Background
Hereditary Cancer Syndromes, including Hereditary Breast and Ovarian Cancer (HBOC) and Lynch Syndrome (LS), can result in various forms of cancer due to germline mutations in cancer predisposition genes. While the major contributory genes for these syndromes have been identified and well-studied (
BRCA1/
BRCA2 for HBOC and
MSH2/
MSH6/
MLH1/
PMS2/EPCAM for LS), there remains a large percentage of associated cancer cases that are negative for germline mutations in these genes, including 80% of women with a personal or family history of breast cancer who are negative for
BRCA1/2 mutations [
1]. Similarly, between 30 and 50% of families fulfill stringent criteria for LS and test negative for germline mismatch repair gene mutations [
2]. Adding complexity to these disorders is the significant overlap in the spectrum of cancers observed between various hereditary cancer syndromes, including many cancer susceptibility syndromes. Some that contribute to elevated breast cancer risk include Li-Fraumeni Syndrome, Cowden syndrome, Hereditary Diffuse Gastric Cancer, Ataxia-telangiectasia, Lynch Syndrome, and Peutz-Jeghers syndrome [
3], while others that contribute to elevated colorectal cancer risk include Familial Adenomatous Polyposis and
MUTYH-associated Polyposis [
4]. Additionally, the risk of cancers observed within these syndromes can vary widely. HBOC not only increases the life-time risk of breast and ovarian cancers, but also for prostate and pancreatic cancers [
3]. Similarly, LS-spectrum cancers include colorectal cancer, as well as uterine (endometrial), ovarian, gastric, and other rare forms of cancer [
5].
Over the past decade, linkage, GWAS, and re-sequencing studies have identified additional genes that contribute to the risk of hereditary cancer [
6]. The discoveries of these genes, which vary in their level of penetrance and risk-contribution, present opportunities for personalized clinical management in patients and families with hereditary cancer.
Over time, multi-gene panels for hereditary cancer continue to grow in complexity, from sequencing only the major contributory genes for HBOC and LS, a total of roughly six genes, now expanding to 27 genes and higher [
2,
7‐
14]. The addition of less characterized genes to hereditary cancer panels poses challenges in interpretation due to lack of available information such as disease prevalence, penetrance, locus/allelic heterogeneity, expressivity, age of onset, and the spectrum of gene-associated cancers. Three recent studies reporting the clinical outcomes of panel-based testing for inherited cancer have utilized cohorts selected for patients fulfilling the National Comprehensive Cancer Network (NCCN) guidelines for breast cancer [
7] or colorectal cancer [
2], or Society of Gynecologic Oncology 5–10% criteria for endometrial cancer [
9]. One of the studies used retrospective history-enriched cases from clinical bio banks with referrals based on known familial pathogenic variants [
7]. In this study, we report our experience with a 27-gene inherited cancer panel on a cohort of 630 consecutive patients referred for testing at our laboratory. Our objectives were: 1. Determine the rates for positive cases and those with variants of uncertain clinical significance (VUS) relative to data published in the recent literature, 2. Examine heterogeneity among the constituent genes on the panel, and 3. Review test uptake in the cohort relative to other reports describing outcomes for expanded panel testing.
Discussion
Table
3 summarizes our panel positive rate in relation to other recently published reports and illustrates the following trends. 1. The overall positive rate across similarly sized gene panels (27+/−2 genes), including the current study, is relatively uniform between 8 and 13% (median = 10.3%). 2. As expected, panels with fewer genes that include
BRCA1/2 tend to have a higher positive rate and a lower VUS rate presumably due to a narrower range of indications guiding test uptake. 3. Cohorts enriched with patients fulfilling either the NCCN guidelines for HBOC [
7], LS [
2] or society of gynecological oncology criteria for endometrial cancer [
9] display lower positive rates for “non-BRCA” and “non-LS genes” relative to the
BRCA1/2 and LS genes.
Table 3
Comparison between the most recent reports on the performance of inherited cancer panels
Current Study | 630 | 27 genes | 10.3% | 32.7% | – | 7.1% | 8.1% | Prospective commercial laboratory cohort unselected for NCCN, or previous BRCA/LS testing tested on a panel that includes BRCA1/2 and LS genes |
Lincoln, et al. 20157 | 735 | 29 genes | 12.4%b | 41.0%c | 23 | 3.5% | 11.3% | NCCN HBOC enriched clinical referral cohort tested on a panel that includes BRCA1/2 and LS genes |
Yurgelun, et al. 20152 | 1260 | 25 genes | 12.4% | 38.0% | 25 | 11.2% | 3.3% | NCCN LS enriched cohort tested on a panel that includes BRCA1/2 and LS genes |
Yurgelun et al. 20178 | 1058 | 25 genes | 8.2% | 31.2% | 25 | 7.2% | 5.1% | Unselected CRC care clinic based cohort tested on a panel that includes BRCA1/2 and LS genes |
Ring, et al. 20169 | 381 | 25 genes | 9.2% | 26.0% | 25 | 8.7% | 3.4% | Society of gynecologic oncology criteria for endometrial cancer enriched cohort tested on a panel that includes BRCA1/2 and LS genes |
Minion et al., 201510 | 911 | 21 genes | 7.4% | 27.1% | 19 | 7.4% | 5.9% | BRCA1/2 test negative cohort with a personal history of HBOC tested on a panel that includes BRCA1/2 and LS genes |
Tung, et al. 201511,d | 1781 | 25 genes | 13.5% | 41.7% | 25 | 4.3% | 13.1% | Commercial laboratory BRCA1/2 referral cohort tested on a panel that includes BRCA1/2 and LS genes |
LaDuca, et al. 201412 | 425 | 22 genes | 9.6% | 23.5% | 20 | 9.6% | 7.8% | Commercial laboratory cohort tested on a cancer panel that did not include BRCA1/2 but included LS genes |
LaDuca, et al. 201412 | 874 | 14 genes | 7.4% | 19.8% | 12 | 7.4% | 7.4% | Commercial laboratory high risk of HBOC selected cohort tested on a panel that did not include BRCA1/2 but included LS genes |
Mannan, et al. 201613 | 141 | 13 genes | 36.2% | 14.8% | 13 | 9.9% | 36.2% | Unrelated patients and families with HBOC tested on a panel that did not include LS genes but included BRCA1/2 |
Susswein, et al. 201614,e | 2056 | 29 genes | 10.2% | 34.7% | 25 | Not inferred | Not inferred | Commercial laboratory referral cohort tested on a panel that includes BRCA1/2 and LS genes |
Recent studies that were enriched for patients fulfilling the NCCN criteria for HBOC [
7] and LS [
2] reported identical overall panel positive rates of 12.4% (Table
3) with
BRCA1/2 and LS genes contributing the major source of positive cases (
BRCA1/2, 66/735, clinical referral cohort and LS, 114/1260 at 9.0% each). As would be expected, the corresponding non-
BRCA (26/735) and non-LS (42/1260) positive rates in these studies were lower at 3.5 and 3.3%, respectively. In further support of this observation, a similar trend of a lower non-LS positive rate was observed in another study reporting a selected cohort enriched of patients with endometrial cancer [
9], which is presumably similar to the NCCN LS-enriched cohort in its characteristics (Table
3). A recent follow-up study has demonstrated an increase in non-LS positive rate to 5.1% when testing was performed on a “NCCN unselected” colorectal cancer clinic based cohort [
8]. As expected, our study, which represents a clinical laboratory referral cohort not enriched for patients preselected to fulfill the NCCN guidelines, yields a proportionately higher positive rate for non-BRCA 7.1% (45/630) and non-LS 8.1% (51/630) genes while yielding a lower positive rate of 3.2% for
BRCA1/2 (20/630) and 2.2% for LS (14/630) genes, respectively. A similar positive rate of 7.4% (67/911) for non-BRCA genes has been recently reported in a cohort of 911 subjects selected on the basis of a personal history of breast and/or ovarian, fallopian, or primary peritoneal cancer, who were previously tested to be negative for
BRCA1/2 genes [
10]. This cohort was also not selected for enrichment based on NCCN guidelines. Several factors influence the test performance reported in each of these studies, such as the type of cohort, if enriched for family history or not, the number and content of genes on a panel, and differences in approaches to variant classification. Out of the 81 DV identified in this study, 42% were detected in either of
BRCA1/2 or LS genes, while 58% were detected in other genes on the panel, demonstrating a higher positive rate for non-BRCA and non-LS genes in our study relative to cohorts enriched for NCCN criteria.
We did not observe a significant enrichment for carriers of
MUTYH in our referral cohort relative to the estimated frequency of 1–2% reported in the general population [
4]. Furthermore, the estimated carrier frequency of
MUTYH in our clinical referral cohort (1/42, 2.4%) is consistent with several recently published reports enriched for patients based on NCCN guidelines for breast and colorectal cancer. The risk for developing colorectal cancer among carriers of
MUTYH is unclear with some studies reporting an increase in colorectal cancer risk among carriers with an affected first-degree relative, [
23‐
25] while others have found no significant increase in the risk of colorectal cancer [
26]. The lack of significant differences in
MUTYH carrier frequency across multiple independent studies, including this report, is consistent with current NCCN guidelines that suggest tailored surveillance based on individual and family risk while not proposing specific recommendations for medical management among
MUTYH carriers [
27].
Out of the 23 patients with a personal history of HBOC-associated cancer who were found to carry a DV, only 35% (n = 8) had a DV in BRCA1/2, while the majority of these positive cases had a DV in non-BRCA genes. Similarly, out of the 8 patients with a personal history of LS-associated cancers who were found to carry a DV, only 38% (n = 3) were in LS-associated genes. The same findings hold true when considering family history. Out of the 56 DV-positive individuals with a family history of HBOC-associated cancers, only 32% (n = 18) were found to have DV in BRCA1/2 and of the 47 DV-positive individuals with a family history of LS-associated cancer, only 21% (n = 10) carry DV in LS genes. These results emphasize the clinical utility of larger gene panels, as the majority of positive cases would have been missed, if they would have been tested solely for mutations in genes that were clinically indicated.
Interestingly, there were 15 individuals in the dataset who had both a personal and a family history of both HBOC- and LS-associated cancer, including 10 patients with ovarian cancer (which falls under both syndromes) and 5 individuals with both breast and colon cancer. Family history for these individuals was extensive and included at least two types of cancer in the family, and up to 7 types of cancer within a single family. However, despite a strong personal and family history of cancer, no DV was detected in this small subset. One possible explanation for this may be that these individuals carry DV in untested genes, further supporting the expansion of hereditary cancer gene panels, although the contribution of environmental factors leading to familial cancer susceptibility cannot be entirely excluded.
In the first 630 patients tested with the 27-gene panel, a DV was detected in 14 genes. After the collection of data for this study, additional DV have been detected in other genes including
BMPR1A,
FAM175A,
MLH1,
RAD51C,
RAD51D,
SMAD4,
STK11, and
TP53. Of note, although
MLH1 represents about 50% of all LS cases [
4], no DV in this gene were detected via the 27-gene panel during the study period, and a limited number have been detected after the conclusion of the study. The reason for the underrepresentation of
MLH1 DV within this panel testing may be attributed to either a low percentage of individuals reporting a personal history of LS in our cohort (10%,
n = 64) or an increased uptake of the specific LS gene panel prior to the consideration of expanded panels such as ours. In our experience, the vast majority of
MLH1 DV were detected by
MLH1 single gene sequencing (60%) or the broader 4 gene LS gene panel testing (35%), while only a handful of
MLH1 DV were detected through the 27-gene panel, providing support for this finding.
We observed a 4-fold increase in the number of DV detected when going from BRCA1/2 genes upwards to the 25 additional genes on the panel. This increase in the positive rate clearly offsets the concomitant 8-fold increase in VUS. The positive rate as each gene set is added continued to increase, without a plateau being observed, suggesting that inclusion of additional genes to the panel offers room for a further increase in the overall positive rate. However, the critical number of genes before the positive rate plateaus towards the asymptote remains to be determined.
Three patients were found to have large deletions: one
APC promoter deletion in a patient with an extensive family history of cancer, a multiple-exon deletion in
BRCA1 in a patient with personal and family history of HBOC-associated cancers, and a large multiple exon deletion in
PMS2 in a patient with a family history of both LS- and HBOC-associated cancer. Other studies testing for large deletion/duplication analysis within hereditary cancer panels have reported a range of positive findings from 0.7–2% [
7,
14,
28] of the entire testing population, which is within range of 0.5% in our cohort.
Fifty out of the 52 DV in our dataset (96%) had concordant classifications in the ClinVar database. Four variants had ClinVar entries with other classifications in addition to pathogenic and likely pathogenic (see Additional file
1).
BRIP1 c.2392C > T (p.Arg798Ter) had a VUS and a pathogenic entry relative to phenotypes of breast cancer and Fanconi Anemia respectively, both contributed by the same submitter. As heterozygous carriers of truncating mutations in
BRIP1 have been reported to be susceptible to breast cancer [
29], this VUS entry represents a pathogenic low penetrance classification that is currently not amenable to ClinVar submission requirements [
30]. Similarly,
CHEK2 c.1555C > T (p.Arg519Ter) has a VUS entry without supporting evidence by a submitter, last evaluated in 2015. However, the corresponding publication by the same submitter cited in ClinVar did not list this variant among the reported VUS in
CHEK2. Therefore, these two variants were not counted as discordant. The third variant,
ATM c.4394 T > C (p.Leu1465Pro) has a VUS entry by a submitter, last evaluated in 2016, while acknowledging the experimental and clinical studies supporting a damaging role of the variant. This relates to differences originating from evidence weighting among submitters contributing to public databases. Lastly,
PMS2 c.2186_2187delTC (p.Leu729Glnfs) has one VUS entry and a likely benign entry, based on its high frequency in control databases (ExAC) and the possibility of this variant coming from
PMS2CL, the
PMS2 pseudogene. We have observed this variant in several patients with either a personal and/or a family history of breast, brain, endometrial, cervical and uterine cancers referred for testing at our laboratory. In all our cases, we confirmed the presence of this variant in
PMS2 gene by long-range PCR. Furthermore, literature evidence describing this variant reports confirmation of this variant by long-range PCR and its segregation in individuals with features of Lynch and Constitutional Mismatch Repair Deficiency [
31] (CMMRD) phenotypes. Recent studies have reported the presence of biallelic
PMS2 mutations in up to 60% of patients with CMMRD [
31] in contrast to LS patients, the large majority of whom are carriers of
MLH1 and
MSH2 mutations. Furthermore, studies in unselected cohorts of colorectal cancer patients have demonstrated a higher prevalence of
PMS2-associated LS than previously thought [
32]. Nevertheless, its presence at a high frequency in ExAC could arise from a sequencing misalignment/pseudogene issues within the ExAC dataset, a higher carrier frequency for CMMRD reflecting the reduced penetrance of heterozygous
PMS2 mutations, or a combination of these factors.
This study was limited by the mode of data collection as the personal and family histories of cancer were ascertained from the self-reported information provided on requisition forms at test intake. Therefore, some of the information provided may not have been exhaustive for personal and family histories, including previous BRCA1/2 or LS testing history and fulfillment of the NCCN guidelines as part of insurance coverage requirements prior to ordering genetic testing. Although these factors could be sources of potential bias in our cohort, they do not seem to inflate our tested BRCA and LS positive rates. The higher positive rates for non-BRCA and non-LS genes in our study relative to previous reports point to some unique features of our dataset. Additional studies would be needed to independently confirm these findings in unbiased datasets. For a subset of our cohort, no reasons for referral were provided. This includes seven individuals in whom a DV was identified. Another limitation is posed by the dynamic nature of variant classifications. All variants reported in this study represent a snapshot in time limited to the classifications obtained within the range of data collection. Concomitantly, the VUS rates reported are subject to some change as variants are re-evaluated at defined intervals. However, this is unlikely to have a significant impact on positive rates as most VUS move towards the normal spectrum as additional information becomes available. Lastly, although the overall sample size of our cohort was comparable to other published reports, the smaller subset of patients within each phenotypic grouping must also be considered as a potential limitation.