CAMLS are myeloid polyploid cells emerging from tumor masses and transit the circulation in a variety of malignancies. They are present at all stages and may be used to track disease progression and evolution. (Adams, PNAS 2014, Moran, JCO Prec Oncol 2022). Prior studies demonstrated that larger CAMLs (> 50 uM) were particularly associated with disease.

The TAPUR (Targeted Agent and Profiling Utilization Registry) study is a phase II prospective, non-randomized, multi-basket, trial. The study aims to identify signals of drug activity when clinically approved drugs are matched to pre-specified genomic targets in patients with advanced cancers, outside of their approved indications. Eligibility requires pre-specified histology and potentially actionable somatic mutations. Patient identification for the treatment cohorts is complicated by the inclusion of a wide range of histology's that transcend the usual disease-based teams.

Introduction: Real-time cancer detection can increase the diagnostic yield of robotic-assisted navigation bronchoscopy (RANB). Needle-based confocal laser endomicroscopy (nCLE) allows real-time microscopic imaging at the needle tip. Aim: assess feasibility, safety and accuracy of RANB nCLE imaging Methods: In suspected peripheral lung cancer patients scheduled for fluoroscopic/radial EBUS guided RANB, the target area was imaged with nCLE before TBNA and biopsies. Reported nCLE malignancy and airway/lung parenchyma criteria were used to assess the optimal sampling location. nCLE videos were compared to the corresponding TBNA and final diagnosis. After 15 minutes training, five blinded raters validated nCLE videos of malignancy and airway/lung parenchyma twice Results: 20 patients underwent nCLE imaging (final diagnosis: n=18 cancer, n=1 granulomatous disease, n=1 undetermined) with mean lesion size 15 mm (range 8-28 mm). No complications occurred. In 17/18 (94%) patients with cancer, nCLE imaging correlated with a final diagnosis of cancer, including 4 patients with malignant nCLE patterns but non-diagnostic TBNA and biopsies. The diagnostic yield of the RANB procedure was 75% (15/20). Blinded raters validated nCLE videos of malignancy and airway/lung parenchyma (260 ratings) with 92% accuracy. The inter-observer agreement (κ=0.79, 95% CI 0.71 to 0.87) and intra-observer reliability (mean±SD κ=0.80±0.05) were substantial. Conclusion: Small lung lesions can safely be imaged by RANB nCLE. Blinded raters accurately and consistently distinguished malignant nCLE videos from airway/lung parenchyma, showing the potential to improve lung cancer diagnostics.

Checkpoint inhibitors have dramatically changed the treatment landscape for advanced NSCLC, however only a fraction of patients (pts) benefit with durable responses. Tumors that possess oncogenic driver mutations may have increased resistance to PD-1/PDL-1 inhibitors based on subgroup data from clinical trials, though little research has been performed using large real-world patient cohorts, nor is it clear whether PD-L1 expression or smoking history is predictive of immunotherapy response in these pts. This retrospective study assessed realworld progression-free survival (rwPFS) in pts with driver-mutated advanced NSCLC and correlated endpoints with PD-L1 expression and smoking history. The nationwide Flatiron Health Electronic Health Record (EHR)-derived deidentified database was used to analyze data from pts with advanced NSCLC with tumors possessing an oncogenic driver mutation treated with single-agent immunotherapy (pembrolizumab, nivolumab, or atezolizumab). Median rwPFS (based on clinician documentation of clinical progression) in months (m) was determined for tumor molecular subtypes (EGFR, ALK, BRAF, and KRAS) and correlated to PD-L1 expression and smoking history. Kaplan-Meier curves characterized rwPFS and comparisons were assessed in the overall cohort and in driver-mutated subgroups using the log-rank test. 1,746 pts with driver-mutated tumors involving alterations in EGFR (n = 458), ALK (n = 65), BRAF (n = 146), and KRAS (n = 1077) received single-agent immunotherapy between 4/23/14 to 2/28/19. Median age was 69 years, 58% were female, and 19% had no smoking history. Median rwPFS varied significantly by tumor mutation subtypes (p < 0.001) with KRAS- (3.3 m, 95% CI 3.0, 3.6) and BRAF-mutated (3.6 m, 95% CI 2.6, 4.7) tumors having longer rwPFS than EGFR- (2.5 m, 95% CI 2.3, 2.6) and ALK-mutated tumors (2.3 m, 95% CI 1.6, 3.1). Percent progression-free at 12 m was 21% for KRAS-, 21% for BRAF-, 8% for EGFR-, and 11% for ALK-mutated tumors. In a subset of 795 patients, PD-L1 expression data were available. Only in KRAS-mutated tumors did rwPFS vary significantly by PD-L1 expression with PD-L1 positive (> 1%) tumors having longer rwPFS than PD-L1 negative tumors (4.2 m vs. 3.0 m, p < 0.001). PD-L1 expression was not associated with rwPFS in EGFR-, ALK-, or BRAF-mutated tumors. However, a history of smoking was associated with longer rwPFS as compared to no smoking history in EGFR- (2.6 m vs. 2.3 m, p < 0.05) and ALKmutated tumors (3.0 m vs. 2.1 m, p < 0.05). In this study, of one of the largest cohorts of pts with driver-mutated NSCLC to date, rwPFS with single-agent checkpoint inhibitors varied significantly between oncogenic mutation subtypes. PD-L1 expression may not be a useful biomarker of immunotherapy response in EGFR-, ALK-, and BRAF-mutated tumors. Moreover, rwPFS was relatively short for EGFR- and ALK-mutated tumors and patients with these molecular subtypes of NSCLC may derive little benefit from single-agent immunotherapy. Continued study in these sub-populations of immuno-chemo +/- VEGF inhibitor therapy combinations that may overcome resistance to PD-1/PD-L1 inhibitors is needed.

Head and neck cancer (HNC) patients have a high rate of work discontinuation while receiving daily fractionated radiation (RT). In this analysis of HNC cancer patients treated with RT we compare baseline characteristics between working patients and not-working/not-retired patients and identify patient and treatment-related factors that can be targeted to reduce time lost from work. Number of days worked in the past week was prospectively recorded at each on treatment visit (OTV) for HNC patients receiving definitive or adjuvant (chemo)radiation from 6/2016-6/2018. Patients not working due to disability/unemployment were eligible; retired patients were excluded. Baseline characteristics were compared. Weekly patient weight (lbs), pain rating (0-10), fatigue (0-10), and ECOG status (0-5) changes were described by group. Weekly ECOG status was also analyzed dichotomously as ECOG 0 compared to ECOG ≥1. Impact of each variable on loss of working days was assessed in both univariate analysis (UVA) and multivariate analysis (MVA). A total of 69 patients were included, 29 of whom were working at the start of RT. Mean age was higher in working patients (58.8 vs 53.7 years, p = 0.004). Distance from cancer center, gender, tobacco use, alcohol use, hospitalizations during RT, presence of feeding tubes, surgery prior to RT, concurrent chemotherapy, p16 status, and location of tumor were not significantly different between working and disabled groups. Changes in weight percentages, pain, fatigue, and ECOG status between last OTV and baseline were not significantly different between the two groups. Differences in percent of patients with ECOG ≥1 in each OTV are described in the table below. In working patients, mean days worked steadily declined, starting at 4.1 days in week 1 to 1.1 in week 7. In UVA, 10% loss in weight from baseline results in 2.6 days/week lost from work (p<0.005). Pain increase by 1 unit results in 0.25 days/week lost, fatigue increase by 1 unit results in 0.28 days/week lost, and ECOG status increase by 1 unit results in 1.48 days/week lost (p<0.005 for all). In MVA, change in pain was significantly associated with days/week lost from work (p = 0.05). Changes in weight, fatigue, and ECOG status were non-significant (p = 0.08, 0.11, 0.07, respectively). The decision to work during radiation is not impacted by many baseline patient characteristics, and less than 50% of HNC patients attempt to work during radiation. In working patients, weight loss and increase in pain, fatigue, and ECOG status are significantly associated with reduction in working days per week. Of these, increase in pain has the most significant impact.Abstract 3869; TableWeek 1Week 2Week 3Week 4Week 5Week 6Week 7Working (%)10.310.317.244.848.355.671.4Not Working (%)35.045.052.555.055.075.077.8p-value0.0240.0030.0050.4690.6310.1170.741 Open table in a new tab

Patients with advanced NSCLC who harbor an EGFR mutation or are never smokers do not benefit from single-agent immunotherapy. Retrospective subgroup analyses from recent phase III trials suggest that immunotherapy-chemotherapy +/- VEGF inhibition may overcome the resistance to immunotherapy seen in these patients, though further prospective research is needed and no checkpoint inhibitor to date is approved in the first-line for EGFR patients after TKI failure. This trial will exclusively examine a population of patients with stage IV non-squamous disease who either have an EGFR exon 19 or 21 mutation or are never smoker wild-types to determine whether the PD-L1 inhibitor atezolizumab in combination with pemetrexed/carboplatin and bevacizumab can improve outcomes. This is a phase II double-arm, multi-center, open-label trial to assess pemetrexed/carboplatin and bevacizumab ± atezolizumab in 117 subjects with stage IV non-squamous NSCLC. Randomization will be 2:1 with twice the number in the + atezolizumab arm, and randomization will be stratified by EGFR mutation status (i.e. EGFR exon 19 or 21 vs. never smoker wild-type) to ensure equal distribution in each arm. Never smoker wild-type is defined as smoking <100 cigarettes in a lifetime and without any EGFR mutation or ALK or ROS1 rearrangement. Patients with EGFR exon 19 or 21 mutations must have progression or intolerance of treatment with prior TKI therapy. All patients must be chemotherapy, immunotherapy, and VEGF inhibitor therapy naïve. Primary endpoint will be progression-free survival (PFS). Secondary endpoints will include overall survival (OS), overall response rate, duration to response, and time to response. Primary objective is to compare PFS between arms. Secondary objectives include a safety analysis in all treated subjects, and comparisons of PFS and OS between arms for the subset of patients with tumors with EGFR exon 19 or 21 mutations. Correlative studies will include evaluation of biomarkers of the signaling network and tumor microenvironment and characterizing the potential contribution of estrogen metabolites. Enrollment for this trial will open in August of this year and accrual will continue for 31 months. Section not applicable - Trial in progress Section not applicable - Trial in progress

Purpose/Objective(s)Longer treatment times for intensity modulated radiotherapy (IMRT) have raised concerns about sublethal damage repair and intrafraction patient motion. In addition, higher monitor units (MU) required for IMRT may theoretically increase the risk of second malignancies. The purpose of this study was to determine if using larger IMRT beamlets (segments) can reduce treatment time and MU without diminishing IMRT plan quality.Materials/MethodsDefinitive IMRT plans were generated using Pinnacle inverse treatment planning software for 10 patients with locally advanced head and neck carcinoma (5 oropharynx and 5 larynx). Target volumes and critical structures were contoured and dose constrained in accordance with RTOG protocol 0522 (70 Gy high dose PTV and 56 Gy elective dose PTV). In every plan, bilateral cervical lymph nodes were treated and 8-9 evenly spaced beams with a maximum of 120 segments were utilized. Historically for IMRT we have used a minimum segment size on the order of 2 cm2, which represents only 1-3% of the size of a typical PTV seen on a beam's-eye-view projection (BEV). For each study patient, 4 IMRT plans were generated; a baseline plan using the following settings: A2 (minimum segment area [A] = 2% of the size of PTV on BEV) with MU3 (minimum of 3 MU per segment), and 3 comparison plans: A25 MU5, A25 MU10, and A50 MU5. All other planning parameters were held constant. We compared dose-volume histogram data, total treatment time, and total MU for each plan using a paired t test.ResultsWhile meeting the RTOG protocol-defined target dose specifications and normal tissue constraints in every case, the A25 MU5 segment settings reduced the average treatment time by 40% (9.2 vs. 15.4 min, p = 0.0001) compared to the baseline A2 MU3 plans. The total MU required for A25 MU5 plans was likewise reduced by 47% (435 vs. 827 MU, p < 0.0001) compared to A2 MU3 plans. The amount of PTV70 receiving D105 (73.5 Gy) was increased 11% (9% to 20%) by the A25 MU5 settings, however the volume receiving D110, D100, D95, D90, and the mean dose varied by < = 1% for PTV70, PTV56, and for non-target tissue. The mean dose to each of the 14 parotid glands eligible for sparing averaged 22.0 Gy (range, 16.3-24.9) using A2 MU3 settings vs. 22.1 Gy (range, 16.5-25.1) for A25 MU3. Plans with further increase in segment size (A50 MU5) or MU per segment (A25 MU10) reduced treatment time by 53-55%, however the dose conformality and homogeneity was reduced such that 10% of PTV70 received D110, 44-49% received D105, and in 1 case the parotid and spinal cord tolerances could not be met.ConclusionsBy increasing the minimum IMRT segment size and MU settings, the total treatment time and total MU for head & neck IMRT can be appreciably reduced without compromising plan quality. Each of these settings appears to have an optimal range, above which the IMRT plan quality will diminish. Purpose/Objective(s)Longer treatment times for intensity modulated radiotherapy (IMRT) have raised concerns about sublethal damage repair and intrafraction patient motion. In addition, higher monitor units (MU) required for IMRT may theoretically increase the risk of second malignancies. The purpose of this study was to determine if using larger IMRT beamlets (segments) can reduce treatment time and MU without diminishing IMRT plan quality. Longer treatment times for intensity modulated radiotherapy (IMRT) have raised concerns about sublethal damage repair and intrafraction patient motion. In addition, higher monitor units (MU) required for IMRT may theoretically increase the risk of second malignancies. The purpose of this study was to determine if using larger IMRT beamlets (segments) can reduce treatment time and MU without diminishing IMRT plan quality. Materials/MethodsDefinitive IMRT plans were generated using Pinnacle inverse treatment planning software for 10 patients with locally advanced head and neck carcinoma (5 oropharynx and 5 larynx). Target volumes and critical structures were contoured and dose constrained in accordance with RTOG protocol 0522 (70 Gy high dose PTV and 56 Gy elective dose PTV). In every plan, bilateral cervical lymph nodes were treated and 8-9 evenly spaced beams with a maximum of 120 segments were utilized. Historically for IMRT we have used a minimum segment size on the order of 2 cm2, which represents only 1-3% of the size of a typical PTV seen on a beam's-eye-view projection (BEV). For each study patient, 4 IMRT plans were generated; a baseline plan using the following settings: A2 (minimum segment area [A] = 2% of the size of PTV on BEV) with MU3 (minimum of 3 MU per segment), and 3 comparison plans: A25 MU5, A25 MU10, and A50 MU5. All other planning parameters were held constant. We compared dose-volume histogram data, total treatment time, and total MU for each plan using a paired t test. Definitive IMRT plans were generated using Pinnacle inverse treatment planning software for 10 patients with locally advanced head and neck carcinoma (5 oropharynx and 5 larynx). Target volumes and critical structures were contoured and dose constrained in accordance with RTOG protocol 0522 (70 Gy high dose PTV and 56 Gy elective dose PTV). In every plan, bilateral cervical lymph nodes were treated and 8-9 evenly spaced beams with a maximum of 120 segments were utilized. Historically for IMRT we have used a minimum segment size on the order of 2 cm2, which represents only 1-3% of the size of a typical PTV seen on a beam's-eye-view projection (BEV). For each study patient, 4 IMRT plans were generated; a baseline plan using the following settings: A2 (minimum segment area [A] = 2% of the size of PTV on BEV) with MU3 (minimum of 3 MU per segment), and 3 comparison plans: A25 MU5, A25 MU10, and A50 MU5. All other planning parameters were held constant. We compared dose-volume histogram data, total treatment time, and total MU for each plan using a paired t test. ResultsWhile meeting the RTOG protocol-defined target dose specifications and normal tissue constraints in every case, the A25 MU5 segment settings reduced the average treatment time by 40% (9.2 vs. 15.4 min, p = 0.0001) compared to the baseline A2 MU3 plans. The total MU required for A25 MU5 plans was likewise reduced by 47% (435 vs. 827 MU, p < 0.0001) compared to A2 MU3 plans. The amount of PTV70 receiving D105 (73.5 Gy) was increased 11% (9% to 20%) by the A25 MU5 settings, however the volume receiving D110, D100, D95, D90, and the mean dose varied by < = 1% for PTV70, PTV56, and for non-target tissue. The mean dose to each of the 14 parotid glands eligible for sparing averaged 22.0 Gy (range, 16.3-24.9) using A2 MU3 settings vs. 22.1 Gy (range, 16.5-25.1) for A25 MU3. Plans with further increase in segment size (A50 MU5) or MU per segment (A25 MU10) reduced treatment time by 53-55%, however the dose conformality and homogeneity was reduced such that 10% of PTV70 received D110, 44-49% received D105, and in 1 case the parotid and spinal cord tolerances could not be met. While meeting the RTOG protocol-defined target dose specifications and normal tissue constraints in every case, the A25 MU5 segment settings reduced the average treatment time by 40% (9.2 vs. 15.4 min, p = 0.0001) compared to the baseline A2 MU3 plans. The total MU required for A25 MU5 plans was likewise reduced by 47% (435 vs. 827 MU, p < 0.0001) compared to A2 MU3 plans. The amount of PTV70 receiving D105 (73.5 Gy) was increased 11% (9% to 20%) by the A25 MU5 settings, however the volume receiving D110, D100, D95, D90, and the mean dose varied by < = 1% for PTV70, PTV56, and for non-target tissue. The mean dose to each of the 14 parotid glands eligible for sparing averaged 22.0 Gy (range, 16.3-24.9) using A2 MU3 settings vs. 22.1 Gy (range, 16.5-25.1) for A25 MU3. Plans with further increase in segment size (A50 MU5) or MU per segment (A25 MU10) reduced treatment time by 53-55%, however the dose conformality and homogeneity was reduced such that 10% of PTV70 received D110, 44-49% received D105, and in 1 case the parotid and spinal cord tolerances could not be met. ConclusionsBy increasing the minimum IMRT segment size and MU settings, the total treatment time and total MU for head & neck IMRT can be appreciably reduced without compromising plan quality. Each of these settings appears to have an optimal range, above which the IMRT plan quality will diminish. By increasing the minimum IMRT segment size and MU settings, the total treatment time and total MU for head & neck IMRT can be appreciably reduced without compromising plan quality. Each of these settings appears to have an optimal range, above which the IMRT plan quality will diminish.

IMRT has greatly enhanced our ability to conform dose to target structures while minimizing dose to normal tissues. This improved conformality mandates that targets be clearly defined and target motion be critically assessed. We set out to analyze breast motion due to respiration using 4D CT scans for women undergoing IMRT breast radiotherapy at Fox Chase Cancer Center (FCCC) and to determine whether our current margins are adequate to cover the entire breast in all phases of the respiratory cycle during normal quiet breathing. 20 women, 10 with left breast cancer and 10 with right breast cancer, who were simulated for breast radiotherapy at FCCC underwent a free breathing 4D CT scan. An average intensity projection scan (ave) was constructed from the 4D dataset. The 4D scan was analyzed in detail in order to identify the two phases of the respiratory cycle which represented the extremes of respiration during normal quiet breathing–maximum inspiration (MI) and maximum expiration (ME). The breast, ipsilateral lung and, for left-sided breast cancers, the heart, were contoured on the average scan as well as on the two phases mentioned above. Breast motion was measured on the 4D scan in three planes: axial, sagittal and coronal. The axial motion was measured at the chest wall, the coronal motion was measured at the inframammary fold and the sagittal motion was measured at the breast surface. These landmarks represented the points of maximum motion observed on the 4D scan in each plane. In order to assess the coverage of the breast and the doses to the normal structures at the extremes of the respiratory cycle, IMRT plans were generated for the ave scan as well as for the MI and ME phases. The IMRT plan generated from the ave scan was used to treat the patient. Each IMRT plan was then analyzed and compared using the parameters which are used to evaluate all breast IMRT plans at FCCC. For the 20 patients analyzed, the mean motion in the axial direction was 1.9 mm (range 0–4 mm), in the coronal direction was 0.7 mm (range 0–2 mm) and in the sagittal direction was 2.0 mm (range 0–4 mm). For 7 patients, the sagittal motion was greatest and for 8 patients, the axial motion was the greatest. For all but 1 patient, the coronal motion was the least. PTVs were defined as breast tissue plus 2 cm superiorly and inferiorly and 7 mm posteriorly. Dose volume parameters were analyzed for the first 10 patients. The median PTV volume for the ave scan was 748 cc, for the MI phase was 733 cc and for the ME phase was 728 cc. The V95 for the PTV for the ave plan 98.4% vs. 96.3% for the MI plan and 96.5% for the ME plan. (The FCCC goal for the V95 is ≥95%). None of the other dose-volume parameters analyzed (V100, V105, D99, D1, mean dose to PTV, min dose to PTV and max dose to PTV) were significantly different between the ave and the MI and ME plans. The V20 for the lung was also analyzed and was not significantly different between the ave, MI and ME plans. 4D CT simulation is unnecessary for breast radiotherapy treatment planning, as the motion of the breast is minimal during normal quiet breathing. The present IMRT methods being utilized at our institution are sufficient and no additional margin to account for respiratory motion is necessary.

Purpose/Objective: A study by the RTOG has shown that single men with head and neck cancer treated with radiation have a worse overall survival when compared with married men, independent of disease presentation and treatment. For women with head and neck cancer, survival was not affected by marital status. The purpose of this study was to examine the effect of marital status on patient characteristics and outcome in women with early stage breast cancer treated with breast conserving therapy. Materials/Methods: The study population consisted of 2,143 patients with T1/T2, pN0-2 breast cancer treated with lumpectomy and radiation at the Fox Chase Cancer Center between 1984 and 2003. Patients treated with mastectomy, male breast cancer, and locally advanced disease were ineligible. The median follow-up was 76 months (range 12–265 months). The patients were divided into four groups based on their marital status---63% were married, 10% were single, 10% were divorced or separated and 18% were widowed. The Cox proportional hazards model was used to determine if marital status was an independent predictor of relapse free survival (RFS) controlling for age, T-stage, pathologic nodal status, margins, ER status, use of tamoxifen or adjuvant chemotherapy and income. Bonferroni adjustment was applied to account for multiple comparisons. Results: As expected, there were significant differences in RFS by age but not by marital status. Women < 40 y/o were more often single than married (p=0.0002), while women > 70 y/o were more often widowed (p<0.0001). Divorced women tended to be younger than married (p=0.0008) women. Widowed women more often had tumors detected by mammography than physical exam alone as compared to married (p=0.0084), single (p=0.0042) or divorced (p=0.0018) women. Widowed women had fewer re-excisions than married (p=0.0004) or divorced (p=0.0018) women, and more likely received no adjuvant therapy compared with married (p<0.0001), single (p<0.0001) or divorced (p<0.0001) women. Marital status was not an independent predictor of relapse-free survival. The hazard ratios and p-values were as follows---widowed vs. married (p=0.8470, HR=0.966), single vs. married (p=0.6459, HR=0.907), and divorced/separated vs. married (p=0.5642, HR=1.118). There was no statistically significant association between marital status and RFS. However, age was found to be a statistically significant predictor of RFS, with women < 40 y/o vs. > 70 y/o (p=0.0109, HR=1.912) doing worse despite similar disease characteristics and treatment factors. Conclusions: There are significant differences in age, method of detection, use of re-excision, and use of systemic therapy based upon marital status in women with early stage breast cancer. However, despite these differences the relapse-free survival for women with early stage breast cancer is not affected by their marital status. Our findings corroborate those of the RTOG in head and neck cancer which found that married women did as well as their unmarried counterparts.

Purpose/Objective: Tamoxifen (tam) has an important role in the adjuvant treatment of breast cancer pts, but the optimal sequencing of tam and radiation therapy (RT) is not well established. Some studies have suggested a potential negative effect of tam on radiosensitivity when administered during RT. Other clinical studies have not reported adverse outcomes with concurrent administration. This study analyzes the impact of the sequencing of tam and RT on outcome in women with early stage I/II breast cancer treated with conservative surgery and RT.Materials/Methods: From 1982 to 2004, 1088 women with stage I/II breast cancer underwent conservative surgery and RT and received adjuvant tam, with or without chemotherapy. 699 pts received tam sequentially after completion of RT (sequential group). 389 pts received concurrent tam during RT (concurrent group). Median age was 57 yrs for the sequential group and 64 yrs for the concurrent group. Median follow-up was 48 months for the sequential group and 86 months for the concurrent group. Cox multivariate regression models (MVA) were used to determine independent predictors of local failure (LF), distant metastases (DM), relapse-free survival (RFS) and overall survival (OS). Ten-yr rates were estimated using Kaplan-Meier methodology and comparisons were made using the log-rank test. Complications were assessed using chi-square and Fishers Exact tests and included breast edema, breast erythema, breast fibrosis, arm edema, pneumonitis, rib fracture and fat necrosis. Cosmesis was scored as excellent/good or fair/poor at each follow-up. Differences in cosmesis were assessed using chi-square test.Results: Statistically significant differences between the two groups were observed for age and use of chemotherapy. There were more pts in the sequential group younger than 49 yrs of age (26% vs 10%,p<0.0001) and more that had chemotherapy (43% vs 14%,p<0.0001) compared to the concurrent group. The ER status, PR status, tumor size and regions of RT were comparable between the two groups. There were no statistically significant differences observed between the sequential group and the concurrent group for the actuarial 10-yr rates of LF (1% vs 2%;p=0.45), DM (5% vs 6%;p=0.48), and RFS (94% vs 92%;p=0.25). Univariate analysis showed a significant difference between the sequential group and the concurrent group for the 10-yr rates of OS (94% vs 89%;p=0.0007). However, after adjusting for age and other prognostic variables in the MVA, the sequencing of tam no longer significantly influenced OS (p=0.21). The sequencing of tam with RT was not found to be a significant predictor of LF (p=0.26), DM (p=0.93), RFS (p=0.56) or OS (p=0.21). In addition, there were no significant differences observed in the rates of any of the complications between the sequential and concurrent groups (all p>0.08). However, there was a statistically significant higher rate of excellent/good cosmesis in the sequential group compared to the concurrent group (95% vs 88%,p=0.0004).Conclusions: The sequencing of tam with RT does not influence disease control or survival outcome in women with early stage breast cancer treated with conservative surgery and RT and tam. To our knowledge, this current study is the largest series to date demonstrating this. While no increased risk of complications was observed when tam was given concurrently with RT, we did observe a small but statistically significant negative effect on long-term cosmesis. We continue to recommend routine sequential sequencing of tamoxifen and radiation, but concurrent administration has acceptable outcomes if required by a clinical trial or delay in time to radiation after surgery or chemotherapy. Purpose/Objective: Tamoxifen (tam) has an important role in the adjuvant treatment of breast cancer pts, but the optimal sequencing of tam and radiation therapy (RT) is not well established. Some studies have suggested a potential negative effect of tam on radiosensitivity when administered during RT. Other clinical studies have not reported adverse outcomes with concurrent administration. This study analyzes the impact of the sequencing of tam and RT on outcome in women with early stage I/II breast cancer treated with conservative surgery and RT. Materials/Methods: From 1982 to 2004, 1088 women with stage I/II breast cancer underwent conservative surgery and RT and received adjuvant tam, with or without chemotherapy. 699 pts received tam sequentially after completion of RT (sequential group). 389 pts received concurrent tam during RT (concurrent group). Median age was 57 yrs for the sequential group and 64 yrs for the concurrent group. Median follow-up was 48 months for the sequential group and 86 months for the concurrent group. Cox multivariate regression models (MVA) were used to determine independent predictors of local failure (LF), distant metastases (DM), relapse-free survival (RFS) and overall survival (OS). Ten-yr rates were estimated using Kaplan-Meier methodology and comparisons were made using the log-rank test. Complications were assessed using chi-square and Fishers Exact tests and included breast edema, breast erythema, breast fibrosis, arm edema, pneumonitis, rib fracture and fat necrosis. Cosmesis was scored as excellent/good or fair/poor at each follow-up. Differences in cosmesis were assessed using chi-square test. Results: Statistically significant differences between the two groups were observed for age and use of chemotherapy. There were more pts in the sequential group younger than 49 yrs of age (26% vs 10%,p<0.0001) and more that had chemotherapy (43% vs 14%,p<0.0001) compared to the concurrent group. The ER status, PR status, tumor size and regions of RT were comparable between the two groups. There were no statistically significant differences observed between the sequential group and the concurrent group for the actuarial 10-yr rates of LF (1% vs 2%;p=0.45), DM (5% vs 6%;p=0.48), and RFS (94% vs 92%;p=0.25). Univariate analysis showed a significant difference between the sequential group and the concurrent group for the 10-yr rates of OS (94% vs 89%;p=0.0007). However, after adjusting for age and other prognostic variables in the MVA, the sequencing of tam no longer significantly influenced OS (p=0.21). The sequencing of tam with RT was not found to be a significant predictor of LF (p=0.26), DM (p=0.93), RFS (p=0.56) or OS (p=0.21). In addition, there were no significant differences observed in the rates of any of the complications between the sequential and concurrent groups (all p>0.08). However, there was a statistically significant higher rate of excellent/good cosmesis in the sequential group compared to the concurrent group (95% vs 88%,p=0.0004). Conclusions: The sequencing of tam with RT does not influence disease control or survival outcome in women with early stage breast cancer treated with conservative surgery and RT and tam. To our knowledge, this current study is the largest series to date demonstrating this. While no increased risk of complications was observed when tam was given concurrently with RT, we did observe a small but statistically significant negative effect on long-term cosmesis. We continue to recommend routine sequential sequencing of tamoxifen and radiation, but concurrent administration has acceptable outcomes if required by a clinical trial or delay in time to radiation after surgery or chemotherapy.

Purpose/Objective: A randomized trial has shown improved disease-free survival from an aromatase inhibitor after completion of 5 years of tamoxifen (tam). The purpose of this study was to further estimate the potential clinical benefits from an aromatase inhibitor by observing outcomes in patients who are event-free 5 years after by breast-conserving surgery, radiation, and tam. Materials/Methods: 471 women with invasive breast cancer were treated by breast-conserving surgery, axillary dissection and radiation. Eligibility included T1-2 disease, use of tam, follow-up ≥ 5 years, no prior breast cancer, and freedom from all events - local-regional or distant recurrence, or contralateral breast cancer (CBC) at 5 years. Patients had tam alone (319) or tam with chemotherapy (152). Patients treated with chemotherapy were more often T2 (30% vs. 18%, p=0.004), node positive (57% vs. 21%, p<0.0001), and under 60 (80% vs. 39%, p<0.0001). No patient during this study period (1982–99) had an aromatase inhibitor. Median follow-up is 8.25 years (range 5–18.5 years). Results: There were 36 events after the initial event-free 5-year period after radiation and tam. There were 10 CBC, and 26 recurrences – 8 local, 1 regional and 17 distant. The 10-year risk of local-regional recurrence was 2.5%. There was no difference in this risk between patients receiving tam and chemotherapy or tam alone (p=0.9). The 10-year risk of CBC was 3.6%. The risk was 7.4% in those having chemotherapy and 2.3% with tam alone (p=0.34). The 10-year risk of distant metastasis was 4.4%: It was 7.6% in those receiving chemotherapy and 3.1% after tam alone (p=0.12). Similarly, the event-free survival rate for all patients was 93%: It was 90% after chemotherapy and tam and 94% after tam alone (p=0.31). These differences in the rates of CBC, distant metastasis and event-free survival are attributable to more adverse characteristics that would have led to use of chemotherapy for initial therapy. The overall survival was 89%, and there was a trend that did not reach statistical significance for improvement for patients receiving chemotherapy versus tam alone (94% versus 88% p=0.15). Multivariate analysis for overall survival was conducted with covariates of age, stage, node status, ER status, menopause status, and use of chemotherapy. Only younger age (< 60 versus 60+, p<0.0001) and lower node status (N0 versus N4+, p=0.0045) (N1-3 versus N4+, p=0.0051) were significant predictors of improved overall survival. Conclusions: Women who were event-free 5 years after completion of breast-conserving surgery, radiation and tam had a 10-year local-regional recurrence of 2.5%. Assuming a 40% reduction of this risk by the addition of an aromatase inhibitor to 5 years of tam, the absolute benefit would be only 1% of patients. Difference in endpoints between patients treated with and without chemotherapy did not reach statistical significance probably due to the smaller numbers in this study compared to larger randomized trials. With a 2% rate of CBC, 3% rate of distant metastases, and 94% rate of event-free survival, low risk patients originally selected for treatment by tam alone could have a marginal 1–2% clinical benefit from the additional therapy. Patients treated with chemotherapy showed trends for a potential greater benefit of 3% or more from an aromatase inhibitor after completing 5 years of tam, because of factors placing them at higher risk than women usually treated by tam alone.

Background High-resolution positron-emission mammography (PEM) is a new device, which allows the imaging of breast tissue. A prospective study was performed to assess the accuracy of PEM in newly diagnosed breast cancer patients. Methods In a prospective multicenter study, 44 women with confirmed breast cancers were imaged with a high-resolution PEM scanner (Naviscan PET Systems, Rockville, MD) with 18F-fluorodeoxyglucose. The images were blindly evaluated and were compared with final pathology. Results The majority of the index lesions were seen on PEM (89%, 39/44). PEM detected 4 of 5 incidental breast cancers, 3 of which were not seen by any other imaging modalities. Of 19 patients undergoing breast-conserving surgery, PEM correctly predicted 6 of 8 (75%) patients with positive margins and 100% (11/11) with negative margins. Conclusion The current PEM device shows promise in detecting breast malignancies and may assist in the planning of adequate partial mastectomy procedures to better ensure negative margins.