Sharon Nofech-Mozes’s research while affiliated with University of Toronto and other places

Background: Targeted next-generation sequencing (NGS) panels are increasingly being utilized to identify actionable gene amplifications (copy number > 4) among solid tumors. Methods: This study validated the analytical performance of two amplicon-based NGS assays, the Oncomine Comprehensive Panel (OCAv3) and the Oncomine Focus Assay (OFA), for detecting gene amplification in formalin-fixed paraffin-embedded (FFPE) tumors of varying cellularity. OCAv3 was assessed for amplification detection in 756 FFPE samples comprising various tumor types. Results: We demonstrated that with standardized quality control metrics, including median absolute pairwise difference score, these assays can achieve a near-perfect positive predictive value, although their sensitivity for detecting amplifications significantly decreased in tumors with cellularity below 30%. Stratifying tumor cellularity into 10–30%, 31–60%, and 61–95% groups revealed significantly higher gene amplification detection rates in the 31–60% and 61–95% groups versus the 10–30% group (20.6% and 26.7% vs. 9.2%, p < 0.0001). When considering all detected gene amplifications, the average amplification calling per sample was nearly five-fold lower in the 10–30% group versus the 61–95% group (0.11 vs. 0.52; p < 0.0001). To further investigate the analytic performance of OCAv3 in detecting ERBB2 amplification, we analyzed a cohort of 121 uterine carcinomas with confirmed ERBB2 status by HER2 IHC or FISH, in which a threshold incorporating amplifications and tumor cellularity achieved 79% sensitivity and 100% specificity, potentially eliminating the need for FISH analysis in 34% of equivocal cases. In a separate validation cohort, similar analytical performance was observed, with the threshold demonstrating consistent sensitivity and specificity. Conclusions: This study highlights the strengths and limitations of amplicon-based NGS assays in detecting amplifications using real-world data.

PURPOSE Ductal carcinoma in situ (DCIS) is routinely treated with adjuvant radiotherapy (RT) after breast-conserving surgery (BCS). The inability to accurately estimate an individual's risk of local recurrence (LR) and invasive LR using clinicopathologic factors (CPF) contributes to the overtreatment of DCIS. We examined the impact of the 12-gene DCIS Score (DS) and the 21-gene Recurrence Score (RS) on the accuracy of predicting LR and invasive LR. METHODS A population-based cohort diagnosed with pure DCIS treated with BCS ± RT from 1994 to 2003 was used. All patients had expert pathology review and assessment of the DS and RS. Predictive models (CPF alone, DS + CPF, and RS + CPF) were developed using multivariable Cox regression analyses to predict 10-year LR and invasive LR risks. Models were evaluated on the basis of c-statistic, –2log likelihood estimate (–2LLE), and Akaike information criterion. Calibration was performed using bootstrap resamples, with replacement. RESULTS The cohort includes 1,226 women treated with BCS; 712 received RT. 194 women (15.8%) experienced ipsilateral LR as a first event; 112 were invasive. Models including the DS or RS performed better in predicting the 10-year risk of LR compared with models on the basis of CPF alone with excellent calibration. The two molecular-based models also performed better in predicting invasive LR compared with the CPF model but the model incorporating the RS did not perform better in the prediction of invasive LR compared with the DS-based model. CONCLUSION Models incorporating the DS or RS more accurately predicted the 10-year risk of LR and invasive LR after BCS compared with models on the basis of CPF alone. Inclusion of the RS, compared with DS, did not improve the prediction of the 10-year risk of invasive LR.

HER2-targeted therapies have transformed the management of advanced or recurrent serous endometrial cancer (EC), leading to an increased clinical demand for HER2 testing. Despite its adoption in select academic centers, the global extent of such tumor testing is unclear. In this study, we report on the initial two-year experience of HER2 testing at a major academic center with a reference gynecologic oncology service and biomarker reference laboratory. All patients who underwent HER2 testing based on physician discretion, reflex HER2 testing, and reference laboratory requests were included. From February 2021 to October 2023, HER2 testing was performed on 192 tumor tissue samples from 180 EC patients. Serous carcinoma constituted 52% of samples, reflecting diagnostic challenges and limited therapeutic options for advanced EC. HER2 positivity was found in 28% of all cases and 30% of p53-aberrant cases. An immunohistochemistry (IHC) score of 3+ was found in 15% of samples, while IHC 2+ was found in 45% (13% IHC 2+/ISH+ and 32% IHC 2+/ISH-). The newly identified 'HER2-low' category comprised 46% of the samples. Heterogeneity was noted in 42% of HER2-positive cases, with complex patterns in 3%. NGS and HER2 IHC-FISH showed a 24% discordance, attributed to intratumoral heterogeneity, tumor cellularity, a small number of amplified cells, and the HER2/CEP17 ratio near the cut-off. This study offers real-world insights into HER2 testing in EC, highlighting the challenges and underscoring the need for standardized guidelines in specimen handling, proficiency testing, and scoring criteria to enhance patient management and therapeutic decision-making.

Objective Preoperative detection of axillary lymph node metastases (ALNMs) from breast cancer is suboptimal; however, recent work suggests radiomics may improve detection of ALNMs. This study aims to develop a 3D CT radiomics model to improve detection of ALNMs compared to conventional imaging features in patients with locally advanced breast cancer. Methods Retrospective chart review was performed on patients referred to a specialty breast cancer center between 2015 and 2020 with US-guided biopsy-proven ALNMs and pretreatment chest CT. One hundred and twelve patients (224 lymph nodes) met inclusion and exclusion criteria and were assigned to discovery (n = 150 nodes) and testing (n = 74 nodes) cohorts. US-biopsy images were referenced in identifying ALNMs on CT, with contralateral nodes taken as negative controls. Positive and negative nodes were assessed for conventional features of lymphadenopathy as well as for 107 radiomic features extracted following 3D segmentation. Diagnostic performance of individual and combined radiomic features was evaluated. Results The strongest conventional imaging feature of ALNMs was short axis diameter ≥10 mm with a sensitivity of 64%, specificity of 95%, and area under the curve (AUC) of 0.89 (95% CI, 0.84-0.94). Several radiomic features outperformed conventional features, most notably energy, a measure of voxel density magnitude. This feature demonstrated a sensitivity, specificity, and AUC of 91%, 79%, and 0.94 (95% CI, 0.91-0.98) for the discovery cohort. On the testing cohort, energy scored 92%, 81%, and 0.94 (95% CI, 0.89-0.99) for sensitivity, specificity, and AUC, respectively. Combining radiomic features did not improve AUC compared to energy alone (P = .08). Conclusion 3D radiomic analysis represents a promising approach for noninvasive and accurate detection of ALNMs.

Introduction Ductal carcinoma in situ (DCIS) is routinely treated with adjuvant radiotherapy (RT) after breast conserving surgery (BCS) in order to reduce the risk of local recurrence (LR) and invasive LR. Nomograms based on clinicopathological features (CPF) and molecular expression assays have been developed in an effort to provide individualized risk estimates and personalize decision-making. However, molecular assays are costly and it remains unclear if they provide more accurate recurrence risk estimates compared to algorithms based on CPF alone. We examined the impact of the 12-Gene DCIS Score (DS) and the 21-Gene Recurrence Score (RS) molecular expression assays, in addition to CPF, on the accuracy of predicting 10-year LR and invasive LR risk compared to predicted estimates based on CPF alone. In addition, we examined if a model including the 21-Gene RS improves the 10-year predicted risks of invasive LR after BCS for DCIS compared to estimates based on the DS+CPF or CPF alone. Methods We used a population-based cohort diagnosed with pure DCIS treated with BCS +/- RT from 1994-2003. All cases had expert pathology review providing contemporary assessment of diagnosis, margin status, margin width, multifocality, presence and extent of comedo necrosis, subtype, nuclear grade, and tumor size. For each case, a representative tissue block or unstained slide was sent to measure the 12-Gene DS and 21-Gene RS. Predictive models were developed using multivariable Cox regression analyses with backward selection and included all CPF, treatment with RT, and interactions. The performance of each model was evaluated based on c-statistic, -2log likelihood estimate (-2LLE), and Akaike information criterion (AIC). Calibration was performed using bootstrap resamples, with replacement. We compared the performance of the best model derived from CPF alone, the 12-Gene DS with CPF, and the 21-Gene RS with CPF on their ability to predict the 10 year risks of LR and invasive LR measured against outcomes observed in the cohort. Results The population-based cohort includes 1226 women, 514 were treated with BCS alone and 712 were treated with BCS + RT. Median age was 56 years. Median follow-up was 10 years. Fifty-two percent of tumors were between 1 and 2.5 cm, 35% were ≤1cm, and 13% were >2.5 cm. Comedo necrosis was present in 68%, and nuclear grade was low, moderate, and high in 7%, 54%, and 39%, respectively. Margins were negative in 90.5% of cases (N=1109). The 12-Gene DS was low, intermediate, and high in 53.5%, 20.9%, and 25.6% and the 21-Gene RS was >25 in 30% of patients. 194 women (15.8%) experienced ipsilateral LR as a first event; 112 were invasive LR. Models including either the DS or RS expression assays performed better in predicting the 10-year risk of LR after BCS compared to the model based on CPFs alone, demonstrating higher c-statistics (0.705, 0.699, and 0.662, respectively), lower AIC and lower -2LLE. The two molecular-based predictive models also performed better in predicting the risk of invasive LR compared to CPF model, although with smaller differences in c-statistics (0.684, 0.683, and 0.667, respectively), AIC or -2LLE. The predictive model based on the 21-Gene RS with CPF did not perform better in the prediction of the 10 year risk of invasive LR compared the 12-Gene DS + CPF model. All models were well calibrated. Conclusion The predictive model based on the 12-Gene DS with CPF more accurately predicted the 10-year risk of LR and invasive LR after BCS compared to model based on CPF alone. Inclusion of the 21-Gene RS with CPF did not improve the prediction of the 10-year risk of LR or invasive LR. This suggests that nomograms that include the 12-Gene assay with CPF provide more accurate individualized estimates of recurrence risk after BCS and can help improve personalized decision-making in the management of DCIS. Citation Format: Ezra Hahn, Rinku Sutradhar, Lawrence Paszat, Lena Nguyen, Danielle Rodin, Sharon Nofech-Mozes, Sabina Trebinjac, Cindy Fong, Eileen Rakovitch. Molecular Expression Assays Improve the Prediction of Local and Invasive Local Recurrence after Breast Conserving Surgery for DCIS [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO4-13-02.

Introduction Women with advanced or recurrent endometrial carcinoma (EC) exhibiting HER2 overexpression or gene amplification are candidates for HER2-targeted therapy. Determining HER2 status involves immunohistochemistry (IHC) and, in equivocal cases, fluorescence in situ hybridization (FISH), which is costly, requires proficiency training and is not widely available. Many well-resourced centers have adopted next generation sequencing (NGS) platforms to detect pathogenic POLE mutations using gene panels that include the ERBB2 gene. Herein, we examine the correlation between HER2 status determined by IHC-FISH and NGS. Methods Retrospective cases tested by IHC/FISH and NGS in a large academic center over 18 months. IHC (Ventana 4B5) and FISH (PathVysionHER2 DNA Probe kit) were evaluated on whole slide and scored according to ISGyP guideline recommendations. NGS used Oncomine Comprehensive Assay v3 on the S5 Prime NGS platform (Thermo Fisher) to detect ERBB2 deletions, single nucleotide variants (SNV) and amplification (figure 1). Results HER2 status was determined by IHC/FISH and NGS in 45 cases: 19 serous, 18 high-grade (unspecified histologic type, HGU), 5 endometrioid 1 carcinosarcoma, 1 undifferentiated and 1 gastric-type. IHC-FISH HER2 was positive in 11/45 (24.4%) of cases, all serous or HGU. NGS identified 3 cases as amplified (concordant) and 1 non-pathogenic SNV. Although NGS had positive predictive value of 100% and negative predictive value of 81.0%, sensitivity was low (27.3%, 3/11). Table 1 describes the discordant cases. Conclusion/Implications IHC/FISH are more sensitive than NGS for identification of HER2 positive cases. Intratumoral heterogeneity, tumour cellularity and HER2/CH17 ratio play significant role in reduced detection of HER2 amplification by NGS. • Download figure • Open in new tab • Download powerpoint Abstract EP204/#941 Figure 1 HER2 testing in endometrial carcinoma View this table: • View inline • View popup • Download powerpoint Abstract EP204/#941 Table 1 IHC/FISH and NGC details in discordant cases

Introduction HER2-targeted therapy improves outcomes of HER2-positive endometrial cancer (EC). A novel class of HER2-directed antibody drug conjugate (ADC) that has demonstrated significant benefit in HER2-Low breast carcinoma, is being evaluated in clinical trials for other solid tumors, including uterine carcinoma with varying levels of HER2 expression. ADC efficacy in HER2-Low cancers may be partially attributed to bystander effect in tumors with heterogenous HER2 expression. We characterize the spectrum of HER2 expression in EC cases tested in a large tertiary care centre. Methods An audit of HER2 testing in EC was conducted, utilizing ISGyP criteria for immunohistochemistry (IHC) scoring. Akin to the DESTINY clinical trials, HER2-P (positive) was defined as IHC score 3+ or FISH-amplified; HER2-L (low) defined as IHC2+ (not-amplified) or IHC1+ and HER2-N (negative) defined as IHC score 0. Results There were 95 primary, 4 local recurrences and 10 metastatic EC HER2 tests, with 33 (30%) HER2-N, 46 (42%) HER2-L and 30 (28%) HER2-P. 60 were serous carcinoma, 29 high-grade EC, 9 carcinosarcomas, 8 Endometrioid and 3 others. Among 79 p53 abnormal cases, 22 (28%), 34 (43%), 23 (29%) were HER2-N, HER2-L and HER2-P, respectively. Table 1 summarizes p53, MMR and ER status by IHC and presence of pathogenic POLE mutations. None of the POLE-mutated or MMR-deficient molecular subgroup cases were HER2-P. Conclusion/Implications The majority of ECs tested by IHC/FISH demonstrated some level of HER2 expression, with 42% meeting the criteria for HER2-L designation, potentially doubling the patients that may be considered for novel ADC therapies compared to the legacy HER2 targeted therapies.View this table: • View inline • View popup • Download powerpoint Abstract EP206/#943 Table 1 Stratification of p53, MMR and ER status by immunohistochemistry, pathogenic POLE mutation analysis via Next Generation Sequencing, by HER2 Status

Simple Summary Often in endometrial cancer (EC), the mismatch repair (MMR) system, which helps fix DNA mistakes, goes haywire because of MLH1 promoter hypermethylation (MLH1-PHM). We set out to find out how many ECs have MLH1-PHM, understand the impact of reflex MLH1-PHM testing and evaluate the associated costs within the publicly funded Canadian healthcare system. We looked at 2504 EC samples and found that in 534 of them (21.4%), the MMR system was deficient due to dual MLH1/PMS2-deficiency. Out of the 418 cases with available data, 404 (96.7%) had MLH1-PHM, while only 14 (3.3%) didn’t. Reflex MLH1-PHM tests cost CAD 231.90 per case, amounting to CAD 123,834.60 for 534 cases, with 30 tests needed per additional candidate for MLH1 germline analysis (CAD 6,957.00 per candidate). This raises a provocative question: can we assume that the majority of the MLH1-deficient ECs are due to PHM and forgo further testing in healthcare systems with finite resources? Abstract MLH1/PMS2 loss due to MLH1 promoter hypermethylation (MLH1-PHM) is the most common cause of mismatch repair (MMR) deficiency in endometrial cancer (EC). This study aimed to determine the proportion of MLH1-deficient EC with PHM, assess the impact of the reflex MLH1-PHM testing strategy, and evaluate the associated costs within the publicly funded Canadian healthcare system. In a cohort of 2504 EC samples, 534 (21.4%) exhibited dual MLH1/PMS2 loss, prompting MLH1-PHM testing. Among 418 cases with available testing results, 404 (96.7%) were MLH1-hypermethylated, while 14 (3.3%) were non-methylated. The incidence of MLH1 non-methylated cases in our cohort was 14/2504 (0.56%) of all ECs, underscoring the prevalence of hypermethylation-driven MLH1/PMS2 loss in ECs universally screened for MMR deficiency. Reflex MLH1-PHM testing incurs substantial costs and resource utilization. Assay cost is CAD 231.90 per case, amounting to CAD 123,834.60 for 534 cases, with 30 tests needed per additional candidate for MLH1 germline analysis (CAD 6957.00 per candidate). This raises a provocative question: can we assume that the majority of the MLH1-deficient ECs are due to PHM and forgo further testing in healthcare systems with finite resources? It is imperative to assess resource utilization efficiency and explore optimized approaches that encompass clinical correlation, family history and judicious utilization of methylation testing to ensure it is provided only to those who stand to benefit from it.