Skip to main content
NCBI home page
American Journal of Cancer Research logoLink to American Journal of Cancer Research
. 2024 Feb 15;14(2):796–808. doi: 10.62347/FSSF9938

Clinical implications of PD-L1 expression and pathway-related molecular subtypes in advanced Asian colorectal cancer patients

Qingqing Qiu 1,*, Dan Tan 1,*, Qiaofeng Chen 1,*, Ru Zhou 1, Xiaokai Zhao 3,4, Wei Wen 3,4, Pengmin Yang 3,4, Jieyi Li 3,4, Ziying Gong 3,4, Daoyun Zhang 3,4, Mingliang Wang 1,2,*
PMCID: PMC10915335  PMID: 38455414

Abstract

The expression level of PD-L1 does not accurately predict the prognosis of advanced colorectal cancer (CRC) patients, but it still reflects the tumor microenvironment to some extent. By stratifying PD-L1 status, gene subtypes in PD-L1 positivity-related pathological pathways were analyzed for their relationship to MSI or TMB to provide more individualized treatment options for CRCs. A total of 752 advanced CRCs were included, and their genomic variance was measured by a targeted next generation sequencing panel in this study. MSI and TMB were both measured by NGS, while PD-L1 expression level was measured using the PD-L1 colon 22C3 pharmDx kit. We found RTK/RAS pathway was positively related to high PD-L1 expression, with BRAF V600E and most KRAS mutations (G12 and G13) subtypes showing a significant correlation. Conversely, the Wnt and p53 pathways were negatively related to high PD-L1 expression, with APC C-terminal alterations and other non-inactivation mutations in TP53 making a primary contribution with significant statistical significance. Major subtypes showing a significantly higher proportion of TMB-H or MSI-H were irrespective of PD-L1 status. These findings demonstrate pathological pathways associated with high PD-L1 expression, suggesting that pathway-induced oncogenic constructive PD-L1 upregulation may be the reason for the corresponding patients’ primary resistance to immune checkpoint inhibitors (ICIs), rather than a lack of pre-existing immune responses.

Keywords: TMB, MSI, programmed death-ligand 1 (PD-L1), advanced colorectal cancer (CRC)

Introduction

Immune checkpoint inhibitors (ICIs) have achieved dazzling clinical efficacy in various advanced solid tumors, providing more treatment options for advanced patients [1]. Nevertheless, Pembrolizumab, one type of PD-1 ICIs, was only recently approved by the FDA for the first-line treatment of mismatch-repair-deficient (dMMR)/microsatellite instability-high (MSI-H) metastatic colorectal cancer (mCRC) patients [2]. Given the unclear response mechanism of patients to immunotherapies, multiple clinical trials, including monotherapy [2,3], combination of ICIs [4], radiotherapy [5], MEK inhibitor [6,7] and anti-angiogenic agents [8], are underway to identify more suitable treatment regimens and prognostic biomarkers for different subtypes of CRC. Generally accepted immunotherapy biomarkers for colorectal cancer include microsatellite instability (MSI) [9], tumor mutation burden (TMB) [10,11], and POLE gene mutation [12]. The proportion of POLE gene mutations in the Asian patient population is relatively small [13]. In comparison, MSI and TMB testing provide a distinct advantage in personalized treatments for Asian patients.

The rate of MSI-H is relatively low in mCRC, at about 5% [14-16], while the proportion of TMB high level is slightly higher with a positivity rate about 10% at a cutoff of 10 muts/Mb [17,18]. Consequently, only a small number of advanced CRCs can benefit from immunotherapy, and other potential biomarkers remain to be explored. In comparison to these two biomarkers, PD-L1 status, with a positivity rate about 10% in CRCs [19], has received inconsistent conclusions regarding its prognostic effect and its ability to predict response to ICIs [20-22]. Although PD-L1 status cannot serve as an individual biomarker for predicting the efficacy of immunotherapy in clinical applications [23,24], it is still closely associated with an active immune microenvironment in some pathways [25,26]. Additionally, the high heterogeneity of CRCs at the genomic level is also one of the reasons for the significant difference in curative effect, suggesting that gene molecular typing may play an important role in the clinical treatment of advanced CRCs. Moreover, meta-analyses have also linked PD-L1 expression to a poor prognosis in colorectal cancer [27,28]. Our study aims to compare gene mutation profiles between PD-L1 positive and negative groups and analyze gene enrichment related to tumor signaling pathways. This exploration seeks to uncover potential mechanisms associated with a poor prognosis. Furthermore, we conducted a detailed stratified analysis of key genes related to PD-L1 expression, along with assessments of TMB and MSI. This comprehensive approach helps identify patient subgroups that may benefit from targeted therapy, immunotherapy or a combination of both, offering novel insights into precision treatment for colorectal cancer.

Methods

Patient and sample characteristics

A total of 809 patients with stage IV advanced colorectal cancer were enrolled in this study at Ruijin Hospital from January 2021 to May 2023. Samples lacking paired blood, without pathological confirming of CRC, and with a tumor content of less than 20% were all excluded, resulting in 752 samples from 752 patients being included. Clinical data, including age and gender, were obtained from medical records. Written informed consent was obtained from all participants, and this study was approved by the institutional review board of our hospital (LWEC2020010).

DNA extraction and library construction

The extraction and purification of blood DNA and tissue DNA are implemented by using the human blood genome DNA extraction kit (Shanghai YunYing) and the human tissue DNA extraction kit (Shanghai YunYing), respectively. NanoDrop2000 (Thermo Fisher Scientific, Waltham, MA, USA) was applied to evaluate the contamination and purity of samples, followed by a storage at -20°C before use. The VAHTS Universal DNA Library Prep Kit for Illumina was used for library preparation. Shanghai YunYing’s optimized probes, which target the exons and some introns of 639 genes related to cancer (Table S4), were applied. Targeted-sequencing was performed on the NextSeq500 platform (Illumina, Carlsbad, CA, USA) strictly following the manufacturer’s protocols.

NGS-based assay and bioinformatical analysis

FastQC software (version 0.11.2) was used for the screening of FASTQ files. Customized Python scripts were conducted to remove the sequence from adaptor or with a quality score below 30. Referring on the human genome GRCh37/hg19, clean reads were mapped by the Burrows-Wheeler Aligner (BWA, version 0.7.7), resulting in corresponding bam files which were realigned and recalibrated by GATK3.5. Picard MarkDuplicates (version 1.35) was used to remove duplicate sequences and reduce potential polymerase chain reaction bias. Single nucleotide variations (SNVs) were detected by VarScan (version 2.3.2) following the criteria: (1) Total reads depth was more than 500; (2) Mutated reads frequency was higher than 1%. Insertion of deletion (indel) was detected by Pindel (version 0.2.5b8) with default parameters. FACTERA (version 1.4.4) was applied to identify structure variation using default parameters.

Mutation difference of somatic SNVs and Indels were compared between tumor samples and matched normal samples by MuTect (version 1.1.4), followed by functional annotations using Verscan2 (version 2.3.9). The number of all somatic, coding, base substitution and indel mutations per megabase were used to calculate tumor mutation burden (TMB). Thus, the TMB per megabase was calculated by the quotient of the counted total mutation number divided by the size of the coding region of the targeted territory (1.1 Mb of coding genome).

MSIsensor [29], a software tool for quantifying MSI in genome sequencing data using tumor samples with or without paired normal ones, were applied to calculate the MSI score of all samples with default parameters. The MSI score was defined as the percentage of unstable microsatellites in all used microsatellites. At least 20 spanning reads and single-nucleotide mutations would be included in each microsatellite site.

PD-L1 expression level was measured by immunohistochemical (IHC) method and assessed by experienced pathologists. According to the PD-L1 IHC 22C3 pharmDx package insert, TPS was calculated to determine the expression level of PD-L1. Briefly, TPS was defined as the number of viable tumor cells displaying partial of complete membrane staining of PD-L1 which is divided by the total number of viable tumor cells and then multiplied by 100%. The TPS value which is 1% or more than 1% is defined as PD-L1 positivity and that less than 1% is grouped as PD-L1 negativity.

Mutations in APC are divided into C-terminal inactivation, N-terminal inactivation and other non-inactivation referring the standard of Mondaca et al [30] and those of TP53 referring Hung-Chih Hsu [31].

Statistical analysis

All statistical analyses were conducted in the R-project (version 4.2.1). The mutation landscape heat maps and the pathological pathways in PD-L1 positive cohorts were depicted by R package “maftools” [32]. R package “ggplot2” [33] and “corrplot” [34] were used to drawn the box plots and triangle heat maps of mutation frequency difference of important gene sites in PD-L1 positive/negative cohorts, respectively. The “fisher.test” function in R was implemented to calculate the significance of proportion difference between groups.

Results

Demographic characteristics of included patients

From 2021 to 2023, 752 patients pathologically diagnosed with colorectal cancer or intestinal cancer were enrolled in the present study. As shown in Table 1, the age and TMB status of the entire enrolled cohort are unrelated to PD-L1 positivity. However, after stratifying by PD-L1 expression level, female patients were more enriched in the PD-L1 positive group. At the same time, MSI-H status was positively related to PD-L1 positivity, suggesting that PD-L1 status may also be a noteworthy indicator similar to MSI. In addition, no statistic shifts are observed in cohort TMB or MSI grouped by age or gender (Table S1), except for the acknowledged positive correlation between MSI and age, which indicates that there is no clear cohort preference for any type of gender or age in terms of MSI or TMB. More detailed characteristics of the cohort are demonstrated in Table 1.

Table 1.

Patient characteristics

Characteristics PD-L1 positive No. (%) (n=127) PD-L1 negative No. (%) (n=625) aP value
Total 127 (100) 625 (100)
    Gender
        Male 70 (55.12) 408 (65.28) 0.04
        Female 57 (44.88) 217 (34.72)
    Age at diagnosis in years
        < 60 63 (49.61) 275 (44.00) 0.3
        ≥ 60 64 (50.39) 350 (56.00)
    MSI state
        MSI-H 15 (11.81) 25 (4.00) 0.002
        Non MSI-H (MSS and MSI-L) 112 (88.19) 600 (96.00)
    Tumor mutation burden status
        TMB-H (≥ 10 muts/Mb) 20 (15.75) 63 (10.08) 0.09
        TMB-L (< 10 muts/Mb) 107 (81.89) 562 (89.92)
Open in a new tab
a

P value are tested by Fisher’s exact Test.

Mutation profiling

The overall genotype landscape, stratified by PD-L1 status, is hierarchically colored depending on the number of mutations in corresponding genes, as shown in Figure 1A and 1B. The top 10 most mutated genes are basically consistent (8/10) in both the PD-L1 positive group and PD-L1 negative group, including TP53, KRAS, APC, PIK3CA, LRP1B, KMT2C, FAT1, FBXW7 (61% vs. 71%, 60% vs. 46%, 49% vs. 66%, 25% vs. 18%, 18% vs. 16%, 17% vs. 21, 14% vs. 13% and 11% vs. 12%; positive vs. negative, respectively). As depicted in Figure 1C and 1D, the co-occurrence of mutations is more common in the PD-L1 negative group. The mutual exclusion between BRAF and KRAS or APC is not affected by the PD-L1 grouping, while the mutual exclusion between KRAS and TP53 is weakened in the PD-L1 positive group.

Figure 1.

Figure 1

Open in a new tab

The water-fall diagram of the Top 25 mutated genes and their variant types in PD-L1 positive (A) and negative (B) advanced CRC tissues based on next generation sequencing data from collected samples. Co-occurrence and mutually exclusive relationships at gene-level are analyzed and plotted by R pack “maftools” for PD-L1 positive (C) and negative (D) cohort. Top 25 genes with the most distinct relationship are shown.

Distinct pathogenic pathways in PD-L1 positive/negative groups

Some studies have mentioned that pathway-induced oncogenic constructive PD-L1 upregulation, rather than a lack of pre-existing immune responses, may be the reason for a patient’s primary resistance to ICIs [26]. Thus, an analysis for the frequency of oncogenic pathway alterations between PD-L1-positive status and PD-L1-negative statuses was conducted, indicating a positive correlation between the RTK/RAS (KRAS and BRAF) signaling pathway and PD-L1 positive expression with statistical significance (P < 0.001, Figure 2A). Additionally, the Wnt and p53 signal pathways were found to be negatively related to PD-L1 expression, showing statistically significant differences (P < 0.001 and P < 0.01, Figure 2A).

Figure 2.

Figure 2

Open in a new tab

A. Frequency of oncogenic pathway alterations by PD-L1 status. B. Count of oncogenic pathway-related gene alterations by PD-L1 status. C. Frequency of concurrent oncogenic alterations by PD-L1 status. Note: * denotes P < 0.05; **, P < 0.01; ***, P < 0.001, Fisher’s exact test.

We further performed a comparative analysis of mutations between PD-L1-positive and PD-L1-negative statuses, revealing a significant difference in the mutation frequency of four oncogenic genes (APC, TP53, BRAF and KFAS; Figure 2B, P < 0.05), which accounted for the corresponding differences in signaling pathways aforementioned. A detailed display of the frequency of concurrent oncogenic alterations in these four important oncogenic genes is demonstrated in Figure 2C.

Subtypic view of important genes mutation frequency by PD-L1 expression hierarchy

Gene molecular typing can more clearly demonstrate the differences in oncogenic pathways caused by PD-L1 status stratification. Therefore, we analyzed the distribution percentage of PD-L1 status among important gene subtypes (Figure 3A) and demonstrated the distribution of mutations and their positions in Figure S1. BRAF-V600E (P < 0.001, compared to wild type), KRAS-G12, and KRAS-G13 (P < 0.05 and P < 0.01, compared to wild type) mutation frequencies both display a distinct positive correlation to PD-L1 positivity, as shown in Table 2; Figure 3A and 3B. Along with more detailed insight into KRAS mutations sites, G13D accounts for the clearly higher G13 mutation frequency in PD-L1 positive group compared to that in the negative group (data not shown).

Figure 3.

Figure 3

Open in a new tab

Comparison of PD-L1 status distribution percentage among important gene subtypes in total patient population (A). The difference of PD-L1 positivity rate in subtypes between pairwise groups are presented by fan plots (B). Red fan diagram denotes lower positivity rate in the corresponding column than that in row and blue one denotes the opposite. Fisher’s exact test is applied in this analysis. Note: * denotes P < 0.05; **, P < 0.01; ***, P < 0.001.

Table 2.

Site view of important genes mutation frequency in PD-L1 positive/negative groups

PD-L1 positive No. (%) (n=127) PD-L1 negative No. (%) (n=625)
aAPC-N 9 (7.09) 63 (10.08)
bAPC-C 50 (39.37) 334 (53.44)
cAPC-other 3 (2.36) 18 (2.88)
Wild type 65 (51.18) 210 (33.60)
dTP53-other 61 (48.03) 357 (57.12)
eTP53-DBD 16 (12.60) 88 (14.08)
Wild type 50 (39.37) 180 (28.8)
BRAF-V600E 18 (14.17) 23 (3.68)
fBRAF-other 6 (4.72) 23 (3.68)
Wild type 104 (81.89) 579 (92.64)
KRAS-G12 51 (40.15) 210 (33.60)
KRAS-G13 19 (14.96) 49 (7.84)
gKRAS-other 10 (7.87) 43 (6.88)
Wild type 51 (40.16) 335 (53.6)
Open in a new tab
a

inactivation mutations on the N-terminal;

b

inactivation mutations on the C-terminal;

c

non-inactivation mutations;

d

mutations outside the protein-binding region;

e

mutations in the protein-binding region;

f

mutations other than V600E;

g

mutations other than G12X nor G13X.

Due to a considerable kinds of mutation sites, unlike oncogenes, which are simply divided into subtypes based on mutation sites, tumor suppressor gene subgroups are divided by mutation regions referring to previous studies [30,31]. APC mutations are categorized by the mutation coordinate into N-terminal inactivation (APC-N), C-terminal inactivation (APC-C), other non-inactivation mutations (APC-other), and wild type [30]. No significant mutation distribution difference is caused by PD-L1 stratification between APC-C and APC-N. However, a distinct difference is observed between APC-C and wild type (P < 0.001, Figure 2, with more APC-C in the PD-L1 negative cohort) as well as between APC-N and wild type (P = 0.051, fisher’s exact test, with more APC-N in the PD-L1 negative cohort). Similarly, TP53 mutations are classified into DNA binding domain and others according to the variance coordinates [31]. Compared with the wild type, an enrichment of TP53 other mutations is observed in the PD-L1 negative cohort (P < 0.05). The above subtypes with higher or lower mutation frequency compared with wild types may indicate some new explanations on the pathway mechanisms related to oncogenesis and progression.

Subtypic view of immunotherapy potential in PD-L1 positive cohort

Considering that TMB and MSI are both approved to be the biomarker of immunotherapy recently [9], we display the TMB and MSI distribution in the subtypes of important genes related to the pathological pathway in PD-L1 positive cohorts (Table S2; Figure 4A and 4B). The distribution of TMB and MSI in the subgroup is basically consistent. PD-L1 positive patients with TP53 and KRAS wild type, as well as patients carrying APC-N, BRAF-other and KRAS-other mutations may benefit from combining immunotherapy drugs with the original treatment schemes. Compared to its wild type, KRAS-other is the only subtype with a higher proportion of TMB-H status (P < 0.01, Figure 4C and 4D). TMB-L and non MSI-H are significantly enriched in patients with KRAS-G12 and KRAS-G13 mutations, which may explain the poor curative effect of ICIs as a monotherapy for such patients. However, the population of PD-L1 positive patients is relatively small, with only one or two patients carrying mutations in some subtypes (such as APC-N, BRAF-other), indicating that more data are required to draw a more accurate conclusion.

Figure 4.

Figure 4

Open in a new tab

Percentage distribution of TMB status in PD-L1 positive patient cohort who carries important oncogenic gene mutations in advanced colon cancer (A) and distribution of MSI status (B). Subtype-level TMB positivity rate difference and corresponding statistical significance between gene subtypes are calculated pairwise in PD-L1 positive cohorts (C), where TMB positivity rate in each subtype is obtained by dividing the number of TMB-positive patients carrying the mutations of the subtype by the total patients of this subtype. Subtype-level MSI positivity rate difference and corresponding statistical significance (D). TMB and MSI expression are categorized into high level and low level by the aforementioned criterion and the distribution difference of TMB and MSI in each pairwise subtype is examined by Fisher’s exact test. Red fan diagram denotes lower positivity rate in the corresponding column than that in row and blue one denotes the opposite. The size of the fan shaped coloring area corresponds to the value of positivity rate difference. Note: * denotes P < 0.05; **, P < 0.01; ***, P < 0.001.

Subtypic view of immunotherapy potential in PD-L1 negative cohort

The distribution of TMB and MSI in the subgroup is also basically consistent in the PD-L1 negative group, with only the distribution in KRAS-G13 mutations showing significant differences in the population stratified by PD-L1 status (Table S3; Figure 5A and 5B). PD-L1 negative patients with wild type of TP53 genes, and patients carrying APC-other, BRAF-other, KRAS-other and KRAS-G13 mutations, significantly represent TMB-H or MSI-H (Figure 5C and 5D), suggesting potential benefit from combining immunotherapy drugs with the original treatment schemes. It is noteworthy that these results are consistent with previous studies on the relationship between TMB and APC subtyping [35] in the PD-L1 positive group, but are in contrast with them in PD-L1 negative groups, which may account for the lack of PD-L1 stratification previously. Coincidentally, these four subtypes (APC-other, TP53 wild type, BRAF-other and KRAS-other) also showed significantly higher levels of TMB or MSI in the PD-L1 positive patient group, suggesting that the immunotherapy potential of related subtypes may not be related to PD-L1 expression. At the same time, the proportion of KRAS-G13 subtypes showed significant differences between the PD-L1 positive and negative groups, suggesting that the tumor microenvironment of these two molecular subtypes and immunotherapy may require attention to the patient’s PD-L1 status.

Figure 5.

Figure 5

Open in a new tab

Percentage distribution of TMB status in PD-L1 negative patient cohort who carries important oncogenic gene mutations in advanced colon cancer (A) and distribution of MSI status (B). Subtype-level TMB positivity rate difference and corresponding statistical significance between gene subtypes are calculated pairwise in PD-L1 negative cohorts (C), where TMB positivity rate in each subtype is obtained by dividing the number of TMB-positive patients carrying the mutations of the subtype by the total patients of this subtype. Subtype-level MSI positivity rate difference and corresponding statistical significance (D). TMB and MSI expression are categorized into high level and low level by the aforementioned criterion and the distribution difference of TMB and MSI in each pairwise subtype is examined by Fisher’s exact test. Red fan diagram denotes lower positivity rate in the corresponding column than that in row and blue one denotes the opposite. The size of the fan shaped coloring area corresponds to the value of positivity rate difference. Note: * denotes P < 0.05; **, P < 0.01; ***, P < 0.001.

Discussion

PD-L1 is one of the widely studied biomarkers for solid tumor, and its expression level reflects the degree of immunosuppression in the tumor microenvironment to some extent. Even though PD-L1 cannot individually predict the efficacy of ICIs in CRC to date [24], more in-depth studies and detailed analyses are still required to elucidate its role in displaying the tumor microenvironment and cell-intrinsic immune programs, as well as the mechanisms behind the poor association between PD-L1 positivity and immunotherapy efficacy [36]. The PD-L1 positivity might result from immune response-induced PD-L1 expression or oncogenic constructive PD-L1 upregulation. The latter commonly demonstrates resistance to PD-1/PD-L1 therapies due to the lack of pre-existing immune response [26]. From the perspective of pathological pathways, we found that the RTK/RAS pathway (KRAS and BRAF mutations) is enriched in PD-L1 positive populations, while Wnt (APC mutations) and p53 (TP53 mutations) are enriched in PD-L1 negative populations in our study. These findings are all related to oncogenic construction [37-40]. In previous research on the RTK/RAS pathways in CRC, KRAS mutations have been associated with down-regulation of immune pathways and a reduced number of tumor infiltrating lymphocytes (TILs) [41], which are considered markers for worse survival at all disease stages [42]. Additionally, a study of triple-negative breast cancer also found that alterations in RTK/RAS signaling were correlated with low TIL numbers, which in turn correlated with worse recurrence-free survival [42]. The association between PD-L1 positivity and the RTK/RAS pathway may be the reason for the poor prognosis in PD-L1 positive CRC cohort.

According to studies on therapeutic valuable targets in the treatment of CRC, it is almost impossible for p53 drugs to be used as monotherapy for cancer treatment [43]. As for the Wnt/β-catenin signaling pathway, several inhibitors have been developed for CRC treatment. However, so far no molecular therapeutics targeting this pathway have been incorporated into oncological practice [37]. A recent noteworthy proof-of-concept clinical trial [44], which combines ICIs with genomic stratification of the RTK/RAS pathway to improve effectiveness in clinically troublesome non-MSI-H subtypes, indicates an intriguing approach: considering the tumor microenvironment indicated by PD-L1 status and the genomic characteristics of the cohort may improve the clinical effect in some combined therapy strategies.

Currently, there are few personalized treatment schemes for PD-L1 negative patients, who account for the majority of total patient population. Considering the extremely effectiveness of ICIs for mCRC patients with MSI-H status [45], screening out genomic subtypes which are enriched in MSI-H or TMB-H status may improve the immunotherapy efficacy of ICIs for the corresponding subtype population. In this study, we found that PD-L1 negative patients with wild type of APC and KRAS genes, as well as those carrying APC-C, TP53-other, and KRAS-G12 mutations demonstrated significantly higher MSI-H or non-conflictingly corresponding TMB-H levels. This indicates a potential for these patient populations to gain more benefits from ICIs treatments. A recent single-arm phase II clinical trial, LCCC1632 (NCT03442569), targeting KRAS/NRAS/BRAF wild-type MSS refractory mCRC demonstrated a promising breakthrough in the combination of ICIs and anti-angiogenic drugs [46,47]. Considering that targeting KRAS has become less difficult recently and one of the FDA-approved KRAS inhibitors, Sotorasib, has shown anticancer activity in patients with KRAS G12C-mutated advanced solid tumors [48-50]. In our study, CRC patients with KARS-G12 exhibited similar low MSI and TMB status to that of wild type, may also benefit from the scheme similar to LCCC1632, which combines take combined-therapy including ICIs for subtypes with MSS status.

Notably, there are several limitations to this study. Firstly, it is a single-center retrospective study without treatment follow-up data. More promising multicenter studies with larger cohorts and long-term follow-up data are needed for a better understanding of the relationship between these gene mutations and ICIs therapy efficacy. Secondly, the tumor tissues used in this study can only depict a limited range of solid tumors in a single time-frame, which is inadequate to characterize the full view because of intra-tumoral and inter-metastatic heterogeneity.

In conclusion, our study developed a potentially valuable approach to stratifying advanced colorectal patient based on PD-L1 expression levels and pathway-related molecular subtypes. We identified one pathological pathway (RTK/RAS) positively related to PD-L1 expression levels and two pathways (Wnt and p53) negatively related to PD-L1 expression. We investigated the detailed distribution preference of PD-L1 in corresponding gene subtypes. Furthermore, we found that patients with APC-other, TP53 wild type, BRAF-other and KRAS-other mutations showed a significantly higher proportion of TMB-H or MSI-H, regardless of PD-L1 status. This suggests that the immunotherapy potential of these subtypes may not be related to PD-L1 expression. Additionally, the high of TMB or MSI in KRAS-G13 subtypes showed significant differences between PD-L1 positive and negative groups, indicating that PD-L1 stated may require attention for patients with these molecular subtypes when considering ICIs treatments. Our study may provide potential implications on the strategy of combining immune checkpoint inhibitors and pathway-targeted therapy.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (grant number. 82272633); the Key Research and Development Program of Zhejiang province [2023C03057]; Zhejiang Leading Talent Entrepreneurship Project [2021R02019]; the Jiaxing Leading Talent Entrepreneurship Project; and the Key Technology Innovation Projects of Jiaxing [2021BZ10004].

Written informed consent was obtained from all patients.

Disclosure of conflict of interest

Wei Wen, Xiaokai Zhao, Jieyi Li, Pengmin Yang, Daoyun Zhang, and Ziying Gong were employed by the Jiaxing Yunying Medical Inspection Co., Ltd., and the Zhejiang Yunying Medical Technology Co., Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Supporting Information

ajcr0014-0796-f6.pdf (397.2KB, pdf)

References

  • 1.Cristescu R, Mogg R, Ayers M, Albright A, Murphy E, Yearley J, Sher X, Liu XQ, Lu H, Nebozhyn M, Zhang C, Lunceford JK, Joe A, Cheng J, Webber AL, Ibrahim N, Plimack ER, Ott PA, Seiwert TY, Ribas A, McClanahan TK, Tomassini JE, Loboda A, Kaufman D. Pan-tumor genomic biomarkers for PD-1 checkpoint blockade-based immunotherapy. Science. 2018;362:eaar3593. doi: 10.1126/science.aar3593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Andre T, Shiu KK, Kim TW, Jensen BV, Jensen LH, Punt CJA, Smith DM, Garcia-Carbonero R, Benavides M, Gibbs P, De La Fouchardiere C, Rivera F, Elez E, Bendell JC, Le DT, Yoshino T, Yang P, Farooqui MZH, Marinello P, Diaz LA. Pembrolizumab versus chemotherapy for microsatellite instability-high/mismatch repair deficient metastatic colorectal cancer: the phase 3 KEYNOTE-177 Study. J. Clin. Oncol. 2020;38:LBA4. [Google Scholar]
  • 3.Le DT, Uram JN, Wang H, Bartlett B, Kemberling H, Eyring A, Azad NS, Laheru D, Donehower RC, Crocenzi TS, Goldberg RM, Fisher GA, Lee JJ, Greten TF, Koshiji M, Kang SP, Anders RA, Eshleman JR, Vogelstein B, Diaz LA. Programmed death-1 blockade in mismatch repair deficient colorectal cancer. J. Clin. Oncol. 2016;34:103. [Google Scholar]
  • 4.Chen EX, Jonker DJ, Loree JM, Kennecke HF, Berry SR, Couture F, Ahmad CE, Goffin JR, Kavan P, Harb M, Colwell B, Samimi S, Samson B, Abbas T, Aucoin N, Aubin F, Koski SL, Wei AC, Magoski NM, Tu D, O’Callaghan CJ. Effect of combined immune checkpoint inhibition vs best supportive care alone in patients with advanced colorectal cancer: the Canadian Cancer Trials Group CO.26 Study. JAMA Oncol. 2020;6:831–838. doi: 10.1001/jamaoncol.2020.0910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Parikh AR, Clark JW, Wo JYL, Yeap BY, Allen JN, Blaszkowsky LS, Ryan DP, Giantonio BJ, Weekes CD, Zhu AX, Van Seventer EE, Matlack L, Foreman B, Ly L, Drapek LC, Ting DT, Corcoran RB, Hong TS. A phase II study of ipilimumab and nivolumab with radiation in microsatellite stable (MSS) metastatic colorectal adenocarcinoma (mCRC) J. Clin. Oncol. 2019;37:3514. [Google Scholar]
  • 6.Bendell JC, Kim TW, Goh BC, Wallin J, Oh DY, Han SW, Lee CB, Hellmann MD, Desai J, Lewin JH, Solomon BJ, Chow LQM, Miller WH, Gainor JF, Flaherty K, Infante JR, Das-Thakur M, Foster P, Cha E, Bang YJ. Clinical activity and safety of cobimetinib (cobi) and atezolizumab in colorectal cancer (CRC) J. Clin. Oncol. 2016;34:3502. [Google Scholar]
  • 7.Bendell JC, Bang YJ, Chee CE, Ryan DP, McRee AJ, Chow LQ, Desai J, Wongchenko M, Yan Y, Pitcher B, Foster P, Cha E, Grossman W, Kim TW. A phase Ib study of safety and clinical activity of atezolizumab (A) and cobimetinib (C) in patients (pts) with metastatic colorectal cancer (mCRC) J. Clin. Oncol. 2018;36:560. [Google Scholar]
  • 8.Hochster HS, Bendell JC, Cleary JM, Foster P, Zhang W, He X, Hernandez G, Iizuka K, Eckhardt SG. Efficacy and safety of atezolizumab (atezo) and bevacizumab (bev) in a phase Ib study of microsatellite instability (MSI)-high metastatic colorectal cancer (mCRC) J. Clin. Oncol. 2017;35:673. [Google Scholar]
  • 9.Fan A, Wang B, Wang X, Nie Y, Fan D, Zhao X, Lu Y. Immunotherapy in colorectal cancer: current achievements and future perspective. Int J Biol Sci. 2021;17:3837–3849. doi: 10.7150/ijbs.64077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Yarchoan M, Hopkins A, Jaffee EM. Tumor mutational burden and response rate to PD-1 inhibition. N Engl J Med. 2017;377:2500–2501. doi: 10.1056/NEJMc1713444. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Panda A, Betigeri A, Subramanian K, Ross JS, Pavlick DC, Ali S, Markowski P, Silk A, Kaufman HL, Lattime E, Mehnert JM, Sullivan R, Lovly CM, Sosman J, Johnson DB, Bhanot G, Ganesan S. Identifying a clinically applicable mutational burden threshold as a potential biomarker of response to immune checkpoint therapy in solid tumors. JCO Precis Oncol. 2017;2017:PO.17.00146. doi: 10.1200/PO.17.00146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lau D, Kalaitzaki E, Church DN, Pandha H, Tomlinson I, Annels N, Gerlinger M, Sclafani F, Smith G, Begum R, Crux R, Gillbanks A, Wordsworth S, Chau I, Starling N, Cunningham D, Dhillon T. Rationale and design of the POLEM trial: avelumab plus fluoropyrimidine-based chemotherapy as adjuvant treatment for stage III mismatch repair deficient or POLE exonuclease domain mutant colon cancer: a phase III randomised study. ESMO Open. 2020;5:e000638. doi: 10.1136/esmoopen-2019-000638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hu H, Cai W, Wu D, Hu W, Dong Wang L, Mao J, Zheng S, Ge W. Ultra-mutated colorectal cancer patients with POLE driver mutations exhibit distinct clinical patterns. Cancer Med. 2021;10:135–142. doi: 10.1002/cam4.3579. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Flecchia C, Zaanan A, Lahlou W, Basile D, Broudin C, Gallois C, Pilla L, Karoui M, Manceau G, Taieb J. MSI colorectal cancer, all you need to know. Clin Res Hepatol Gastroenterol. 2022;46:101983. doi: 10.1016/j.clinre.2022.101983. [DOI] [PubMed] [Google Scholar]
  • 15.Harada S, Morlote D. Molecular pathology of colorectal cancer. Adv Anat Pathol. 2020;27:20–26. doi: 10.1097/PAP.0000000000000247. [DOI] [PubMed] [Google Scholar]
  • 16.Martelli V, Pastorino A, Sobrero AF. Prognostic and predictive molecular biomarkers in advanced colorectal cancer. Pharmacol Ther. 2022;236:108239. doi: 10.1016/j.pharmthera.2022.108239. [DOI] [PubMed] [Google Scholar]
  • 17.Chalmers ZR, Connelly CF, Fabrizio D, Gay L, Ali SM, Ennis R, Schrock A, Campbell B, Shlien A, Chmielecki J, Huang F, He Y, Sun J, Tabori U, Kennedy M, Lieber DS, Roels S, White J, Otto GA, Ross JS, Garraway L, Miller VA, Stephens PJ, Frampton GM. Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden. Genome Med. 2017;9:34. doi: 10.1186/s13073-017-0424-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Chabanon RM, Pedrero M, Lefebvre C, Marabelle A, Soria JC, Postel-Vinay S. Mutational landscape and sensitivity to immune checkpoint blockers. Clin Cancer Res. 2016;22:4309–4321. doi: 10.1158/1078-0432.CCR-16-0903. [DOI] [PubMed] [Google Scholar]
  • 19.Yarchoan M, Albacker LA, Hopkins AC, Montesion M, Murugesan K, Vithayathil TT, Zaidi N, Azad NS, Laheru DA, Frampton GM, Jaffee EM. PD-L1 expression and tumor mutational burden are independent biomarkers in most cancers. JCI Insight. 2019;4:e126908. doi: 10.1172/jci.insight.126908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Yang L, Xue R, Pan C. Prognostic and clinicopathological value of PD-L1 in colorectal cancer: a systematic review and meta-analysis. Onco Targets Ther. 2019;12:3671–3682. doi: 10.2147/OTT.S190168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Wang S, Yuan B, Wang Y, Li M, Liu X, Cao J, Li C, Hu J. Clinicopathological and prognostic significance of PD-L1 expression in colorectal cancer: a meta-analysis. Int J Colorectal Dis. 2021;36:117–130. doi: 10.1007/s00384-020-03734-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Alexander PG, McMillan DC, Park JH. A meta-analysis of CD274 (PD-L1) assessment and prognosis in colorectal cancer and its role in predicting response to anti-PD-1 therapy. Crit Rev Oncol Hematol. 2021;157:103147. doi: 10.1016/j.critrevonc.2020.103147. [DOI] [PubMed] [Google Scholar]
  • 23.Le DT, Uram JN, Wang H, Bartlett BR, Kemberling H, Eyring AD, Skora AD, Luber BS, Azad NS, Laheru D, Biedrzycki B, Donehower RC, Zaheer A, Fisher GA, Crocenzi TS, Lee JJ, Duffy SM, Goldberg RM, de la Chapelle A, Koshiji M, Bhaijee F, Huebner T, Hruban RH, Wood LD, Cuka N, Pardoll DM, Papadopoulos N, Kinzler KW, Zhou S, Cornish TC, Taube JM, Anders RA, Eshleman JR, Vogelstein B, Diaz LA Jr. PD-1 blockade in tumors with mismatch-repair deficiency. N Engl J Med. 2015;372:2509–2520. doi: 10.1056/NEJMoa1500596. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Overman MJ, Lonardi S, Leone F, McDermott RS, Morse MA, Wong KYM, Neyns B, Leach JL, Garcia Alfonso P, Lee JJ, Hill A, Lenz HJ, Desai J, Moss RA, Cao ZA, Ledeine JM, Tang H, Kopetz S, Andre T. Nivolumab in patients with DNA mismatch repair deficient/microsatellite instability high metastatic colorectal cancer: update from CheckMate 142. J. Clin. Oncol. 2017;35:519. [Google Scholar]
  • 25.Peng QH, Wang CH, Chen HM, Zhang RX, Pan ZZ, Lu ZH, Wang GY, Yue X, Huang W, Liu RY. CMTM6 and PD-L1 coexpression is associated with an active immune microenvironment and a favorable prognosis in colorectal cancer. J Immunother Cancer. 2021;9:e001638. doi: 10.1136/jitc-2020-001638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Yi M, Niu M, Xu L, Luo S, Wu K. Regulation of PD-L1 expression in the tumor microenvironment. J Hematol Oncol. 2021;14:10. doi: 10.1186/s13045-020-01027-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Zhao T, Li Y, Zhang J, Zhang B. PD-L1 expression increased by IFN-gamma via JAK2-STAT1 signaling and predicts a poor survival in colorectal cancer. Oncol Lett. 2020;20:1127–1134. doi: 10.3892/ol.2020.11647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Shen Z, Gu L, Mao D, Chen M, Jin R. Clinicopathological and prognostic significance of PD-L1 expression in colorectal cancer: a systematic review and meta-analysis. World J Surg Oncol. 2019;17:4. doi: 10.1186/s12957-018-1544-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Han X, Zhang S, Zhou DC, Wang D, He X, Yuan D, Li R, He J, Duan X, Wendl MC, Ding L, Niu B. MSIsensor-ct: microsatellite instability detection using cfDNA sequencing data. Brief Bioinform. 2021;22:bbaa402. doi: 10.1093/bib/bbaa402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Mondaca S, Walch H, Nandakumar S, Chatila WK, Schultz N, Yaeger R. Specific Mutations in APC, but not alterations in DNA damage response, associate with outcomes of patients with metastatic colorectal cancer. Gastroenterology. 2020;159:1975–1978. e1974. doi: 10.1053/j.gastro.2020.07.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Hsu HC, You JF, Chen SJ, Chen HC, Yeh CY, Tsai WS, Hung HY, Yang TS, Lapke N, Tan KT. TP53 DNA binding domain mutations predict progression-free survival of bevacizumab therapy in metastatic colorectal cancer. Cancers (Basel) 2019;11:1079. doi: 10.3390/cancers11081079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Mayakonda A, Lin DC, Assenov Y, Plass C, Koeffler HP. Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Res. 2018;28:1747–1756. doi: 10.1101/gr.239244.118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wickham H. ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag; 2016. [Google Scholar]
  • 34.Simko V, Wei T. R package ‘corrplot’: visualization of a Correlation Matrix (Version 0.92) 2021 [Google Scholar]
  • 35.Peng H, Ying J, Zang J, Lu H, Zhao X, Yang P, Wang X, Li J, Gong Z, Zhang D, Wang Z. Specific mutations in APC, with prognostic implications in metastatic colorectal cancer. Cancer Res Treat. 2023;55:1270–1280. doi: 10.4143/crt.2023.415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Shen X, Zhao B. Efficacy of PD-1 or PD-L1 inhibitors and PD-L1 expression status in cancer: meta-analysis. BMJ. 2018;362:k3529. doi: 10.1136/bmj.k3529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Cheng X, Xu X, Chen D, Zhao F, Wang W. Therapeutic potential of targeting the Wnt/beta-catenin signaling pathway in colorectal cancer. Biomed Pharmacother. 2019;110:473–481. doi: 10.1016/j.biopha.2018.11.082. [DOI] [PubMed] [Google Scholar]
  • 38.Zheng CC, Liao L, Liu YP, Yang YM, He Y, Zhang GG, Li SJ, Liu T, Xu WW, Li B. Blockade of nuclear beta-catenin signaling via direct targeting of RanBP3 with NU2058 induces cell senescence to suppress colorectal tumorigenesis. Adv Sci (Weinh) 2022;9:e2202528. doi: 10.1002/advs.202202528. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Liao H, Li X, Zhao L, Wang Y, Wang X, Wu Y, Zhou X, Fu W, Liu L, Hu HG, Chen YG. A PROTAC peptide induces durable beta-catenin degradation and suppresses Wnt-dependent intestinal cancer. Cell Discov. 2020;6:35. doi: 10.1038/s41421-020-0171-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Zhang W, Sun R, Zhang Y, Hu R, Li Q, Wu W, Cao X, Zhou J, Pei J, Yuan P. Cabazitaxel suppresses colorectal cancer cell growth via enhancing the p53 antitumor pathway. FEBS Open Bio. 2021;11:3032–3050. doi: 10.1002/2211-5463.13290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Lal N, White BS, Goussous G, Pickles O, Mason MJ, Beggs AD, Taniere P, Willcox BE, Guinney J, Middleton GW. KRAS mutation and consensus molecular subtypes 2 and 3 are independently associated with reduced immune infiltration and reactivity in colorectal cancer. Clin Cancer Res. 2018;24:224–233. doi: 10.1158/1078-0432.CCR-17-1090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Sillo TO, Beggs AD, Morton DG, Middleton G. Mechanisms of immunogenicity in colorectal cancer. Br J Surg. 2019;106:1283–1297. doi: 10.1002/bjs.11204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Hassin O, Oren M. Drugging p53 in cancer: one protein, many targets. Nat Rev Drug Discov. 2023;22:127–144. doi: 10.1038/s41573-022-00571-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Tian J, Chen JH, Chao SX, Pelka K, Giannakis M, Hess J, Burke K, Jorgji V, Sindurakar P, Braverman J, Mehta A, Oka T, Huang M, Lieb D, Spurrell M, Allen JN, Abrams TA, Clark JW, Enzinger AC, Enzinger PC, Klempner SJ, McCleary NJ, Meyerhardt JA, Ryan DP, Yurgelun MB, Kanter K, Van Seventer EE, Baiev I, Chi G, Jarnagin J, Bradford WB, Wong E, Michel AG, Fetter IJ, Siravegna G, Gemma AJ, Sharpe A, Demehri S, Leary R, Campbell CD, Yilmaz O, Getz GA, Parikh AR, Hacohen N, Corcoran RB. Combined PD-1, BRAF and MEK inhibition in BRAF(V600E) colorectal cancer: a phase 2 trial. Nat Med. 2023;29:458–466. doi: 10.1038/s41591-022-02181-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Lenz HJ, Overman MJ, Van Cutsem E, Limon ML, Wong MK, Hendlisz A, Aglietta M, Garcia-Alfonso P, Neyns B, Gelsomino F, Cardin DB, Dragovich T, Shah U, McCraith SM, Wang A, Lei M, Yao J, Jin L, Lonardi S. First-line (1L) nivolumab (NIVO) + ipilimumab (IPI) in patients (pts) with microsatellite instability-high/mismatch repair deficient (MSI-H/dMMR) metastatic colorectal cancer (mCRC): 64-month (mo) follow-up from CheckMate 142. J. Clin. Oncol. 2023;41:3550. [Google Scholar]
  • 46.Lee MS, Loehrer PJ, Imanirad I, Cohen S, Ciombor KK, Moore DT, Carlson CA, Sanoff HK, McRee AJ. Phase II study of ipilimumab, nivolumab, and panitumumab in patients with KRAS/NRAS/BRAF wild-type (WT) microsatellite stable (MSS) metastatic colorectal cancer (mCRC) J. Clin. Oncol. 2021;39:7. [Google Scholar]
  • 47.NIH-ClinicalTrials.gov. PhII Trial Panitumumab, Nivolumab, Ipilimumab in Kras/Nras/BRAF Wild-type MSS Refractory mCRC. 2020. https://classic.clinicaltrials.gov/ct2/show/ NCT03442569.
  • 48.Skoulidis F, Li BT, Dy GK, Price TJ, Falchook GS, Wolf J, Italiano A, Schuler M, Borghaei H, Barlesi F, Kato T, Curioni-Fontecedro A, Sacher A, Spira A, Ramalingam SS, Takahashi T, Besse B, Anderson A, Ang A, Tran Q, Mather O, Henary H, Ngarmchamnanrith G, Friberg G, Velcheti V, Govindan R. Sotorasib for lung cancers with KRAS p.G12C mutation. N Engl J Med. 2021;384:2371–2381. doi: 10.1056/NEJMoa2103695. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Blair HA. Sotorasib: first approval. Drugs. 2021;81:1573–1579. doi: 10.1007/s40265-021-01574-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Fakih MG, Kopetz S, Kuboki Y, Kim TW, Munster PN, Krauss JC, Falchook GS, Han SW, Heinemann V, Muro K, Strickler JH, Hong DS, Denlinger CS, Girotto G, Lee MA, Henary H, Tran Q, Park JK, Ngarmchamnanrith G, Prenen H, Price TJ. Sotorasib for previously treated colorectal cancers with KRAS(G12C) mutation (CodeBreaK100): a prespecified analysis of a single-arm, phase 2 trial. Lancet Oncol. 2022;23:115–124. doi: 10.1016/S1470-2045(21)00605-7. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ajcr0014-0796-f6.pdf (397.2KB, pdf)

Articles from American Journal of Cancer Research are provided here courtesy of e-Century Publishing Corporation

RESOURCES