If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Pathogenic or likely pathogenic germline variants (PGVs) were found in 13.6% of uterine cancer patients tested.
•
Nearly 40% of PGVs were found outside of Lynch syndrome genes and PTEN.
•
The prevalences of PGVs in patients <50 years and ≥ 50 years at the time of testing were similar (15.1% vs 13.2%).
•
Sixty percent of PGVs were associated with FDA-approved therapies and 35% with precision therapy clinical trials.
Abstract
Objective
Hereditary uterine cancer (UC) is traditionally associated with pathogenic/likely pathogenic germline variants (PGVs) in Lynch syndrome genes or PTEN; however, growing evidence supports a role for other genes that may reveal new clinical management options. In this study we assessed the prevalence and potential clinical impact of PGVs identified in UC patients referred for comprehensive germline genetic testing that combined testing for Lynch syndrome, PTEN, and other cancer predisposition genes.
Methods
Prevalence of PGVs in patients referred to a single clinical lab for germline genetic testing with an indication of uterine or endometrial cancer were retrospectively assessed and compared by syndrome type, patient age at testing, and self-reported ancestry. Potential clinical actionability of PGVs was based on established guidelines for clinical management, targeted therapies, and clinical trial eligibility.
Results
PGVs were detected in 13.6% of the cohort (880/6490). PGVs were most frequently observed in Lynch syndrome genes (60.4%) and PTEN (1.5%), with 38.1% in another cancer predisposition gene (i.e., CHEK2, BRCA1/BRCA2). PGV prevalence was similar for patients <50 years and those ≥50 years (15.1% vs 13.2%). Nearly all PGVs (97.2%) were associated with guideline-recommended management, including cascade testing; 60.5% were associated with FDA-approved therapies; and 35.2% were associated with clinical trials.
Conclusions
Focusing germline testing on Lynch syndrome genes and PTEN and limiting testing to patients <50 years of age at diagnosis may overlook a substantial proportion of UC patients who harbor actionable PGVs. Universal comprehensive genetic testing of UC patients could benefit many patients and at-risk family members.
]. Identifying patients with hereditary risk for UC through germline genetic testing can facilitate enhanced surveillance and potential risk reduction for future cancers, as well as access to gene-targeted treatments and clinical treatment trials. It can also benefit at-risk relatives who choose to be screened for the cancer predisposition syndrome identified in the index patient.
Most hereditary cases of UC are associated with Lynch syndrome (LS). Although this autosomal dominant cancer syndrome is commonly associated with colorectal cancer, gynecological cancers can be the sentinel cancer in half of cases among women [
]. LS is typically caused by pathogenic/likely pathogenic germline variants (PGVs) in four mismatch repair (MMR) genes: MLH1, MSH2, MSH6, and PMS2. Deletions in the EPCAM gene, which silence MSH2, also contribute to a small proportion of LS cases. Compared with the general population, patients with LS have a 20–30 times greater risk of developing endometrial cancer (EC), the most common type of UC [
]. Separately, a small proportion of hereditary UC cases are associated with PTEN-hamartoma tumor syndrome (PHTS), also known as Cowden syndrome, a cancer syndrome caused by PGVs in the PTEN gene [
]. Although LS and PHTS are traditionally associated with hereditary UC, evidence is growing for the role of other genes in increasing UC risk. PGVs in homologous recombination repair (HRR) genes BRCA1 and BRCA2 have been associated with increased risk for UC in some populations [
High incidence of BRCA1–2 Germline mutations, previous breast Cancer and familial Cancer history in Jewish patients with uterine serous papillary carcinoma.
Eur. J. Surg. Oncol.: J. Eur. Soc. Surg. Oncol. Br. Assoc. Surg. Oncol.2006; 32: 1097-1100
]. Current guidelines from the National Comprehensive Cancer Network (NCCN) recommend that providers screen all EC patients for LS using immunohistochemistry chemistry (IHC) or microsatellite instability (MSI) testing [
]. However, the recommendations restrict germline genetic testing for LS to patients according to a complicated algorithm that combines abnormal tumor screening, specific family history of cancer, and/or a UC diagnosis before 50 years of age. The NCCN also recommends PTEN testing for women with EC if they exhibit other signs of PHTS [
]. However, because these criteria have been shown to exclude some patients with hereditary cancer risk, the incidence and spectrum of genetic drivers of UC risk may not be fully understood [
]. In addition, the restrictive guidelines may prevent at-risk patients from accessing recommended clinical interventions such as checkpoint inhibitors or clinical trial-related interventions that require a genetic diagnosis in specific genes [
]. Finally, many UC patients are diagnosed at early cancer stages and have high survival rates, and those with hereditary cancer syndromes may go on to develop other cancers later in life [
]. The failure to test UC patients for PGVs in hereditary cancer syndrome genes is a missed opportunity to undertake enhanced cancer prevention measures in these patients. Finally, any missed opportunity to identify hereditary cancer risk in a UC patient is compounded by the missed opportunity for cascade testing in family members. Cascade testing can identify LS or other genetic cancer risk in healthy individuals and thereby enable implementation of enhanced screenings and other preventive care before a cancer emerges [
The prevalence of PGVs in UC patients remains unclear, as a wide range of prevalences (4.5%–23%) have been reported in cohorts with limited sample sizes, eligibility for testing (including specificity of UC type), and number of cancer predisposition genes tested [
]. With the growing opportunities for precision therapy in UC, in addition to primary and secondary prevention strategies, an in-depth understanding of the genetic contributors to increased risk of UC is warranted. Here, we describe the prevalence of PGVs in cancer risk genes and potential clinical actionability among a large cohort of patients receiving comprehensive germline genetic testing for UC.
2. Methods
2.1 Study design
Under independent review board (IRB) approval (WCG IRB protocol number 1167406), the study retrospectively reviewed 6,490 patients who were referred for clinical germline genetic testing with an indication of UC (or EC, specifically) between October 2014 and August 2018. Clinical genetic testing was performed at a commercial laboratory (Invitae, San Francisco, CA, USA), which is certified for clinical use and patient reporting under Clinical Laboratory Improvement Amendments (CLIA) and accredited by the College of American Pathologists (CAP). De-identified, clinician-reported patient information from test requisition forms, including age at the time of testing, indication for testing, and self-reported ancestry, was analyzed.
All patients underwent clinical germline genetic testing for PGVs in cancer predisposition genes. The decision to order testing and the specific genes to be tested were determined by the ordering clinician, who could request individual genes or a CLIA-certified next-generation sequencing-based hereditary cancer panel, as previously described [
]. Inclusion in the study required that at least one LS gene (i.e., MSH6, MSH2, MLH1, PMS2, or EPCAM) or PTEN be included in their testing. Patients with a known familial variant were excluded.
2.2 Genetic testing
Full-gene sequencing and deletion/duplication analysis, including coding exons and 10 base pairs (up to 20 bp in BRCA1 and BRCA2) of adjacent intronic sequence on either side of the coding exons, was performed to detect single nucleotide variants, insertions and deletions < 15 bp in length, and exon-level deletions and duplications, as previously described [
], a refinement of variant interpretation guidelines from the American College of Medical Genetics and Genomics (ACMG) and the Association for Molecular Pathology (AMP) [
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.
Genet. Med.: Off. J. Am. Coll. Med. Genet.2015; 17: 405-424
]. With Sherloc, variants were classified as pathogenic (P), likely pathogenic (LP), variant(s) of uncertain significance (VUS), likely benign, or benign. P or LP variants (i.e., PGVs) and VUS were reported to clinicians.
Patient results were categorized based on variant classification and inheritance of conditions associated with observed variants. Patients were defined as having a positive result if they had at least one PGV in a gene associated with dominant inheritance or two PGVs in a gene associated with recessive inheritance. Patients with a single PGV in a gene associated with recessive inheritance that would have no potential impact on UC patient care (i.e., patients with non-actionable recessive findings) were excluded from the positives and all subsequent analyses in the study. Patients were defined as having an uncertain result if they had a VUS in the absence of any PGV. Finally, patients were defined as having a negative result if they had no PGV or VUS.
2.3 Actionability of genetic testing results
Actionable testing results comprised three categories based on the potential implications of a positive result in associated genes: 1) U.S. Food and Drug Administration (FDA)-approved therapies included results in genes that grant eligibility for immune checkpoint inhibitors (i.e., MLH1, MSH2, MSH6, PMS2, EPCAM); 2) Clinical trial eligibility included results in genes that grant eligibility for clinical trials of experimental cancer therapies, such as PARP inhibitors (i.e., ATM, BAP1, BARD1, BLM, BRCA1, BRCA2, BRIP1, CHEK2, FANCC, FANCL, NBN, PALB2, PTEN, RAD50, SLX4, WRN); 3) Established management recommendations included results in genes that are recommended by NCCN guidelines to be followed with family cascade testing (any cancer-predisposition gene with dominant inheritance) or enhanced cancer surveillance or risk-reduction [i.e., APC, ATM, BAP1, BARD1, BLM, BRCA1, BRCA2, BRIP1, CDH1, CDKN2A, CHEK2, DICER1, EPCAM, FANCC, FANCL, FH, FLCN, KIT, MEN1, MLH1, MSH2, MSH6, MUTYH (biallelic), NF1, PALB2, PMS2, PTEN, RAD51C, RAD51D, RET, SDHA, SDHB, SDHC, SLX4, SMAD4, TMEM127, TP53, WRN].
2.4 Analyses
Cancer type was determined by computational searches of the following ICD-10 codes and terms in the free-text responses on test order forms: (1) For UC, C54.0, C54.2, C54.3, C54.8, C54.9, C55, Z85.42, “uterine”, “uterine ca”, “uterine cancer”; (2) For EC, C54.1, “endometrial”, “endometrial ca”, and “endometrial cancer”. Cases with multiple results were manually reviewed.
Differences in PGV frequency by patient age and ancestry were evaluated using Chi-square tests (https://www.socscistatistics.com/tests/chisquare). P values <0.05 indicate statistical significance. Difference in age-at-onset of patients with PGVs in Lynch Syndrome/PTEN versus those with PGV in other hereditary cancer syndrome genes was by unpaired t-test (GraphPad https://www.graphpad.com/quickcalcs/ttest1/?format=C). Ancestry was determined by patient or clinician report on genetic testing order forms, which included preset options and the ability to enter free text.
3. Results
3.1 Study cohort
In the study cohort of 6490 patients, the mean age at time of testing was 61 years (range, 16–104 years) (Table 1). Patients under age 50 represented 18.6% (1208/6490) of the cohort. The cohort was predominantly White (69.5%), with smaller representation from other self-reported ancestries, including Hispanic (6.2%), Black/African American (4.9%), Asian or Pacific Islander (4.7%), and Ashkenazi Jewish (2.9%). Nearly two-thirds (4223/6490) of patients had endometrial cancer, while 34.9% (2266/6490) had uterine cancer. One patient had both endometrial and uterine cancer indicated. Patients were tested for a median of 42 genes (range, 1–220 genes). Most patients (85.6%; 5553/6490) had >12 genes tested, and 64.7% (4200/6490) had >24 genes tested.
Excludes patients with a single pathogenic or likely pathogenic germline variant (PGV) in a gene associated with recessive inheritance that would have no impact on uterine cancer patient care (DIS3L2, FH p.Lys477dup, MSH3 (monoallelic), MUTYH (monoallelic), NTHL1 (monoallelic), or VHL p.Arg200Trp). These patients are included in the column “Patients without a PGV.”
(N = 880) No. (%)
Patients without a PGV (N = 5610) No. (%)
P Value
Age at testing, years
Mean (range)
61 (16–104)
59.5 (16–95)
62 (17–104)
0.0001
<50, n (%)
1208 (18.6)
182 (20.7)
1026 (18.3)
0.0936
≥50, n (%)
5282 (81.4%)
698 (79.3)
4584 (86.8)
Self-reported ancestry, n (%)
Ashkenazi Jewish
186 (2.9)
18 (2.1)
168 (3.0)
0.1282
Asian or Pacific Islander
302 (4.7)
49 (5.6)
253 (4.5)
0.1684
Black or African American
320 (4.9)
43 (4.9)
277 (4.9)
1
Hispanic
403 (6.2)
61 (6.9)
342 (6.1)
0.3293
Multiple
330 (5.1)
49 (5.6)
281 (5.0)
0.4586
Native American
24 (0.4)
6 (0.7)
18 (0.3)
0.1260
White
4509 (69.5)
588 (66.8)
3921 (69.9)
0.0701
Other
60 (0.9)
9 (1.0)
51 (0.9)
0.70531
Unknown
356 (5.5)
57 (6.5)
299 (5.3)
0.1754
Excludes patients with a single pathogenic or likely pathogenic germline variant (PGV) in a gene associated with recessive inheritance that would have no impact on uterine cancer patient care (DIS3L2, FH p.Lys477dup, MSH3 (monoallelic), MUTYH (monoallelic), NTHL1 (monoallelic), or VHL p.Arg200Trp). These patients are included in the column “Patients without a PGV.”
Overall, 13.6% of patients (880/6490) had a positive result (Fig. 1A ). This excluded 72 patients (1.1% of the cohort) who harbored a single PGV in a gene that is associated with a recessively inherited condition and lacks any current management or treatment implications for UC patients (i.e., DIS3L2, FH p.Lys477dup, MSH3 (monoallelic), MUTYH (monoallelic), NTHL1 (monoallelic), VHL p.Arg200Trp). Most of these recessive, non-actionable findings were in MUTYH (62 patients, 1.0% of cohort). All subsequent analyses excluded all 72 patients with recessive, non-actionable findings.
Fig. 1Genetic testing results in uterine cancer patients. Percentage of positive, negative and uncertain results identified in 6490 uterine cancer patients. Positive patients (n = 880) had at least one pathogenic or likely pathogenic germline variant (PGV), excluding those patients with a PGV in a gene associated with a recessive condition without clinical management implications. Uncertain patients (n = 1954) had a variant of uncertain significance (VUS) in the absence of any PGV. Negative patients (n = 3656) had no PGV or VUS, or only an unactionable PGV associated with a recessive condition (A). Percentage of the 880 positive patients with a PGV in a Lynch syndrome gene (MSH6, MSH2, MLH1, PMS2, or EPCAM) (n = 532), PTEN (n = 13), or other known cancer predisposition gene (n = 335). Patients with more than one PGV are only counted once (B).
In total, 998 PGVs were identified in 953 patients, with 910 patients harboring one PGV, 41 patients harboring two PGVs, and 2 patients harboring three PGVs. PGVs were most frequent in MSH6 (23.4%; 234/998), MSH2 (13.0%; 130/998), PMS2 (10.0%; 100/998), and CHEK2 (9.4%; 94/998) (Fig. 2). One or more VUS in the absence of any PGV were observed in 30.1% (1954/6490) of patients.
Fig. 2Pathogenic or likely pathogenic germline variants in uterine cancer patients. The 130 MSH2 variants were observed in 129 patients, as one patient was homozygous for MSH2 variants. MUTYH (biallelic) indicates the patients were homozygous or compound heterozygous for two pathogenic or likely pathogenic germline variants (PGVs) in MUTYH; this pairing was observed in four patients. All other results represent a heterozygous (single PGV) finding, including in WRN and FANCC, which are associated with recessively inherited conditions. Genes with fewer than four PGVs were excluded from the figure for brevity. See Supplemental Table 1 for all observed PGVs by gene.
Among the 880 positive patients, 60.3% (532/880) had a PGV in a LS gene (Fig. 1B), corresponding to 8.2% of the total cohort (532/6490). Most LS findings were in MSH6 (43.8%; 234/532) (Fig. 3). Patients with a PGV in PTEN represented 1.5% of all positive UC patients (13/880) and 0.2% (13/6490) of the entire cohort. More than one-third of positive patients (38.1%; 335/880) had a PGV in a cancer predisposition gene not associated with LS or PHTS. These non-LS/non-PHTS positives accounted for 5.2% (335/6490) of the entire cohort. The non-LS/non-PHTS results were found in 52 cancer-predisposition genes, most frequently CHEK2 (10.5% of all positives, 93/880), BRCA2 (5.9%, 52/880), BRCA1 (4.6%, 41/880), and ATM (4.2%, 37/880).
Fig. 3Lynch syndrome and PHTS genes with pathogenic or likely pathogenic variants in uterine cancer patients. One patient with a pathogenic or likely pathogenic germline variant (PGV) in both MSH6 and PMS2 is counted in each gene's column. MSH2 and PMS2 results each include one patient with two PGVs in the same gene. The phase of the two MSH2 PGVs is unknown. The PMS2 PGVs are confirmed to be in trans; the prognostic and management implications for this patient would differ from the patients with a single PGV. Results also include patients who had a PGV in both a Lynch syndrome gene or PTEN and another gene: MSH6 + CHEK2 (2 patients) and MSH6 + MUTYH (1 patient); MSH2 + NF1 (1 patient) and MSH2 + BRIP1 (1 patient); and PMS2 + CHEK2 (2 patients), PMS2 + BRCA2 (2 patients), PMS2 + RAD50 (1 patient), PMS2 + MUTYH (1 patient), PTEN + RB1 (1 patient), and PTEN + MUTYH (1 patient).
The frequency of PGVs was similar in patients <50 years (15.1%; 182/1208) and those ≥50 years of age (13.2%; 698/5282) at the time of testing) (p = 0.0936). Patients <50 years of age at testing comprised 20.7% (182/880) of patients with a PGV (Table 1). The remaining 79.3% (698/880) of PGV-positive patients were ≥ 50 years old. Patients with a PGV in a LS gene or PTEN were on average significantly younger than patients with other PGVs by 4.7 years (mean, 57.7 vs 62.4 years; p = 0.0001).
3.3 Actionable findings
Of all patients with a diagnostic or carrier PGV, 97.2% (855/880) were eligible for NCCN and/or peer-reviewed consensus management recommendations based on their germline genetic testing results, including cascade testing of at-risk family members (Table 2). In addition, 60.5% (532/880) were potentially eligible for FDA-approved therapies, including PD-1 or PD-L1 checkpoint inhibitors. Finally, 35.2% (310/880) were potentially eligible for clinical treatment trials based on positive findings in the HRR pathway, including BRCA1 and BRCA2.
Table 2Clinical utility of germline mutations observed in patients with uterine cancer.
Most patients with actionable findings had PGVs in LS genes (60.3%; 532/880); however, 38.1% (335/880) of patients with actionable findings had a PGV outside of the LS genes and PTEN. Thus, 5.2% of all UC patients in this cohort (335/6490) had an actionable finding in a gene that was not a LS gene or PTEN. No significant difference in the frequency of actionable PGVs was observed between patients <50 years and those ≥50 years of age at time of testing (14.9% vs 13.1%; p = 0.0926).
4. Discussion
To the best of our knowledge, this study analyzed the largest reported cohort of patients with UC who underwent germline genetic testing with a comprehensive multigene panel. Overall, 13.6% of the 6490 patients tested harbored a PGV, and 97% of these PGV-positive patients were eligible for guideline-recommended management. PGVs in LS genes are critical for identifying patients who could be eligible for PD-1/PD-L1 inhibitors and are at risk for additional LS-associated cancers (e.g., ovarian cancer). Additionally, clinical trial eligibility was available for 35.2% of patients with positive findings, including results in BRCA1/BRCA2 and other HRR genes that could render patients eligible for PARP inhibitors.
Various frequencies of PGV-positive patients have been reported in other recent studies of germline genetic testing in UC patients (4.5%–23.2%) [
]. Differences in cohort selection and which genes were tested likely contribute to the variation. Whereas our cohort selection was agnostic to UC type, most studies of the genetic etiology of UC are limited to patients with EC [
]. In addition, despite the wide range in number of genes tested in our cohort (1–220 genes), most patients had testing of at least 25 genes (median, 42 genes). In contrast, many other studies tested with more restrictive gene panels [
]. Our inclusive testing enabled a comprehensive view of cancer predisposition PGVs in UC patients referred for genetic testing, among whom 5.2% harbored a PGV in a gene otherwise not associated with LS or PHTS. Although this study was unable to stratify genetic testing results by UC subtype, other work has indicated a markedly different genetic landscape for rare UC subtypes when compared with common subtypes of UC [
Germline mutations of SMARCA4 in small cell carcinoma of the ovary, Hypercalcemic type and in SMARCA4-deficient undifferentiated uterine sarcoma: clinical features of a single family and comparison of large cohorts.
]. For example, a rare and aggressive form of uterine sarcoma is associated with SMARCA4 PGVs, and serous or serous-like EC has been associated with BRCA1 PGVs [
High incidence of BRCA1–2 Germline mutations, previous breast Cancer and familial Cancer history in Jewish patients with uterine serous papillary carcinoma.
Eur. J. Surg. Oncol.: J. Eur. Soc. Surg. Oncol. Br. Assoc. Surg. Oncol.2006; 32: 1097-1100
]. Genetic testing that is limited to LS genes and PTEN would miss many of these important patient results. Finally, our study used a convenience sample of patients referred for genetic testing with an indication of UC or EC, and therefore may have been enriched for patients with PGVs. This may in part explain the higher prevalence of PGVs observed in our cohort (13.6%) compared to a recent study by Levine et al. which reported a PGV prevalence of 10.1% among an unselected cohort of EC patients [
That a hereditary predisposition to cancer can be found in a substantial proportion of UC patients highlights the importance of germline testing in early detection and cancer prevention. A study of >11,000 UC patients in Taiwan who were followed for 30 years found that 4.8% developed at least one second primary malignancy, including ovarian, colon, and breast cancers [
]) leaving many survivors at increased risk for second cancers, particularly those with hereditary cancer predisposition syndromes. In a recent study, Lincoln et al. found that 11.2% of patients with any type of cancer had PGVs identified only after presenting with a second primary cancer that could have been detected earlier or prevented [
], a clear indication that increased genetic testing and awareness among patients and their providers is needed. Identification of cancer predisposition syndromes following a UC diagnosis can be particularly beneficial for individuals with PHTS, whose lifetime risk of developing breast cancer (up to 50%) may be at least twice that of developing UC (19%–28%) [
Although this study did not evaluate which patients in the cohort met NCCN guidelines for germline genetic testing, evidence from small cohorts suggests that current clinical criteria would fail to identify some patients who harbor actionable variants [
]. Moreover, 79.3% of all diagnostic PGVs in this study were in patients ≥50 years old, none of whom would be eligible for genetic testing under strict adherence to NCCN guidelines. Others have also noted the risk of missing hereditary cases of UC when screening is restricted to those under age 50 [
]. This body of evidence suggests that a change in practice should be considered. Simplification of testing guidelines could help increase access to genetic testing and reduce the frequently missed opportunities for genetic testing that have been reported even among high-risk women [
], some hospital systems and healthcare groups are moving toward a more universal testing approach for UC patients, similar to that for ovarian and pancreatic cancer patients [
]. For example, medical policy for UnitedHealthcare, an insurance provider, currently states that multi-gene germline genetic testing is proven and medically necessary for individuals with a personal history of a cancer associated with LS (including EC), as well as individuals with a first-degree relative with an LS-associated cancer.[40].
Our data showed that approximately 8% of UC patients referred for genetic testing had LS. Germline testing all individuals with a new diagnosis of UC may be a more efficient method of identifying patients with LS compared to tumor screening by MSI/IHC. Screening patients for LS by MSI/IHC requires confirmation of suspected positives by germline genetic testing, thereby adding cost and time to diagnosis [
]. Moreover, as demonstrated in this study, germline genetic testing can reveal non-LS cancer syndromes that would not be detected by MSI/IHC. However, a germline-only approach would have its own limitations, including missed opportunities to determine eligibility for immunotherapy and provide prognostic information not available in germline results [
]. Therefore, the field lacks clarity regarding the best strategy for screening UC tumors. Future studies exploring patient outcomes and healthcare expenditures among patients who receive different approaches of care (including paired somatic and germline genetic testing) may be needed to fill this gap in understanding.
A strength of this study was the large cohort of patients with a broad range of UC indications (including EC and more rare conditions such as papillary serous UC), which bolsters the generalizability of our estimates of the incidence of PGVs in UC. However, our study was limited by the fact that personal and family history data were not consistently reported by ordering clinicians and could not be confirmed by direct review of the medical records. Age at diagnosis, for instance, was not required. Because we could not assess this information across patients, age at testing was used as a proxy. Furthermore, because pathology records were infrequently available for review, we could not analyze our results based on UC/EC subtype, MMR status, or IHC status. Given UC and EC are often used interchangeably even among experts, it would be beneficial to conduct studies with detailed pathology information to better understand the yield of PGVs in EC and non-EC subtypes of UC. In addition, the cohort may have been enriched for patients with a higher probability of having a PGV since UC patients referred for germline testing are more likely to have suggestive personal and/or family histories of cancer (i.e., selection bias by ordering providers). However, even if the rates of PGVs would be lower in an unselected population, the clinical utility of uncovering these findings is still estimated to be high. Moreover, although patients with PGVs in MUTYH were excluded from our analyses on actionable results, NCCN guidelines would recommend early initiation of colorectal cancer screening for any of the 62 patients harboring a PGV in MUTYH if they had a relative with colorectal cancer [
]. Therefore, there may be some underestimation in clinical utility among this cohort. Finally, future work should explore the impact of cancer predisposition PGVs on the development of second cancers in UC survivors.
5. Conclusion
In this study evaluating 6490 UC patients referred for genetic testing, 13.6% of patients carried PGVs in cancer predisposition genes, 38.1% of which were in genes not traditionally associated with hereditary UC. >95% of patients with PGVs were potentially eligible for approved precision therapies and/or clinical treatment trials. Overall, our data support the clinical utility of universal, comprehensive germline genetic testing for UC patients, including those over age 50, given the impact these results can have on gene-directed cancer treatment, secondary cancer surveillance and prevention, and family cascade testing. This study provides further evidence suggesting universal germline testing of patients with solid tumors, and subsequent cascade testing in family members, may be a powerful approach for identifying cancer patients and healthy individuals with increased cancer risk due to hereditary cancer syndromes.
Prior presentation
Data from this study was presented at the Annual Scientific Meeting of the Australia New Zealand Gynaecological Oncology Group in Melbourne on March 26, 2022.
Contribution to authorship
SM, SMN, SY, STM, and EDE investigated and conceptualized the study. SY curated the data and BH, STM, and SMN formally analyzed the data. BH, STM, SMN, SR, and EDE wrote the original draft manuscript and all authors reviewed and edited the manuscript.
Declaration of Competing Interest
Authors BH, SMN, SR, STM, and EDE are employees and shareholders of Invitae. Authors SM and SY are shareholders of Invitae.
High incidence of BRCA1–2 Germline mutations, previous breast Cancer and familial Cancer history in Jewish patients with uterine serous papillary carcinoma.
Eur. J. Surg. Oncol.: J. Eur. Soc. Surg. Oncol. Br. Assoc. Surg. Oncol.2006; 32: 1097-1100
Germline mutations of SMARCA4 in small cell carcinoma of the ovary, Hypercalcemic type and in SMARCA4-deficient undifferentiated uterine sarcoma: clinical features of a single family and comparison of large cohorts.
Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology.
Genet. Med.: Off. J. Am. Coll. Med. Genet.2015; 17: 405-424
Excludes patients with a single pathogenic or likely pathogenic germline variant (PGV) in a gene associated with recessive inheritance that would have no impact on uterine cancer patient care (DIS3L2, FH p.Lys477dup, MSH3 (monoallelic), MUTYH (monoallelic), NTHL1 (monoallelic), or VHL p.Arg200Trp). These patients are included in the column “Patients without a PGV.”
00334-1&id=gr1.jpg){kind=link}
00334-1&id=gr2.jpg){kind=link}
00334-1&id=gr3.jpg){kind=link}