrisk PD, 37% (n = 19) high-risk PD, and 24% (n = 11) DP patients (low vs. high-risk PD p < 0.001, low-risk PD vs. DP p = 0.004, high-risk PD vs DP p = 0.25).DFA1 and DFA3 levels correlated with POPF with AUC of 0.865 and 0.861, respectively.B/C POPF occurred in 16% (n = 21) of combined PD patients (high+low risk PD), and ROC analysis excluded POPF in this combined PD cohort with 100% sensitivity if DFA1 136 (Units/L) or DFA3 93, 90% if DFA1 568 or DFA3 164, and 80% if DFA1 820 or DFA3 172.If using the historical DFA1 <5000 cutoff, B/C POPF could only be excluded with 57% sensitivity among combined PD patients.B/C POPF was excluded with 100% sensitivity if DFA1 661 or DFA3 141 in low-risk PD patients, 100% sensitivity if DFA1 136 or DFA3 93 in high-risk PD patients, and with 90% sensitivity if DFA1 <468 or DFA3 <169 in high-risk PD patients.B/C POPF was excluded with 100% sensitivity if DFA1 49 or DFA3 26, and with 90% sensitivity if DFA1 730 or DFA3 90, in DP patients. Conclusion:The sensitivity with which DFA1 and DFA3 predict postoperative pancreatic leak is different for each preoperatively defined POPF risk pathway.The traditional cutoff of DFA1 <5000 U/L did not appropriately rule out B/C leak in our PD patients, with a sensitivity of only 57%.Rather than attempting to apply a general cutoff, we propose using institution-specific DFA1 and DFA3 cutoffs tailored to specific POPF-risk pathways.

The objective of this study is to perform a patterns of care analysis for the treatment of FIGO Stage I-II cervical cancer over a 26 year time period. We anticipate seeing decreased use of surgery in patients with larger tumor size, higher stage, and positive nodes, thus decreasing serious side effects and lowering the cost of care in these patients who will require radiation therapy regardless of surgical intervention. A total of 20,290 women with clinical stage IA through IIB cervical cancer with known tumor characteristics were identified in the Surveillance, Epidemiology and End Results (SEER) program from 1983 to 2009. Of these, 12,993 women had known treatment status (radiation alone, surgery alone, or surgery plus radiation) and were included in the final analysis. Of the 12,993 women identified, 30% received both surgery plus radiation, 46% received surgery alone, and 24% received radiation alone. Over the time period evaluated, a decline in percentage of patients receiving surgery alone was noted, with an accompanying increase in percentage of patients receiving radiation alone, and a moderate decline in percentage of patients receiving surgery plus radiation. No change was seen in the odds ratio of a node positive patient or of a patient with parametrial involvement receiving surgery plus radiation over the time period; however, a significant decrease in the odds ratio of a patient with tumor size >4 cm receiving surgery plus radiation therapy was seen (OR 0.96; 95% CI, 0.95, 0.97; p < 0.0001). On multivariate analysis, factors significantly associated with choice of therapy included year of diagnosis, age, race, SCC histology, and tumor size. Smaller tumor size (1-4 cm) was associated with decreased odds ratio of radiation alone versus surgery plus radiation, while tumors measuring 5+ cm had an increased odds ratio of radiation alone. Node positivity was associated with decreased odds ratio of receiving radiation alone (0.33, p < .0001). Equivalent survival results with surgery versus radiation have been demonstrated for early stage cervical cancer, with associated increased toxicity when both therapies are used for tumors with high risk features. In this large review, a decline in surgery alone was seen with a moderate decline in surgery plus radiation therapy, and an increase in radiation alone. Based on published data, this should decrease the risk of serious side effects associated with surgery plus radiation, and also decrease costs via de-escalating treatment and causing fewer serious side effects to manage. Patients with tumor >5 cm are more likely to appropriately receive definitive radiation. Further improvement in patient quality of life and cost effectiveness can be realized by consistent use of PET/CT to prospectively identify patients with positive nodes for definitive radiation therapy.

The clinical applications of stereotactic body radiotherapy or stereotactic ablative radiotherapy (SABR) for the treatment of primary and metastatic tumours of different organ sites have been expanding rapidly in the recent decade. SABR requires advanced technology in radiotherapy planning and image guidance to deliver a highly conformal ablative dose precisely to targets (or tumours) in the body. Although this treatment modality has shown promising results with regard to tumour control, some serious complications have been observed and reported. In order to achieve a favourable therapeutic ratio, strategies to mitigate the risk of complications must be in place. This overview will summarise the reported serious complications caused by SABR and strategies to mitigate the risk will be discussed.

Standard therapy for locally advanced cervical cancer includes intracavitary brachytherapy. Tandem and ovoids (TO) are traditionally used in the US, but when the vaginal fornices are obliterated by tumor, ovoid placement is difficult. Practice patterns surveys in Canada reveal 9% of practitioners there use tandem only. This study compares the dosimetry of TO versus tandem (T) alone insertions with respect to the dose to ICRU points A, B, bladder and rectum, volumetric dose to bladder and rectum and volume of tissue receiving various dose gradients. One hundred eight brachytherapy cases from 27 women with stage II-IV cervical cancer were treated consecutively for cervical cancer with HDR brachytherapy with 4 applicator insertions. The prescribed dose of each insertion was 7 Gy to ICRU point A with additional goals of minimizing bladder and rectal doses. Patients were treated with either TO or T alone at physician discretion based on the clinical evaluation of the vaginal fornices at the time of the procedure. Parameters compared were ICRU point A and B doses, volume receiving 95%, 85%, 50%, and 20% of the prescription dose, ICRU bladder and rectal point doses, and maximal dose to 2 cc of bladder and rectum (D2 cc). The Mann-Whitney U test was used to compare these parameters. The mean dose to ICRU point A was not significantly different between the TO and T applicators, 681 cGy vs 670 cGy (p = 0.09). The mean dose to ICRU point B was significantly higher with the TO applicator than T alone, 200 Gy vs 177 cGy (p = 10-7). Dose distribution was different between the two applicators, with significantly increased volume receiving 95%, 85%, 50%, and 20% of the prescription dose using the TO applicator. The mean V95 was 109.5 cc for TO vs 96.4 cc for T (p = 0.001) and the mean V20 was 1041.7 cc for TO vs 923.9 cc for T (p = 0.009). There was no significant difference in ICRU rectal point doses between the applicators, 361.8 cGy TO vs 375.1 cGy T (p = 0.7). However, there were significantly increased doses to the ICRU bladder point, and D2 cc of the rectum and bladder with the T alone. The bladder point dose was 294.9 cGy TO vs 360.6 cGy T (p = 0.02), bladder D2 cc 403.7 cGy TO vs 632.7 cGy T (p = 0.001) and rectal D2 cc 289.7 cGy TO vs 224.8 cGy T (p = 0.01). TO and T applicator insertions for cervix brachytherapy provide equivalent doses to the ICRU point A, however, the volumetric distribution is significantly different between the 3 channel applicator and single channel tandem. The TO applicator provides increased volumetric dose within the patient, but allows better sparing of the rectum and bladder. In trained providers, T may be considered when the vaginal fornices prohibit appropriate placement of ovoid applicators to provide equivalent Point A doses.

BACKGROUND:Preoperative chemoradiotherapy (CRT) improves outcomes in patients with locally advanced but resectable adenocarcinoma of the esophagus. ACOSOG Z4051 evaluated CRT with docetaxel, cisplatin, and panitumumab (DCP) in this patient group with a primary end point of a pathologic complete response (pCR) ≥35%.PATIENTS AND METHODS:From 15 January 2009 to 22 July 2011, 70 patients with locally advanced but resectable distal esophageal adenocarcinoma were enrolled. Patients received docetaxel (40 mg/m(2)), cisplatin (40 mg/m(2)), and panitumumab (6 mg/kg) on weeks 1, 3, 5, 7, and 9 with RT (5040 cGy, 180 cGy/day × 28 days) beginning week 5. Resection was planned after completing CRT. PCR was defined as no viable residual tumor cells. Secondary objectives included near-pCR (≤10% viable cancer cells), toxicity, and overall and disease-free survival. Adverse events were graded using the CTCAE Version 3.0.RESULTS:Five of 70 patients were ineligible. Of 65 eligible patients (59 M; median age 61), 11 did not undergo surgery, leaving 54 assessable. PCR rate was 33.3% and near-pCR was 20.4%. Secenty-three percent of patients completed DCP (n = 70) and 92% completed RT. 48.5% had toxicity ≥grade 4. Lymphopenia (43%) was most common. Operative mortality was 3.7%. Adult respiratory distress syndrome was encountered in two patients (3.7%). At median follow-up of 26.3 months, median overall survival was 19.4 months and 3-year overall survival was 38.6% (95% confidence interval 24.5% to 60.8%).CONCLUSIONS:Neoadjuvant CRT with DCP is active (pCR + near-pCR = 53.7%) but toxicity is significant. Further evaluation of this regimen in an unselected population is not recommended.CLINICALTRIALSGOV IDENTIFIER:NCT00757172.

135 Background: Elderly pts with comorbidities have limited Rx options for localized EC. High EGFR expression correlates with poor response to RT. We examined the safety and efficacy of erlotinib, an oral EGFR tyrosine kinase inhibitor with RT in these pts.METHODSPts older than 65 yrs ineligible for platinum based Rx with carcinoma of the thoracic esophagus or gastroesophageal junction received erlotinib 150mg PO QD for one year starting D1 of RT [50.4 Gy D1-28 (M-F) at 1.8 Gy per fraction]. Assessments- mucosal response by EGD 4-8 wks post RT; RECIST response by CT q3mo.ENDPOINTSPrimary- overall survival (OS), secondary- quality of life (QOL) using the FACT-E QOL tool, progression free survival (PFS) and toxicity; correlative- pre-treatment tumor EGFR and pEGFR expression by immunohistochemistry (IHC).RESULTSThe study was closed after 17 of planned 35 pts were included due to poor accrual. Baseline characteristics: median age 78 yrs (66-91); gender M/F: 11/6; ECOG PS 0/1/2=2/12/3; stage I-1, II=5, III=7 and IVa= 4; histology: adenocarcinoma 16, squamous cell 1; dysphagia at baseline 13/17 (76%). Median OS was 7.3 months (95% CI: 4.5-22.3) with 12 pts dead (5 alive, 3 still on treatment). Reason for coming off study (n=14): disease progression (n=6), toxicity (n=5), withdrew consent, completed one year of treatment, death from unrelated cause (n=1 each). There were 2 mucosal CRs and one residual carcinoma in situ, 3 partial endoscopic responses of the 9 pts who had post RT endoscopy. Estimated PFS is 5.3 months (95% CI: 2.4-11). Sites of progression- distant 3, locoregional 6, unknown 5 and too early 3. Estimated one year survival is 26 %, 3 pts lived >12 months. Smoking status- current/past/never 3/12/2. Treatment related toxicities (any grade n=or>5) were, rash (16), fatigue (16), diarrhea (11), lymphopenia (10), anorexia (7) and dehydration (6) and 21 grade 3/4 toxicities occurred. IHC results- EGFR, neg: 1/16; pos: 15/16 and pEGFR, neg: 2/16; pos:14/16.CONCLUSIONSFor elderly pts with localized EC and no chemotherapy options, erlotinib monotherapy with RT is a tolerable therapy with modest activity. Further studies are needed to define the role of EGFR inhibition with RT in EC. No significant financial relationships to disclose.

To determine the maximum tolerable dose of bortezomib (MTD) given intravenously once weekly with paclitaxel and radiation therapy in patients with locally advanced, non-metastatic pancreatic or biliary tract cancer. An open label, dose escalation, phase I clinical trial evaluated the safety of three dose levels of intravenously (IV) administered bortezomib (0.7 mg/m2, 1.0 mg/m2, and 1.3 mg/m2/dose) with paclitaxel 50 mg/m2 IV, both administered weekly for 5 weeks with conventionally fractionated radiation therapy (RT). RT was 45 Gy to primary tumor and regional lymphatics. Gross disease received an additional 5.4 Gy boost. All patients had histologically confirmed, unresectable, non-metastatic pancreatic, or biliary tract cancer. Dose limiting toxicity (DLT) was the primary endpoint. A DLT was defined as any grade ≥3 non-hematologic toxicity, grade 2 neuropathy with pain, neutropenia lasting > 7 days or any toxicity requiring treatment to be delayed for ≥ 2 weeks. A total of 12 patients were screened for inclusion. 2 patients were ineligible and did not receive trial therapy. Trial therapy was discontinued in a single patient experiencing a severe transfusion reaction to paclitaxel. All patients remaining on trial received >90% of planned therapy. No DLT was appreciated in cohort 1. A single DLT was appreciated in cohort 2 (grade 3 urinary tract infection; dehydration; hypotension.) Common adverse events not meeting criteria for DLT included lymphopenia (9), febrile neutropenia (1), pulmonary embolus (1), DVT (1), diarrhea (1), and rash (1). This trial was closed prior to initiation of cohort 3 due to poor accrual. Median survival was 14 months (range 3-39 months). Eight patients have died. All surviving patients experienced disease progression. Weekly bortezomib up to a maximum dose of 1 mg/m2 delivered concurrently with weekly paxclitaxel and external beam radiation was tolerable in a small cohort of patients with unresectable pancreatic/biliary tract cancer. Due to poor accrual, this trial failed to establish the MTD. Additional investigation is warranted to determine the MTD, safety profile, and efficacy associated with this treatment regimen.

Purpose: To describe the histologic and volumetric changes in normal liver tissue after stereotactic body radiotherapy (SBRT) for liver metastases.Methods and Materials: Pre- and post-SBRT imaging studies were analyzed to evaluate the effect of SBRT on normal liver volume (NLV) in 15 patients treated in a prospective clinical trial. Two other patients underwent exploratory surgery after SBRT and histologic analyses of the irradiated liver were performed to characterize the pathologic effects of SBRT.Results: In the 15 patients studied quantitatively, the total NLV had decreased transiently at 2-3 months after SBRT and then began to regenerate at 3-8 months after SBRT. The median NLV reduction at the maximal observed effect was 315 cm(3) (range, 125-600) or 19% (range, 13-33%). Among the several dosimetric parameters evaluated, the strongest linear correlation was noted for the NLV percentage receiving 30 Gy as a predictor of maximal NLV reduction (r(2) = 0.72). The histologic changes observed 2 and 8 months after SBRT demonstrated distinct zones of tissue injury consistent with localized veno-occlusive disease.Conclusion: The well-demarcated focal parenchymal changes after liver SBRT (demonstrated both radiographically and histologically) within the high-dose zone are consistent with a threshold dose-induced set of phenomena. In contrast, the more global effect of NLV reduction, which is roughly proportional to whole organ dose parameters, resembles more closely an effect determined from radiobiologically parallel architecture. These observations suggest that modeling of normal tissue effects after liver SBRT might require different governing equations for different classes of effects. (C) 2009 Elsevier Inc.

To report outcome and toxicity in patients with unresectable cholangiocarcinoma (CCA) receiving definitive treatment at a single institution. Patients with unresectable CCA evaluated at the University of Colorado Cancer Center (UCCC) from 2001-05 were identified through a search of the medical records on an IRB approved protocol. Unresectable disease was determined by a liver surgeon and/or GI oncologic surgeon after review of radiographic imaging or exploratory surgery. Demographic, disease, treatment, and toxicity data were obtained for all patients. Kaplan-Meier analyses were used to calculate actuarial 1-year overall survival (OS), median survival (MS), and progression-free survival (PFS). Survival was calculated from date of diagnosis. Progression was defined by imaging and, when possible, confirmed histologically. Seventy-eight patients with CCA were identified during the study period. A total of 92% (72/78) had pathologically proven CCA. Of these patients, 16 with unresectable non-metastatic disease receiving definitive treatment formed the study cohort. Primary disease location was extrahepatic (EH) 50% (8/16), intrahepatic (IH) 44% (7/16), and both IH/EH 6% (1/16). All patients received concurrent chemoradiation (CRT). Sixty-three percent (10/16) received ≥54 Gy. Median EBRT dose was 50.4 Gy (range, 45-54 Gy). Target volume included tumor, portahepatis, and celiac axis with appropriate margin. Eight patients received HDR endoluminal brachytherapy boost delivered through percutaneous biliary catheters placed by interventional radiology. The median HDR dose was 18 Gy (range, 13-25 Gy) in 3 fractions. The most common CRT regimen used was weekly gemcitabine 75 mg/m2 and paclitaxel 40 mg/m2 (63%). Other CRT regimens included single agent 5-FU, capecitabine, and gemcitabine. Actuarial 1-year OS was 62.5%. The MS and PFS were 16.5 months (range, 3-30 m) and 10.5 months (range, 0-23 m), respectively. All patients experienced disease progression and had died at time of analysis. Nineteen percent (3/16) of patients treated with CRT were hospitalized with GI bleed. Other recorded toxicity included nausea (5), gastritis (3), biliary obstruction (2), and DVT (1). Results from UCCC are comparable to other published series. Patients with unresectable CCA continue to have a poor prognosis despite aggressive combined modality treatment. Further toxicity and dosimetric analysis is warranted to analyze the etiology of adverse events in patients receiving high-dose radiation treatment along with chemotherapy for locally advanced CCA. Future prospective studies aimed at improving outcome are needed.

Purpose/Objective(s)Although typical radiographic changes after liver SBRT have been well described[1], the effect of liver SBRT on total normal liver volume (NLV) has not been well studied quantitatively. Here we analyze the effect of liver SBRT on NLV to determine whether dosimetric parameters can predict the maximum observed NLV reduction following liver SBRT.Materials/MethodsWe studied 15 patients who were treated on an IRB-approved prospective multi-institutional Phase I/II trial of SBRT for liver metastases[2,3]. The nominal SBRT dose to the planning target volume ranged from 36–60 Gy in 3 fractions. The pre-SBRT NLV (NLVpre) included the entire liver minus gross tumor volume; the post-SBRT NLVpost was measured on f/u imaging scans in the same way. A critical volume model constraint was applied in all cases so that ≥700 cc of NLVpre received <15 Gy. The maximum absolute and percent NLV reduction was then analyzed in relation to location of lesion (peripheral v. perihilar) and selected NLV dosimetric parameters: mean dose, NLV receiving 15, 30, or 40 Gy (V15, V30, V40); and percent NLC receiving those doses (V15%, V30%, and V40%).ResultsOf the 15 patients evaluated in this series there were 8 males and 7 females with a median age of 53 yrs (range 40–83 yrs). Median NLVpre was 1483 cc (range, 880–2562 cc). Typically, the NLVpost decreased transiently and then began to regenerate 2–8 months after SBRT. The median absolute reduction in NLV (NLVpost-NLVpre) was 346 cc (range, 125–600 cc). The median percent NLV reduction was 19%(range, 13–33%). Among the variables analyzed, the strongest linear correlation was noted for V30% as a predictor of maximum %NLV reduction (Fig. 1a). Mean NLV dose was somewhat less well correlated (Fig. 1b). There was no correlation between %NLV reduction and location of lesion.ConclusionsPatients receiving liver SBRT experienced transient reduction in normal liver volume following treatment. The observed relationship between mean liver dose and percent reduction in NLV, coupled with the lack of effect of lesion location within the liver, are consistent with radiobiologically parallel hepatic architecture. While there have been no adverse clinical consequences correlated with this observation, it is important to be aware of this effect, especially when treatment is planned for patients with small NLVpre in the range of 1000 cc or less. Here it is prudent to maintain low V30% and mean NLV dose to avoid excess NLV reduction following SBRT where patients might be at risk of liver insufficiency due to low NLV. Purpose/Objective(s)Although typical radiographic changes after liver SBRT have been well described[1], the effect of liver SBRT on total normal liver volume (NLV) has not been well studied quantitatively. Here we analyze the effect of liver SBRT on NLV to determine whether dosimetric parameters can predict the maximum observed NLV reduction following liver SBRT. Although typical radiographic changes after liver SBRT have been well described[1], the effect of liver SBRT on total normal liver volume (NLV) has not been well studied quantitatively. Here we analyze the effect of liver SBRT on NLV to determine whether dosimetric parameters can predict the maximum observed NLV reduction following liver SBRT. Materials/MethodsWe studied 15 patients who were treated on an IRB-approved prospective multi-institutional Phase I/II trial of SBRT for liver metastases[2,3]. The nominal SBRT dose to the planning target volume ranged from 36–60 Gy in 3 fractions. The pre-SBRT NLV (NLVpre) included the entire liver minus gross tumor volume; the post-SBRT NLVpost was measured on f/u imaging scans in the same way. A critical volume model constraint was applied in all cases so that ≥700 cc of NLVpre received <15 Gy. The maximum absolute and percent NLV reduction was then analyzed in relation to location of lesion (peripheral v. perihilar) and selected NLV dosimetric parameters: mean dose, NLV receiving 15, 30, or 40 Gy (V15, V30, V40); and percent NLC receiving those doses (V15%, V30%, and V40%). We studied 15 patients who were treated on an IRB-approved prospective multi-institutional Phase I/II trial of SBRT for liver metastases[2,3]. The nominal SBRT dose to the planning target volume ranged from 36–60 Gy in 3 fractions. The pre-SBRT NLV (NLVpre) included the entire liver minus gross tumor volume; the post-SBRT NLVpost was measured on f/u imaging scans in the same way. A critical volume model constraint was applied in all cases so that ≥700 cc of NLVpre received <15 Gy. The maximum absolute and percent NLV reduction was then analyzed in relation to location of lesion (peripheral v. perihilar) and selected NLV dosimetric parameters: mean dose, NLV receiving 15, 30, or 40 Gy (V15, V30, V40); and percent NLC receiving those doses (V15%, V30%, and V40%). ResultsOf the 15 patients evaluated in this series there were 8 males and 7 females with a median age of 53 yrs (range 40–83 yrs). Median NLVpre was 1483 cc (range, 880–2562 cc). Typically, the NLVpost decreased transiently and then began to regenerate 2–8 months after SBRT. The median absolute reduction in NLV (NLVpost-NLVpre) was 346 cc (range, 125–600 cc). The median percent NLV reduction was 19%(range, 13–33%). Among the variables analyzed, the strongest linear correlation was noted for V30% as a predictor of maximum %NLV reduction (Fig. 1a). Mean NLV dose was somewhat less well correlated (Fig. 1b). There was no correlation between %NLV reduction and location of lesion. Of the 15 patients evaluated in this series there were 8 males and 7 females with a median age of 53 yrs (range 40–83 yrs). Median NLVpre was 1483 cc (range, 880–2562 cc). Typically, the NLVpost decreased transiently and then began to regenerate 2–8 months after SBRT. The median absolute reduction in NLV (NLVpost-NLVpre) was 346 cc (range, 125–600 cc). The median percent NLV reduction was 19%(range, 13–33%). Among the variables analyzed, the strongest linear correlation was noted for V30% as a predictor of maximum %NLV reduction (Fig. 1a). Mean NLV dose was somewhat less well correlated (Fig. 1b). There was no correlation between %NLV reduction and location of lesion. ConclusionsPatients receiving liver SBRT experienced transient reduction in normal liver volume following treatment. The observed relationship between mean liver dose and percent reduction in NLV, coupled with the lack of effect of lesion location within the liver, are consistent with radiobiologically parallel hepatic architecture. While there have been no adverse clinical consequences correlated with this observation, it is important to be aware of this effect, especially when treatment is planned for patients with small NLVpre in the range of 1000 cc or less. Here it is prudent to maintain low V30% and mean NLV dose to avoid excess NLV reduction following SBRT where patients might be at risk of liver insufficiency due to low NLV. Patients receiving liver SBRT experienced transient reduction in normal liver volume following treatment. The observed relationship between mean liver dose and percent reduction in NLV, coupled with the lack of effect of lesion location within the liver, are consistent with radiobiologically parallel hepatic architecture. While there have been no adverse clinical consequences correlated with this observation, it is important to be aware of this effect, especially when treatment is planned for patients with small NLVpre in the range of 1000 cc or less. Here it is prudent to maintain low V30% and mean NLV dose to avoid excess NLV reduction following SBRT where patients might be at risk of liver insufficiency due to low NLV.

Purpose/Objective: Extracranial stereotactic radiosurgery (ESR) is being widely investigated, and encouraging tumor control rates have been observed. However, variations in dose prescriptions between institutions and inherent inter-patient variability render it difficult to construct accurate dose-response relationships. These limitations could impede efforts to conduct multi-center clinical trials. Equivalent uniform dose (EUD) formalism was applied to describe the differences in biologic effect among clinically applied methods of ESR. Additionally, since the typical dose per fraction in ESR is much higher than conventional radiotherapy and most cells of average radiosensitivity will be killed, the implications of observed tumor control rates with respect to the fraction of highly resistant cells are explored by analyzing tumor control probability (TCP). Materials/Methods: EUD=2 Gy(ln[(1/V) ΣVbin(SF2)∧Dbin/2 Gy])/(ln(SF2)) where V = total volume, Vbin = volume of a bin within the dose-volume histogram (DVH), Dbin = dose to the bin, and SF2 = surviving fraction after 2 Gy. An SF2 of 0.5 was assumed. First, for a 12 Gy of ESR prescribed to the 80% isodose line, the effect of percent "hotspot" was considered by modifying the DVH from 0% "hotspot", ie uniform dose, up to an 80% "hotspot" of maximum dose. Next, differences in EUD resulting from the application of different technical styles of ESR (7 non-coplanar parabolic fields, Karolinska margin reduction, arcs, or coplanar flat fields) to the same tumor were determined. TCP = ΠBCPj BCP = exp(-N ∗ SF) where BCPj = bin control probability of the jthbin, N = clonogen number in the bin, SF = surviving fraction. Results: Variation in "hotspot" from 0 to 80% increased EUD from 12 to 14.8 Gy. ESR planning for three 12 Gy fractions to the same 30 cc lung tumor yielded wide variation in EUD according to the technique of ESR, ranging from 13.2 to 16.8 Gy. A hypothetical hypoxic fraction (HF) of cells were considered to be a subpopulation 10 times more radioresistant than the rest. Increase in HF corresponds to lower estimates of TCP as illustrated below. Conclusions: The EUD of ESR DVHs varies greatly when there are differences in the percent "hotspot", even when the same nominal dose is prescribed to the same isodose line. For a multi-center clinical trials of ESR, it is advisable that dose be expressed by radiobiological dose equivalency formalism to ensure evaluable dose-response relationships. The rate of tumor control observed for similar size tumors treated on a tightly regulated phase II trial could provide an estimate of the fraction of highly resistant cells due to hypoxia or other mechanisms, thus indicating any role for hypoxic cell sensitization or other adjuvant radiosensitization.