Abstract
Background/Aim: Multi-cancer genome profiling (multi-CGP) testing intends to predict the therapeutic efficacy of anticancer medication treatments for eligible patients as part of “precision cancer care.” The number of cases in which a new treatment was applied based on multi-CGP testing has been reported to be between 10% and 20% for all patients in Japan. This study aimed to determine the significance of multi-CGP testing in Japan by analyzing clinical data from multi-CGP testing in various solid cancers at our Hospital. Patients and Methods: A total of 230 patients examined by one of three tests for multi-CGP including NCC Oncopanel, FoundationOne CDx, and FoundationOne Liquid were retrospectively enrolled. Adequate treatment for each patient was discussed at the expert panel meeting according to the results from the genome profiling tests. Results: The most frequent cancer types enrolled in this study were pancreas cancer, bowel cancer, and biliary cancer. Of the 230 cases, 106 (46%) were druggable cases, and 21 (9.1%) were administered medication. Partial response (PR) effect was found in 7 (33.3%) of the 21 cases, of which 3 were biliary cancer and 3 had a BRCA2 mutation. Of all the 21 cases, 7 (33.3%) had the maximum treatment benefit of PR. Three cases of biliary tumors were found in the 7 PR cases within the 21 cases. Conclusion: Of 230 patients, 21 were administered medication following multi-CGP testing data, especially frequent in biliary tumor patients. Multi-CGP testing might be particularly beneficial to patients with biliary tumors in Japan.
Precision medicine, which provides individually appropriate treatment according to personal variants, has been recently introduced in cancer care. Multi-cancer genome profiling (multi-CGP) testing, provides cancer tumor profiling tests under the national health insurance system (1) as part of the precision cancer medicine plan since 2019. Multi-CGP testing aims to predict the therapeutic efficacy of novel anticancer drug treatments for eligible patients who are ineligible for standard therapy (2). NCC Oncopanel has been initiated since December 2019 at the Osaka Metropolitan University Hospital, and the FoundationOne CDx (3) has been initiated since April 2020. Additionally, the FoundationOne Liquid CDx test (4) has been initiated since December 2021. The number of cases in which a new treatment was applied based on multi-CGP testing, which collectively analyzes many genetic abnormalities (mutations, base substitutions, insertions/deletions, copy number abnormalities, rearrangements, fusions, etc.), has been reported at various frequencies between 10% and 20% in all patients in Japan (5-7). Conversely, these cases have higher frequency overseas than that in Japan; however, the usefulness of multi-CGP testing remained not fully clarified (8, 9) in Japan. In this study, we analyzed clinical data from multi-CGP testing in various solid cancers to clarify the significance of multi-CGP testing at our Hospital.
Patients and Methods
Patients. A total of 237 patients were enrolled between December 2019 and January 2023 at the Osaka Metropolitan University Hospital. Since 7 patients died before results were reported, we analyzed 230 patients in this study. Patient criteria for the analysis were as follows, the patient shows good performance status (PS 0 or 1) and the standard treatment for the patient has almost terminated.
Multi-CGP testing. Patients were examined by the NCC Oncopanel examination and were tested for 114 gene variants (5), by the FoundationOne CDx examination for 324 gene variants (3), or the FoundationOne Liquid CDx examination which extracts circulating tumor DNA from the blood for 324 gene variants (4). The type of panel testing was selected according to tumor size and sample type. Subsequently, Expert panel was performed after C-CAT (Center for Cancer Genomics and Advanced Therapeutics in Japan) data reported. We considered pathological significance about gene variant using Clinvar (https://www.ncbi.nlm.nih.gov), prior to Varsome (https://varsome.com).
Microsatellite status and tumor mutational burden. MSI status (stable vs. high) was determined using 114 intronic homopolymer repeat loci with adequate coverage on the panel testing. These sequences were analyzed for length variability and compiled into an overall score using principal components analysis (10). Tumor mutation burden (TMB) is a major factor in the selection of immune checkpoint therapies (11). The high tumor mutational burden (TMB-high) is defined by over 10 mutations/megabase (Mut/Mb) (12).
Expert panel. In the Expert Panel the attending physician or physicians meet with several experts representing various fields, including an oncologist, physicians in charge of clinical trials, genetic counselors, pathologists, and physicians specializing in clinical genetics, who convene after results for a patient are obtained from the NCC Oncopanel report, the FoundationOne CDx report, FoundationOne Liquid CDx report, or the C-CAT report. The possibility of a hereditary tumor and the best treatment based on the cancer gene profiles were discussed at the meeting of the expert panel, whose final report was then prepared and returned to the attending physician.
Ethics statement. This study was approved by the Osaka Metropolitan University Hospital Certified Review Board (Permission number: 3925) and was carried out according to the guidelines of the Committee. Informed consent was obtained in writing from all patients included in the study. And this study has been conducted according to the principles of the declaration of Helsinki. All data in this study were handled in accordance with the Japanese personal information protection law.
Results
Clinicopathological features of patients with refractory tumors. Table I summarizes the characteristics of 230 patients. The male:female ratio was 118:112 for 230 patients. The medium age was 62 (range=12-88) years, metastasis was detected in 205 patients. NCC Oncopanel was employed for 17 patients, FoundationOne CDs for 182 patients, and FoundationOne CDs Liquid for 31 patients. The average period from order panel testing to the expert panel was 29.7 days.
Clinicopathologic features of 230 patients.
The most frequent cancer types examined by panel testing were pancreatic cancer (N=42, 18.3%), colon cancer (N=41 cases, 17.8%), and biliary cancer (N=27, 11.7%). Test submission sites were primary tumors in 141 (61.3%) patients, liquid in 31 (13.5%) and metastasis in 58 patients. The metastasis with the highest frequency was in the liver (15 cases, 6.5%), peritoneum (N=15, 6.5%), lung (N=9, 3.9%), and lymph nodes (N=7, 3.0%). The pathological classification determined 202 (87.8%) epithelial and 22 (10%) non-epithelial cancer cases. The 31 genes used in this investigation were those with a frequency of occurrence of pathological variations of ≥3.0%.
Genomic profile of patients. Genes with the high frequency of pathogenic variant were TP53 (N=133, 57.8%), followed by KRAS (N=65 cases, 28.3%), CDKN2A (N=49, 21.3%), and APC (N=45, 19.6%) (Figure 1A). The cancer type with the highest frequency of pathogenic variants was pancreatic, subsequently bowel, biliary tract, and breast. The most frequently occurring genes were KRAS followed by TP53 in the pancreas, APC followed by TP53 in the bowel, and TP53 followed by CDKN2A in the biliary tract (Figure 1B).
Gene mutations were included in the study. (A) Genes with a frequency of abnormalities of ≥3% in tested patients at the Osaka Metropolitan University Hospital. (B) Gene mutation frequency by organ.
Druggable cases were defined as cases that could be treated by administration of drugs, including approved drugs, investigational new drugs, and off-label drugs in Japan. This study reported 106 druggable cases out of 230 cases (Figure 2A). The approved druggable cases for other cancer types (off-label use) were 89, the Patient-Proposed Healthcare Services cases were 60, and the approved drug cases were 31 in the 106 druggable cases (Figure 2B). Of the 106 druggable cases, 15 were pancreas and bowel, 14 were biliary tract, and 11 were breast cancer (Figure 2C).
Druggable cases and administered medication cases in clinical sequencing. (A) Druggable cases in which cancer gene testing led to medicated cases. (B) Venn diagram of the distribution of 3 types of the druggable pattern, approved drug other cancer types (off-label use), approved drugs, and patient-proposed healthcare services. (C) Frequency of druggable cases by organ.
Treatment cases and drop-out cases. Of the 106 druggable cases, 21 cases received approved treatment (Table II). Of these, a partial response (PR) was found in 7 cases (case number #1 - #7), stable disease (SD) in 4 cases (case #8-#11), and progressive disease (PD) in 10 cases (case #12-#21). Four of 7 PR cases were biliary tumor cases (case #3, #4, #5, and #7). Pathological gene variants most common among the 21 cases were TMB-high in 6 cases (case #7, #9, #14, #18, #19, #20) and BRCA2 mutation in 4 cases (case #1, #2, #6, #8). All 4 cases with BRCA2 mutation showed PR or SD. On the other hand, 87 of the 106 cases dropped out (Table III), mainly because of the use of alternative therapies or the ineligibility for clinical studies, such as non-fulfillment of eligibility criteria, being outside the registration period, or the poor condition of patients (N=11, 13%). Most patients with advanced-stage disease are not eligible for many clinical studies.
Outcomes of 21 cases administered drugs.
Reasons for rejection of the genotype-matched therapy in 87 patients of 106 druggable cases.
Discussion
This study analyzed 230 patients at our Hospital with the aim to determine the significance of multi-CGP testing in Japan using clinical data from 3 types of multi-CGP panels. Among the 3 types of panel tests, FoundationOne CDx examination was more frequently (79.1%) performed than the NCC Oncopanel (7.4%). This is because more genes were paneled on FoundationOne CDx in comparison with NCC Oncopanel.
The average period from orders for the panel testing to the expert panel was taken 29.8 days in our study, similar to a previous report from Japan (6, 11). This period seems to be longer compared to that reported from overseas (4, 11). The delay might be due to the expert panel which is essential in Japan and is typical for clinical settings (13, 14). Artificial intelligence assistance might contribute to shortening the period. Some patients could not be administered medication because of their worsening performance status. Because multi-CGP testing takes two or three months before results are announced, multi-CGP testing might be necessary to be performed at earlier stages. Development of molecular-targeted drugs may be also necessary to increase candidate treatment cases by multi-CGP testing.
The tumor type with the highest frequency of pathogenic variants was the pancreas, followed by bowel, biliary tract, and breast. Genes with a high frequency of pathogenic variants were TP53, KRAS, CDKN2A, and APC. Tumor types with a high frequency of pathogenic variants include the pancreas, followed by the bowel, biliary tract, and breast, while certain approved drugs against these pathogenic variants are available so far. Developing useful drugs against these major pathogenic variants will soon be necessary.
This study revealed 106 (46.0%) druggable cases of all 230 cases. However, 87 of 106 cases were excluded because of ineligibility in clinical studies, such as non-fulfillment of eligibility criteria, outside the registration period, or poor condition of patients. Most patients with advanced stage are not suitable for many clinical studies. Currently, multi-CGP testing is applied to patients with advanced solid tumors who fail to respond to standard therapies in Japan (13). Some reports indicate that cancer genome tests should be performed before first-line treatment (14-16). Our study indicated that multi-CGP testing may help in the early examination, which will increase opportunities for “precision cancer care” based on genetic variants because cancer gene profiling testing takes time to finalize reports.
The cancer types with the highest frequency from panel testing were pancreatic cancer, colon cancer, and biliary cancer, and druggable cases were the most common in pancreatic and colon tumors, followed by biliary and breast tumors. Of the 106 druggable cases, 15 were both pancreatic and colon cases, 14 were biliary tract, and 11 were breast cancer cases. No case was administered medication for patients with colon cancer although bowel tumor was most common in the druggable cases.
Additionally, the frequency of druggable cases was higher in breast tumors (84.6%) compared to other tumor types, but only one breast tumor case was administered medication. The approved drug cases were only 34 of the 106 druggable cases. These findings indicate that developing novel molecular-targeted drugs might be urgently necessary, especially in colon cancer and breast cancer. Approved drugs outside Japan may also be recommended to be part of clinical trials in Japan.
Of the 106 druggable cases, 21 received special action, meaning that only 21 (9.1%) of all 230 cases were administered medication. The main reasons for the low medication rate were due to the late stage of patients and limited access to drugs/clinical trials. These findings suggested that multi-CGP testing should be performed at an earlier clinical stage. Of all 21 administered medication cases, 7 (33.3%) had the maximum treatment benefit of PR. Three cases of biliary tumors were found in the 7 PR cases within 21 administered medication cases. Pathological gene variants cases that caused action in druggable cases were TMB-high and BRCA2 mutation, with all BRCA2 mutations showing PR or SD. Currently, therapeutic agents have been developed for unresectable biliary tract FGFR2 fusion gene-positive or FGFR2 rearrangement-positive intrahepatic cholangiocarcinoma (17). These data indicate that patients with biliary tumors might have beneficial benefits for multi-CGP testing in Japan. Recently, SOX17 has reported to be correlated with tumor stages and patient survival (18). SOX17 is currently listed in a new multi-CGP testing in Japan. Also, EV-H19, which has been reported to be a novel biomarker for diagnosing androgen receptor-axis-targeted resistance in castration resistant prostate cancer patients may negatively reflect androgen receptor signaling activity in prostate cancer tissue (19). Our study showed that there was only 1 PR case of 4 Druggable prostate cancer cases (Table II), suggesting that EV-H19 might be added in the panel as companion marker for prostate cancer. Addiction of companion genes to the multi-CGP testing may improve the significance of multi-CGP testing. Since the druggable cases were less than 40% in ovarian cancer, esophageal cancer, pancreatic cancer, and bowel cancer (Table II), multi CGP testing might be improved by multi-omics analyses including the characteristic genes for these cancer types.
In conclusion, we reported 230 cases of multi-CGP testing in our hospital. Of 230 patients, 21 (9.1%) were administered medication, which was especially frequent for biliary tumor patients. Multi-CGP testing is particularly beneficial to patients with biliary tumors in Japan.
Acknowledgements
This study was partially founded by KAKENHI (Grant-in-Aid for Scientific Research) 21H03008. We thank Drs. (Osaka City Graduate School of Medicine, Osaka, Japan), and Chiaki Kawano (Genetic Counseling Unit, Osaka Metropolitan University, Graduate School of Medicine, Osaka, Japan) who contributed to this study.
Footnotes
Conflicts of Interest
The Authors declare no conflicts of interest.
Authors’ Contributions
RA and MY designed and summarized data and co-wrote the manuscript. RA and MY contributed equally. HN, KK, SK, YY, and SN collected the data. All Authors read and approved the final manuscript.
- Received October 13, 2023.
- Revision received November 25, 2023.
- Accepted November 29, 2023.
- Copyright © 2024, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved
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).