Prevention of metachronous colorectal cancer in patients with Lynch syndrome

Identifying patients with Lynch syndrome using a universal tumor screening program in an integrated healthcare system

Introduction

Lynch syndrome (LS) is associated with an increased risk of colorectal (CRC) and endometrial (EC) cancers. Universal tumor screening (UTS) of all individuals diagnosed with CRC and EC is recommended to increase identification of LS. Kaiser Permanente Northwest (KPNW) implemented a UTS program for LS among individuals newly diagnosed with CRC in January 2016 and EC in November 2016. UTS at KPNW begins with immunohistochemistry (IHC) of tumor tissue to determine loss of mismatch repair proteins associated with LS (MLH1, MSH2, MSH6, and PMS2)., IHC showing loss of MLH1 is followed by reflex testing (automatic testing) to detect the presence of the BRAF V600E variant (in cases of CRC) and MLH1 promoter hypermethylation to rule out likely sporadic cases.

Materials and methods

Individuals newly diagnosed with CRC and EC were identified between the initiation of the respective UTS programs and July 2018. Electronic medical records were reviewed to extract patient data related to UTS, including IHC and reflex testing results, date of referrals to the genetics department, and results of germline genetic testing for LS.

Results

313 out of 362 individuals diagnosed with CRC and 61 out of 64 individuals diagnosed with EC who were eligible were screened by IHC for LS. Most (47/52 or 90%, including 46/49 CRC and 1/3 EC) individuals that were not screened by IHC only had a biopsy sample available. Fourteen individuals (3.7% overall, including 13/313 CRC and 1/61 EC) received an abnormal result after reflex testing and were referred for genetic counseling. Of these, 10 individuals (71% overall, including 9/13 CRC and 1/1 EC) underwent germline genetic testing for LS. Five individuals diagnosed with CRC were found to have pathogenic variants. in PMS2 (n = 3), MLH1 (n = 1), and MSH6 (n = 1). No pathogenic variants were identified in individuals diagnosed with EC.

Conclusions

UTS identified individuals at risk for LS. Most individuals who screened positive for LS had follow-up germline genetic testing for LS. The consistent use of biopsy samples is an opportunity to improve UTS.

Preoperative diagnosis of Lynch syndrome with DNA mismatch repair immunohistochemistry on a diagnostic biopsy

BACKGROUND

DNA mismatch repair immunohistochemistry on tumor tissue is a simple, readily available, and cost-effective method of identifying patients with Lynch syndrome in the postoperative setting. The aim of the study was to assess whether the mismatch repair status of a colorectal cancer can be confirmed by mismatch repair immunohistochemistry on preoperative biopsy.

DESIGN

Germline positive patients with Lynch syndrome were identified from a prospectively collected Familial Cancer Clinic database. Preoperative colorectal cancer biopsy specimens were obtained from the source pathology provider to generate a cohort of matched preoperative and postoperative specimens. The specimens were sectioned and stained for 4 mismatch repair proteins (MLH1, MSH2, MSH6, PMS2). An age-matched cohort to compare specimens was selected from Bethesda positive but mismatch repair immunohistochemistry negative patients. All slides were reviewed by a single blinded pathologist. The Wilson method was used to calculate a true underlying proportion of patients for whom the preoperative result matched the postoperative test result with a 95% confidence interval.

RESULTS

Of 128 germline positive mutation carriers, 40 patients (mean age 41, SD 11.3) had colorectal resections. Thirty-three preoperative specimens were retrievable and were matched with biopsies from 33 controls. The germline mutations included in the study were 8 MLH1, 19 MSH2, 3 MSH6, and 2 PMS2. In patients where germline positive status was known, sensitivity was 100% (95% CI 89.2-100) and specificity was 100% (95% CI 89.2-100). Identical sensitivity and specificity were observed in 33 age-matched patients. The sensitivity of the endoscopic biopsy in predicting germline status was 94.9% (95% CI 80.4-98.3).

CONCLUSION

The mismatch repair disease status of a colorectal cancer can be reliably confirmed by mismatch repair immunohistochemistry on a diagnostic colorectal cancer biopsy sample before definitive surgery. Ascertaining a diagnosis of Lynch syndrome before definitive surgery can influence surgical planning.