Purpose/Objective(s) Radiation therapy (RT) remains a staple in the treatment of women with breast cancer (BC). Locoregional disease recurrence remains a substantial clinical concern that compromises survival in both estrogen receptor-positive (ER+) and triple-negative (TN) BC with > 3 involved lymph nodes. Currently, cyclin-dependent kinase 4 and 6 inhibitors (CDK4/6i) abemaciclib and ribociclib are indicated for the adjuvant treatment of women with locally advanced ER+ BC. We previously demonstrated that concurrent CDK4/6i and RT radiosensitizes in vitro and in vivo models of ER+ BC. The optimal treatment sequencing, however, remains unclear, and this information would inform future clinical trial design. Materials/Methods ER+ MCF-7 and T47D cells were treated in vitro with FDA-approved CDK4/6i abemaciclib or ribociclib in three treatment sequences: concurrent, adjuvant, and neoadjuvant (with a 48-hour window) and clonogenic survival assays were used to determine the most effective sequence. Neoadjuvant treatment consisted of drug for 48 hours then RT, concurrent consisted of drug 1 hour prior to RT, and adjuvant treatment consisted of RT followed by drug 48 hours later. IC50 values for CDK4/6i were used. Results Concurrent administration of abemaciclib and ribociclib led to clinically meaningful radiosensitization, with radiation enhancement ratios (rER) ranging from 1.62-2.23 with abemaciclib: 1.62-2.23 and 2.21-3.24 with ribociclib. Similar effects were observed with adjuvant CDK4/6i (rER abemaciclib: 1.52-1.93, ribociclib: 1.74-2.33). In contrast, no significant radiosensitization was noted with the neoadjuvant sequence as compared to the controls (rER: abemaciclib: 0.93-1.01, ribociclib: 0.96-1.13). Mechanistically, this effect was not dependent on cell cycle rearrangements and was mediated, at least in part, by impaired homologous recombination. Conclusion Our data suggest that RT is most effective when used concurrently with, or prior to, CDK4/6i administration. These data will inform ongoing in vivo studies as part of a federally funded SPORE grant and provide the basis for treatment sequencing in the recently opened phase IB safety and efficacy trial that is treating women with locally advanced ER+ BC who are at high risk of locoregional recurrence (NCT05996107). Ongoing efforts are exploring the efficacy and sequencing of models of TNBC as well.

Purpose/Objective(s) Most women with breast cancer (BC) receive radiation therapy (RT) as part of the standard of care, however its efficacy remains inadequate for those with locally advanced disease. Thus, more effective radiosensitization strategies are needed. We performed a radiosensitizer screen to identify potential mediators of RT resistance and identify novel therapeutic targets. Materials/Methods We screened 130 clinically available drugs for radiosensitization using the Genomics of Drug Sensitivity in Cancer database and analyzed their effectiveness based on clonogenic survival with IC50 values. MDM2 inhibitors navtemadlin (AMG-232) and alrizomadlin (APG-115) were used in p53 wild-type and mutant models of estrogen receptor (p53-WT:MCF-7, p53-MT:T47D) and ER-negative BC (p53-WT: CAL-51, p53-MT: MDA-MB-231). Alterations to DNA damage response were assessed with gamma-H2AX Immunofluorescence. Flow Cytometry with Annexin V staining was used to evaluate apoptosis and with Propidium Iodide staining for cell cycle progression. β-gal staining was used to evaluate senescence. In vivo efficacy of MDM2 inhibition and RT combination was evaluated with CAL-51 and CAL-51 p53 CRISPR xenograft models. Results An MDM2 inhibitor was nominated as a potent radiosensitizer (R2 = 0.43, P < 0.01). MDM2 was significantly overexpressed after RT compared to no RT in p53-WT cells. Cell growth was decreased with navtemadlin and alrizomadlin in p53-WT cells (IC50s:264-592nM) but not in p53-MT cells (IC50s > 10 μm). MDM2 inhibition with either drug radiosensitized p53-WT cells (rERs = 1.81-2.85) but not p53-MT cells (rERs = 1.00-1.03), regardless of hormone receptor status, or p53 CRISPR cells (rERs = 1.06-1.12). MDM2 inhibition in combination with RT demonstrated delayed DNA damage repair compared to RT alone. P53-WT cells demonstrated increased G1 cell cycle arrest with RT and combination therapy. Additionally, in vivo experiments showed MDM2 inhibition, using both genetic and pharmacological (navtemadlin) inhibition in combination with RT led to a significantly increased time to tumor tripling in CAL51-WT cells, but not in isogenic CAL-51 p53 CRISPR models (tripling time 31 days in WT vs not reached in RT+MDM2 KO or navtemadlin treated groups, P < 0.01; 26 days vs. 30 days in WT vs RT+navtemadlin treated p53 CRISPR CAL-51 models, P > 0.4. Mechanistically, combination MDM2 inhibition with navtemadlin and RT resulted in significantly increased apoptosis and senescence in p53-WT cell lines (CAL-51, ZR-75, and CAMA-1 cells % apoptotic: DMSO 6-9%, combination 31-54%, P < 0.01; senescent cells: DMSO 3-5%, combination 37-64%, P < 0.01). Conclusion These results demonstrate the combination of RT and MDM2 inhibition may be an effective therapeutic strategy in patients with p53-WT BC which represent the majority of BCs, regardless of hormone receptor status. Clinical trial development is currently underway to test this in women with locally advanced p53-WT BCs at high risk of locoregional recurrence.

Our analysis identifies cancer diagnosis, time to surgical consultation, time to MRI order, and ambulation status as influencing the time to inpatient radiation consultation and spine MRI orders. These variables should be taken into consideration to optimize care for inpatients with spine metastases and cord compression.

Qualitative and quantitative analyses of ARRTs based on CTCAE v4.0 show discordance in pain scores and radiation dermatitis. The underscoring of pain amongst B patients suggests possible cultural hesitancy to report pain or implicit biases in evaluating pain suggesting the need for alternative ways to discuss and score pain in B patients. Our data suggests that CTCAE does not effectively assess radiation dermatitis in B patients. There are no descriptors of hyperpigmentation in grading of radiation dermatitis in CTCAE, which may lead to underreporting in dark-skinned individuals. Better assessment tools are needed to document ARRTs to ensure appropriate evaluation and treatment across all racial groups.

Purpose: Expression of the androgen receptor (AR) has been identified as a driver of tumor growth in triple negative breast cancers (TNBC), and previous work has nominated AR inhibition as a strategy for radiosensitization in AR+ TNBC. Despite its role in radioresistance in AR+ TNBC, the mechanistic role of AR and specifically its role in mediating DNA damage repair in response to radiation therapy (RT) remains unknown. Methods: Nuclear fractionation experiments were performed to assess cellular localization of AR protein in AR+ TNBC cell lines (ACC-422, MDA-MB-453). Cells were cultured in media containing hormones (FBS) with treatment of enzalutamide (ENZA), apalutamide (APA), or darolutamide (DARO). Cells were alternatively cultured in media containing charcoal stripped serum (CSS) without hormones with R1881 stimulation. RNA-sequencing was performed to compare AR+ TNBC cells treated with CSS or R1881 stimulation alone or in combination with ionizing radiation. Reverse phase protein arrays were performed in cells treated with ENZA, RT, or combination treatment. Results: While stimulation with R1881 was sufficient to induce nuclear translocation of AR in MDA-MB-453 cells, AR inhibition with ENZA, APA, or DARO blocked AR nuclear translocation under CSS or FBS growth conditions. When cells were treated with R1881+RT, AR nuclear translocation was induced at similar or greater levels compared to R1881 alone in MDA-MB-453 and ACC-422 cells. Combination treatment of RT with ENZA in the presence of hormones reduced AR nuclear localization (39% reduction in MDA-MB-453 cells and 32% reduction in ACC-422 cells) compared to RT alone. These results suggest that decreased promoter region binding, and gene expression upregulation may be a mechanism of radiosensitization with AR inhibition. In addition, transcriptomic analyses demonstrated at least 979 genes differentially expressed in multiple models. Pathway analyses in these models showed common affected pathways included ECM-receptor interaction, PPAR-gamma activation, PI3K-Akt signaling pathway, and the MAPK/ERK signaling pathway. Proteomic analysis in the same cell lines identified apoptosis, DNA damage, and cell cycle pathway changes after RT when AR-signaling was blocked. Common affected pathways in combined analyses identified PI3K-Akt and MAPK/ERK signaling pathway changes that may be responsible for this radiosensitizing phenotype. Conclusions: Our data suggest that AR inhibition in AR+ TNBC is sufficient to inhibit AR nuclear translocation suggesting that AR may play a nuclear role in response to RT to promote DNA repair and radioresistance. We identify potential pathways, including ECM-receptor interaction, PI3K-Akt signaling pathway, and the MAPK/ERK signaling pathway that may be regulated by AR in response to RT and therefore may be responsible for radioresistance. Citation Format: Anna R. Michmerhuizen, Andrea M. Pesch, Benjamin C. Chandler, Lynn M. Lerner, Connor Ward, Leah Moubadder, Stephanie The, Breanna McBean, Caleb Cheng, Lori J. Pierce, Corey W. Speers. Multiomics analysis to uncover the mechanism of radiosensitization of AR-positive triple negative breast cancers with AR inhibition [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3307.

Purpose/Objective(s) Randomized trials support a supplemental radiation dose (Boost) to the lumpectomy (L) cavity region after whole breast irradiation (WBI), providing a 20-30% relative reduction of in-breast recurrence (IBR); with the disadvantage that it extends treatment duration ⁓ 1 week. Hypofractionated WBI (H-WBI) in ≤ 3 weeks is used after L to deliver adjuvant WBI with acceptable toxicity and comparable IBR as conventional WBI (C-WBI) 50 Gy in 2 Gy fractions (F). NRG RTOG 1005 sought to determine whether a boost delivered concomitantly with H-WBI over 15 F is non-inferior for IBR compared to boost delivered sequentially after C-WBI in patients (pts) considered at High Risk of IBR. Materials/Methods Protocol-specified High Risk pts post L with stages 0, I & II breast cancer were randomized to C-WBI 50 Gy in 25 F or 42.7 Gy in 16 F plus sequential boost of 12 Gy in 6 F or 14 Gy in 7 F (Arm I) or H-WBI 40 Gy in 15 F plus concomitant boost of 8 Gy in 15 F of 0.53 Gy per day (Arm II). Radiation was target based 3-dimensional conformal radiation therapy (3DCRT) or intensity modulated radiation therapy (IMRT) and quality review (QA) was required. Stratification was by age (<50 vs ≥ 50), adjuvant chemotherapy (Y vs N), ER status (+ vs -) and histologic grade (1, 2 v 3). The primary endpoint is IBR. Assuming Arm I 5‐year IBR of 1.59%, 90% CI upper bound hazard ratio (HR) of 2.12 and 1‐sided significance level = 0.05, 2150 pts with at least 46 IBR events provide >80% power to conclude non‐inferiority. IBR was analyzed comparing the cause-specific hazards. Adverse Events (AE) were graded with NCI CTCAE v4. Physician-reported NRG-RTOG Global Cosmetic Score (GCS) was grouped as excellent/good vs fair/poor, and arms compared with chi-square. Results 2262 of 2354 randomized pts were eligible (Arm I n=1124; Arm II n=1138). Median age 55 years, 34% Stage II, 52% grade 3, 30% ER-, 17% close/+ margins consistent with a "High Risk" population. Radiation was 3DCRT 81%, IMRT 19%, and the QA was per protocol 81% and 88% on Arm I vs II, respectively. With a median follow-up of 7.3 years and 56 IBR events, the 5 and 7-year IBR were 2.0% and 2.2% on Arm I and 1.9% and 2.6% on Arm II. The non-inferiority comparison (Arm I reference level) resulted in a HR (90% CI): 1.32 (0.84, 2.05) and p = 0.039, thus meeting non-inferiority. No differences in AEs noted between arms, with low rates of ≥ grade 3 treatment-related AEs, 3.3% vs 3.5% for Arm I vs II, respectively (p=0.79). No difference in 3-year excellent/good cosmesis by arm: 86% for Arm I vs 84% Arm II (p=0.61). Conclusion Concomitant boost with H-WBI results in non-inferior IBR compared to sequential boost after C-WBI in high-risk cases and reduces overall treatment time. Using target based 3DCRT or IMRT, there are no differences in toxicity or cosmetic outcome for concomitant v sequential boost or the WBI fractionation regimen.

Purpose/Objective(s) Since the publication of ACOSOG Z0011 in 2010, and the subsequent analysis of its radiation field design in 2014, there has been controversy about the extent of regional radiotherapy (RT) for patients with early-stage breast cancer found to have limited axillary nodal disease at time of breast conserving surgery (BCS). Furthermore, the MA.20 and EORTC 22922/10925 trials, both published in 2015, found that the addition of regional nodal RT to the internal mammary nodal chain (IMN) and upper axillary nodes including the supraclavicular region (SCV) demonstrated a reduction in regional and distant recurrences. Given the varying acceptable options for regional RT in this patient population, we sought to characterize the practice patterns at our institution. Materials/Methods Retrospective data from 2001-2021 were reviewed from a single provider to eliminate inter-provider variation. Included patients had estrogen receptor positive, HER2-negative breast cancer, and were found to have 1-3 positive axillary lymph nodes (LN) at time of upfront BCS. Univariate analyses were conducted to describe variations in treatment patterns. Results When divided by dates of diagnoses from 2001-2010 (n=82), 2011-2014 (n=44), and 2015-2021 (n=88) based on the publication of the landmark trials listed above, several significant trends were seen. There was a significant shift in radiation target, with a decrease in treating breast alone in favor of treating the breast + low axilla (Table 1). For patients in whom nodal irradiation was used, there was a significant increase in the inclusion of the IMN (0%, 9.1%, 25.0%; p<0.001). There was an increase in the utilization of sentinel LN biopsy resulting in a decrease in the median number of LN removed (11, 5, 3, respectively; p<0.001). This resulted in an increase in the mean ratio of positive LN to total LN removed (0.19, 0.39. 0.51, respectively; p<0.001). For patients treated from 2011-2021, those who received radiation to the SCV +/- IMN had larger tumors (mean 2.42 vs 1.75 cm, p=0.003), higher rates of ≥2 positive LN (40.8% vs 8.43%, p<0.001), and more lymphovascular invasion (60.5% vs 27.4%, p=0.002) compared to those who received radiation to the breast +/- low axilla. Conclusion Radiation target volumes have evolved over time in response to the publication of multiple randomized trials comparing different surgical approaches and radiation techniques. There is still significant heterogeneity, especially in treating patients with low-risk, node-positive breast cancer. Thus, future randomized control trials should consider enrolling patients across a broad range of radiation target volumes to align with contemporary treatment patterns.

Purpose/Objective(s) In 2020, 5-year results of the FAST-Forward trial and 10-year results for both the FAST trial and APBI-IMRT-Florence trials using ultrahypofractionated radiation (UHFR) were all published, each detailing equivalent outcomes for partial and whole breast UHFR schedules compared to more conventional fractionation regimes in early-stage breast cancer. Due to concerns over viral exposure, these treatment regimens were adopted in an effort to minimize patient exposure while achieving equivalent oncologic control. The purpose of this study was to compare the use of UHFR schedules for breast cancer patients before COVID-19, at the height of the pandemic and after widespread vaccination and analyze factors predictive of use. Materials/Methods Between October 2019 and February 2022, 733 patients with stage 0 or I breast cancer were offered radiation at our institution following breast conserving surgery. 43 patients were excluded (35 declined radiation and 8 received partial breast re-irradiation). We analyzed target volume (whole vs partial breast) and fractionation schemes (1 week vs 3 or more weeks) for 3 time periods of interest: pre-COVID-19 (October 2019-March 2020), height of COVID-19 (March 2020-March 2021) and post-COVID-19 vaccine (March 2021-February 2022). Fisher's exact test was used to compare frequencies of treatment parameters across groups and time periods. Logistic regression was used to determine factors associated with UHFR use. R (v4.1.2) was used for analysis. Results The use of UHFR increased during the post-vaccine period as compared to pre-Covid-19 and the pre-vaccine time periods (p<0.001), with 4.6%, 9.2% and 23%, receiving UHFR respectively. On univariate analysis in the pre-vaccine time period, patients with grade 2 and 3 cancers were less likely to get UHFR (p = 0.003 and p = 0.005, respectively). Receiving radiation from an academic attending was associated with higher use of UHFR (p = 0.0004). On multivariate (MVA) patients with grade 2 and 3 disease remained less likely to get UHFR (p 0.001 and 0.004). Treatment by an academic attending remained a significant factor for use of UHFR (p=0.0003). In the post-vaccine time period, on univariate analysis, negative ER status and grade 3 tumors were associated with less use of UHFR (p = 0.04 and p = 0.0001, respectively). On MVA, only grade 3 was associated with less UHFR use (p = 0.0002). Conclusion During COVID there was a significant increase in the use of UHFR regimens for treatment of early-stage breast cancer patients at our institution that persisted and increased after the introduction of the vaccine despite return to pre-COVID patient volumes. Uptake of these new regimens was initially higher among academic radiation oncologists, however with education of community physicians and implementation of treatment planning scorecards, widespread adoption was seen.

At a median follow-up of 205 months (17 yrs), overall survival is 75% and disease-free survival is 86.4%. There were 2 local recurrences yielding a local recurrence free survival of 95.5%. One of the patients with a local recurrence previously had a stage IIA (T1N1M0), grade 2 infiltrating ductal carcinoma, ER 90%, PR 50%, HER2 negative. The recurrence was stage IIA, (T1N1M0) grade 2, invasive lobular carcinoma, triple negative, considered a new primary. The two patients who experienced a local recurrence were successfully salvaged and are disease free. There were no cases of radiation pneumonitis. There was no significant change in the diffusing capacity for carbon monoxide either immediately after radiotherapy (P = 0.51) or with extended follow-up (P = 0.63) CONCLUSION: With long-term follow-up, the use of concurrent chemotherapy with low dose radiation as adjuvant treatment for node positive breast cancer has continued to demonstrate efficacy with excellent disease-free survival, overall survival and local control with acceptable toxicity.

e12547 Background: Clinical trials have demonstrated radiation therapy (RT) significantly reduces local recurrence following BCS, but that omission of RT does not compromise survival in the majority of women with early stage, low risk breast cancer. Criteria for omission of RT have been based on clinical factors such as age, stage, tumor size, surgical margins and estrogen receptor (ER) status. The utility of Oncotype DX RS in determining benefit of RT is not well defined. Methods: The National Cancer Database (NCDB) was queried for women ages 50-69 with T1N0M0, grade 1-2, ER+, Her2- breast cancer who underwent BCS with negative margins and had Oncotype DX RS of 0-18. Overall survival (OS) was estimated using the Kaplan-Meier method and compared between patients who received RT and endocrine therapy (ET) versus ET alone using logrank analysis. Propensity matching was performed to reduce the impact of potential confounders and balance sample bias. Cox proportional hazards regression was used to identify predictors of OS. Results: A total of 13,648 women met inclusion criteria. The median age was 60 years. 13,389 women had adjuvant RT+ET, while 259 women had ET alone. Five year OS was 98.6% in patients who underwent RT+ET compared to 95.5% in those that had ET alone (p = 0.0012). Propensity-matching by age, Charlson Deyo Comorbid Condition score, tumor size, Oncotype RS, and race. Five year OS in the propensity matched cohort was 99.6% for women receiving RT+ET, and 98.3% for ET alone, which was not significantly different (p = 0.095). On multivariate analysis receipt of radiotherapy was not predictive of survival. Age and comorbidity score were the only significant predictors of survival. Conclusions: Patients who receive adjuvant RT with low risk, early stage ER+/Her2- breast cancer had higher OS than women who received ET alone on univariate analysis. However, results from both multivariate analysis and propensity score matching suggest no survival benefit to the addition of RT. Prospective studies are underway assessing omission of RT on the basis of multigene assays rather than clinical features alone. [Table: see text]

572 Background: The innovation of sentinel lymph node biopsy (SLNB) has allowed many patients with invasive breast cancer to forego ALND. However, the benefit of ALND is unclear in patients with pathologic N1 disease detected on SLNB following neoadjuvant chemotherapy, particularly in patients who receive adjuvant regional nodal irradiation. Methods: The National Cancer Database (NCDB) was queried for women ages 18-75 with cT1-3N1, and ypT0-T3N1M0 invasive breast cancer who underwent definitive surgical resection with axillary staging and also received adjuvant RNI. Patients treated from 2012 – 2015 were included to allow for appropriate coding of extent of axillary surgery. Overall survival (OS) was estimated using the Kaplan-Meier method and compared between patients who received SLNB alone and ALND with or without SLNB utilizing log rank testing. Propensity matching was performed to reduce the impact of potential confounders and balance sample bias. Cox proportional hazards regression was used to identify predictors of overall survival. Results: A total of 1411 women were identified who met inclusion criteria. The median age was 52 (23-75) years. 206 (15%) women had SLNB alone and 1205 (85%) had ALND with or without SLNB. Five year OS was 73% in patients who underwent ALND compared to 76% in those who had SLNB alone (p =0.39). Following propensity matching by age, race, Charlson Deyo Comorbid Condition score, pT stage, grade, ER status, and HER2 status, 5 year OS was 79% in patients who underwent SLNB alone vs. 69% in patients who had ALND performed (p = 0.33). On Cox regression analysis, none of the variables predicted for 5 year OS. Conclusions: ALND in addition to RNI did not improve survival in patients with cT1-3N1M0 and ypT0-3N1M0 breast cancer compared to SLNB and RNI. We await results of the Alliance 011202 randomized trial for prospective validation of ALND omission in a similar subset of patients. [Table: see text]

Purpose/Objective(s)As long-term cardiac effects of breast cancer treatment become better defined, it is important to minimize cardiac exposure from radiation. We undertook this study to compare normal tissue exposure (heart, left anterior descending artery (LAD), and lung) and target volume coverage (whole breast and tumor bed) using free breathing (FB) 3-D conformal planning, IMRT, and deep inspiration breath hold (DIBH) using the assisted breathing control (ABC) device. A cost effectiveness analysis is also included.Materials/MethodsFollowing IRB approval, we reviewed 10 consecutive patients who received radiation for left sided breast cancer treated with ABC. Patients were selected for the ABC device if any part of the heart would receive greater than 25 Gy with standard 3-D planning. For these 10 patients we created FB 3-D conformal plans as well as IMRT plans to compare target coverage and normal tissue avoidance. All patients were planned with a whole breast dose of 46 Gy with a 16 Gy tumor bed boost. Average values were calculated for each parameter.ResultsThe following numbers (illustrated in the Table) represent the means for each parameter for 3D, IMRT, DIBH respectively: V20 Lt lung: 12.3, 19.5, 12.1%; V10 Lt lung: 17.4, 35.4, 19.1%; mean lung dose: 3.7, 7.8, 3.6 Gy; V25 heart: 2.4, 6.1, 0.2%; mean heart dose: 3.8, 8.6, 2.8 Gy; V25 LAD: 55, 9.84, 1.78%; Mean LAD dose: 33.4, 16.2, 9.2 Gy; Breast volume receiving 95% of Rx: 86.0, 93.4, 85.5%; Tumor bed volume receiving full Rx dose: 96.1, 91.7, 95.8%.The typical sums of technical and professional charges for a course of breast radiation are: $72,593, $117,927 and $73,902 for 3D, IMRT and DIBH, respectively.ConclusionsPoster Viewing Abstract 2069; TableCoverage comparison for 3D, IMRT and DIBHV20 lt lungV10 lt lungMean lung doseV25 heartMean heart doseV25 LADMean LAD doseBreast volume receiving 95% of RxTumor bed volume receiving full Rx3D conformal12.3%17.4%3.7 Gy2.4%3.8 Gy55.0%33.4 Gy86.0%96.1%IMRT19.5%35.4%7.8 Gy6.1%8.6 Gy9.84%16.2 Gy93.4%91.7%DIBH12.1%19.1%3.6 Gy0.2%2.8 Gy1.78%9.2 Gy85.5%95.8% Open table in a new tab Purpose/Objective(s)As long-term cardiac effects of breast cancer treatment become better defined, it is important to minimize cardiac exposure from radiation. We undertook this study to compare normal tissue exposure (heart, left anterior descending artery (LAD), and lung) and target volume coverage (whole breast and tumor bed) using free breathing (FB) 3-D conformal planning, IMRT, and deep inspiration breath hold (DIBH) using the assisted breathing control (ABC) device. A cost effectiveness analysis is also included. As long-term cardiac effects of breast cancer treatment become better defined, it is important to minimize cardiac exposure from radiation. We undertook this study to compare normal tissue exposure (heart, left anterior descending artery (LAD), and lung) and target volume coverage (whole breast and tumor bed) using free breathing (FB) 3-D conformal planning, IMRT, and deep inspiration breath hold (DIBH) using the assisted breathing control (ABC) device. A cost effectiveness analysis is also included. Materials/MethodsFollowing IRB approval, we reviewed 10 consecutive patients who received radiation for left sided breast cancer treated with ABC. Patients were selected for the ABC device if any part of the heart would receive greater than 25 Gy with standard 3-D planning. For these 10 patients we created FB 3-D conformal plans as well as IMRT plans to compare target coverage and normal tissue avoidance. All patients were planned with a whole breast dose of 46 Gy with a 16 Gy tumor bed boost. Average values were calculated for each parameter. Following IRB approval, we reviewed 10 consecutive patients who received radiation for left sided breast cancer treated with ABC. Patients were selected for the ABC device if any part of the heart would receive greater than 25 Gy with standard 3-D planning. For these 10 patients we created FB 3-D conformal plans as well as IMRT plans to compare target coverage and normal tissue avoidance. All patients were planned with a whole breast dose of 46 Gy with a 16 Gy tumor bed boost. Average values were calculated for each parameter. ResultsThe following numbers (illustrated in the Table) represent the means for each parameter for 3D, IMRT, DIBH respectively: V20 Lt lung: 12.3, 19.5, 12.1%; V10 Lt lung: 17.4, 35.4, 19.1%; mean lung dose: 3.7, 7.8, 3.6 Gy; V25 heart: 2.4, 6.1, 0.2%; mean heart dose: 3.8, 8.6, 2.8 Gy; V25 LAD: 55, 9.84, 1.78%; Mean LAD dose: 33.4, 16.2, 9.2 Gy; Breast volume receiving 95% of Rx: 86.0, 93.4, 85.5%; Tumor bed volume receiving full Rx dose: 96.1, 91.7, 95.8%.The typical sums of technical and professional charges for a course of breast radiation are: $72,593, $117,927 and $73,902 for 3D, IMRT and DIBH, respectively. The following numbers (illustrated in the Table) represent the means for each parameter for 3D, IMRT, DIBH respectively: V20 Lt lung: 12.3, 19.5, 12.1%; V10 Lt lung: 17.4, 35.4, 19.1%; mean lung dose: 3.7, 7.8, 3.6 Gy; V25 heart: 2.4, 6.1, 0.2%; mean heart dose: 3.8, 8.6, 2.8 Gy; V25 LAD: 55, 9.84, 1.78%; Mean LAD dose: 33.4, 16.2, 9.2 Gy; Breast volume receiving 95% of Rx: 86.0, 93.4, 85.5%; Tumor bed volume receiving full Rx dose: 96.1, 91.7, 95.8%.The typical sums of technical and professional charges for a course of breast radiation are: $72,593, $117,927 and $73,902 for 3D, IMRT and DIBH, respectively. ConclusionsPoster Viewing Abstract 2069; TableCoverage comparison for 3D, IMRT and DIBHV20 lt lungV10 lt lungMean lung doseV25 heartMean heart doseV25 LADMean LAD doseBreast volume receiving 95% of RxTumor bed volume receiving full Rx3D conformal12.3%17.4%3.7 Gy2.4%3.8 Gy55.0%33.4 Gy86.0%96.1%IMRT19.5%35.4%7.8 Gy6.1%8.6 Gy9.84%16.2 Gy93.4%91.7%DIBH12.1%19.1%3.6 Gy0.2%2.8 Gy1.78%9.2 Gy85.5%95.8% Open table in a new tab

Abstract Abstract #3124 Background: Single agent weekly docetaxel (D) is an active agent in the treatment of metastatic breast cancer (MBC) with response rates of 29% - 53%. Erlotinib (OSI-774, Tarceva®) is a tyrosine kinase inhibitor directed against EGFR, which is overexpressed in 30-40% of breast cancers, making EGFR an attractive treatment target. This study was designed to assess the combination of D and E in previously untreated recurrent and/or MBC. Methods: Adult patients with histologically confirmed MBC without prior chemotherapy for recurrence or metastases were eligible. Treatment plan was: D (initially 35 mg/m2 intravenous infusion weekly x 3 every 4 weeks) and E 150 mg orally daily uninterrupted. In patients with responding or stable disease, E was continued in 4 week cycles following a minimum of 6 cycles of D and E. Estimates of overall survival (OS) and progression free survival (PFS) were made by Kaplan-Meier method and the difference between groups by log-rank test. Tumor EGFR expression by immunohistochemisty and ER/PR was correlated with OS and PFS. Results: 39 female pts were enrolled between 12/02 and 8/06. The median age was 51 yrs (range 28-78). The median number of cycles of D and E received was 4 (range 1-26) and of E following D and E was 11 (range 2-18). EGFR, ER/PR and Her-2/neu status was determined on 35/39 patients. EGFR: 23 positive, 12 negative. ER/PR: 25 positive, 10 negative. Her-2/neu: 2 positive, 33 negative. Ten pts. were not evaluable for survival or response due to toxicity occurring within the first cycle. Best responses (n=29) ; PR 11(39%), SD 4 (14%), PD 13 (45%), and clinical benefit (PR+SD) 15 (54%). Median PFS was 8 mos (95% CI: 4.4-12.2). PFS for EGFR negative tumors appeared better than EGFR positive tumors (12 mos PFS 33% vs. 23%) but was not significant (p = 0.53). There was no difference in OS between these groups (p=0.38). PFS and OS for ER/PR positive pts was significantly higher than ER/PR negative pts 6 mos PFS 67% vs. 25% (p= 0.009) and 2 yr OS 53.9% vs. 12.5% (p=0.015). , All patients were included for toxicity assessment (n-39). The first 26 pts received planned D dose of 35mg/m2. Because of non-hematologic toxicity, trial was subsequently modified to start D at 25 mg/m2. Grade 3 or 4 Leukopenia was seen in 15% pts. Principal non-hematologic grade 3-4 toxicities included anorexia, diarrhea, and fatigue (18% pts). Conclusions: Combination therapy for advanced breast cancer with docetaxel and erlotinib shows promising activity with favorable response compared with other studies. There was no significant association with EGFR expression and PFS, however this combination is more favorable for ER positive patients. Randomized trials for ER positive disease is warranted to further investigate the efficacy of this combination compared to single agent docetaxel. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 3124.

Purpose/Objective(s)Sentinel lymph node (SLN) biopsy has become standard treatment for patients presenting with early-stage breast cancer. Patients with positive SLNs routinely undergo axillary lymph node (ALN) sampling though the rate of additional positive nodes varies. Patients deemed N0 at the time of surgery may be found to have isolated tumor cell clusters [N0 (i+)] or micrometastases (N1mi) by IHC. The clinical significance and recurrence rate of low volume SLN disease remains unclear.Materials/MethodsWe retrospectively identified 1,414 patients that underwent SLN biopsy at our institution between 1996 and 2007. Of these, 1,265 were either N0(i-) or had metastases greater than 2 mm and were excluded, leaving a cohort of 149 patients. For the analysis of subsequent ALN positivity, 20/149 were excluded as additional ALNs were not sampled, leaving 129 patients. For the analysis of recurrence, 39/149 were excluded due to: Tis (2), neoadjuvant chemotherapy (4), Stage III disease (2), subsequent positive ALNs on further dissection (5), or inadequate follow-up (26), leaving 110 patients; 60 N0(i+) and 50 N1mi. Our institution considered patients with N0(i+) disease node negative for adjuvant treatment decisions. Data was analyzed for baseline differences between the groups in terms of age, T-Stage, grade, ER/PR status, Her2 status, LVI, and type of adjuvant therapy.ResultsIn 129 patients with further ALN sampling, the rate of further ALN positivity was 5% (6/129); 4% and 5% for N0(i+), and N1mi, respectively. Patients in the N1mi group were more likely to receive chemotherapy (p = 0.002) including a taxane (p = 0.0012). With a median follow-up of 44 months, 8 (7.3%) patients recurred. Six patients (10%) in the N0(i+) group experienced a recurrence (3 local, 3 distant), whereas 2 (4%) in the N1mi group experienced a recurrence (both local). This difference approached statistical significance based on univariate (log–rank) analysis (p = 0.08). All recurrences in the N0(i+) group occurred within 36 months of completion of adjuvant therapy whereas recurrences in the N1mi group occurred after 60 months. Using a Wilcoxon test giving greater weight to early recurrences, there was an increased risk of recurrence in the N0(i+) group compared to the N1mi group (p = 0.0251). Age <56.5 and Grade 3 tumors were also associated with increased risk of recurrence (p = 0.049 and p = 0.044).ConclusionsRates of subsequent +ALNs were low in N0(i+) and N1mi patients, comparing favorably to the known false neg rate of SLN biopsy. Additional axillary surgery or radiation, which increases lymphedema risk, is probably unnecessary. Patients with N0(i+) had higher rates of recurrence compared to patients with N1mi reflecting success of the more aggressive adjuvant therapy given to the N1mi group and should be considered for N0(i+) particularly in younger patients with Grade 3 tumors. Prospective randomized trials are needed to determine who may benefit from more aggressive adjuvant therapy. Purpose/Objective(s)Sentinel lymph node (SLN) biopsy has become standard treatment for patients presenting with early-stage breast cancer. Patients with positive SLNs routinely undergo axillary lymph node (ALN) sampling though the rate of additional positive nodes varies. Patients deemed N0 at the time of surgery may be found to have isolated tumor cell clusters [N0 (i+)] or micrometastases (N1mi) by IHC. The clinical significance and recurrence rate of low volume SLN disease remains unclear. Sentinel lymph node (SLN) biopsy has become standard treatment for patients presenting with early-stage breast cancer. Patients with positive SLNs routinely undergo axillary lymph node (ALN) sampling though the rate of additional positive nodes varies. Patients deemed N0 at the time of surgery may be found to have isolated tumor cell clusters [N0 (i+)] or micrometastases (N1mi) by IHC. The clinical significance and recurrence rate of low volume SLN disease remains unclear. Materials/MethodsWe retrospectively identified 1,414 patients that underwent SLN biopsy at our institution between 1996 and 2007. Of these, 1,265 were either N0(i-) or had metastases greater than 2 mm and were excluded, leaving a cohort of 149 patients. For the analysis of subsequent ALN positivity, 20/149 were excluded as additional ALNs were not sampled, leaving 129 patients. For the analysis of recurrence, 39/149 were excluded due to: Tis (2), neoadjuvant chemotherapy (4), Stage III disease (2), subsequent positive ALNs on further dissection (5), or inadequate follow-up (26), leaving 110 patients; 60 N0(i+) and 50 N1mi. Our institution considered patients with N0(i+) disease node negative for adjuvant treatment decisions. Data was analyzed for baseline differences between the groups in terms of age, T-Stage, grade, ER/PR status, Her2 status, LVI, and type of adjuvant therapy. We retrospectively identified 1,414 patients that underwent SLN biopsy at our institution between 1996 and 2007. Of these, 1,265 were either N0(i-) or had metastases greater than 2 mm and were excluded, leaving a cohort of 149 patients. For the analysis of subsequent ALN positivity, 20/149 were excluded as additional ALNs were not sampled, leaving 129 patients. For the analysis of recurrence, 39/149 were excluded due to: Tis (2), neoadjuvant chemotherapy (4), Stage III disease (2), subsequent positive ALNs on further dissection (5), or inadequate follow-up (26), leaving 110 patients; 60 N0(i+) and 50 N1mi. Our institution considered patients with N0(i+) disease node negative for adjuvant treatment decisions. Data was analyzed for baseline differences between the groups in terms of age, T-Stage, grade, ER/PR status, Her2 status, LVI, and type of adjuvant therapy. ResultsIn 129 patients with further ALN sampling, the rate of further ALN positivity was 5% (6/129); 4% and 5% for N0(i+), and N1mi, respectively. Patients in the N1mi group were more likely to receive chemotherapy (p = 0.002) including a taxane (p = 0.0012). With a median follow-up of 44 months, 8 (7.3%) patients recurred. Six patients (10%) in the N0(i+) group experienced a recurrence (3 local, 3 distant), whereas 2 (4%) in the N1mi group experienced a recurrence (both local). This difference approached statistical significance based on univariate (log–rank) analysis (p = 0.08). All recurrences in the N0(i+) group occurred within 36 months of completion of adjuvant therapy whereas recurrences in the N1mi group occurred after 60 months. Using a Wilcoxon test giving greater weight to early recurrences, there was an increased risk of recurrence in the N0(i+) group compared to the N1mi group (p = 0.0251). Age <56.5 and Grade 3 tumors were also associated with increased risk of recurrence (p = 0.049 and p = 0.044). In 129 patients with further ALN sampling, the rate of further ALN positivity was 5% (6/129); 4% and 5% for N0(i+), and N1mi, respectively. Patients in the N1mi group were more likely to receive chemotherapy (p = 0.002) including a taxane (p = 0.0012). With a median follow-up of 44 months, 8 (7.3%) patients recurred. Six patients (10%) in the N0(i+) group experienced a recurrence (3 local, 3 distant), whereas 2 (4%) in the N1mi group experienced a recurrence (both local). This difference approached statistical significance based on univariate (log–rank) analysis (p = 0.08). All recurrences in the N0(i+) group occurred within 36 months of completion of adjuvant therapy whereas recurrences in the N1mi group occurred after 60 months. Using a Wilcoxon test giving greater weight to early recurrences, there was an increased risk of recurrence in the N0(i+) group compared to the N1mi group (p = 0.0251). Age <56.5 and Grade 3 tumors were also associated with increased risk of recurrence (p = 0.049 and p = 0.044). ConclusionsRates of subsequent +ALNs were low in N0(i+) and N1mi patients, comparing favorably to the known false neg rate of SLN biopsy. Additional axillary surgery or radiation, which increases lymphedema risk, is probably unnecessary. Patients with N0(i+) had higher rates of recurrence compared to patients with N1mi reflecting success of the more aggressive adjuvant therapy given to the N1mi group and should be considered for N0(i+) particularly in younger patients with Grade 3 tumors. Prospective randomized trials are needed to determine who may benefit from more aggressive adjuvant therapy. Rates of subsequent +ALNs were low in N0(i+) and N1mi patients, comparing favorably to the known false neg rate of SLN biopsy. Additional axillary surgery or radiation, which increases lymphedema risk, is probably unnecessary. Patients with N0(i+) had higher rates of recurrence compared to patients with N1mi reflecting success of the more aggressive adjuvant therapy given to the N1mi group and should be considered for N0(i+) particularly in younger patients with Grade 3 tumors. Prospective randomized trials are needed to determine who may benefit from more aggressive adjuvant therapy.

Sequencing of adjuvant chemotherapy prior to radiotherapy decreases the risk of systemic recurrence, but delays in breast radiotherapy have been associated with lower rates of local control. We assessed the feasibility and efficacy of placing radiotherapy earlier in the treatment course. We performed a prospective, single-arm study of women with early-stage node-positive breast cancer who underwent breast-conserving surgery and adjuvant therapy consisting of 4 cycles of doxorubicin (60 mg/m2) and cyclophosphamide (600 mg/m2) followed by 4 cycles of paclitaxel (175 mg/m2) delivered every 3 weeks. Radiotherapy was delivered concurrent with the first two cycles of paclitaxel and consisted of 39.6 Gy/22 fxs to the breast with a tumor bed boost of 14 Gy/7 fxs. Supraclavicular nodes were treated in patients with 4 or more positive axillary nodes or extranodal extension. Axillary fields were used for patients with inadequate dissections or gross nodal disease. The diffusing capacity for carbon monoxide (DLCO) was measured as a proxy for functional lung volume. Pulmonary function tests with DLCO were obtained before and after radiation, after completing chemotherapy, and regularly for 5 years. The CT-based planning information was used to measure the volume of lung that received 20 Gy (V20) and the volume of breast tissue encompassed by the prescribed isodose line. Breast cosmesis scores were assigned prior to radiation and at follow-up using RTOG criteria. Forty-four women entered the trial. Two patients did not complete the treatment protocol; one developed chemotherapy toxicity before initiation of radiotherapy and one withdrew from study after cycle 1 of paclitaxel. Median follow-up is 62 months (range, 18 - 84 months). The 5-year actuarial rate of disease-free survival is 86% and overall survival is 93%. There were no local failures. No patients developed radiation pneumonitis. Median V20 was 152 cm3 (range, 29 - 341 cm3) and median %V20 was 5.7% (range, 1.4 - 13%). The DLCO decreased in 17 patients immediately following radiation, but there was no statistically significant change immediately post-radiotherapy (p = 0.51) or with extended follow-up (p = 0.74). Two patients developed acute Grade 3 skin toxicity (median, mode Grade 2). There was no statistically significant change in overall late cosmesis scores (p = 0.80). There was no correlation between V20 and change in DLCO (p = 0.10), or between overall late cosmesis scores and volume of irradiated breast tissue (p = 0.50). Concurrent paclitaxel chemotherapy and radiotherapy after breast-conserving surgery shortened total treatment time, provided excellent local control, was well tolerated, had no significant pulmonary toxicity, and had acceptable cosmetic outcomes.

To analyze the effects of postoperative radiation therapy on breast reconstruction following modified radical mastectomy for breast cancer. Between 1988 and 2006, 73 women were treated with post-mastectomy radiation therapy (RT) after breast reconstruction. Their RT records were retrospectively reviewed. Information regarding comorbid conditions, systemic therapy, reconstruction method, and RT technique was obtained and correlated with acute RT toxictiy, lymphedema, fat necrosis, telangiectasias, cosmetic outcome, and revision of reconstruction. Cosmesis was scored prospectively pre-RT and at follow-up using RTOG criteria. The median duration of follow-up was 46 months (4 mo–11 yrs). The median age was 47 years (24–79 yrs). 62 patients received chemotherapy (15 neoadjuvant, 10 neoadjuvant + adjuvant, and 37 adjuvant). Reconstruction was immediate in 71 patients and delayed by >1 year for 2. All received RT following reconstruction with opposed tangents to a median dose of 50.4 Gy (6–56 Gy). 13 had a boost (8–16 Gy), 64 had supraclavicular fields (SCV) treated with a median dose of 46 Gy, and 19 had a posterior axillary boost (PAB) to with median mid-plane dose of 48 Gy. 40 patients had 1 cm (23) or 1/2 cm (17) of bolus applied to the chest wall every other day. Median time from reconstruction to RT was 7.6 months (2 mo–14 years). 37 had transverse rectus abdominus flap (TRAM), 25 had tissue expander (TE), 4 had latissimus dorsi flap (LD), and 7 had LD with a TE. 9 had revisions prior to RT (infection 3; skin necrosis 3; capsular contracture 3; poor cosmesis 1), and 13 had revisions following XRT (infection 2; skin necrosis 1; leaking implant 1; poor cosmesis 9). TRAM was less likely than TE to require a revision after XRT (9% vs. 35%; p = 0.016) and overall (16% vs. 48%; p = 0.006). Pre-RT cosmesis was excellent (27%), good (44%), fair (17%), or poor (13%). Post-RT cosmesis was excellent (22%), good (54%), fair (13%), or poor (11%). Pre and post-RT cosmesis did not differ significantly between reconstruction types. After RT, 16 developed fat necrosis, 7 developed telangiectasias, and 8 developed lymphedema. TRAMs were more likely to develop fat necrosis (38%; p = 0.0004). Fat necrosis, lymphedema, and telangiectasias did not reflect post-RT cosmetic score or predict for revision. Patients who received chemotherapy had worse cosmesis before RT, which persisted after RT (p = 0.037). There was a trend formore revisions in women receiving chemotherapy (34% vs. 9%; p = 0.1). No single agent or regimen was correlated with revision rate or cosmetic outcome. Hypertension, diabetes, alcohol use, smoking, bolus, RT dose, RT energy, RT boost, SCV, and PAB were not associated with cosmesis or revisions. Moist or dry desquamation and extent of RT skin erythema did not affect cosmesis or revisions. Time between reconstruction and RT did not predict for cosmetic outcome, RT toxicity, or revisions. Postmastectomy RT after breast reconstruction is well tolerated with 75% having good or excellent cosmetic outcomes. Neither acute RT toxicity nor fat necrosis predicted poorer cosmetic outcome or need for future revision. TE was more likely than TRAM reconstruction to require future surgical revision. Patients with TRAM reconstruction were more likely to experience fat necrosis, but this did not affect cosmesis.

Purpose/Objective(s)Glomus tumors (GT) are rare low grade tumors originating from the paraganglionic tissue in the skull base. Their treatment by stereotactic radiotherapy demands high accuracy in defining the tumor margins. Despite high contrast enhancement and typical pattern in magnetic resonance imaging (MRI) of GTs, the exact delineation can be challenging. As GTs show high expression levels of somatostatin receptor subtypes, 18F-Octreotate PET is supposedly useful for improving target volume delineation in radiotherapy planning of GTs. Up to now 18F-Octreotate PET is used for diagnostic purposes in a semi quantitative manner only. Aim of this study was the definition of a quantitative segmentation threshold based on phantom measurements and its validation for tumor volume determination in patients with GT.Materials/MethodsTo identify the segmentation threshold, a 20 cm diameter cylindric phantom containing 18F-solution filled spheres of different sizes, was used. Background activity was chosen analogous to the mean values measured in the contralateral petrous portion of examined patients. Phantom measurements were performed in three different tumor to background ratios. 9 Patients with 11 tumors (6 untreated, 5 operated tumors) were examined by 18F-octreotate-PET (Biograph 16 PET/CT Tomograph or ECAT HR + PET Tomograph, Siemens Medical Solutions). After image reconstruction (OSEM with 4 iterations, 8 subsets, attenuation and scatter correction) the tumor volume was determined, using a segmentation algorithm implemented in the TrueD software (Siemens Medical Solutions). Patients were also imaged by MRI including a contrast enhanced 3-D MRI dataset (Philipps ACS Gyroscan, 1.5 T). All images were transferred to the radiotherapy treatment planning software (BrainScan 5.21, BrainLab) and tumor volumes were determined by two independent observers (one radiologist, one radiooncologist) using the contrast enhanced T1-weighted images without knowledge of the PET result. Additional MRI sequences were observed if necessary.ResultsPhantom studies revealed that 32% of SUVmax is the appropriate threshold for the determination of the tumor volume at the given, very low background of 18F-Octreotate-PET in the petrous portion of the temporal bone. Using this threshold the mean tumor volumes determined by PET and MRI were 8,6 ± 9 cm3 and 10 ± 10 cm3. We found a significant correlation of the volumes determined by PET compared to the volumes determined by MRI (r = 0.97, p = 0.0001). Correlation was better for larger tumors than for smaller tumors. In two postoperative cases 18F-Octreotate-PET was positive although a tumor could not be distinguished from normal tissue in MRI.ConclusionsThis study shows that a segmentation threshold determined in phantom studies can be applied for gross tumor volume definition in patients with GTs. 32% of the maximal standardized uptake value (SUVmax) is an appropriate threshold in the described setting (scanner, software, reconstruction algorithm) for target volume delineation of GTs by 18F-Octreotate-PET. In addition we showed, that 18F-Octreotate-PET improves radiation therapy planning especially if an accurate discrimination between tumor and normal tissue by MRI is difficult. Purpose/Objective(s)Glomus tumors (GT) are rare low grade tumors originating from the paraganglionic tissue in the skull base. Their treatment by stereotactic radiotherapy demands high accuracy in defining the tumor margins. Despite high contrast enhancement and typical pattern in magnetic resonance imaging (MRI) of GTs, the exact delineation can be challenging. As GTs show high expression levels of somatostatin receptor subtypes, 18F-Octreotate PET is supposedly useful for improving target volume delineation in radiotherapy planning of GTs. Up to now 18F-Octreotate PET is used for diagnostic purposes in a semi quantitative manner only. Aim of this study was the definition of a quantitative segmentation threshold based on phantom measurements and its validation for tumor volume determination in patients with GT. Glomus tumors (GT) are rare low grade tumors originating from the paraganglionic tissue in the skull base. Their treatment by stereotactic radiotherapy demands high accuracy in defining the tumor margins. Despite high contrast enhancement and typical pattern in magnetic resonance imaging (MRI) of GTs, the exact delineation can be challenging. As GTs show high expression levels of somatostatin receptor subtypes, 18F-Octreotate PET is supposedly useful for improving target volume delineation in radiotherapy planning of GTs. Up to now 18F-Octreotate PET is used for diagnostic purposes in a semi quantitative manner only. Aim of this study was the definition of a quantitative segmentation threshold based on phantom measurements and its validation for tumor volume determination in patients with GT. Materials/MethodsTo identify the segmentation threshold, a 20 cm diameter cylindric phantom containing 18F-solution filled spheres of different sizes, was used. Background activity was chosen analogous to the mean values measured in the contralateral petrous portion of examined patients. Phantom measurements were performed in three different tumor to background ratios. 9 Patients with 11 tumors (6 untreated, 5 operated tumors) were examined by 18F-octreotate-PET (Biograph 16 PET/CT Tomograph or ECAT HR + PET Tomograph, Siemens Medical Solutions). After image reconstruction (OSEM with 4 iterations, 8 subsets, attenuation and scatter correction) the tumor volume was determined, using a segmentation algorithm implemented in the TrueD software (Siemens Medical Solutions). Patients were also imaged by MRI including a contrast enhanced 3-D MRI dataset (Philipps ACS Gyroscan, 1.5 T). All images were transferred to the radiotherapy treatment planning software (BrainScan 5.21, BrainLab) and tumor volumes were determined by two independent observers (one radiologist, one radiooncologist) using the contrast enhanced T1-weighted images without knowledge of the PET result. Additional MRI sequences were observed if necessary. To identify the segmentation threshold, a 20 cm diameter cylindric phantom containing 18F-solution filled spheres of different sizes, was used. Background activity was chosen analogous to the mean values measured in the contralateral petrous portion of examined patients. Phantom measurements were performed in three different tumor to background ratios. 9 Patients with 11 tumors (6 untreated, 5 operated tumors) were examined by 18F-octreotate-PET (Biograph 16 PET/CT Tomograph or ECAT HR + PET Tomograph, Siemens Medical Solutions). After image reconstruction (OSEM with 4 iterations, 8 subsets, attenuation and scatter correction) the tumor volume was determined, using a segmentation algorithm implemented in the TrueD software (Siemens Medical Solutions). Patients were also imaged by MRI including a contrast enhanced 3-D MRI dataset (Philipps ACS Gyroscan, 1.5 T). All images were transferred to the radiotherapy treatment planning software (BrainScan 5.21, BrainLab) and tumor volumes were determined by two independent observers (one radiologist, one radiooncologist) using the contrast enhanced T1-weighted images without knowledge of the PET result. Additional MRI sequences were observed if necessary. ResultsPhantom studies revealed that 32% of SUVmax is the appropriate threshold for the determination of the tumor volume at the given, very low background of 18F-Octreotate-PET in the petrous portion of the temporal bone. Using this threshold the mean tumor volumes determined by PET and MRI were 8,6 ± 9 cm3 and 10 ± 10 cm3. We found a significant correlation of the volumes determined by PET compared to the volumes determined by MRI (r = 0.97, p = 0.0001). Correlation was better for larger tumors than for smaller tumors. In two postoperative cases 18F-Octreotate-PET was positive although a tumor could not be distinguished from normal tissue in MRI. Phantom studies revealed that 32% of SUVmax is the appropriate threshold for the determination of the tumor volume at the given, very low background of 18F-Octreotate-PET in the petrous portion of the temporal bone. Using this threshold the mean tumor volumes determined by PET and MRI were 8,6 ± 9 cm3 and 10 ± 10 cm3. We found a significant correlation of the volumes determined by PET compared to the volumes determined by MRI (r = 0.97, p = 0.0001). Correlation was better for larger tumors than for smaller tumors. In two postoperative cases 18F-Octreotate-PET was positive although a tumor could not be distinguished from normal tissue in MRI. ConclusionsThis study shows that a segmentation threshold determined in phantom studies can be applied for gross tumor volume definition in patients with GTs. 32% of the maximal standardized uptake value (SUVmax) is an appropriate threshold in the described setting (scanner, software, reconstruction algorithm) for target volume delineation of GTs by 18F-Octreotate-PET. In addition we showed, that 18F-Octreotate-PET improves radiation therapy planning especially if an accurate discrimination between tumor and normal tissue by MRI is difficult. This study shows that a segmentation threshold determined in phantom studies can be applied for gross tumor volume definition in patients with GTs. 32% of the maximal standardized uptake value (SUVmax) is an appropriate threshold in the described setting (scanner, software, reconstruction algorithm) for target volume delineation of GTs by 18F-Octreotate-PET. In addition we showed, that 18F-Octreotate-PET improves radiation therapy planning especially if an accurate discrimination between tumor and normal tissue by MRI is difficult.