Background:

Tumor mutation burden (TMB) is the total exonic mutation count per megabase of tumor DNA. Recent advances in precision medicine occasionally detect Lynch syndrome (LS) by germline sequencing for mismatch-repair (g.MMR) genes but not using TMB. The current study analyzes the utility of TMB in detecting LS.

Methods:

Whole-exome sequencing (ion-semiconductor sequencing) was performed for somatic and germline DNA from 2,501 various cancer patients to detect TMB and g.MMR sequencing. MMR IHC was conducted when high TMB (≥10) was detected in LS-related cancers with an additional condition of wild-type BRAF in colorectal cancers. Target sequencing and multiplex ligation-dependent probe amplification (MLPA) were further performed for g.MMR genes in MMR-deficient cancers (TMB-based g.MMR target sequencing). We compared universal sequencing and TMB-based target sequencing in their sensitivity for detecting LS.

Results:

LS was detected in 16 (0.6%) of the 2,501 patients: 1.1% (9/826) of colorectal cancer patients, 16.2% (6/37) of endometrial cancer patients, and 14.3% (1/7) of small intestine cancer patients. TMB-based g.MMR target sequencing (81.3%) showed superior sensitivity for detecting LS than universal g.MMR sequencing (56.3%; P = 0.127) but missed 3 LS patients (1 with a low-TMB cancer, 1 with a BRAF-mutant colorectal cancer, and 1 with an MMR-proficient cancer). Ion-semiconductor sequencing could detect single-nucleotide substitutions but not large deletions. POL-mutated cancers showed extremely high TMBs (48.4–749.2).

Conclusions:

g.MMR target sequencing, combined with TMB, somatic BRAF mutation, and MMR IHC is an effective strategy for detecting LS.

Impact:

TMB can be a biomarker for detecting LS in precision medicine.

Tumor mutation burden (TMB) is defined as the total count of single-nucleotide variations, insertions, and deletions (indels) per each megabase (Mb) in the coding region of tumor DNA and is also a predictive biomarker for the response to immune-checkpoint inhibitors (1–3). Today, next-generation sequencing (4) is being increasingly applied in the oncological field to genetically characterize tumors and select the most suitable treatment agent or “precision medicine.” TMB can be measured not only through whole-exome sequencing (WES) but also by multigene panel testing using NGS on fresh-frozen (5) or formalin-fixed paraffin-embedded (FFPE) cancer tissues (6). TMB degree varies among human cancers, tending to be higher in the type of lung cancer developed in smokers, cutaneous squamous carcinoma, melanoma, and colorectal carcinoma; in contrast, it is lower in non–small cell-type lung cancer, breast cancer, pancreatic cancer, and childhood cancers (6–8). TMB is also high in cancers associated with microsatellite instability (MSI), which is caused by the promoter methylation of MLH1 (often coupled with BRAF V600E mutation in colorectal cancer; refs. 4, 9) and by mutations of DNA mismatch-repair (MMR) genes (6). TMB is further higher in cancers exhibiting somatic and germline mutation of DNA polymerase proofreading genes, including POLE and POLD1 (10–12).

Lynch syndrome (LS) is an autosomal dominant cancer predisposition syndrome caused by germline variants in MMR genes (MLH1, MSH2, MSH6, and PMS2), accounting for 1% to 4% of colorectal cancer (13–16) and 2% to 6% of endometrial cancers (EC; refs. 17–19). Variant carriers are at risk of early-onset colorectal cancer (15, 16), EC (17, 18), upper tract urothelial cancers (20, 21), and a spectrum of other associated tumors (22), and, therefore, the detection of g.MMR variant carriers is crucial in reducing the number of cancer-related deaths via the surveillance of target organs.

The standard diagnosis of LS begins with the selection of high-risk individuals by reviewing their and their family's cancer history based on the Amsterdam II criteria or revised Bethesda guidelines (14). These patients are then further evaluated by the IHC of MMR proteins as well as MSI analysis in their tumors. A final diagnosis is made based on deleterious g.MMR gene variants detectable via germline DNA sequencing in MSI-high and/or MMR-deficient cancer patients (13, 14). However, these guidelines and criteria were established only for colorectal cancer patients, with limited sensitivity for LS. Therefore, recently, the strategy for LS detection has shifted toward universal screening using MMR IHC and/or MSI analysis for all or for age-specific conditions of LS-related cancers (18, 19, 23–25). However, to date, TMB has not been used as a marker for screening LS.

This study analyzed the utility of TMB analysis using WES to detect LS in cancer patients, especially LS-associated cancer patients.

Patients and samples

In January 2014, the Shizuoka Cancer Center started a project with high-tech omics-based patient evaluation (Project HOPE) that performed genome-wide exome sequencing in germline and somatic DNAs obtained from various cancer patients (26). From January of 2014 through June of 2016, 2,501 cancer patients who agreed on this study and had a tumor large enough for the genomic analysis participated in the current study, within total 8,549 cancer patients who underwent surgical resection for their cancers in the Shizuoka Cancer Center (Table 1). Hence, almost all participants presented with good performance status (PS: 0–1) and had cancers in the resectable stage. All cancers were pathologically confirmed using surgical samples.

Table 1.

Cancer types and demographics of 2,501 cancer cases.

Number of casesGender (n)Age (y.o.)
Cancer origin/typeResected casesaCases in this studyMaleFemaleMean ± SD
Colon 1,347 826 502 324 66.3 ± 11.9 
Lung 760 420 247 173 69.4 ± 8.9 
Stomach 936 283 206 77 69.7 ± 10.5 
Head and neck 868 195 143 52 64.4 ± 12.8 
Breast 957 181 180 57.7 ± 12.9 
Liver 333 112 88 24 69.5 ± 9.4 
Pancreasb 215 60 34 26 71.3 ± 8.6 
Skin 460 42 23 19 61.5 ± 21.8 
Brain 191 40 24 16 56.7 ± 15.0 
Ovary 311 37 37 59.3 ± 13.7 
Endometrium 239 37 37 61.8 ± 12.2 
Biliary tract 160 30 20 10 69.0 ± 6.9 
Esophagus 118 27 23 67.9 ± 8.9 
Kidney 195 27 11 16 62.7 ± 14.1 
Soft tissue 206 24 17 58.3 ± 22.1 
Uterus cervix 413 15 15 51.3 ± 15.9 
Bone 104 29.9 ± 21.6 
Small intestine 52 64.0 ± 13.5 
GISTc 42 38 22 16 63.4 ± 12.7 
NET/NECd 41 31 19 12 64.7 ± 10.9 
Lymphomae 62.7 ± 9.1 
Others 598 59 27 32 55.3 ± 17.1 
Total 8,549 2,501 1,417 1,084 65.7 ± 12.7 
Number of casesGender (n)Age (y.o.)
Cancer origin/typeResected casesaCases in this studyMaleFemaleMean ± SD
Colon 1,347 826 502 324 66.3 ± 11.9 
Lung 760 420 247 173 69.4 ± 8.9 
Stomach 936 283 206 77 69.7 ± 10.5 
Head and neck 868 195 143 52 64.4 ± 12.8 
Breast 957 181 180 57.7 ± 12.9 
Liver 333 112 88 24 69.5 ± 9.4 
Pancreasb 215 60 34 26 71.3 ± 8.6 
Skin 460 42 23 19 61.5 ± 21.8 
Brain 191 40 24 16 56.7 ± 15.0 
Ovary 311 37 37 59.3 ± 13.7 
Endometrium 239 37 37 61.8 ± 12.2 
Biliary tract 160 30 20 10 69.0 ± 6.9 
Esophagus 118 27 23 67.9 ± 8.9 
Kidney 195 27 11 16 62.7 ± 14.1 
Soft tissue 206 24 17 58.3 ± 22.1 
Uterus cervix 413 15 15 51.3 ± 15.9 
Bone 104 29.9 ± 21.6 
Small intestine 52 64.0 ± 13.5 
GISTc 42 38 22 16 63.4 ± 12.7 
NET/NECd 41 31 19 12 64.7 ± 10.9 
Lymphomae 62.7 ± 9.1 
Others 598 59 27 32 55.3 ± 17.1 
Total 8,549 2,501 1,417 1,084 65.7 ± 12.7 

aTotal number of surgically resected cases during the study period.

bPancreas including ductal adenocarcinoma and intraductal papillary neoplasm (IPMC: 5, IPMA: 1).

cStomach: 22, colon: 5, small intestine: 9, esophagus: 1, omentum: 1.

dLung: 10, pancreas: 10, small intestine: 3, liver: 3, head and neck: 1, esophagus: 1, uterus cervix: 1, biliary tract: 1, and colon: 1.

eColon: 1, brain: 1, and bone: 1.

At the initial hospital visit, patients and their families filled out questionnaires concerning disease history, family history, and lifestyle aspects. The nurses reconfirmed the content of the questionnaires by conducting interviews for about 20 to 30 minutes with each patient. LS-associated cancers were defined to be limited to colorectal cancer, EC, renal pelvic and ureteral cancers, and small intestine cancer (13). The conditions of the revised Bethesda guidelines (14) were evaluated in cases with colorectal cancer. None of the patients or their families had been diagnosed with LS before study entry. The Institutional Review Board of the Shizuoka Cancer Center ethically approved this study (IRB no. 2019-24), and all procedures were conducted in accordance with the Helsinki Declaration. Written informed consent for this study was obtained from all participants prior to study entry.

WES of somatic and germline DNA

In this study, WES was performed to analyze somatic and germline DNA (27). Somatic DNA analyses included TMB, total exonic mutation numbers, mutations of BRAF, 4 MMR genes, and DNA polymerase proofreading genes (POLE and POLD1). Germline DNA analyses were limited to g.MMR genes to detect pathogenic variants and diagnose LS. A pair of DNA samples extracted from blood and surgically resected fresh-frozen cancer tissues were collected at each surgery and stored at −70°C. Fresh surgical specimens were diagnosed by a pathologist, and each tumor was histologically estimated to weigh >100 mg and contain >50% concentration tumor cells. DNA was extracted using a QIAmp Kit (QIAGEN) according to the manufacturer's instructions, with a process of proteinase K addition. WES was performed with 100 ng of template DNA using an Ion Torrent AmpliSeq Exome RDY Panel kit (Thermo Fisher Scientific) and the Ion Torrent Proton semiconductor DNA sequencer (Thermo Fisher Scientific) following the manufacturer's recommended protocol as written previously (5, 26, 27). The threshold of WES was set to 30%, and <30% of an allele ratio was not judged to be germline variants (28). Loss of heterozygosity (29) was analyzed in the LS-associated cancers with high TMB (hTMB; ≥10 mutations/Mb; refs. 10, 30). LOH analysis was performed using somatic copy-number alteration analysis using sequencing and SNP array data (SAAS-CNV; ref. 31). Pathogenicity of the detected variant was judged based on the consensus recommendation of American College of Medical Genetics (32). Variant of single-nucleotide substitution or small insertion/deletion was judged by using the ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar/). Null variants (nonsense, frameshift, canonical ± 1 or 2 splice sites, initiation codon, single or multiexon deletion) in a gene, where loss of function is a known mechanism of disease, were basically judged as pathogenic (32), also refereeing to the general population data (minor allele frequency ≥0.5% was judged as nonpathogenic), computational and predictive data, functional data, and so on.

Comparison of two approaches for detecting germline MMR variants

We compared the detection rates of LS (g.MMR variant carrier) of universal g.MMR gene sequencing using the above-mentioned ion-semiconductor sequencing (universal g.MMR sequencing approach; ref. 27) and g.MMR target sequencing coupled with multiplex ligation-dependent probe amplification (MLPA; hTMB-based g.MMR target sequencing approach). The latter approach was implemented in the cases of LS-related cancers (13) with high levels of TMB, MMR deficiency (dMMR) based on IHC, and an additional condition of negative somatic BRAF V600E mutation in cases of colorectal cancer.

For the hTMB-based g.MMR target sequencing approach, we defined a tumor with 10 mutations per Mb as hTMB, as per The Cancer Genome Atlas (TCGA; ref. 10). MMR IHC or MSI were recommended for all colorectal cancer patients with hTMB and without somatic BRAF mutation and were implemented when written informed consent was obtained. When either MMR IHC or MSI was revealed to be abnormal, target sequencing and MLPA for g.MMR genes were additionally examined after obtaining additional informed consent. For g.MMR target sequencing, genomic DNA was isolated from peripheral blood via a QIAamp DNA Blood Kit (QIAGEN). DNA was then amplified via PCR using oligonucleotide primers for MLH1, MSH2, MSH6, and PMS2, and then a library was prepared using the Nextera XT System (Illumina, Inc.). Samples were pooled for multiplexed sequencing, which was performed on MiSeq (Illumina) by a 150-bp paired-end sequence using the MiSeq Reagent Kit v2 (Illumina). Sequenced reads were mapped against reference sequences, and variants were identified by Genomics Workbench Software v7.5 (QIAGEN) and confirmed by Omixon Target Software (Omixon). All variants, except for polymorphism, detected by g.MMR target sequencing were again confirmed via Sanger sequencing. MLPA analysis was performed using the Salsa P003 kit for the MLH1 and MSH2 genes (MRC-Holland) and was performed according to the manufacturer's instructions. Amplification products were electrophoresed on the ABI 3130xl genetic analyzer (Applied Biosystems). Data analysis for detecting copy-number variations was performed using Coffalyser.net software (MRC-Holland).

IHC of the MMR protein

Thin-sliced sections (3-μm thickness) of FFPE cancer tissues from the surgical specimen were prepared for the evaluation of MMR protein expression (27). After deparaffinization with xylene for 15 minutes and stepwise de-xylene treatment with ethanol, heat treatment was applied using Epitope Retrieval Solution 2 (Leica Biosystems) at 95°C for 20 minutes to activate antigenicity. As a primary antibody against each MMR protein, the anti-hMLH1 antibody (Clone ES05, ×50 dilution, Dako), anti-hMSH2 antibody (Clone FE11, ×50 dilution, Dako), anti-hMSH6 antibody (Clone EP49, ×50 dilution, Dako), and anti-hPMS2 antibody (Clone EP51, ×25 dilution, Dako) were used. The secondary antibody was reacted at room temperature for 8 minutes using Bond Polymer Refine Detection (Leica Biosystems). Then, color was developed using diaminobenzidine (SIGMA) for 10 minutes at room temperature.

When the MMR protein was not expressed or severely repressed in the cancer tissue, contrasting the diffuse expression in the nonneoplastic tissue, the tumor was determined to have a negative MMR expression. This histologic evaluation was conducted by the expert pathologist (T. Oishi).

MSI analysis

MSI was analyzed after the acquisition of a separate written informed consent for the diagnosis of LS (27). MSI analysis was entrusted to Falco Holdings Co., Ltd. and performed using the MSI Analysis System, Version 1.2 (Promega Corporation) following the manufacturer's recommended protocol. MSI status was evaluated by comparing normal and tumor DNAs using 5 nearly monomorphic mononucleotide microsatellite markers (BAT25, BAT26, NR21, NR24, and MONO27). A high frequency of microsatellite instability (MSI-H) was defined when the tumor DNA demonstrated instability in 2 or more markers, whereas a low frequency of MSI (MSI-L) was defined when only one marker was instable and microsatellite stability when a null marker showed instability.

Gene-expression profiling

Gene-expression profiling was performed to compare the MLH1 gene expressions of BRAF V600E wild-type and BRAF V600E-mutant-hTMB colorectal cancers. Gene-expression profiling was analyzed by following the previously described protocol (33).

Statistical analysis

Statistical analyses were performed using JMP ver.12.2.0 statistical software (SAS Institute Japan Ltd.). A Fisher exact test or χ2 test was used to assess the categorical variables. Student t test or Mann–Whitney U test were used to analyze the continuous variables. A probability value <0.05 was considered statistically significant.

TMB in various human cancers

Depth of coverage in the current WES for somatic DNA was 121.7 ± 850.9 (average ± SD) by ion-proton semiconductor sequencing. TMB values and hTMB ratios by cancer type are shown in Table 2. The average TMB levels were highest in ECs (65.3 ± 155.0) followed by skin cancers (13.4 ± 39.4), small intestine cancers (8.9 ± 14.8), colorectal cancers (7.8 ± 24.2), and gastric cancers (7.2 ± 14.0). Meanwhile, the ratio of hTMB (≥ 10 mutations/Mb) was highest in ECs (40.5%), followed by small intestine cancers (28.6%), skin cancers (16.7%), gastric cancers (14.5%), and lung cancers (11.7%). The average TMB of LS-associated cancers (10.3 ± 41.0) was significantly higher than that in other cancer types (4.0 ± 10.0; P < 0.0001). The TMB of lung cancers (4.3 ± 7.6) and neuroendocrine tumor/carcinomas (3.6 ± 4.1) was lower than the average of total cancers; nevertheless, the hTMB ratio was at a high level in these cancers (11.7% and 12.9%, respectively).

Table 2.

TMB score and rate of hTMB in 2,501 cancer cases.

Cancer typenTMB (mean ± SD)P valueRatio of hTMBa % (n)P value
Lynch syndrome–associated cancer types      
 Colon 826 7.8 ± 24.2   7.0% (58)   
 Endometrium 37 65.3 ± 155.0 10.3 ± 41.0B  40.5% (15) 8.6%D  
 Small intestine 8.9 ± 14.8   28.6% (2)   
 Ureter —   —   
Other cancer types      
 Lung 420 4.3 ± 7.6   11.7% (49)   
 Stomach 283 7.2 ± 14.0   14.5% (41)   
 Head and neck 195 3.2 ± 4.2   4.1% (8)   
 Breast 181 1.5 ± 1.8   0.6% (1)   
 Liver 112 4.2 ± 5.4   1.8% (2)   
 Pancreas 60 0.5 ± 0.7     
 Skin 42 13.4 ± 39.4  <0.0001B, C 16.7% (7)  0.145D, E 
 Brain 40 1.5 ± 0.9     
 Ovary 37 2.5 ± 5.1 4.0 ± 10.0C  2.7% (1) 7.0%E  
 Biliary tract 30 2.2 ± 2.1     
 Esophagus 27 2.6 ± 1.5     
 Kidney 27 1.6 ± 0.7     
 Soft tissue 24 1.3 ± 1.4     
 Uterus cervix 15 3.2 ± 4.3   7.7% (1)   
 Bone 1.2 ± 1.1     
 GIST 38 0.7 ± 0.3     
 NET/NEC 31 3.6 ± 4.1   12.9% (4)   
 Lymphoma 4.6 ± 2.9     
Others 59 0.8 ± 0.9     
Total 2,501 6.2 ± 25.7   7.6% (189)   
Cancer typenTMB (mean ± SD)P valueRatio of hTMBa % (n)P value
Lynch syndrome–associated cancer types      
 Colon 826 7.8 ± 24.2   7.0% (58)   
 Endometrium 37 65.3 ± 155.0 10.3 ± 41.0B  40.5% (15) 8.6%D  
 Small intestine 8.9 ± 14.8   28.6% (2)   
 Ureter —   —   
Other cancer types      
 Lung 420 4.3 ± 7.6   11.7% (49)   
 Stomach 283 7.2 ± 14.0   14.5% (41)   
 Head and neck 195 3.2 ± 4.2   4.1% (8)   
 Breast 181 1.5 ± 1.8   0.6% (1)   
 Liver 112 4.2 ± 5.4   1.8% (2)   
 Pancreas 60 0.5 ± 0.7     
 Skin 42 13.4 ± 39.4  <0.0001B, C 16.7% (7)  0.145D, E 
 Brain 40 1.5 ± 0.9     
 Ovary 37 2.5 ± 5.1 4.0 ± 10.0C  2.7% (1) 7.0%E  
 Biliary tract 30 2.2 ± 2.1     
 Esophagus 27 2.6 ± 1.5     
 Kidney 27 1.6 ± 0.7     
 Soft tissue 24 1.3 ± 1.4     
 Uterus cervix 15 3.2 ± 4.3   7.7% (1)   
 Bone 1.2 ± 1.1     
 GIST 38 0.7 ± 0.3     
 NET/NEC 31 3.6 ± 4.1   12.9% (4)   
 Lymphoma 4.6 ± 2.9     
Others 59 0.8 ± 0.9     
Total 2,501 6.2 ± 25.7   7.6% (189)   

ahTMB is defined as >10 of TMB in tumor DNA.

Demographics of patients with colorectal cancer and EC with hTMB

Patient demographics were compared between the hTMB and low-TMB (lTMB) groups for patients with 2 LS-associated cancer types, colorectal cancer (Table 3) and EC (Table 4), as the current study analyzed only one patient with hTMB small intestine cancer and null patients with renal pelvis or ureter cancer. hTMB cancer patients accounted for 7.0% of the 826 colorectal cancer patients and 40.5% of the 37 EC patients, and their incidences were significantly different (P < 0.0001). Average age of the hTMB cancer patients was almost equal to that of the lTMB cancer patients with colorectal cancer (65.1 vs. 66.1 years; Table 3), but the same comparison was nearly significant in patients with EC (57.5 vs. 64.7 years, P = 0.076; Table 4). The incidence of the patients meeting the revised Bethesda criteria was significantly higher in the hTMB colorectal cancer group (34.5%) than in the lTMB group (21.7%; P = 0.024). Also, in the ECs, the incidence of either a personal or family history of LS-associated cancers was higher in the hTMB cancer group than in the lTMB cancer group; however, this was not statistically significant (40.0% vs. 18.2%, P = 0.142; Table 4). TMB levels were higher in the hTMB ECs (157.7) than the hTMB colorectal cancers (67.6); however, this was not a significant difference (P = 0.924, by Mann–Whitney U).

Table 3.

Demographics and tumor findings in colorectal cancer cases based on the level of TMB.

hTMB group (n = 58)lTMB group (n = 768)P value
Patient findings    
 Proportion 7.0% 93.0%  
 Gender (male/female) 30/28 472/296 0.143 
 Age (y.o.) 65.1 ± 15.5 66.1 ± 11.6 0.537 
 Brinkman indexa 384.9 ± 521.2 398.1 ± 481.6 0.841 
 Rev. Bethesda criteriab 34.5% 21.6% 0.024 
Tumor findings    
 TMB (/mb) 67.6 ± 67.3 3.3 ± 1.4 <0.0001 
 Number of mutationc 2,221.8 ± 2,194.3 108.3 ± 44.9 <0.0001 
 POLE or POLD1 exonuclease domain mutant 17.2% (10/58) 0% (0/768) <0.0001 
 Average MLH1 expression BRAF V600E mutant −2.0 ± 0.8 — <0.0001 
 BRAF V600E wild −0.1 ± 0.8 —  
 MSI  66.7% (22/33) NA — 
 dMMR  69.7% (23/33) NA — 
hTMB group (n = 58)lTMB group (n = 768)P value
Patient findings    
 Proportion 7.0% 93.0%  
 Gender (male/female) 30/28 472/296 0.143 
 Age (y.o.) 65.1 ± 15.5 66.1 ± 11.6 0.537 
 Brinkman indexa 384.9 ± 521.2 398.1 ± 481.6 0.841 
 Rev. Bethesda criteriab 34.5% 21.6% 0.024 
Tumor findings    
 TMB (/mb) 67.6 ± 67.3 3.3 ± 1.4 <0.0001 
 Number of mutationc 2,221.8 ± 2,194.3 108.3 ± 44.9 <0.0001 
 POLE or POLD1 exonuclease domain mutant 17.2% (10/58) 0% (0/768) <0.0001 
 Average MLH1 expression BRAF V600E mutant −2.0 ± 0.8 — <0.0001 
 BRAF V600E wild −0.1 ± 0.8 —  
 MSI  66.7% (22/33) NA — 
 dMMR  69.7% (23/33) NA — 

aBrinkman index = number of cigarettes smoked/day × number of years smoked.

bIncidence of the patients meeting the revised Bethesda criteria; dMMR: mismatch-repair deficiency.

cMutation number within the coding region of tumor DNA.

Table 4.

Demographics and tumor findings in endometrial cancer cases based on the level of TMB.

hTMB group (n = 15)lTMB group (n = 22)P value
Patient findings   
 Proportion 40.5% 59.5%  
 Age (y.o) 57.5 ± 13.3 64.7 ± 10.8 0.076 
 Brinkman indexa 19.3 ± 64.6 26.8 ± 63.1 0.728 
 Incidence of personal or family history of LS-associated cancer (n40.0% (6) 18.2% (4) 0.142 
Tumor findings    
 TMB (/mb) 157.7 ± 215.4 2.3 ± 1.7 0.002 
 Number of mutationb 5,213.8 ± 7,115.7 75.7 ± 57.7 0.002 
 POLE or POLD1 exonuclease domain mutant 46.7% (7/15) 0% (0/22) 0.001 
 MSI 6/12 (50.0%) NA — 
 dMMR 9/12 (75.0%) NA — 
hTMB group (n = 15)lTMB group (n = 22)P value
Patient findings   
 Proportion 40.5% 59.5%  
 Age (y.o) 57.5 ± 13.3 64.7 ± 10.8 0.076 
 Brinkman indexa 19.3 ± 64.6 26.8 ± 63.1 0.728 
 Incidence of personal or family history of LS-associated cancer (n40.0% (6) 18.2% (4) 0.142 
Tumor findings    
 TMB (/mb) 157.7 ± 215.4 2.3 ± 1.7 0.002 
 Number of mutationb 5,213.8 ± 7,115.7 75.7 ± 57.7 0.002 
 POLE or POLD1 exonuclease domain mutant 46.7% (7/15) 0% (0/22) 0.001 
 MSI 6/12 (50.0%) NA — 
 dMMR 9/12 (75.0%) NA — 

aBrinkman index = number of cigarettes smoked/day × number of years smoked.

bMutation number within the coding region of tumor DNA; dMMR: mismatch-repair deficiency.

BRAF mutation, MMR expression, and MSI in LS-associated hTMB cancers

IHC of 4 MMR proteins (MLH1, MSH2, MSH6, and PMS2) and MSI analysis were conducted for 47 hTMB cancers (34 colorectal cancers, 1 small intestine cancer, and 12 ECs) but not performed in 24 hTMB colorectal cancers (BRAF V600E mutation positive: 21, uncontacted: 2, and disagreed: 1) and 3 hTMB ECs (disagreed in all 3). Hence, 6 cases of BRAF wild hTMB colorectal cancer and 3 cases of hTMB EC were candidates for g.MMR target sequencing, but did not undergo the gene examination (Supplementary Tables S1 and S2).

Deficiency in MMR expression was confirmed in 69.7% (23/33) of hTMB colorectal cancers, of which 95.7% (22/23) showed high levels of MSI (Table 3), except for one MSH6-repressed colorectal cancer [in germline MSH6 (g.MSH6) p.W413S (c.1238G > C) variant] with low levels of MSI (C-25 in Supplementary Table S1). MLH1 gene-expression levels were significantly lower in hTMB colorectal cancers with BRAF V600E mutations (−2.0 ± 0.8) than those without (−0.1 ± 0.8; P < 0.0001; Table 3).

Meanwhile, dMMR was recognized in 75% (9/12) of hTMB ECs, of which 66.7% (6/9) were positive for MSI (Table 4). All 4 BRAF mutations recognized in the 3 hTMB ECs were not registered in ClinVar and HGMD, and unknown for pathogenicity. Discrepancy of MMR IHC and MSI was recognized in 25% (3 of 12) of hTMB ECs; g.MSH6 variants were confirmed in these 3 EC cases with repressed MSH6 expression but without MSI (E-4, 8, and 9 in Supplementary Table S2). The prevalence of MSI and dMMR was not significantly different between the colorectal cancers and ECs (P = 0.31 and P = 0.73, respectively; Tables 3 and 4).

Detection of Lynch syndrome

Of the 2,501 patients with surgically resected cancers, the g.MMR variant was detected in 16 (0.6%) patients; all 16 patients had LS-related cancers [9 cases of colorectal cancer (1.1%; 9/826), 6 cases of EC (16.2%; 6/37), and one case of small intestine cancer (14.3%; 1/7); Table 5]. A g.MLH1 variant (exon 5 deletion) could not be detected either by universal sequencing or by the TMB-based method, because the variant was a large deletion, undetectable by ion-semiconductor sequencing, and this colon cancer did not have hTMB (TMB = 8.5). However, the detection rate (sensitivity) of the g.MMR variant largely differed between universal g.MMR sequencing (56.3%, 9/16) and TMB-based target g.MMR sequencing (81.3%, 13/16; P = 0.127, Table 5), as in the following.

Table 5.

Total g.MMR variants detected in 2,501 cancer patients: Comparison of strategies for g.MMR variant detection.

Detection of g.MMR variant
Case No.g.MMR variantUniversal sequencingTMB-based-targeted sequencingOthers
Colorectal cancer (n = 826) C-51 MLH1 (p.E23*) Yes Yes  
 C-48 MLH1 (IVS6+2T>C) Yes Yes  
 C-47 MLH1 (p.A681T) Yes Yes  
  MLH1 (exon5 del)a No No Yes 
 C-12 MLH1 (exon13–19 del) No Yes  
 C-29 MSH2 (exon7 dup) No Yes  
 C-9 MSH6 (p.E376*)b Yes Yes  
 C-25 MSH6 (p.W413S)c Yes No  
 C-16 MSH6 (p.R482*) Yes Yes  
   Frequency of detected g.MMR variants 
   0.7% (6/826) 0.85% (7/826) 0.1% (1/826) 
Endometrial cancer (n = 37) E-14 MSH2 (p.K235Ffs*9)b No Yes  
 E-10 MSH2 (p.E290Dfs*2) No Yes  
 E-4 MSH6 (p.E376*)b Yes Yes  
 E-8 MSH6 (p.W413S)c Yes Yes  
 E-9 MSH6 (p.F1088Sfs*2) No Yes  
  MSH6 (IVS6+1G>T)d Yes No  
   Frequency of detected g.MMR variants 
   8.1% (3/37) 13.5% (5/37)  
Small intestine cancer (n = 7) S-1 PMS2 (p.I18Sfs*34) No Yes  
   Frequency of detected g.MMR variants 
   0% (0/7) 14.3% (1/7) N/A 
Incidence of detected g.MMR variant (n = 2,501)   0.4% (9/2,501) 0.5% (13/2,501) 0.04% (1/2,501) 
Detection rate of g.MMR variant (n = 16)   56.3% (9/16)e 81.3% (13/16)e 6.3% (1/16) 
Detection of g.MMR variant
Case No.g.MMR variantUniversal sequencingTMB-based-targeted sequencingOthers
Colorectal cancer (n = 826) C-51 MLH1 (p.E23*) Yes Yes  
 C-48 MLH1 (IVS6+2T>C) Yes Yes  
 C-47 MLH1 (p.A681T) Yes Yes  
  MLH1 (exon5 del)a No No Yes 
 C-12 MLH1 (exon13–19 del) No Yes  
 C-29 MSH2 (exon7 dup) No Yes  
 C-9 MSH6 (p.E376*)b Yes Yes  
 C-25 MSH6 (p.W413S)c Yes No  
 C-16 MSH6 (p.R482*) Yes Yes  
   Frequency of detected g.MMR variants 
   0.7% (6/826) 0.85% (7/826) 0.1% (1/826) 
Endometrial cancer (n = 37) E-14 MSH2 (p.K235Ffs*9)b No Yes  
 E-10 MSH2 (p.E290Dfs*2) No Yes  
 E-4 MSH6 (p.E376*)b Yes Yes  
 E-8 MSH6 (p.W413S)c Yes Yes  
 E-9 MSH6 (p.F1088Sfs*2) No Yes  
  MSH6 (IVS6+1G>T)d Yes No  
   Frequency of detected g.MMR variants 
   8.1% (3/37) 13.5% (5/37)  
Small intestine cancer (n = 7) S-1 PMS2 (p.I18Sfs*34) No Yes  
   Frequency of detected g.MMR variants 
   0% (0/7) 14.3% (1/7) N/A 
Incidence of detected g.MMR variant (n = 2,501)   0.4% (9/2,501) 0.5% (13/2,501) 0.04% (1/2,501) 
Detection rate of g.MMR variant (n = 16)   56.3% (9/16)e 81.3% (13/16)e 6.3% (1/16) 

aA variant of MLH1 (exon 5 del) was not detected either by the hTMB approach or by the universal g.MMR sequencing approach as the variant was a large deletion and the TMB was 8.5 (<10/Mb). This variant was detected in cases of family history, MSI, and MMR IHC using the MLPA method.

bPathogenicity of the two newly detected variants was evaluated on the basis of ACMG guideline.

cPathogenicity of one probable pathogenic variant was evaluated on the basis of ACMG guideline.

dAn MSH6 (IVS6+1G>T) variant was detected only by universal g.MMR sequencing but not by the TMB-based approach, as the analyzed endometrial cancer was not developed via the MMR pathway. However, her previous colorectal cancer was positive for MSH6 deficiency and MSI.

eP = 0.13.

Universal g.MMR sequencing

Depth of coverage in current WES for germline DNA was 122.1 ± 860.7 (average ± SD). Universal exome sequencing detected 9 g.MMR variants (0.4%) in the 2,501 cancer cases: 6 colorectal cancers and 3 ECs (Table 5). These variants were all single-nucleotide substitutions within the exome and a few bases into the intron site. A case of EC with an MSH6 (IVS6+1G>T) variant (Table 5), which was detected via universal g.MMR sequencing (ion-proton method), could not be detected via TMB-based g.MMR target sequencing, as this EC was negative for MMR protein repression and MSI. Meanwhile, this EC patient previously developed a colorectal cancer showing MSH6 protein deficiency and high-level MSI.

Target sequencing and MLPA of g.MMR in cases with hTMB, BRAF wild-type, and MMR-deficient cancers

The target sequencing and MLPA of g.MMR genes was performed on d.MMR, TMB-high, and LS-associated cancers as well as with an additional condition of wild-type BRAF in cases of colorectal cancer. Depth of coverage for g.MMR target sequencing was 7,143.4 ± 2,944.0 (average ± SD). This strategy detected LS in 13 cases (0.5%) in the 2,501 cancer cases: 8 colorectal cancers, 5 ECs, and 1 small intestine cancer (Table 5). This method detected a short to large range of deletions and duplications, such as MLH1 (exon13–19 del), MSH2 (p.E290Dfs*2), PMS2 (p.I18Sfs*34), and MSH2 (exon7 dup), which could not be discovered with ion-semiconductor sequencing. However, this TMB-based approach missed 2 LS patients, as an EC that developed in 1 patient [MSH6 (IVS6+1G>T)] was not associated with the MMR pathway, and a colorectal cancer developed in another patient showed moderate levels of TMB (8.5; Table 5).

Germline MMR sequencing was performed for 30 hTMB LS-associated cancers with repressed or lost MMR protein expression (20 colorectal cancers, 1 small intestine cancer, 9 ECs; Supplementary Tables S1 and S2). Six cases of BRAF V600E-mutant hTMB colorectal cancer were exceptionally analyzed for g.MMR sequencing, and one pathogenic variant (MSH6 p.W413S, C-25 in Supplementary Table S1) was detected. Within the remaining 24 hTMB cases sequenced for g.MMR, the pathogenic variant was detected in 13 cases (54.2%; 7 colorectal cancers, 1 small intestine cancer, and 5 ECs; Supplementary Tables S1 and S2). The results of MMR IHC and g.MMR sequencing were largely concordant, except for one case (C-44) that had a benign MLH1 variant (p.Q701K; Supplementary Table S1); otherwise, MLH1 lost cancers that can be affected by MLH1 methylation.

In 14 MMR-deficient hTMB cancers in the LS patients (Supplementary Tables S1 and S2), somatic deleterious MMR gene mutations were detected in 9 cancers (64.3%: 6 of 8 colorectal cancers, 1 of 1 small intestine cancer, and 2 of 5 of ECs), by either of point mutation, deletion, frameshift, or LOH.

Pathogenic evaluation of novel g.MMR variants

In this study, 2 novel germline variants [MSH2 (p.K235Ffs*9) and MSH6 (p.E376*); Table 5] were deemed to be pathogenic (PVS1) based on the ACMG guidelines (32), as they were thought to create nonsense mutations. One probable pathogenic variant (MSH6 p.W413S), which was seen in one colorectal cancer case (C-25 in Supplementary Table S1) and one EC case (E-8 in Supplementary Table S2), was deemed to be likely pathogenic, as per the ACMG guidelines, and its repressed MSH6 expression.

Today, TMB has been garnering increasing attention in cancer precision medicine due to its critical role as a key biomarker in predicting the response to immune-checkpoint inhibitors (1–3). Cristescu and colleagues (34) defined hTMB as >102.5 mutations/exome for predicting good responders to immune-checkpoint inhibitors; however, the Cancer Genome Atlas (10, 30) defined hTMB as ≥10 mutations/Mb for detecting MSI-high and POL-mutant cancers (LS detection rate: 81.3%). If we set hTMB as ≥20 mutations/Mb, ≥50 mutations/Mb, or ≥100 mutations/Mb, detection rate of LS was 75.0% (12/16), 25.0% (4/16), or 6.3% (1/16), respectively. Hence, excessively high cutoff level of TMB will miss LS patients. Actually, most of our hTMB cancer groups (TMB ≥10) exhibited hypermutators and ultramutators and shared several demographic features, such as frequent LS-associated cancers and skin cancer (Table 2), modestly increasing cases meeting the revised Bethesda criteria (Table 3), and cases with personal and family histories of LS-related cancers (Table 4). Although TMB is affected by mutagenic exposures, such as smoking for lung cancer (35, 36) and ultraviolet light for skin cancer (5, 35), current results demonstrated no correlation between Brinkman index and TMB in cases with LS-related cancer (Tables 3 and 4). Aging also increases the level of TMB mildly (37), but it rather showed negative correlation in current cases of endometrial cancer (Table 4), probably due to the high proportion of POL-mutant cancers (38). To comprehend these findings, the hTMB LS-associated cancer patients were thought to be good candidates for LS screening, and we detected LS in 54.2% of the patients tested with target g.MMR sequencing and MLPA. The sensitivity of this strategy (81.3%) was superior to universal g.MMR sequencing using the ion-semiconductor method (56.3%), although this difference did not reach statistical significance (Table 5).

Formerly, the diagnosis of LS had been made based on the selection of high-risk individuals based on the patients' demographics and tumor histologic findings, according to the Amsterdam II criteria or revised Bethesda guidelines (14). However, due to its low detection rate, the Evaluation of Genomic Application in Practice and Prevention (EGAPP) working group does not currently recommend using this strategy for LS identification (39). In the past decade, universal screening has been attempted for LS-related cancer patients or those with age-specific conditions, specifically using MSI analysis and/or MMR IHC in their tumors. MMR IHC seems to have some advantage over MSI, as cancers with MSH6 or PMS2 mutations have shown repressed protein expression but sometimes lack MSI (refs. 40, 41; Supplementary Table S2: E-4, 8, and 9 in the current study). However, such universal screening is not fully diffused, even in the USA, and is limited to several cancer centers and some departments at core hospitals (42). In familial cancer clinics, blood relatives are simply diagnosed by examining known g.MMR variants. However, new LS probands need to be diagnosed based on deleterious g.MMR variants, with prior screening of an MMR deficiency or microsatellite instability in their cancers, besides with an additional condition of MLH1 promoter methylation or BRAF mutation in cases of colorectal cancers (43, 44).

Major cancer centers in the USA have demonstrated pathogenic g.MMR variants in 0.5% (45) to 0.7% (6) of advanced cancer patients who underwent clinical sequencing for precision medicine. The universal g.MMR sequencing seemed to be most sensitive as it directly determines g.MMR genes; however, several issues appeared in this study. First, short-read NGS analysis poorly detects large indels, especially when using ion-semiconductor sequencing (46). As described in Table 5, ion-semiconductor sequencing for g.MMR genes detected only single-nucleotide variants but missed large deletions and duplications (6 of 16 variants, 37.5%) that were detectable via target sequencing using the MLPA method. To date, WES has dramatically improved the diagnostic yield; however, WES cannot fully delineate certain DNA changes (e.g., structural rearrangement, copy-number variants, tandem-repeat expansions, and variants within the intronic and regulatory regions; refs. 47, 48). Whole-genome sequencing or paired genomic and transcriptomic NGS can solve these issues (48). Recent advance in long-read sequencing technologies, such as single-molecule real-time sequencing developed by Pacific Biosciences and nanopore sequencing by Oxford Nanopore Technologies, enable to detect large genetic alterations (1 bp to 1 million bp) and cover the role of MLPA in addition to NGS, but with high cost and high error rate (5%–15%; ref. 49). Hence, within the context of precision medicine, g.MMR sequencing with NGS alone is not optimal for LS detection and should be coupled with TMB. In the place of TMB, MSI can be another option. However, as previously mentioned, universal MSI analysis misses g.MSH6 and g.PMS2 variants that are often seen in ECs (50) and could be detected using the hTMB-based approach (Supplementary Table S2). The setting of hTMB can be adjusted, as current condition (hTMB: ≥10) missed one g.MLH1 variant with moderate levels of TMB (8.5; Table 5). The BRAF wild-type condition in colorectal cancer cases is effective and recommended in the guidelines for detecting LS (13, 44); however, a BRAF mutation does not always indicate the existence of MLH1 methylation (4) and can also occur in LS cancer patients (C-25 in Supplementary Table S1). Even with MLH1 hypermethylation, LS cannot be completely denied (51). The last pitfall is that not all tumors in LS patients develop through a mismatch-repair pathway, as shown by an EC case with a g.MSH6 IVS6+1G>T variant (Table 5), which could only be detected via universal g.MMR sequencing and not by either MSI, TMB, or MMR IHC. When predicted findings are obtained in cases with multiple LS-associated cancers, additional MMR IHC should be considered.

In precision medicine, germline DNA sequencing is not always obtained upon the first step, but tumor DNA sequencing is usually done. Hence, the prediction of germline-derived mutation has been pursued in somatic DNA information for detecting hidden inherited cancers (28). This process is routinely done in precision medicine prior to genetic counseling and further attempts at family surveillance. Mandelker and colleagues (28) compared the mutations of somatic and germline DNAs from inherited cancer-associated genes in 17,152 unselected cancer cases (MSK-IMPACT), predicting germline-derived somatic mutations with high sensitivity. They refined their candidates for the germline test into a group with 62.7% sensitivity (653 true germline variants out of 1,042 candidate cancer cases) based on variant allele frequency, minor allele frequency, patient's age, and cancer type. However, by current ion-semiconductor sequencing, only 64.3% (9/14) of LS patients demonstrated somatic mutations in the corresponding MMR genes (Supplementary Tables S1 and S2), indicating not only a lower positive predictive value but a low sensitivity to detect LS patients. Undetected somatic MMR alteration can be explained by methylation of MLH1 (52), MSH2 (53), and/or deletion of EPCAM (54). We also need to bear in mind that some of structural genomic rearrangement cannot be detected by NGS analysis (55). In this sense, TMB-based target g.MMR sequencing is thought to be an efficient strategy for detecting LS in the setting of precision medicine. When cancer DNA demonstrates high levels of TMB, MMR IHC should be applied after checking for BRAF wild status, followed by a germline examination for the corresponding MMR gene, conducted through genetic counseling.

The current study included several limitations. The authors presented all registered cases during the study period (2014–2017), but the studied cancer patients could be divided by their entry conditions: patients with surgically resectable cancer and those with tumors large enough to receive 5 × 5 × 5 mm3 study material without disturbing the pathologic diagnosis. Therefore, the number of cases entered by cancer type was not consistent among various cancer types (Table 1). TMB-high cancers are biologically less aggressive (10, 56), current cases were diagnosed within the operable stage and therefore may have higher incidence of high MSI and hTMB than entire cases of colorectal and ECs. Universal g.MMR sequencing was done using the ion-semiconductor sequencing method, which is useful for single-nucleotide substitutes and short indels, but may have a drawback in its detection of a large range of nucleotide alterations (46). The evaluation of pathogenicity for each germline variant was conducted based on the HGMD and ClinVar databases during the study period and is changeable in the future.

In conclusion, the current study demonstrated the critical role of TMB in detecting LS by its increasing sensitivity to simple universal NGS sequencing within the setting of cancer precision medicine. This strategy can also detect PPAP (11, 57) and constitutional mismatch-repair deficiency syndrome (CMMRD; ref. 58) effectively. By further applying somatic MLPA for MMR genes, patients with Lynch-like syndrome (59, 60) can be confirmed. TMB-based cancer sequencing is also recommended due to its treatment effects, such as predicting the effect of immune-checkpoint inhibitors (3, 61).

No disclosures were reported.

Y. Kiyozumi: Conceptualization, data curation, methodology, and writing–original draft. H. Matsubayashi: Conceptualization, data curation, formal analysis, methodology, writing–original draft, writing–review and editing. S. Higashigawa: Formal analysis. Y. Horiuchi: Formal analysis. N. Kado: Resources. Y. Hirashima: Resources. A. Shiomi: Resources. T. Oishi: Resources and formal analysis. S. Ohnami: Formal analysis and investigation. K. Ohshima: Formal analysis and investigation. K. Urakami: Formal analysis and investigation. T. Nagashima: Data curation and formal analysis. K. Yamaguchi: Supervision and project administration.

This study is economically supported by Shizuoka Prefecture.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Rizvi
H
,
Sanchez-Vega
F
,
La
K
,
Chatila
W
,
Jonsson
P
,
Halpenny
D
, et al
Molecular determinants of response to anti-programmed cell death (PD)-1 and anti-programmed death-ligand 1 (PD-L1) blockade in patients with non-small-cell lung cancer profiled with targeted next-generation sequencing
.
J Clin Oncol
2018
;
36
:
633
41
.
2.
Carbone
DP
,
Reck
M
,
Paz-Ares
L
,
Creelan
B
,
Horn
L
,
Steins
M
, et al
First-line nivolumab in stage IV or recurrent non-small-cell lung cancer
.
N Engl J Med
2017
;
376
:
2415
26
.
3.
Gargiulo
P
,
Della Pepa
C
,
Berardi
S
,
Califano
D
,
Scala
S
,
Buonaguro
L
, et al
Tumor genotype and immune microenvironment in POLE-ultramutated and MSI-hypermutated endometrial cancers: new candidates for checkpoint blockade immunotherapy?
Cancer Treat Rev
2016
;
48
:
61
8
.
4.
Nagasaka
T
,
Sasamoto
H
,
Notohara
K
,
Cullings
HM
,
Takeda
M
,
Kimura
K
, et al
Colorectal cancer with mutation in BRAF, KRAS, and wild-type with respect to both oncogenes showing different patterns of DNA methylation
.
J Clin Oncol
2004
;
22
:
4584
94
.
5.
Hatakeyama
K
,
Nagashima
T
,
Urakami
K
,
Ohshima
K
,
Serizawa
M
,
Ohnami
S
, et al
Tumor mutational burden analysis of 2,000 Japanese cancer genomes using whole exome and targeted gene panel sequencing
.
Biomed Res
2018
;
39
:
159
67
.
6.
Schrader
KA
,
Cheng
DT
,
Joseph
V
,
Prasad
M
,
Walsh
M
,
Zehir
A
, et al
Germline variants in targeted tumor sequencing using matched normal DNA
.
JAMA Oncol
2016
;
2
:
104
11
.
7.
Vogelstein
B
,
Papadopoulos
N
,
Velculescu
VE
,
Zhou
S
,
Diaz
LA
 Jr
,
Kinzler
KW
. 
Cancer genome landscapes
.
Science
2013
;
339
:
1546
58
.
8.
Yarchoan
M
,
Hopkins
A
,
Jaffee
EM
. 
Tumor mutational burden and response rate to PD-1 inhibition
.
N Engl J Med
2017
;
377
:
2500
1
.
9.
Deng
G
,
Bell
I
,
Crawley
S
,
Gum
J
,
Terdiman
JP
,
Allen
BA
, et al
BRAF mutation is frequently present in sporadic colorectal cancer with methylated hMLH1, but not in hereditary nonpolyposis colorectal cancer
.
Clin Cancer Res
2004
;
10
:
191
5
.
10.
Cancer Genome Atlas
N
. 
Comprehensive molecular characterization of human colon and rectal cancer
.
Nature
2012
;
487
:
330
7
.
11.
Palles
C
,
Cazier
JB
,
Howarth
KM
,
Domingo
E
,
Jones
AM
,
Broderick
P
, et al
Germline mutations affecting the proofreading domains of POLE and POLD1 predispose to colorectal adenomas and carcinomas
.
Nat Genet
2013
;
45
:
136
44
.
12.
Spier
I
,
Holzapfel
S
,
Altmuller
J
,
Zhao
B
,
Horpaopan
S
,
Vogt
S
, et al
Frequency and phenotypic spectrum of germline mutations in POLE and seven other polymerase genes in 266 patients with colorectal adenomas and carcinomas
.
Int J Cancer
2015
;
137
:
320
31
.
13.
Vasen
HF
,
Moslein
G
,
Alonso
A
,
Bernstein
I
,
Bertario
L
,
Blanco
I
, et al
Guidelines for the clinical management of Lynch syndrome (hereditary non-polyposis cancer)
.
J Med Genet
2007
;
44
:
353
62
.
14.
Pinol
V
,
Castells
A
,
Andreu
M
,
Castellvi-Bel
S
,
Alenda
C
,
Llor
X
, et al
Accuracy of revised Bethesda guidelines, microsatellite instability, and immunohistochemistry for the identification of patients with hereditary nonpolyposis colorectal cancer
.
JAMA
2005
;
293
:
1986
94
.
15.
Ricker
CN
,
Hanna
DL
,
Peng
C
,
Nguyen
NT
,
Stern
MC
,
Schmit
SL
, et al
DNA mismatch repair deficiency and hereditary syndromes in Latino patients with colorectal cancer
.
Cancer
123
:
3732
43
.
16.
Perez-Carbonell
L
,
Ruiz-Ponte
C
,
Guarinos
C
,
Alenda
C
,
Paya
A
,
Brea
A
, et al
Comparison between universal molecular screening for Lynch syndrome and revised Bethesda guidelines in a large population-based cohort of patients with colorectal cancer
.
Gut
2012
;
61
:
865
72
.
17.
Najdawi
F
,
Crook
A
,
Maidens
J
,
McEvoy
C
,
Fellowes
A
,
Pickett
J
, et al
Lessons learnt from implementation of a Lynch syndrome screening program for patients with gynaecological malignancy
.
Pathology
2017
;
49
:
457
64
.
18.
Ferguson
SE
,
Aronson
M
,
Pollett
A
,
Eiriksson
LR
,
Oza
AM
,
Gallinger
S
, et al
Performance characteristics of screening strategies for Lynch syndrome in unselected women with newly diagnosed endometrial cancer who have undergone universal germline mutation testing
.
Cancer
2014
;
120
:
3932
9
.
19.
Adar
T
,
Rodgers
LH
,
Shannon
KM
,
Yoshida
M
,
Ma
T
,
Mattia
A
, et al
Universal screening of both endometrial and colon cancers increases the detection of Lynch syndrome
.
Cancer
2018
;
124
:
3145
53
.
20.
Ju
JY
,
Mills
AM
,
Mahadevan
MS
,
Fan
J
,
Culp
SH
,
Thomas
MH
, et al
Universal Lynch syndrome screening should be performed in all upper tract urothelial carcinomas
.
Am J Surg Pathol
2018
;
42
:
1549
55
.
21.
Metcalfe
MJ
,
Petros
FG
,
Rao
P
,
Mork
ME
,
Xiao
L
,
Broaddus
RR
, et al
Universal point of care testing for Lynch syndrome in patients with upper tract urothelial carcinoma
.
J Urol
2018
;
199
:
60
5
.
22.
Dunlop
MG
,
Farrington
SM
,
Carothers
AD
,
Wyllie
AH
,
Sharp
L
,
Burn
J
, et al
Cancer risk associated with germline DNA mismatch repair gene mutations
.
Hum Mol Genet
1997
;
6
:
105
10
.
23.
Canard
G
,
Lefevre
JH
,
Colas
C
,
Coulet
F
,
Svrcek
M
,
Lascols
O
, et al
Screening for Lynch syndrome in colorectal cancer: are we doing enough?
Ann Surg Oncol
2012
;
19
:
809
16
.
24.
Moreira
L
,
Balaguer
F
,
Lindor
N
,
de la Chapelle
A
,
Hampel
H
,
Aaltonen
LA
, et al
Identification of Lynch syndrome among patients with colorectal cancer
.
JAMA
2012
;
308
:
1555
65
.
25.
Bruegl
AS
,
Djordjevic
B
,
Batte
B
,
Daniels
M
,
Fellman
B
,
Urbauer
D
, et al
Evaluation of clinical criteria for the identification of Lynch syndrome among unselected patients with endometrial cancer
.
Cancer Prev Res (Phila)
2014
;
7
:
686
97
.
26.
Yamaguchi
K
,
Urakami
K
,
Ohshima
K
,
Mochizuki
T
,
Akiyama
Y
,
Uesaka
K
, et al
Implementation of individualized medicine for cancer patients by multiomics-based analyses—the Project HOPE
.
Biomed Res
2014
;
35
:
407
12
.
27.
Kiyozumi
Y
,
Matsubayashi
H
,
Horiuchi
Y
,
Higashigawa
S
,
Oishi
T
,
Abe
M
, et al
Germline mismatch repair gene variants analyzed by universal sequencing in Japanese cancer patients
.
Cancer Med
2019
;
8
:
5534
43
.
28.
Mandelker
D
,
Donoghue
M
,
Talukdar
S
,
Bandlamudi
C
,
Srinivasan
P
,
Vivek
M
, et al
Germline-focused analysis of tumour-only sequencing: recommendations from the ESMO Precision Medicine Working Group
.
Ann Oncol
2019
;
30
:
1221
31
.
29.
Lawrence
MS
,
Stojanov
P
,
Polak
P
,
Kryukov
GV
,
Cibulskis
K
,
Sivachenko
A
, et al
Mutational heterogeneity in cancer and the search for new cancer-associated genes
.
Nature
2013
;
499
:
214
8
.
30.
Campbell
BB
,
Light
N
,
Fabrizio
D
,
Zatzman
M
,
Fuligni
F
,
de Borja
R
, et al
Comprehensive analysis of hypermutation in human cancer
.
Cell
2017
;
171
:
1042
56
.
e10
.
31.
Zhang
Z
,
Hao
K
. 
SAAS-CNV: a joint segmentation approach on aggregated and allele specific signals for the identification of somatic copy number alterations with next-generation sequencing data
.
PLoS Comput Biol
2015
;
11
:
e1004618
.
32.
Richards
S
,
Aziz
N
,
Bale
S
,
Bick
D
,
Das
S
,
Gastier-Foster
J
, et al
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
2015
;
17
:
405
24
.
33.
Ohshima
K
,
Hatakeyama
K
,
Nagashima
T
,
Watanabe
Y
,
Kanto
K
,
Doi
Y
, et al
Integrated analysis of gene expression and copy number identified potential cancer driver genes with amplification-dependent overexpression in 1,454 solid tumors
.
Sci Rep
2017
;
7
:
641
.
34.
Cristescu
R
,
Mogg
R
,
Ayers
M
,
Albright
A
,
Murphy
E
,
Yearley
J
, et al
Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy
.
Science
2018
;
362:eaar3593
.
35.
Alexandrov
LB
,
Nik-Zainal
S
,
Wedge
DC
,
Aparicio
SA
,
Behjati
S
,
Biankin
AV
, et al
Signatures of mutational processes in human cancer
.
Nature
2013
;
500
:
415
21
.
36.
Ono
A
,
Terada
Y
,
Kawata
T
,
Serizawa
M
,
Isaka
M
,
Kawabata
T
, et al
Assessment of associations between clinical and immune microenvironmental factors and tumor mutation burden in resected nonsmall cell lung cancer by applying machine learning to whole-slide images
.
Cancer Med
2020
;
9
:
4864
75
.
37.
Chalmers
ZR
,
Connelly
CF
,
Fabrizio
D
,
Gay
L
,
Ali
SM
,
Ennis
R
, et al
Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden
.
Genome Med
2017
;
9
:
34
.
38.
Hino
H
,
Shiomi
A
,
Kusuhara
M
,
Kagawa
H
,
Yamakawa
Y
,
Hatakeyama
K
, et al
Clinicopathological and mutational analyses of colorectal cancer with mutations in the POLE gene
.
Cancer Med
2019
;
8
:
4587
97
.
39.
Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group
. 
Recommendations from the EGAPP Working Group: genetic testing strategies in newly diagnosed individuals with colorectal cancer aimed at reducing morbidity and mortality from Lynch syndrome in relatives
.
Genet Med
2009
;
11
:
35
41
.
40.
Shia
J
. 
Immunohistochemistry versus microsatellite instability testing for screening colorectal cancer patients at risk for hereditary nonpolyposis colorectal cancer syndrome. Part I. The utility of immunohistochemistry
.
J Mol Diagn
2008
;
10
:
293
300
.
41.
Goel
A
,
Nagasaka
T
,
Spiegel
J
,
Meyer
R
,
Lichliter
WE
,
Boland
CR
. 
Low frequency of Lynch syndrome among young patients with non-familial colorectal cancer
.
Clin Gastroenterol Hepatol
2010
;
8
:
966
71
.
42.
Beamer
LC
,
Grant
ML
,
Espenschied
CR
,
Blazer
KR
,
Hampel
HL
,
Weitzel
JN
, et al
Reflex immunohistochemistry and microsatellite instability testing of colorectal tumors for Lynch syndrome among US cancer programs and follow-up of abnormal results
.
J Clin Oncol
2012
;
30
:
1058
63
.
43.
Rubenstein
JH
,
Enns
R
,
Heidelbaugh
J
,
Barkun
A
,
Clinical Guidelines
Committee
. 
American Gastroenterological Association Institute guideline on the diagnosis and management of Lynch syndrome
.
Gastroenterology
2015
;
149
:
777
82
;
quiz e16–7
.
44.
Giardiello
FM
,
Allen
JI
,
Axilbund
JE
,
Boland
CR
,
Burke
CA
,
Burt
RW
, et al
Guidelines on genetic evaluation and management of Lynch syndrome: a consensus statement by the US Multi-Society Task Force on colorectal cancer
.
Dis Colon Rectum
2014
;
57
:
1025
48
.
45.
Meric-Bernstam
F
,
Brusco
L
,
Daniels
M
,
Wathoo
C
,
Bailey
AM
,
Strong
L
, et al
Incidental germline variants in 1000 advanced cancers on a prospective somatic genomic profiling protocol
.
Ann Oncol
2016
;
27
:
795
800
.
46.
Loman
NJ
,
Misra
RV
,
Dallman
TJ
,
Constantinidou
C
,
Gharbia
SE
,
Wain
J
, et al
Performance comparison of benchtop high-throughput sequencing platforms
.
Nat Biotechnol
2012
;
30
:
434
9
.
47.
Ma
M
,
Ru
Y
,
Chuang
LS
,
Hsu
NY
,
Shi
LS
,
Hakenberg
J
, et al
Disease-associated variants in different categories of disease located in distinct regulatory elements
.
BMC Genomics
2015
;
16
Suppl 8
:
S3
.
48.
Gonorazky
HD
,
Naumenko
S
,
Ramani
AK
,
Nelakuditi
V
,
Mashouri
P
,
Wang
P
, et al
Expanding the boundaries of RNA sequencing as a diagnostic tool for rare Mendelian disease
.
Am J Hum Genet
2019
;
104
:
466
83
.
49.
Midha
MK
,
Wu
M
,
Chiu
KP
. 
Long-read sequencing in deciphering human genetics to a greater depth
.
Hum Genet
2019
;
138
:
1201
15
.
50.
Moller
P
,
Seppala
TT
,
Bernstein
I
,
Holinski-Feder
E
,
Sala
P
,
Gareth Evans
D
, et al
Cancer risk and survival in path_MMR carriers by gene and gender up to 75 years of age: a report from the Prospective Lynch Syndrome Database
.
Gut
2018
;
67
:
1306
16
.
51.
Kientz
C
,
Prieur
F
,
Clemenson
A
,
Joly
MO
,
Stachowicz
ML
,
Auclair
J
, et al
MLH1 promoter hypermethylation: are you absolutely sure about the absence of MLH1 germline mutation? About a new case
.
Fam Cancer
2020
;
19
:
11
4
.
52.
Westwood
A
,
Glover
A
,
Hutchins
G
,
Young
C
,
Brockmoeller
S
,
Robinson
R
, et al
Additional loss of MSH2 and MSH6 expression in sporadic deficient mismatch repair colorectal cancer due to MLH1 promoter hypermethylation
.
J Clin Pathol
2019
;
72
:
443
7
.
53.
Kang
Z
,
Zhu
Y
,
Zhang
QA
,
Dong
L
,
Xu
F
,
Zhang
X
, et al
Methylation and expression analysis of mismatch repair genes in extramammary Paget's disease
.
J Eur Acad Dermatol Venereol
2019
;
33
:
874
9
.
54.
Kang
SY
,
Park
CK
,
Chang
DK
,
Kim
JW
,
Son
HJ
,
Cho
YB
, et al
Lynch-like syndrome: characterization and comparison with EPCAM deletion carriers
.
Int J Cancer
2015
;
136
:
1568
78
.
55.
Morak
M
,
Steinke-Lange
V
,
Massdorf
T
,
Benet-Pages
A
,
Locher
M
,
Laner
A
, et al
Prevalence of CNV-neutral structural genomic rearrangements in MLH1, MSH2, and PMS2 not detectable in routine NGS diagnostics
.
Fam Cancer
2020
;
19
:
161
7
.
56.
Cancer Genome Atlas Research Network;
Kandoth
C
,
Schultz
N
,
Cherniack
AD
,
Akbani
R
,
Liu
Y
, et al
Integrated genomic characterization of endometrial carcinoma
.
Nature
2013
;
497
:
67
73
.
57.
Briggs
S
,
Tomlinson
I
. 
Germline and somatic polymerase epsilon and delta mutations define a new class of hypermutated colorectal and endometrial cancers
.
J Pathol
2013
;
230
:
148
53
.
58.
Carethers
JM
,
Stoffel
EM
. 
Lynch syndrome and Lynch syndrome mimics: the growing complex landscape of hereditary colon cancer
.
World J Gastroenterol
2015
;
21
:
9253
61
.
59.
Mensenkamp
AR
,
Vogelaar
IP
,
van Zelst-Stams
WA
,
Goossens
M
,
Ouchene
H
,
Hendriks-Cornelissen
SJ
, et al
Somatic mutations in MLH1 and MSH2 are a frequent cause of mismatch-repair deficiency in Lynch syndrome-like tumors
.
Gastroenterology
2014
;
146
:
643
6
-
e8
.
60.
Rodriguez-Soler
M
,
Perez-Carbonell
L
,
Guarinos
C
,
Zapater
P
,
Castillejo
A
,
Barbera
VM
, et al
Risk of cancer in cases of suspected Lynch syndrome without germline mutation
.
Gastroenterology
2013
;
144
:
926
32
e1
;
quiz e13–4
.
61.
Le
DT
,
Uram
JN
,
Wang
H
,
Bartlett
BR
,
Kemberling
H
,
Eyring
AD
, et al
PD-1 blockade in tumors with mismatch-repair deficiency
.
N Engl J Med
2015
;
372
:
2509
20
.