Abstract 808: Analysis of 1151 prospective research biopsies: Factors influencing tumor yield across patients and biopsy cores

Abstract

Abstract Background: While important for the initial diagnosis and treatment of malignancies, biopsies have an increasing role in clinical trials for correlative science to improve the molecular understanding of cancer. However, research biopsy specimens can be insufficient for analysis and their scientific impact on drug development and treatment has been questioned. The purpose of our study was to identify which inter- and intrapatient factors predict research biopsy tissue quantity and adequacy. Methods: Biopsies were obtained from patients enrolled in clinical trials across multiple cancer histologies as part of the Adaptive Patient-Oriented Longitudinal Learning and Optimization (APOLLO) program at MD Anderson Cancer Center. For each trial, enrolled patients typically underwent a pre-treatment biopsy, with 60% undergoing on/post treatment biopsies. Biopsy cores were either paraffin embedded or flash frozen and sent for pathology review to determine tumor area (core length times core width multiplied by proportion of core area occupied by tumor cells) as well as percentage of tumor contained in the core (fraction of core containing cellular components of tumor, including malignant and peri-tumor non-malignant cells, reported as a percentage). Results: A total of 1151 research biopsies were performed on 661 patients from which 5126 cores were evaluated (median 5 cores per biopsy). 5 or more cores were sent for pathology review in 753 biopsies (65.4%) with 4 sent in 232 cases (20.2%). Tumor area decreased across increasing core position (p<0.001); each of the first two biopsy cores contained significantly higher tumor area than the fourth and fifth cores (p<0.001 for each comparison). However, tumor percentage did not vary by core position (p=0.180). 687 biopsies (60.7%) contained tumor cells in each core, with 628 of these (91.4%) containing at least 10% tumor in each core. 141 biopsies (12.3%) contained all inadequate cores (cores with <10% tumor) while 382 biopsies (33.2%) contained a mix of adequate and inadequate cores. 1348 cores in total (26.3%) contained less than 10% tumor, and this did not vary by core position (p=0.224). In assessing factors related to mean tumor area per core per biopsy, lesion size, BMI, tumor histology, biopsy location, and needle gauge were significantly associated on univariate analysis (p < 0.025 for each variable). In particular, appendiceal, colorectal, and thyroid malignancies were associated with lower than average yield, while neuroendocrine tumors were associated with the highest yield. Each of these factors remained significant on multivariate analysis. Conclusions: Core position, tumor histology, and biopsy location, in addition to expected factors such as needle gauge and lesion size, appear to influence tumor tissue yield. Moreover, there is notable within-biopsy variability with respect to core adequacy which argues for real-time quality assessment of individual cores during biopsies. A substantial number of cores contain tumor amounts inadequate for molecular analysis. Despite a high rate of low tumor cellularity in at least one core of many patients, when multiple cores were taken by an experienced team, relatively few patients had inadequate samples across all cores. Citation Format: Deepak Bhamidipati, Anuj Verma, Dipen M. Maru, Dawen Sui, Grace Mathew, Wenhua Lang, Juan F. Posadas, Joshua P. Hein, Scott Kopetz, Andrew Futreal, Ignacio I. Wistuba, J. Jack Lee, Michael J. Overman, Alda L. Tam. Analysis of 1151 prospective research biopsies: Factors influencing tumor yield across patients and biopsy cores [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 808.