Supplementary Online Content Hansford S, Kaurah P, Li-Chang H, et al. Hereditary diffuse gastric cancer syndrome: CDH1 mutations and beyond. JAMA Oncol. Published online February 12, 2015. doi:10.1001/jamaoncol.2014.168. eMethods 1. CDH1 Gene Screening eMethods 2. Penetrance Analysis eMethods 3. Multiplexed Sequencing Analysis Using Illumina’s TruSeq Custom Amplicon Assay eMethods 4. Polymerase Chain Reaction Protocol and Primer Sequences eMethods 5. Secondary Analysis of Panel Data eMethods 6. Immunohistochemistry Protocol and Description of Antibodies Used eReferences. Supplemental Methods References eTable 1. CDH1 Mutations Described to Date eTable 2. Families in Penetrance Analysis eTable 3. Genes Selected for Custom Panel Sequencing Based on Association With Upper Gastrointestinal Syndromes eTable 4. Novel and Rare Missense Mutations Detected in HDGC Families Using Custom UGI Gene Panel and Predicted Impact From in Silico Methods eTable 5. Summary of Families With CDH1 Mutations in This Study eTable 6. Cumulative Risk % (95% CI) and Relative Risk of Gastric and Breast Cancer in CDH1 Germline Mutation Carriers eFigure 1. Description of Current Techniques Used in Discovering Pathogenic, Germline Susceptibility Variants eFigure 2. HDGC Mutational Profile eFigure 3. Photomicrographs of the Tumor of the Proband in Family 25 With the CTNNA1 Germline Mutation eFigure 4. Schematic View of Genes With Pathogenic and Likely Pathogenic Mutations Identified Using Panel-Based Sequencing eReferences This supplementary material has been provided by the authors to give readers additional information about their work. © 2015 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/01/2021 eMethods 1. CDH1 gene screening Genomic DNA was extracted from peripheral blood, saliva or paraffin embedded sections using the Gentra Puregene DNA purification Kit (Qiagen), Oragene DNA kit (DNA Genotek) or DNA FFPE tissue kit (Qiagen), respectively, according to the manufacturers’ protocols. Screening for mutations was performed by direct sequencing of all exons and multiplex ligation-dependant probe amplification with previously described primers and reaction conditions (1,2). In short, 16 individual PCR reactions were performed on each sample for a full mutational screen of all 16 exons and splice junctions of the CDH1 gene. Amplified PCR products were bi-directionally sequenced on the Applied Biosystems 3130xl Genetic Analyzer (Life Technologies) using the Big Dye v3.1 Terminator Cycle Sequencing Kit (Life Technologies). Samples with no significant point mutations or minor insertions/deletions were tested for copy number variations using multiplex ligation dependent probe amplification (MRC-Holland). eMethods 2. Penetrance analysis We used the pedigree information to estimate the penetrance of CDH1 using the MENDEL program (3). Ascertainment of families was on the basis of multiple affected members with an index case that tested positive for a truncating CDH1 mutation. Thus the conditional likelihood of the pedigree was maximized given the phenotype of the family and genotype of the index case at ascertainment, where the phenotype of the pedigree is the phenotype of all individuals in the family (vital status, current age/age at death, cancer status and age at diagnosis) and their relationships: L(Pedigree|Ascertained Pedigree) = L(pedigree) / L(Ascertained pedigree) We parameterized the model in terms of log relative risk for GC and BC in mutation carriers compared with population risks. This assumes that cancer incidences in mutation carriers is the same for all families, whatever their ethnic origin. Non-carriers of the deleterious mutation in each family were assumed to develop the disease according to the population incidence rates. The GC relative risks (RR) were estimated separately for men and women and allowed to vary with age using 2 age groups: 10–39 and 40-79. The RR of breast cancer was estimated for Downloaded From: https://jamanetwork.com/ on 10/01/2021 women only and assumed to be constant with age. Other models with either 3 or 4 age groups for GC risk and 2 or 3 age groups for BC risk were also tested. The model selected was the most parsimonious model that fits the data well. The CDH1 mutant allele was assumed to be rare in the general population with a frequency of 0.001. The cumulative risk or penetrance was calculated from the cumulative incidence λ(t): n (t) iktk exp(k ) k1 Where ik is the incidence in the kth age band of length tk and βk is log relative risk in the kth age band. Cumulative risk of penetrance and 95% confidence intervals were calculated from the cumulative incidence A(t), where: n Ʌ(t)= ∑iktkexp(βk) k=1 n n 2 2 1/2 var Ʌ(t)= ∑i kt kvar(βk)exp(2βk) + 2 ∑ikijtkTkTj[ var(βk) var (βj)] exp(βk) (βj)corr((βk, βj) k=1 k=1 th th Where ik = incidence in the k age-band of length Tk and βk = ln (RR) in k age band. The cumulative risk F(t) is given by: F(t) = 1 - exp(-Ʌ(t)) With 95% CI = 1- exp(-Ʌ(t)) ± 1.96 √var Ʌ(t) eMethods 3. Multiplexed sequencing analysis using TruSeq Custom Amplicon assay. Genomic DNA Quantification. DNA samples were quantified to ensure minimum input requirements of 250ng were met for sequencing (Qubit dsDNA broad range kit). All samples were of sufficient quality and quantity and were included in the downstream multiplexed sequencing assay. Multiplexed sequencing analysis was performed using the TruSeq custom amplicon (TSCA) assay (Illumina, San Diego, CA) on 115 germline DNA samples divided into 7 runs, to sequence simultaneously with additional samples of interest from concurrent projects. An average of 42 samples were multiplexed per run. Custom-pooled oligonucleotides were Downloaded From: https://jamanetwork.com/ on 10/01/2021 hybridized to individual samples to generate template libraries and sample-specific indices were then added to each template library by PCR. A bead-based normalization technique eliminated the need for downstream quantification methods. Samples were then pooled prior to loading onto the MiSeq Platform for simultaneous amplification, cluster generation and bi-directional sequencing of 251 cycles per amplicon. Seven germline DNA samples with known pathogenic CDH1 mutations were included in the assay to determine the reliability of the data analysis software. eMethods 4. Polymerase-chain reaction protocol and primer sequences. PCR was performed with variant-specific primers designed using Integrated DNA Technologies online software (CTNNA1 (N71*) F5’-TTGTAAGGTTTACTGGGTCTTCA-3’ and R5’- GGTTAACTAAACCCATGCATCAA-3’; CTNNA1 (R129X) F5’- TGAAAACTCTTAAACTAAATTTGTGC-3’ and R5’-AAAACATCTCTGGTCCATTGAGA-3’; BRCA2 (N1287*) F5’-GTGGGTTTGCAATTTATAAAGCAG-3’ and R5’- TAACTTACCAGAAGCTTGTTTCC-3’; SDHB (S163P) F5’-GCTGAGGTGATGATGGAATCT-3’ and R5’-ACCACACTCCTGGCAATCATCT-3’; STK11 (F354L) F5’-GAGGAGCTGGGTCGGAAA- 3’ and R5’-TGGCCGAGTCAGCAGAG-3’; PALB2 (V398*) F5’- GAAAGTGAGATTCTAAGTCAACCTAAG-3’ and R5’-TTCTTGACATCCAAATGACTCTG-3’; MSR1 (R293X) F5’-AGTACCTTGACAGATGACTAACC-3’ and R5’- CCCTACACATGTACCTGGATG-3’; ATM E1267* F5’- TGTAAAACGACGGCCAGTGCTACTGAACAAGGTCCCATTT-3’ and R5’- CAGGAAACAGCTATGACCCAGTCCTCTTGAATCTGATTAGC-3; ATM R521* F5’- GAGGTCAAACCTAGAAAGCTCA-3’ and R5’-GTGTGTGTCTGTGTGTGTTTATC-3’; ATM Y2791* F5’ - GCTGAATGATCATCAAATGCTCT- 3’ and R5’ - ATGGCTTATTAAAGCTGACAGC- 3’). High Fidelity Taq Polymerase (Life Technologies, Carlsbad, CA) was used during PCR according to manufacturer’s instructions. Post-sequencing cleanup was performed using ExoSAP (Affymetrix, Santa Clara, CA) prior to tagging with M13 primer sequences and Bigdye Downloaded From: https://jamanetwork.com/ on 10/01/2021 fluorescence (Life Technologies) and sequencing on a 3130xl Genetic Analyzer (Applied Biosystems, Carlsbad, CA). eMethods 5. Secondary Analysis of panel data All sequences in the raw data files from 25 cases were first trimmed by 225 base pairs to account for sequence masking done by the MiSeq software in the event that amplicon regions were shorter than 250 base pairs. Samples were then aligned (Burrows-Wheeler Alignment tool-0.5.9)1 to a custom genome then repositioned for variant calling on the full set of loci comprised by the amplicon coordinates. Reads with greater than 5 mismatches were filtered out, and those remaining were called for SNVs using SNVmix2 (4,5) after passing thresholds for base quality (>10) and mapping quality (>20). Each SNV position was tested using binomial exact test, with the background distribution defined by the reference allele frequency from the entire amplicon coordinate list. The resulting p-value was corrected for the number of total tests using the Benjamini Hochberg FDR4 procedure and any resulting q-value <0.00001 were considered TRUE mutations. Indels were called using SAMtools5 pileup, separating lines containing short indels and filtered on a minimum SNP quality score (>300), number of variant reads (>3) and variant “allele” frequency (≥0.1). Coding indels were annotated using information in Ensembl release 67. eMethods 6. Immunohistochemistry protocol and description of antibodies used. Formalin-fixed paraffin-embedded tissue blocks containing whole sections of tumor were obtained from a surgical resection specimen corresponding to the proband of families 16 and 25, found to carry truncating CTNNA1 mutations. Sections (4μm in thickness) were deparafinized, rehydrated, and stained using the semi-automated Ventana Discovery® XT System (Ventana Medical Systems, Tucson, AZ, USA). Antigen retrieval was performed using Cell