Chylothorax and chylous ascites during RET tyrosine kinase inhibitor therapy.

Abstract

9080 Background: Spontaneous, atraumatic chylous effusions are rare. Investigators have observed a higher than anticipated incidence of chylothorax and chylous ascites in patients (pts) treated with RET tyrosine kinase inhibitors (TKIs). A systematic analysis of the occurrence of chylous effusions during RET TKI therapy and management strategies was thus performed. Methods: In this multicenter, retrospective study, the frequency of biochemically confirmed chylothorax or chylous ascites in pts treated with multikinase inhibitors (MKIs) with anti-RET activity or selective RET TKIs was determined. Clinicopathologic features and management of pts with chylous effusions were assessed. Results: A pan-cancer cohort of 7517 pts treated with at least 1 of 17 MKIs and selective RET TKIs and an independent cohort of 96 pts treated with the selective RET TKIs, selpercatinib or pralsetinib, were identified. Across cohorts, chylous effusions were identified in 22 pts and were most common with selpercatinib (7%; 15/217), followed by the MKIs agerafenib (4%; 1/24), cabozantinib (0.3%; 3/918), and lenvatinib (0.3%; 3/1185). Chylous effusions were not noted in 28 pts treated with pralsetinib. The distribution of malignancies included lung adenocarcinoma (54%) medullary thyroid carcinoma (23%), renal cell carcinoma (19%), and desmoplastic small round cell tumor (4%). Of the 22 pts, 12 had chylothorax, 5 had chylous ascites, and 5 had both. The cumulative incidence of chylous effusions from TKI initiation at 12 months was 3.09%. Median fluid triglyceride level was lower in chylothorax than in chylous ascites [397 mg/dL (IQR 282-4000) vs. 3786 mg/dL (IQR 676-6596), p = 0.035]. Median pleural fluid triglyceride level was higher with selpercatinib compared to MKIs [4,000 mg/dL (IQR 356-4425) vs. 287 mg/dL (IQR 216-395); p = 0.017]. Malignant cells were identified in the effusions from 12% (2/17) and 10% (1/10) of pts with chylothoraces and chylous ascites, respectively. Median time to disease progression from radiographic index and biochemical index across the full cohort was 1.5 years (IQR: 0.6-2.4) and 1.0 year (IQR: 0.1-1.2), respectively. Anatomic chyle leak was not identified in 6 pts who underwent lymphangiography. After initial drainage, additional drainage procedures were required in all cases with chylothorax and 50% of cases with chylous ascites. Chylous effusions prompted TKI dose reduction in 47% (7/15) of pts treated with selpercatinib and 14% (1/7) treated with MKI; none discontinued TKI due to chylous effusions. Conclusions: Chylous effusions can emerge on treatment with certain MKIs or selective RET TKIs. Recognition of this potential side effect is key to prevent misattribution of worsening effusions to progressive malignancy and to motivate a better understanding of its biology and management.