Abstract
Background/Aim: As an increasing number of drugs are approved for targeted cancer therapy, comprehensive genomic profiling of cancer patients is frequently conducted to identify potentially relevant genetic variants. Different sequencing technologies are used for this purpose; targeted gene panels cover a selected set of biomarker genes and hotspot regions, while whole genome sequencing (WGS) delivers genome-wide data. This comparison study aimed at evaluating whether one method performs superiorly to the other regarding the detection of targetable variants. Patients and Methods: To evaluate the performance of the two sequencing technologies, we compared the results of targeted sequencing using the Ion Torrent Oncomine Comprehensive Assay Plus (OCA-Plus) panel to Illumina WGS reports in 11 patients diagnosed with pancreatic cancer (PC). All pathogenic and likely pathogenic variants, including those relevant for targeted therapy, reported by WGS and OCA-Plus, were included in the final comparison. Results: Both techniques identified common driver mutations implicated in PC with high concordance (81%) across all variants. For variants relevant to targeted therapy, a 100% concordance between the technologies was observed. Conclusion: A comparable number of variants were reported by WGS and OCA-Plus, and all genetic variants relevant for targeted therapy were identified by both technologies. Thus, WGS does not provide substantial additional information in this patient group.
Next-generation sequencing (NGS) is increasingly used in the diagnostics of cancer to identify crucial prognostic and predictive biomarkers (1). Furthermore, it facilitates the detection of genetic variants that could be targeted with personalized therapies, such as BRAF mutations susceptible to BRAF inhibitors in malignant melanoma or amplifications/activating mutations in the ERBB2 gene, which can be targeted in breast cancer (2, 3). In a large cohort study of 5,000 patients across 24 cancer types, 26% harbored genetic variants potentially relevant for targeted therapy (4). However, the distribution of actionable variants varies across tumor types, leading to an increasing number of anti-cancer drugs targeting specific molecular alterations in various tumor types to be approved by the U.S. Food and Drug Administration and European Medicines Agency (5, 6). Several prospective clinical trials, such as the Danish ProTarget basket trial, are underway to evaluate the efficacy and safety of off-label use of targeted therapies, with extensive molecular profiling used to identify potential drug matches (7). Thus, NGS provides comprehensive genomic information that allows treatment decision to be guided by the tumor’s genomic landscape rather than the tissue of origin.
Various technologies can be used to identify actionable variants in cancer tissue; these technologies include Sanger sequencing which generates data from a single gene, NGS targeted gene panels which provide data from a selection of genes; whole exome sequencing which obtains data from all coding regions of the human genome, and Whole Genome Sequencing (WGS) which provides data across the entire genome. Thus, WGS has the advantage of potentially identifying novel genetic variants outside hotspot regions covered by gene panels, thus providing broader insight into genetic changes associated with cancer pathogenesis, which is highly favorable in exploratory, research-oriented settings (8). However, the enormous amount of data generated by WGS can make data analysis and variant interpretation time consuming. Consequently, in silico-filtering of WGS data is often applied to focus on specific cancer-related genes exclusively, similar to the approach used with large gene panels, when used in clinical diagnostics (9). Additionally, WGS typically has a lower sequencing depth (>50×) compared to other technologies, making it challenging to identify relevant variants in samples with a high proportion of normal tissue (10, 11). In contrast, targeted gene panels rely on the enrichment of relevant genes and hotspot regions, enabling a higher sequencing depth (usually >200-1,000×). This allows for the detection of variants even in samples with a low tumor cell proportion, such as those analyzed for circulating tumor DNA, or DNA of suboptimal quality and quantity (9, 12). Another advantage of targeted gene panels is their shorter turnaround time and lower cost compared to WGS, making them well-suited for widespread use in a clinical workflow.
In this study, we aimed to compare WGS and targeted sequencing in detecting likely pathogenic/pathogenic, and potentially targetable variants, in patients diagnosed with pancreatic cancer (PC), a disease characterized by late-stage diagnosis and high mortality rates (13). Treatment strategies typically involve surgery combined with chemotherapy for resectable tumors, while patients with advanced or metastatic disease receive chemotherapy (14). The genetic landscape of PC is characterized by frequent mutations in the oncogene KRAS and tumor suppressors TP53, SMAD4 and CDKN2A, along with variants in less frequently mutated genes (15–18). Additionally, a subset of tumors harbors mutations in the homologous recombination repair genes, such as BRCA1, BRCA2, PALB2, and ATM, which could potentially be targeted with Poly ADP-Ribose Polymerase (PARP) inhibitors (19).
To compare the performance of the two methods, we conducted targeted sequencing using Ion Torrent Oncomine™ Comprehensive Assay Plus (OCA-Plus) on 11 patients diagnosed with PC and compared the results to corresponding WGS reports. The OCA-Plus panel covers 501 genes associated with cancer, some of which are covered in full coding sequence, while others are covered in specific hotspot regions. Additionally, the panel includes 49 fusion genes involved in inter- and intragenic fusions. The genes in the panel were selected based on information from databases, peer-reviewed literature, and data from the pharmaceutical industry, covering all biomarker genes relevant to targeted therapy (20).
Patients and Methods
Patients. The retrospective method validation study included eleven patients diagnosed with pancreatic adenocarcinoma. The patients were diagnosed at the Department of Pathology at Herlev Hospital or at Rigshospitalet, and the initial WGS genomic profiling was conducted at the Department of Genomic Medicine at Rigshospitalet. Sequencing using the Oncomine Comprehensive Assay Plus was carried out at the Department of Pathology at Herlev Hospital. The study was carried out following the guidelines of the International Conference of Harmonization of Good Clinical Practice and the Declaration of Helsinki. The study was approved by the health authority, Danish Patient Safety Authority (ref: 3-3013-2793/1).
WGS reports. The WGS reports included results from Illumina WGS and RNA- sequencing, as well as results from HD CytoScan analysis. Fresh tumor biopsies were used for the somatic analysis, while the germline analysis was performed on blood samples. The average sequencing depth was at least 50× for tumor samples and at least 30× for germline samples, with at the minimum 95% of the genome covered by at least 10×. Sequencing reads were mapped to the hg38 reference genome. Germline variants were identified using GATK while Haplotype caller or Mutect2 were used for somatic variant calling. Somatic variants with a variant allele frequency (VAF) of at least 10% were reported. However, for samples with a low abundance of tumor tissue, relevant variants below the VAF-cutoff were also analyzed. Germline variants were reported in the following genes, based on the list from the American College of Medical Genetics and Genomics (ACMG) with the addition of selected clinically relevant genes: APC, ATM, BMPR1A, BRCA1, BRCA2, CHEK2, MAX, MEN1, MLH1, MSH2, MSH6, MUTYH, NF2, PALB2, PMS2, PTEN, RAD51C, RB1, RET, SDHAF2, SDHB, SDHC, SDHD, SMAD4, STK11, TMEM127, TP53, VHL, and WT1 (21). Germline structural variants, including copy number variations (CNVs), were identified using the variant callers Manta, Delly, Lumpy and CNVnator. Variants in coding regions and flanking intron sequences (+/− 20 bp) were analyzed. Classification was performed using guidelines from the ACMG and the Association for Molecular Pathology, and only likely pathogenic and pathogenic germline variants present in at least 20% of sequencing reads were reported. A germline analysis was not conducted for two patients (patient 4 and 5), as no blood samples were available. Illumina RNA-sequencing was conducted on RNA isolated from fresh tumor tissue, and the FusionMap v.10.0.1.29 pipeline was used for the fusion analysis (22).
OCA-Plus library preparation and sequencing. Diagnostic formalin-fixed-paraffin-embedded (FFPE) needle biopsies were used for further analysis with OCA-Plus at Herlev Hospital. The biopsies used for OCA-Plus and WGS analyses were conducted either at the same time or within a period of two weeks for nine patients, while for two patients, the biopsies were collected several months apart. FFPE needle biopsies were examined by a pathologist to assess the quality and to determine the tumor cellularity percentage to be above the cutoff of 30% for all biopsies before sequencing. Areas with a high percentage of cancer cells in the FFPE block were selected using a needle puncher, and DNA/RNA was extracted using the Maxwell RSC DNA/RNA FPPE kits (Promega, Madison, WI, USA) and the Maxwell RSC instrument (Promega), following the manufacturer’s instructions. After extraction, the DNA/RNA concentration was determined using the Qubit dsDNA HS assay kit or RNA HS assay kit (Thermo Fisher Scientific, Waltham, MA, USA) on a Qubit Fluorometer (Thermo Fisher Scientific). DNA was treated with Uracil-DNA glycosylase heat labile (Thermo Fisher Scientific) to remove deaminated bases while RNA was converted to complementary DNA using the SuperScript VILO kit (Thermo Fisher Scientific). NGS sequencing libraries were generated using the OCA-Plus assay (Thermo Fisher Scientific), which contains 2-pool DNA and RNA library preparation panels. Inputs of 5-40 ng DNA and 20 ng RNA were used for multiplex PCR amplification according to the manufacturer’s instructions (manual nr. MAN0018490). Finished libraries were diluted to concentrations of 50 pM and pooled appropriately, aiming for >20,000,000 reads for DNA and >1,000,000 reads for RNA. Template preparation and loading on the Ion 550™ Chip were done using the Ion Chef™ Instrument and Ion 550™ Kit - Chef (Thermo Fisher Scientific) according to the manufacturer’s instructions (manual nr. MAN0017275). Following chip loading, the sequencing was performed with Ion Torrent S5™ XL instrument (Thermo Fisher Scientific) according to the manufacturer’s instructions (manual nr. MAN0010811).
For one patient (patient 1), additional DNA purified from blood and a fresh tumor biopsy was obtained from the Department of Genomic Medicine, Rigshospitalet, for confirmatory analysis.
Data analysis. The raw sequencing DAT files were transferred to Ion Torrent Suite Software v5.18.1 (Thermo Fisher Scientific), where signal processing, base calling, removal of barcodes, adapter trimming and alignment to the hg19 reference genome, were carried out. The resulting BAM-files were then transferred to the IonReporter v5.20 software. Quality parameters such as read number, coverage uniformity, and read length histograms provided by the Ion Torrent Suite Software were evaluated for all samples. If the samples met the quality standards, cohesive DNA- and RNA samples were paired and a locally developed workflow (Oncomine Comprehensive Plus – w2.3 – DNA and Fusions – Single Sample_MAPD_0.95_PPA_1.0) was initiated using the extended filter algorithm from Oncomine™ (Oncomine Extended v.5.18v2). In one case, a missed variant was rescued using a locally developed filter chain [TML_Version_4 (5.6)]. While the workflow defines the specific quality parameters the variants must pass to be included, the subsequent filtering step determines which of the variants will be presented in the final analysis results.
Following analysis, quality control (QC) parameters such as fusion sample QC and the percentage of total bases with a coverage greater than 100× were assessed (with a minimal requirement of 90%). Variants that passed the workflow and filtering were manually inspected using Integrative Genomics Viewer version 5.01 to ensure that variants were correctly interpreted and to exclude potential technical artifacts. Variants were excluded as technical artifacts if they were only present in incomplete reads or showed read strand bias compared to the wild type allele. Correct variants were then classified using the search engine Varsome, which provides data from the ClinVar database and the Genome Aggregation Database, as well as references to relevant literature. Additional information was obtained from the OncoKB knowledge base and the TP53 database.
Results
We conducted a comparison between the results of Illumina WGS reports and results of targeted sequencing using the OCA-Plus assay on 11 patients diagnosed with PC to examine how well they aligned in detecting potentially targetable variants. For six patients (patients 3, 4, 6, 7, 8, and 9), the different biopsies used for OCA-Plus and WGS were collected simultaneously. Three patients (patients 1, 2, and 5) had the supplementary biopsy for the WGS analysis performed approximately two weeks after the initial biopsy, and for two patients (patient 10 and 11), the biopsies used for the two analyses were conducted months apart. In the final comparison, single nucleotide variants, multiple nucleotide variants, insertion and deletion variants, and gene fusions, classified as either likely pathogenic or pathogenic were included. In two cases, a variant of uncertain significance (VUS) was noted in the WGS report (EGFR p.Q791H in patient 5 and TP53 p.I195S in patient 4). In these cases, the variant was only included in the final comparison if classified as a likely pathogenic or pathogenic variant in the OCA-Plus analysis.
Concordant variants. In total, 36 likely pathogenic, or pathogenic variants were reported, with 29 of these ( 81%) being concordant between WGS and OCA-Plus (Table I). KRAS was the most frequently mutated gene, with all patients harboring pathogenic missense variants in either codon 12 or 61 of the gene. Seven out of 11 patients had mutations in TP53, either through frameshift deletions or missense mutations, and there was a 100% agreement between WGS and OCA-Plus for variants in both KRAS and TP53. However, there was one instance of discrepancy in classification; the TP53 variant in patient 4 was classified as a VUS in the WGS report but as a likely pathogenic variant in the OCA-Plus analysis. Additionally, both technologies identified pathogenic or likely pathogenic variants in the following patients: patient 4 (RNF43 and SF3B1), patient 5 (TET2), patient 6 (RNF43), patient 10 (GNAS and TGFBR2), and patient 11 (KMT2C and STK11). Notably, there was a discrepancy in the location of the TGFBR2 variant in patient 10, as it was reported at codon 553 in the OCA-Plus analysis and at codon 528 in the WGS analysis due to the use of two different reference transcripts (NM_001024847.2 and NM_003242.6, respectively), despite being identified by both technologies.
Overview of pathogenic and likely pathogenic variants reported in the 11 patients. Regular font: Reported by both technologies; bold: Only reported by Whole Genome Sequencing; italics: only reported by Oncomine Comprehensive Assay Plus.
Discordant variants. Seven variants were reported by only one of the technologies. For the variants identified differently by WGS and OCA-Plus, the raw data was re-examined. In patient 1, WGS identified a likely pathogenic splice site variant in the RB1 gene that was not detected by the OCA-Plus assay. The specific genetic locus in question was covered by one primer amplicon in the OCA-Panel (Low_r249_1.5666), but it had zero end-to-end reads in this sample. As a result, the variant was excluded by the workflow. Additionally, the WGS report identified a nonsense variant in the CDKN2A gene in patient 6 that was not found in the OCA-Plus analysis. Upon reviewing the raw data, we noted that the variant was present with a variant allele frequency (VAF) of 31%, which closely matched the VAF reported by WGS (VAF of 32%). However, due to the very low coverage of the region, with only 12 end-to-end reads of the amplicon covering the locus (CDKN2A_3.6085), the OCA-Plus workflow sorted the variant out. The OCA-Plus analysis found a nonsense mutation in the KMT2C gene that was not reported in the WGS analysis of this patient. At the review, the variant was confirmed by the WGS analysis pipeline; however, it was excluded from the report as it was irrelevant to the treatment of the patient. Similarly, a truncating variant in the FANCC gene was identified by the OCA-Plus analysis in patient 7 but was not mentioned in the WGS report. As germline variants are only reported in a specific set of genes defined by WGS pipeline, the FANCC variant was not included in the report despite being identified by the WGS technology. Conversely, the WGS analysis identified a splice site variant in the CDKN2A gene that was not detected by OCA-Plus in this patient. The variant was reported with a VAF of 5.8% in the WGS report, indicating a low tumor cellular percentage in the sample. Although the variant was present in the raw data from OCA-Plus, it was not called by the workflow as the amplicon (CDKN2A_3.6085) was covered by 21 end-to-end reads only. In patient 9, the OCA-Plus analysis identified a gene fusion between the MTAP gene and CDKN2B_AS1, which was not reported by WGS. However, the WGS report presented a biallelic CDKN2A deletion in the patient, and insight into the raw data indicated that Manta called a larger deletion involving MTAP, CDKN2A, CDKN2B and CDKN2B_AS1. Thus, the genetic event was also identified by WGS but defined as a deletion rather than a fusion. In patient 10, a missense variant in the RNF43 gene was presented in the WGS report but was missed by OCA-Plus analysis. However, the variant was present in the raw data and was called when the filter chain was changed to a locally developed filter chain, indicating that the variant was filtered out by the Oncomine Extended v.5.18v2 filter.
Variants relevant to targeted treatment. Both WGS and OCA-Plus identified frameshift deletions in the BRCA1 and BRCA2 genes in patient 8 and 2, respectively. The WGS report also presented a pathogenic germline variant in the ATM gene in patient 1, which was not initially detected by OCA-Plus. Upon further investigation, it was discovered that the specific ATM locus was sufficiently covered by the OCA-Plus panel in the sample, and a 1bp deletion was observed in a small fraction of the reads. To gain more clarity, an additional germline and somatic OCA-Plus sequencing of the patient was performed using DNA from the fresh tumor biopsy and blood sample used for the WGS analysis of the patient. The variant was identified in the additional germline OCA-Plus sequencing with a VAF of 50%, but it was not detected in the supplementary somatic OCA-Plus sequencing of the tumor. The Department of Genomic Medicine confirmed that the variant was detected exclusively by their germline WGS pipeline. Upon inspection of the raw data from the somatic WGS, it was observed that the variant was present in three of the 138 reads at the locus, likely representing a small fraction of normal tissue in the biopsy. However, this was insufficient for the variant to be called, as was the case with the initial somatic OCA-Plus analysis of the sample. Based on the WGS report analysis and a copy number variation analysis, it was revealed that the tumor did not lose the chromosome carrying the ATM gene variant but rather reversed the mutation. In conclusion, both WGS and OCA-Plus successfully identified the variant in the germline samples but found that the mutation was not present in the tumor of the patient, indicating a 100% concordance between WGS and OCA-Plus regarding the detection of variants in genes relevant to targeted therapy.
Discussion
This study aimed to compare WGS and targeted sequencing using OCA-Plus for detecting likely pathogenic/pathogenic variants, particularly those variants relevant to potential targeted therapy, in patients with PC. The goal was to determine whether one method performed better than the other in this specific setting. Both technologies identified variants in commonly mutated genes in PC, such as KRAS and TP53, with a 100% concordance between the two technologies for these two genes. The high frequency of KRAS (100% of patients) and TP53 mutations (64% of patients) in the cohort align with published findings that these genes are mutated in >90% and ~70% of pancreatic adenocarcinomas, respectively (23). KRAS missense mutations occur at specific codons (12, 13, and 61) in 99% of the samples analyzed, and these hotspots are covered by multiple amplicons in the OCA-Plus panel (24). In contrast, pathogenic mutations in the TP53 gene are more evenly distributed throughout the whole gene, with a predominance of missense variants in the DNA-binding domain, although truncating mutations are also reported (25, 26).
In our comparison study, we observed seven discordant variants. Four were reported by WGS exclusively, while three were reported only by OCA-Plus. For the variants missed by OCA-Plus, inadequate coverage of the region of interest was the cause in three out of four cases, indicating insufficient coverage of certain genetic regions by the panel. This OCA-Plus related issue has been addressed by our research group before and solved by the introduction of an additional low coverage rescue filter (27). Furthermore, OCA-Plus missed a variant in RNF43, likely due to settings in the filter chain, possibly related to the global population frequency of the variant. When the filter chain was changed to a locally developed one, the variant was detected. Our research group has previously demonstrated that using a locally developed filter improves the detection of clinically relevant variants compared to the use of default and extended filters provided by Oncomine™ (28).
Further investigation of the KMT2C and FANCC variants that were not reported by WGS revealed that they were identified by the technology but not included in the final reports. The FANCC variant was excluded as germline variants are not reported in this gene, and the KMT2C variant was not reported as it was considered irrelevant to therapy. Additionally, a fusion between MTAP and CDKN2B_AS1 was not reported by WGS. CDKN2B_AS1, also known as ANRIL, is a non-coding RNA transcript involved in repressing the expression of CDKN2A and CDKN2B. Inactivation of the tumor suppressor CDKN2A is common in several cancers and can occur due to point mutations, promoter methylation, or deletion of the genomic locus that also contains the CDKN2B gene. Fusions between MTAP and CDKN2B_AS1 at specific break points lead to the loss of CDKN2A and CDKN2B, thereby promoting tumorigenesis (29, 30). Importantly, the WGS report presented that the patient harbored a biallelic CDKN2A deletion which was not presented in the comparison as copy number variations were not included in the study. Insight into the raw data revealed a larger deletion affecting MTAP, CDKN2A, CDKN2B and CDKN2B_AS1, indicating that the discordance regarding this variant appears to be caused by different definitions of the genetic event. However, the WGS report did not describe any loss of the MTAP gene, which could be of relevance to targeted therapy as a synthetic lethality drug targeting the PRMT5 enzyme in MTAP-null cells is currently undergoing clinical testing (31).
Two patients had loss-of-function mutations in the BRCA1 or BRCA2 genes, both losses were identified by both the germline WGS and the OCA-Plus analysis. These genes are involved in the repair of double-stranded breaks (DSB) as a part of the homologous recombination repair pathway, a complex mechanism involving numerous proteins. Therefore, mutations in BRCA1/2 or other genes in this pathway, such as ATM or RAD51, can lead to homologous recombination deficiency (HRD), which can be indicative of response towards PARP-inhibitors (32). During repair of single-stranded breaks (SSB), the Poly (ADP-ribose) polymerase (PARP1) enzyme plays an important role by participating in base excision repair. Once inhibited by PARP-inhibitors, the enzyme will be trapped around the SSBs, leading to accumulation of more SSBs and generation of DSBs. In cells with HRD, such as BRCA1/2 mutated tumors, the increasing amount of DSBs cannot be repaired, and the critically high level of genomic instability will lead to apoptosis. Thus, the PARP-inhibitors induce synthetic lethality in these cells (32, 33). Initially, germline mutations in the BRCA genes were associated with a higher risk of ovarian and breast cancer. However, BRCA mutations have also been associated with both familial and non-familial cases of PC, occurring in 5-10% of patients with PC (19). For carriers of germline BRCA-mutations, the ‘second hit’ contributing to tumorigenesis is typically the loss of heterozygosity (34, 35).
The WGS report identified a likely pathogenic frameshift variant in ATM in patient 1. Loss-of-function variants in this gene are associated with HRD, which could make the patient a candidate for targeted treatment with PARP-inhibitors. However, because the variant was not identified by the somatic OCA-Plus sequencing, supplementary analyses were performed. After additional germline- and somatic OCA-Plus sequencing and examination of raw data from both WGS analyses, it was found that the variant was present only in normal tissue and not in the tumor. Consistent with this finding, the variant was classified as germline in the WGS report. It was noted that genes with germline variants were excluded from the somatic WGS analysis pipeline. Since the tumor did not contain the loss-of-function variant in ATM, the patient would not have benefitted from PARP-inhibitor treatment, highlighting the importance of comparing the findings in blood and tissue. In summary, there was a 100% agreement between the two technologies for variants related to targeted therapy. It is important to consider that PC has a relatively low mutational load and a limited profile of mutations relevant to targeted therapy. Therefore, the outcome might have been different if the comparison was done on a tumor type with a more diverse molecular profile.
The WGS analyses were performed on DNA/RNA extracted from fresh tissue and blood, since these types of biological material typically yield high quality and quantity of nucleic acids. However, cancer histopathological diagnosis typically relies on FFPE-tissue, which generally contains lower quality and quantity of DNA/RNA. Despite these challenges, targeted sequencing can achieve sufficient coverage in most cases. FFPE-tissue offers advantages such as preserving cellular morphology, long-term storage, and compatibility with various analytical techniques, such as immunohistochemistry and NGS (36). In addition to the benefit of performing the genetic analysis on the diagnostic FFPE material, thereby minimizing the need for additional invasive biopsies, the use of FFPE-tissue for genetic analyses allows a pathologist to estimate the tumor cellular percentage in the sample before sequencing. Furthermore, the VAF of identified variants can be compared to the tumor burden of the sample in the subsequent data analysis. While none of the WGS analyses were inconclusive due to a high proportion of normal tissue, some reports noted that an excessive amount of normal tissue could disrupt the analysis of chromosomal changes. Apart from the requirement for high-quality DNA/RNA, another challenge in using WGS for routine diagnostics is the time-consuming data mining process, which leads to prolonged turnaround time compared to targeted sequencing, even with in silico filtering of WGS data. Given the poor prognosis of PC, a rapid test result without the need for additional biopsies is preferred in this patient group. Moreover, WGS analysis and data storage are associated with significantly higher costs than targeted sequencing. Therefore, from a clinical perspective, WGS may be less optimal, as it provides information comparable to that from targeted sequencing with OCA-Plus in this setting. While WGS advances our understanding of cancer biology, its broader use needs to be supported by demonstrated survival benefits and a thorough cost-benefit assessment (8).
Conclusion
The study compared Illumina WGS and OCA-Plus targeted sequencing for detecting clinically actionable variants in patients with PC. The methods showed 100% concordance in the identification of targetable variants, suggesting that WGS provides minimal additional benefit in this context. Considering the higher cost, longer turnaround time, and potential need for extra biopsies with WGS, targeted gene panels may offer a more practical approach to identifying actionable variants in this patient group.
Acknowledgements
The Authors would like to thank the Department of Genomic Medicine, Rigshospitalet, Denmark, for providing additional genetic material for a supplementary analysis with OCA-Plus, for being helpful in the interpretation of genetic variants, and for assisting with raw data evaluation of discordant variants.
Footnotes
Authors’ Contributions
TSP and EH conceived the project and IMC was responsible for patient inclusion. The experimental design was conducted by TAH, EH, and TSP. LLK performed the microscopy and selection of patient biopsies. TAH performed all laboratory work, and data analysis and variant interpretation was performed by TAH and TSP. EH and TSP supervised the project and TAH prepared the manuscript. All Authors contributed to the edition of the manuscript and have approved the final version.
Conflicts of Interest
The Authors have no competing interests to declare in relation to this study.
Funding
The Authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
- Received December 16, 2024.
- Revision received December 31, 2024.
- Accepted January 3, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.