Multiple Gastrointestinal Polyps in Patients Treated with BRAF Inhibitors

Abstract

Purpose: BRAF inhibitors (BRAFi) extend survival in BRAF-mutant melanoma but can promote the growth of Ras-mutant neoplasms. This study determined if gastrointestinal polyps found in BRAFi-treated patients harbored Ras mutations.

Experimental design: Colonic and gastric polyps were identified and resected from BRAFi-treated melanoma patients. Next-generation sequencing (NGS) was performed on polyps. The ability of BRAFi to promote polyp formation was functionally characterized in Apc Min(+/-) mice. MAPK and β-catenin pathway activity was assessed by immunohistochemistry in mouse and human polyps.

Results: Fourteen patients treated with BRAFi underwent endoscopy to assess for polyps. Seven out of 7 patients >40 years of age and treated for >2 years were found to have colonic tubular adenomas with 4 out of the 7 patients having 5 or more polyps. One patient presented with bleeding from hyperplastic gastric polyps that recurred 6 months after BRAFi rechallenge. NGS performed on polyps found no mutations in MAPK pathway genes, but found APC mutations in all tubular adenomas. A significant increase in the number of polyps was observed in BRAFi-treated compared with control-treated Apc Min(+/-) mice (20.8 ± 9.2 vs 12.8 ± 0.1; P = 0.016). No polyps were observed in BRAFi-treated wild-type mice.

Conclusions: BRAFi may increase the risk of developing hyperplastic gastric polyps and colonic adenomatous polyps. Due to the risk of gastrointestinal bleeding and the possibility of malignant transformation, further studies are needed to determine whether or not endoscopic surveillance should be recommended for patients treated with BRAFi.

Figure 1. Multiple colonic adenomas and recurrent gastric polyps in a BRAF mutant melanoma patient treated with long term BRAF inhibitor

Representative endoscopic or histological H&E images of (A) one of six gastric hyperplastic polyps; Magnification: 10X (middle); 100X (right) ; (B) one of 5 colonic tubular adenomas; Magnification: 10X (middle); 200X (right) (C) Recurrent gastric polyps 6 months after resection of original polyps and rechallenge with vemurafenib.

Figure 2. 18-Fluorodeoxyglocuse (FDG) avid colonic polyp mistaken for melanoma metastases

(A) Cross sectional image of a FDG PET/CT obtained from a BRAF mutant melanoma patient with a complete response to dabrafenib after 5 years of therapy. The intense uptake of radiotracer in a colonic lesion (arrow) resulted in (B) a colonoscopy which demonstrated a 15 mm colonic tubular adenoma.

Figure 3. Effects of BRAF inhibitor therapy on the number of intestinal polyps in APCMin/+ mice

(A). Representative photographs of proximal and distal small intestine. Arrows indicate scored polyps. (B) Visual scoring of polyps in resected proximal (PSI) and distal (DSI) small intestines in 15 APC/min mice fed control chow compared to 15 APC/min mice fed chow containing the equivalent of 60 mg/kg/day PLX4720 (C) Representative hematoxylin and eosin stained sections of APCMin/+ polyps in mice fed control (CONT) and PLX4720 (PLX) chow; Original magnification 100X. (D) Histological scoring of these GI tracts. *p<0.05.

Figure 4. Immunohistochemistry against phospho-ERK and beta-catenin in mouse and human polyps

(A-B) Representative images and % nuclear staining in Apc Min +/− mouse polyps (A) phospho-ERK; (B) Beta –catenin; ns: not significant. (C-D) Representative images of human tubular adenoma from a BRAF mutant melanoma patient treated with vemurafenib (C) phospho-ERK (D) Beta-catenin . Original magnification 100X (A-B); 200X (C-D).