Very Low Penetrance of Loss-of-function Variants in PMS2 for All Lynch Syndrome Relevant Cancers In Large US Biobanks
Cancer Genetics and Therapeutics
-
Primary Categories:
- Clinical Genetics
-
Secondary Categories:
- Clinical Genetics
Introduction:
Lynch syndrome is a common inherited genetic disorder most often associated with colorectal cancer. It is caused by pathogenic variants in the mismatch repair genes MLH1, MSH2, MSH6 and PMS2. CDC Tier 1 genomic screening, which includes Lynch syndrome, is performed on hundreds of thousands of healthy individuals annually. When returning Lynch syndrome results to healthy individuals, a key challenge is that these genes have differences in penetrance and confer risk to various types of cancers. For example, pathogenic variants in MLH1 and MSH2 are known to have higher penetrance than those in MSH6 and PMS2, and women heterozygotes for MSH6 are particularly at risk of endometrial cancer. Further compounding these challenges, current risk estimates come from studies biased towards previously diagnosed patients. Our aim was to measure the actual clinical impact of having a pathogenic variant in PMS2 or one of the other Lynch syndrome genes in a population setting.
Methods:
We studied 144,852 participants from 7 US health systems in the Helix Research Network. Genetic and longitudinal phenotype information from electronic health records were available for each participant. We used the ACMG/AMP variant interpretation rubric to identify individuals who are heterozygotes for pathogenics variants in MLH1, MSH2, MSH6 or PMS2. Only exons 1 to 10 were considered for PMS2 to avoid any potential false positive calls due to the pseudogene closely resembling exons 11-15 of PMS2. Lynch-associated cancer diagnoses considered for this analysis were colorectal, endometrial, ovarian, stomach, small bowel, kidney and bladder cancer.
Results:
Of the participants screened, 467 individuals (0.32%) were heterozygous for a pathogenic variant in one of the Lynch Syndrome genes: 41 in MLH1, 30 in MSH2, 216 in MSH6 and 180 in PMS2. When assessing the presence of any Lynch syndrome-relevant cancer, those heterozygous for a pathogenic variant in MLH1 or MSH2 had the highest risk and earliest age of onset. Compared to individuals with benign or no variants in the four genes (n=127,246), the Hazard Ratio (HR) for MLH1 = 27.5 (95% CI:14.3-53.0) and HR for MSH2 = 21.9 (9.8-48.8). Penetrance was intermediate for MSH6: HRMSH6 = 7.5 (5.0-11.2). On the other hand, penetrance was very low for those heterozygous for a pathogenic variant in PMS2: HRPMS2 = 2.0 (0.9-4.5). The probability of being diagnosed with any Lynch syndrome-relevant cancer by the age of 60 was 1.2% in controls, 44.8% in MLH1 heterozygotes, 24.8% in MSH2 heterozygotes, 10.0% in MSH6 heterozygotes, and only 1.5% in PMS2 heterozygotes. Very low penetrance for PMS2 was also observed when looking at individual cancer types or when restricting PMS2 pathogenic variants more specifically to those predicted to be loss-of-function.
Conclusion:
Our results show that, at the population level, pathogenic variants in PMS2 have very low penetrance for cancers associated with Lynch syndrome. They, therefore, question whether PMS2 should be included in the list of genes for CDC Tier 1 genomic screening for Lynch Syndrome. Lastly, this study highlights the importance of measuring the risk in the general population when returning genetic results for preventative screening.
Lynch syndrome is a common inherited genetic disorder most often associated with colorectal cancer. It is caused by pathogenic variants in the mismatch repair genes MLH1, MSH2, MSH6 and PMS2. CDC Tier 1 genomic screening, which includes Lynch syndrome, is performed on hundreds of thousands of healthy individuals annually. When returning Lynch syndrome results to healthy individuals, a key challenge is that these genes have differences in penetrance and confer risk to various types of cancers. For example, pathogenic variants in MLH1 and MSH2 are known to have higher penetrance than those in MSH6 and PMS2, and women heterozygotes for MSH6 are particularly at risk of endometrial cancer. Further compounding these challenges, current risk estimates come from studies biased towards previously diagnosed patients. Our aim was to measure the actual clinical impact of having a pathogenic variant in PMS2 or one of the other Lynch syndrome genes in a population setting.
Methods:
We studied 144,852 participants from 7 US health systems in the Helix Research Network. Genetic and longitudinal phenotype information from electronic health records were available for each participant. We used the ACMG/AMP variant interpretation rubric to identify individuals who are heterozygotes for pathogenics variants in MLH1, MSH2, MSH6 or PMS2. Only exons 1 to 10 were considered for PMS2 to avoid any potential false positive calls due to the pseudogene closely resembling exons 11-15 of PMS2. Lynch-associated cancer diagnoses considered for this analysis were colorectal, endometrial, ovarian, stomach, small bowel, kidney and bladder cancer.
Results:
Of the participants screened, 467 individuals (0.32%) were heterozygous for a pathogenic variant in one of the Lynch Syndrome genes: 41 in MLH1, 30 in MSH2, 216 in MSH6 and 180 in PMS2. When assessing the presence of any Lynch syndrome-relevant cancer, those heterozygous for a pathogenic variant in MLH1 or MSH2 had the highest risk and earliest age of onset. Compared to individuals with benign or no variants in the four genes (n=127,246), the Hazard Ratio (HR) for MLH1 = 27.5 (95% CI:14.3-53.0) and HR for MSH2 = 21.9 (9.8-48.8). Penetrance was intermediate for MSH6: HRMSH6 = 7.5 (5.0-11.2). On the other hand, penetrance was very low for those heterozygous for a pathogenic variant in PMS2: HRPMS2 = 2.0 (0.9-4.5). The probability of being diagnosed with any Lynch syndrome-relevant cancer by the age of 60 was 1.2% in controls, 44.8% in MLH1 heterozygotes, 24.8% in MSH2 heterozygotes, 10.0% in MSH6 heterozygotes, and only 1.5% in PMS2 heterozygotes. Very low penetrance for PMS2 was also observed when looking at individual cancer types or when restricting PMS2 pathogenic variants more specifically to those predicted to be loss-of-function.
Conclusion:
Our results show that, at the population level, pathogenic variants in PMS2 have very low penetrance for cancers associated with Lynch syndrome. They, therefore, question whether PMS2 should be included in the list of genes for CDC Tier 1 genomic screening for Lynch Syndrome. Lastly, this study highlights the importance of measuring the risk in the general population when returning genetic results for preventative screening.