Sam Yoon, MD
Chief, Division of Surgical Oncology
Columbia University Irving Medical Center
New York, New York, United States
Disclosure: Attis Lab (Individual(s) Involved: Self): I own a small percentage of this private company, Ownership Interest
TDOs were generated from endoscopic biopsies prior to the initiation of chemotherapy and treated with the two most common chemotherapy regimens: FOLFOX (leucovorin, 5-fluorouracil, and oxaliplatin) and FLOT (additional docetaxel). Cell viability and proliferation were assayed after 3-6 days. Following neoadjuvant chemotherapy, pts underwent surgical resection, and percent pathological necrosis within the tumor was determined by an experienced pathologist.
Results: Successful TDOs were obtained from 13 of 24 (54%) LAGAs. Failure to generate TDOs was due to contamination (n=3, 13%), early senescence (n=3, 13%) and late senescence (n=5, 21%). Four of 13 pts with successful TDOs did not undergo neoadjuvant chemotherapy. By H&E staining, there were significant similarities in tumor morphology including degree of gland formation between TDOs and tumors. There was also a very high correlation in expression of immunohistochemical markers of GA including CEA, E-cadherin, MUC1, MUC2, CDX2, and CK7. There were significant differences in chemosensitivity of TDOs to FOLFOX and FLOT. For the 9 evaluable pts, the Pearson correlation coefficient between chemosensitivity of the TDO to FOLFOX (n=2) or FLOT (n=7) and the percent tumor necrosis in the surgically resected LAGA was 0.87. For the two LAGAs with 50% or less pathologic necrosis, the corresponding TDOs were clearly resistant outliers on the chemosensitivity curves (Fig.). Four pts’ LAGA had 95-100% pathologic necrosis, and 2 of the corresponding TDOs were clearly sensitive outliers.
Conclusions: TDOs from pts with LAGAs in many respects mimic the original tumors from which they are derived and may be used to reliably predict resistance to neoadjuvant chemotherapy.