
==== Front
Thorac Cancer
Thorac Cancer
10.1111/(ISSN)1759-7714
TCA
Thoracic Cancer
1759-7706
1759-7714
John Wiley & Sons Australia, Ltd Melbourne

39081050
10.1111/1759-7714.15408
TCA15408
Original Article
Original Article
Homologous recombination deficiency status predicts response to immunotherapy‐based treatment in non‐small cell lung cancer patients
Gao et al.
Gao Ai https://orcid.org/0000-0001-7526-6792
1
Wang Xin 1
Wang Jing 1
Zhong Diansheng 1 zhongdsh@hotmail.com

Zhang Linlin 1 zllcaroline@tmu.edu.cn

1 Department of Medical Oncology Tianjin Medical University General Hospital Tianjin China
* Correspondence
Linlin Zhang and Diansheng Zhong, Department of Medical Oncology, Tianjin Medical University General Hospital, No. 154, Anshan Dao, Heping District, Tianjin 300052, China.
Email: zllcaroline@tmu.edu.cn and zhongdsh@hotmail.com

30 7 2024
9 2024
15 25 10.1111/tca.v15.25 18421853
11 6 2024
05 3 2024
04 7 2024
© 2024 The Author(s). Thoracic Cancer published by John Wiley & Sons Australia, Ltd.
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the terms of the http://creativecommons.org/licenses/by-nc-nd/4.0/ License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made.

Abstract

Background

Homologous recombination deficiency (HRD) is a biomarker that predicts response to ovarian cancer treatment with poly (ADP‐ribose) polymerase (PARP) inhibitors or breast cancer treatment with first‐line platinum‐based chemotherapy. However, there are few studies on the prognosis of lung cancer patients treated with immune checkpoint inhibitor (ICI) therapy using HRD as a biomarker.

Methods

We studied the relationship between HRD status and the effectiveness of first‐line ICI‐based therapy in EGFR/ALK wild‐type metastatic non‐small cell lung cancer patients (NSCLC) patients.

Results

This study included 22 treatment naïve NSCLC patients. The HRD score ranged from −26.37 to 92.34, with an average of 24.57. Based on analysis of the progression‐free survival (PFS) data from the included NSCLC patients, threshold traversal was carried out. HRD (+) was defined as an HRD score of 31 or higher. Kaplan–Meier PFS survival analysis showed prolonged median PFS (mPFS) in NSCLC patients with HRD (+) versus HRD (−) (N/A vs. 7.0 ms, log‐rank p = 0.029; HR 0.20, 95% CI: 0.04–0.96, likelihood‐ratio p = 0.03). In patients with PD‐L1 TPS ≥50% and HRD score ≥31 (co‐status high), the mPFS was temporarily not reached during the follow‐up period. In patients with PD‐L1 TPS <1% and HRD score <31, the mPFS was 3 ms. Cox regression analysis showed that the hazard ratio of the co‐status was 0.14 (95% CI: 0.04–0.54), which was a good prognostic factor, and the prognostic effect of co‐status was better than that of HRD score alone.

Conclusion

The HRD status can be identified as an independent significance in NSCLC patients treated with first‐line ICI‐based therapy.

In this study of 22 EGFR/ALK wild‐type metastatic NSCLC who received standard first‐line ICI‐based therapy, the presence of HR deficiency (defined as a HRD score ≥31) and the concept of PD‐L1 and HRD score combined as a co‐status are important predictors of response to standard chemotherapy combined with ICI therapy.

HRD
immunotherapy
non‐small cell lung cancer
National Nature Science Foundation of China 10.13039/501100001809 82103045 82000113 Beijing Xisike Clinical Oncology Research Foundation 10.13039/100018904 Y‐XD202001‐0332 Y‐zai2021/ms‐0247 Tianjin Pharmaceutical Young and Middle aged Research ProjectTJYX2023‐12 source-schema-version-number2.0
cover-dateSeptember 2024
details-of-publishers-convertorConverter:WILEY_ML3GV2_TO_JATSPMC version:6.4.8 mode:remove_FC converted:02.09.2024
Gao A , Wang X , Wang J , Zhong D , Zhang L . Homologous recombination deficiency status predicts response to immunotherapy‐based treatment in non‐small cell lung cancer patients. Thorac Cancer. 2024;15 (25 ):1842–1853. 10.1111/1759-7714.15408

Ai Gao is the first author.
==== Body
pmcINTRODUCTION

Lung cancer, particularly non‐small cell lung cancer (NSCLC), remains the leading cause of cancer‐related death worldwide. 1 Immune checkpoint inhibitor (ICI) therapy targeting programmed death receptor 1 (PD‐1) and its ligand PD‐L1 has a significant impact on the prognosis of lung cancer patients, and it is now a recognized treatment. 2 , 3 These therapies have been approved for the first‐line treatment, either alone in tumors with PD‐L1 expression of ≥50% expression or in combination with chemotherapy regardless of receptor status. 4 Despite these benefits, only a limited number of NSCLC patients benefit from ICIs. The reason for this phenomenon may be primary resistance to the growth of super‐progressive tumors and immune‐related adverse events. 5 , 6 Therefore, it is of utmost importance to identify robust predictors that can pinpoint patients who are more likely to respond to anti‐PD‐(L)1 immunotherapy. The expression of PD‐L1 in cancer cells determined by immunohistochemistry is the only biomarker approved by FDA for ICI therapy response. 7 , 8 However, the measurement is influenced by the sites and quality of biopsy samples, antibodies used for detection, and quantification method. 9 , 10 Crucially, while a high PD‐L1 result generally indicates the suitability of ICI therapy, a negative result or low positivity does not necessarily preclude the use of ICI therapy. 11 Therefore, there is an urgent need for biomarkers that can be used to complement immunohistochemical (IHC) staining in order to make informed clinical decisions regarding immunotherapy, especially in patients with PD‐L1 low‐positive or negative status.

The DNA damage response (DDR) system is designed to protect cells from acquiring genome alterations while also monitoring both exogenous and endogenous DNA damage. Multiple mechanisms are involved in the DDR system, including direct repair (DR), base excision repair (BER), mismatch repair (MMR), nucleotide excision repair (NER), homologous recombination repair (HRR) and nonhomologous end‐joining (NHEJ). 12 Defective DDR pathways may promote the growth of cancer. 13 Previous studies revealed that NSCLC patients with more DDR gene mutations had a higher tumor mutational burden (TMB). 14 Additionally, emerging evidence suggests that certain mutations in the DDR pathway are associated with the therapeutic effectiveness of ICI treatment. 15 , 16 , 17 , 18 Among them, several studies have found that loss of homologous recombination (HR) pathway function is related to the efficacy of ICI therapy. 19 , 20 , 21 , 22 Wang and colleagues suggested that comutations of the HRR and MMR pathways (HRR‐MMR) as well as HRR and BER pathways (HRR‐BER) may be useful biomarkers for ICI immunotherapy. 23 This suggests that changes in HR pathway genes are related to ICI therapy response. However, they do not provide specific criteria to guide the change of HR gene in ICI therapy.

The HRD score is obtained by estimating the degree of potential genomic scar caused by HR deficiency by evaluating genome‐wide loss of heterozygosity (LOH), telomeric allelic imbalance (TAI), and large‐scale state transitions (LST) using targeted somatic next‐generation DNA sequencing. Currently, the FDA has approved the evaluation of patients' HR pathway function by detecting and combining LOH, TAI, and LST scores (called HRD score). 24 The HRD score showed a clear correlation with sensitivity to platinum‐based chemotherapy and efficacy of poly (ADP‐ribose) polymerase (PARP) inhibitors in ovarian and breast cancer. 25 These findings indicate that the HRD score can reflect the loss of HR pathway function in tumors.

In a recent study, Zhou and colleagues performed whole‐exome sequencing (WES) on 14 patients receiving immuno‐neoadjuvant therapy, analyzed the genetic changes in the samples, and linked HR gene deactivation and the resulting HRD event to improved immuno‐neoadjuvant treatment outcomes in NSCLC patients. They validated their findings using public cohorts. 26 This result suggests that the HRD score may be a potential biomarker or prognostic factor in the treatment of lung cancer. Therefore, we speculate that it may be of great significance to explore the role of the HRD score in predicting the response to first‐line ICI‐based treatment in NSCLC patients.

Herein, we performed genome‐wide SNP analysis on 22 patients receiving first‐line ICI treatment combined with chemotherapy, with the aim of evaluating the HRD score, which is defined as the unweighted sum of LOH, TAI, and LST. We tested the predictive ability of the HRD score threshold and analyzed the genetic changes of the samples, aiming to identify indicators for first‐line ICI‐based treatment in metastatic NSCLC patients.

METHODS

Description of clinical studies

This study was a prospective, single‐arm, exploratory clinical study. Patients with a diagnosis of NSCLC were included if their age were 18 years or older, with a life expectancy >3 months, and at least one measurable lesion according to RECIST 1.1 criteria. They were chemotherapy naive and EGFR/ALK wild‐type metastatic NSCLC patients (stage IIIB/C or IV unresectable or locally treatable). Patients were excluded from the study if they had previously received any of the following therapies: anti‐PD‐1, anti‐PD‐L1, or anti‐PD‐L2 drugs, or drugs targeting another stimulatory or coinhibitory T cell receptor (e.g., CTLA‐4, OX‐40, CD137). Patients were also excluded if they received systemic treatment with traditional Chinese medicine or immunomodulatory drugs (including thymosin, interferon, interleukin, excluding local use for controlling pleural effusion) with antilung cancer indications within 2 weeks before the first dose, or had undergone major surgical treatment within 3 weeks before the first dose. A total of 10 adenocarcinoma, nine squamous cell carcinoma and three other NSCLC patients were accrued from 2021 to 2022 at Department of Medical Oncology of Tianjin Medical University General Hospital who received first‐line ICI therapy combined with chemotherapy. Table 1 and Table S1 shows the demographics of the recruited participants as well as the treatment features. Before treatment, tissue samples could be provided for the HRD score and immunotherapy markers such as PD‐L1, TMB, and MSI. NSCLC patients received ICIs combined with platinum‐based chemotherapy. The efficacy was evaluated every two cycles. If the disease was controlled (CR + PR + SD) and the adverse reactions were tolerable, the drug could be administered continuously to 2 years, continued until the loss of clinical benefit or toxicity could not be tolerated or evaluated as PD. The disease status was determined using RECIST1.1/iRECIST. RECIST1.1 was the main evaluation standard. At the same time, the iRECIST standard was adopted for confirmation. That is to say, patients with disease progression were judged according to RECIST1.1 standard, and further confirmed according to iRECIST standard, so as to decide whether to take drugs for further observation (Figure S1).

TABLE 1 Patient clinical and demographic data.

	HRD (+)	HRD (−)	p‐value	
Gender			0.6471	
Male	8	7		
Female	3	4		
Age			0.429	
Median	64	67		
Clinical disease stage				
IV	11	11		
Histologic diagnosis			0.6646	
LUSC	4	5		
Nsq‐NSCLC	7	6		
Smoking history			0.6471	
Smoker	7	8		
Nonsmoker	4	3		
PD‐L1 expression			0.6699	
TPS ≥ 1%	5	7		
TPS < 1%	6	4		
Mutation			0.3754	
HRR gene mutation	5	3		
Wild‐type	6	8		
Abbreviations: HRD, homologous recombination deficiency; HRR, homologous recombination repair; LUSC, lung squamous cell carcinoma; Nsq‐NSCLC, nonsquamous‐non‐small cell lung cancer; TPS, tumor proportion score.

The primary endpoint was the correlation between the HRD score and ORR. The second endpoint was the correlation between the HRD score and PFS.

HRD (+): HRD scores 31 or more. Co‐status high: HRD score ≥31 and TPS >50%, co‐status low: HRD score <31 and TPS <1%, co‐status medium: patients did not belong to the above two categories.

Targeted hybridization capture and sequencing

Genome‐wide SNP data were collected using the HRD panel, a proprietary hybridization enrichment panel (Roche, Basel, Switzerland) that targets 93 200 SNPs scattered across the human genome. A proprietary capture chip (BGI Genomics, Shenzhen) was used to enrich all coding exons and intron‐exon boundaries (20 base pairs) of homologous recombination repair (HRR), which included genes related to gynecological oncology and hereditary risk‐associated genes. The MGISEQ‐2000 platform (MGI Tech Co., Ltd.) was used to sequence enriched DNA samples with 100‐bp pair‐end reads. For the HRD panel, the average sequencing depth of tissue samples must be greater than 150x, and for the HRR genes, the average sequencing depth of tissue and blood samples must be at least 500x. The following 24 HRR pathway genes were included in the definition of HRR mutation positive: ATM, BRCA1, BRCA2, ATR, BARD1, BLM, BRIP1, CHEK2, MRE11A, NBN, PALB2, RAD51C, RAD51D, RBBP8, SLX4, XRCC2, FANCA, FANCC, FANCD2, FANCM, FANCG, FANCL. Based on a review of the literature, these genes were chosen since they were likely to affect the HRR pathway when altered.

HRD event quantification

The HRD score was generated by a genomic scar analysis algorithm allele‐specific gene‐scar analysis tool for diagnosis (ASGAD). 27 The LOH, TAI, and LST in gDNA isolated from FFPE tumor tissue specimens can be measured using ASGAD. Meanwhile, the algorithm also considered the variations in ploidy and purity of tumor tissue. The raw sequence data were filtered and mapped to the human genome (hg19) using BWA aligner 0.7.17. Local alignment optimization, variant calling and annotation were performed using GATK toolkit 3.2, and VarScan. Variants in the ExAC, 1000 Genomes, and dbSNP datasets with a population frequency greater than 0.1% were eliminated from further study. The remaining variants were annotated using VEP software and interpreted according to the Genetic Variation Annotation Standards and Guidelines (2015 edition) issued by the American College of Medical Genetics (ACMG) for germline mutation, and the “Cancer mutation interpretation of guidelines and standards (2017 edition)” for somatic mutation, respectively. Gene variants were named according to the Human Genome Variation Society [HGVS]; http://www.hgvs.org/.

Immunohistochemical staining and antibody of PD‐L1

Immunohistochemical detection of PD‐L1 was performed using 22C3 antibody detected by Tianjin Huada Medical Laboratory Co., Ltd., approved as a tumor marker for NSCLC immunotherapy. The expression level of PD‐L1 in patients was evaluated by combining tumor and immune cell staining. The tumor proportion score (TPS), that is, the percentage of the number of tumor cells with positive staining in the number of all tumor cells, was used so as to predict the survival benefit of patients using immunotherapy based on the test results.

Compliance with ethical standards

The experimental protocol was approved by the Human Ethics Committee of Tianjin Medical University General Hospital. Written informed consent was obtained from individual or guardian participants. The authors declare no competing interests.

Statistical analysis

All statistical analysis was carried out with R version 3.6.1 (R Core Team, 2013) with an α of 0.05. Fisher's exact test and one‐way analysis of variance (ANOVA) were used as statistical methods in this study. Every p‐value that was reported was two‐sided. The threshold for statistical significance was p < 0.05.

RESULTS

Establishing a threshold for the combined HRD score

The clinical trial cohort that determined the threshold of combined HRD score came from 22 EGFR/ALK wild‐type metastatic NSCLC patients. Table S1 shows the patient demographic and clinical characteristics of the patients with evaluable HR status. To quantify HRD events, three HRD‐related events including TAI, LST and LOH were compared between the lung squamous cell carcinoma (LUSC)/nonsquamous‐non‐small cell lung cancer (Nsq‐NSCLC) group and male/female groups. Overall, there was no statistically significant difference in the distribution of LOH, TAI, LST and HRD score between the LUSC/Nsq‐NSCLC groups and male/female groups (Figure 1a,b). Given that tobacco smoke is a known carcinogen, which is highly correlated with high scores of TAI, LST, LOH and HRD in lung adenocarcinoma and head and neck cancer, but there is no significant difference in bladder or LUSC, 28 we also compared patients with a history of smoking and patients without a history of smoking. This showed no significant increase in LOH, TAI, LST and HRD score for smokers relative to nonsmokers (Figure 1c), which may be due to our small sample size and the fact that nine of our patients had LUSC. These results indicated that HRD score related indicators had little correlation with pathological subtypes of lung cancer, patient gender and smoking status in our cohort, larger studies are needed to confirm the findings.

FIGURE 1 Establishing a threshold for the combined homologous recombination deficiency (HRD) score. (a) Loss of heterozygosity (LOH), telomeric allelic imbalance (TAI), large‐scale state transitions (LST) and HRD score distribution between male/female groups. (b) LOH, TAI, LST and HRD score distribution between lung squamous cell carcinoma (LUSC)/nonsquamous‐ non‐small cell lung cancer (Nsq‐NSCLC) groups. (c) LOH, TAI, LST and HRD score distribution between smokers/nonsmokers. (d) HRD score distribution in the training set.

The distribution of HRD scores in the training set is shown in Figure 1d, ranging from −26.37 to 92.38, with an average HRD score of 24.57. The HRD threshold was chosen because of its great sensitivity in diagnosing HR deficit in lung cancer. The threshold was set by traversing the candidate HRD scores, and the score corresponding to the best HR was used as the optimal threshold, consequently HRD (+) was defined as HRD scores of 31 or more. A total of 11 of 22 (50%) tumors were HRD (+), the other 11 patients were HRD (−). This is consistent with a previous study where the median HRD score was 24 in lung adenocarcinoma and 34 in LUSC patients. 28

HRD score association with HRR mutation

The higher frequency of high HRD scores is consistent with the higher proportion of BRCA1/2‐mutated tumors in breast cancer patients. 24 An increase in LOH due to HRR gene mutation has also been confirmed in NSCLC patients. 26 However, there was no significant difference in the HRD score among the HRR gene mutation groups in ovarian cancer. 29 We next investigated whether the higher frequency of high HRD scores was consistent with the higher proportion of HRR mutation in our cohort. An HRR gene panel test was successfully performed in all 22 patients. All mutations were previously identified in the BIC database, or classified as harmful based on their nature (nonsense, frameshift, alternate splicing, or deletion). The specific HRR mutations identified in the 24 HRR genes were: BRCA1, BRCA2, BRIP1, RAD54L, RAD51D, RAD51C, CDK12, ATM, BARD1, CHEK2, FANCL, PALB2. A total of eight patients had at least one mutation in a candidate HRR gene (Figure 2a). A total of four patients had a pathogenic or suspected pathogenic mutation in at least one HRR gene, three of which were HRD (+), and these cases had a mean HRD score of 37.17 compared to 28.72 for cases without evidence of biallelic alteration (p = 0.8047; Figure 2b). Subsequently, we investigated the connection between the HRD score and candidate HRR pathway gene mutations (Figure 2c). The HRD score did not differ significantly between the HRR gene mutation and wild‐type groups. As BRCA 1/2 is the key gene of the HR pathway, we studied the relationship between the HRD score and BRCA gene mutations. Of these, 18.2% (4/22) had apparent biallelic alterations in BRCA, with a mean HRD score of 19.06 compared to 32.75 for patients without evidence of biallelic alteration (p = 0.7867; Figure 2d), with no significant difference. Therefore, we hypothesize that HRR gene mutations do not necessarily lead to an increase in HRD score in NSCLC patients.

FIGURE 2 Homologous recombination deficiency (HRD) score association with homologous recombination repair (HRR) mutation. (a) Deletion status of 12 homologous recombination (HR) core pathway genes in our cohort. (b) HRD score distribution between mutation/wildtype groups. (c) HRD score distribution between mutation/wild‐type groups. (d) HRD score distribution between BRCA1/2 mutation/wildtype groups.

Association of HRD score with response to first‐line ICI‐based therapy

A recent study linked HR gene deactivation and the consequent HRD event to better results of immuno‐neoadjuvant therapy in NSCLC. 26 Next, we verified whether HRD score can be used as a quantitative marker of HR injury, so as to predict the efficacy of first‐line chemotherapy combined with ICI therapy in the treatment of patients with advanced NSCLC. Patients enrolled were categorized as SD, PR and PD using the definition described in the method. The distributions of all passing HRD scores by clinical response class for our cohort are shown as box plots in Figure 3a. The disease control rates (DCR) group showed higher HRD scores than the non‐DCR group with a significant difference (median: 33.18 vs. 20.33, p = 0.0093, Figure 3b), indicating that a high HRD score predicts better therapeutic effect. The objective response rate (ORR) was 45.45% (5/11) in HRD(+) versus 27.27% (3/11) in HRD(−) patients (Figure 3c), and the DCR was 100% (11/11) in HRD(+) versus 54.55% (6/11) in HRD(−) patients (Figure 3d), although the difference was not significant. We next sought to investigate the correlation between HRD and PFS in the validation cohort by performing a Kaplan–Meier survival analysis. As shown in Figure 3e, disease progression occurred in all 10 patients with HRD (−) with a median PFS of 7 months. The median PFS of patients with HRD (+) was not reached at the end of follow‐up, which was markedly longer than the median PFS of patients with HRD (−) (N/A vs. 7.0 ms, log‐rank p = 0.029; HR 0.20, 95% CI: 0.04–0.96, likelihood‐ratio p = 0.03). Taken together, these data suggest that HRD (+) is an indicator of good treatment efficacy in this population of patients.

FIGURE 3 Association of homologous recombination deficiency (HRD) score with response to first‐line immune checkpoint inhibitor‐based therapy. (a) HRD score distribution between PR, SD and PD groups. (b) HRD score distribution between disease control rate (DCR) and non‐DCR groups. (c) Objective response rate (ORR) distribution between HRD+ and HRD− groups. (d) DCR distribution between HRD+ and HRD− groups. (e) Progression‐free survival by HRD status.

Association of HRR mutation with response to ICI‐based therapy

In previous studies, HRR mutation was significantly associated with response to platinum‐based chemotherapy in breast cancer patients 24 and enriched in better‐response NSCLC patients. 26 Next, we attempted to investigate the correlation between HRR mutation and ICI therapy combined with chemotherapy. We defined HRR mutation as patients with at least one mutation in the pathogenic or suspected pathogenic core HR gene, totaling four cases. Patients with no pathogenic or suspected pathogenic core HR mutations were defined as the wild‐type group, with 18 patients in total. The ORR was 0 (0/4) in the HRR mutant group versus 44.44% (8/18) in the HRR wild‐type group (Figure 4a), and the DCR was 100% (4/4) in HRR mutant patients versus 72.22% (13/18) in HRR wild‐type patients (Figure 4b), and the difference was not significant. The PFS in the validation cohort was analyzed by a Kaplan–Meier survival analysis. As shown in Figure 4c, the median PFS of HRR wild‐type patients was not reached at the end of follow‐up, and that of wild‐type patients was 8 months (N/A vs. 8 ms, log‐rank p = 0.67; HR 1.38). In addition, patients with BRCA1/2 mutations did not significantly vary from those without BRCA1/2 mutations in terms of PFS, ORR, and DCR (Figure S2). We also combined HRR status and PD‐L1 expression as a co‐status marker to predict the efficacy of chemotherapy combined with ICI in NSCLC patients, but no significant difference was seen between the two groups (Figure 4d,e). These results indicate that HRR mutation is not a good predictor of treatment outcomes.

FIGURE 4 Correlation analysis between homologous recombination deficiency (HRD) score combined with PD‐L1 expression and prognosis. (a) Objective response rate (ORR) distribution between HRR+ and HRR− groups. (b) Disease control rate (DCR) distribution between HRR+ and HRR− groups. (c) Progression‐free survival by HRR status. (d) ORR distribution between coHRR status high, medium and low groups. (e) Progression‐free survival by coHRR status.

Correlation analysis between HRD score combined with PD‐L1 expression and prognosis

The expression of PD‐L1 in cancer cells determined by immunohistochemistry is the only biomarker of FDA‐approved immunotherapy response. Therefore, we explored whether PD‐L1 expression and HRD score could be used as an additional marker to predict the efficacy of ICI‐based therapy in NSCLC patients. A total of 22 patients were enrolled, who had HRD score and PD‐L1 expression results at the same time. Among the 12 patients with high expression of PD‐L1 (TPS ≥1%), there was no significant difference in PFS and ORR between HRD+ and HRD− patients (Figure 5a,b). Among the 10 patients with low expression of PD‐L1 (TPS <1%), HRD score had a better prognosis effect on lung cancer patients, and the hazard ratio was 0.078 (95% CI: 0.08–0.77), indicating again that HRD score was a better prognosis factor (Figure 5c,d). In addition, patients were divided into three groups based on the results: (1) co‐status high: HRD Score ≥31 and TPS >50%, a total of three patients; (2) co‐status low: HRD score <31 and TPS <1%, a total of four patients; (3) co‐status medium: a total of 16 patients did not belong to the above two categories. A Kaplan–Meier survival curve was drawn using R 3.6.0. According to the analysis results of the above survival curve, HRD score combined with PD‐L1 expression has a good prognosis for lung cancer patients. During the follow‐up period, the mPFS of the co‐status high subgroup was temporarily not reached. The mPFS of the subgroups with negative HRD score and low PD‐L1 expression (TPS <1%) was 3 months. Cox regression analysis showed that the hazard ratio of co‐status was 0.14 (95% CI: 0.04–0.54), which was a good prognostic factor. In conclusion, the prognostic effect of the co‐status was better than that of HRD score alone, and for patients with low PD‐L1 expression, HRD score was an index for further stratification analysis.

FIGURE 5 Association of homologous recombination repair (HRR) mutation with response to immune checkpoint inhibitor‐based therapy. (a) Objective response rate (ORR) distribution between HRD+ and HRD− groups in PD‐L1 high expression groups. (b) Progression‐free survival by homologous recombination deficiency (HRD) status in PD‐L1 high expression groups. (c) ORR distribution between HRD+ and HRD− groups in PD‐L1 low expression groups. (d) Progression‐free survival by HRD status in PD‐L1 low expression groups. (e) ORR distribution between co‐status high, medium and low groups. (f) Progression‐free survival by co‐status.

In this study of 22 EGFR/ALK wild‐type metastatic NSCLC patients who received standard first‐line ICI‐based therapy, the presence of HR deficiency (defined as a HRD score ≥31) and the concept of PD‐L1 and HRD score combined as a co‐status are important predictors of response to standard chemotherapy combined with ICI therapy.

DISCUSSION

In our study, we defined HRD (+) as HRD score of 31 or above. However, in breast and ovarian cancer, HR deficiency was defined as high HRD score (above the HRD threshold, >42) and/or mutant tumor BRCA1/2. 24 , 29 Based on the median HRD score, a recent study classified LUAD and LUSC tumor samples from TCGA into subgroups with high and low HRD scores. 30 They found that the median HRD score was 24 in 501 LUADs and 34 in 482 LUSCs. Our samples included LUADs and LUSCs, and the threshold was set by traversing the HRD scores of the candidates. It is reasonable to define HRD (+) as HRD scores more than 31 and take the score corresponding to the best HR as the optimal threshold.

Interestingly, in the present study, tumor HRR mutation status alone was not a significant predictor of ICI‐based therapy. Although it has been shown in breast and ovarian cancer that the higher frequency of high HRD scores is consistent with the higher proportion of BRCA1/2‐mutated tumors, this has not been shown for any other cancer type. A recent study used a selected list of 20 core HR genes to investigate the LOF changes of HR pathway to clarify the driving genomic changes of HRD high genotypes. 30 They combined the copy number deletion, promoter methylation and somatic mutation information, and found that LOF of BRCA1 and WRN constituted the main contributor to the HRD‐high genotypes in various cancer types, including ovarian serous cystadenocarcinoma, breast invasive carcinoma, lung adenocarcinoma, LUSC and bladder urothelial carcinoma. In our study, LOF alteration of BRCA1 and WRN was not detected in the samples, and we found that other HR gene mutations did not lead to an increase in HRD score of NSCLC patients. In NSCLC, an increase in HRD score may be caused by completely different mechanisms; therefore, in our study, HRR mutation was not a good predictor of treatment outcomes.

Deficiencies in the HR system are well‐known targets for cancer treatment strategies, including surgical intervention, 31 radiation therapy, 32 chemotherapy, 33 and targeted therapy. 34 Previous studies have found that HRD‐generated genomic alteration such as LOH, TAI, and LST have a substantial correlation with patient survival, 35 chemotherapy response, 36 and PARP therapeutic efficacy 37 in cancer. According to recent research, mutations in two main HR pathway genes, BRCA1 and BRCA2, establish a distinct tumor microenvironment and increase breast cancer response to ICI treatment. 38 However, in other cancer immunotherapy, especially in lung cancer immunotherapy, HRD and the resulting genomic scar are rarely concerned. Our study associated the resulting HRD event with the improved outcomes of immunotherapy in NSCLC. However, little has been reported about the mechanism.

The potential mechanism by which HRD status predicts the outcome of combination therapy can be explained as follows: (1) Immune‐sensitive microenvironment: a recent study showed that tumors with high HRD score have increased leukocyte infiltration and lymphocyte fraction, as well as an immune‐sensitive microenvironment. 30 The authors further confirmed that the HRD‐high genotype can predict immune treatment responders in both the TCGA‐BRCA and GSE87049 cohorts comprising over 1000 breast cancer patients. The anti‐PD1/ CTLA4‐treated murine BRCA models showed that HRD‐high tumors harbored cellular immunity‐driven TME throughout the ICI therapy course and in the nontreated original tumors, while anti‐PD1/CTLA4‐sensitive mouse tumor models had a significantly higher HRD score. Therefore, changes in immune‐sensitive TME with increased immune cell fractionation will be shown in lung cancer with high HRD, so as to clarify the mechanism by which increased HRD score leads to better immunotherapy effect in lung cancer. (2) STING pathway: HRD is a genetic instability phenotype that occurs when the DNA repair mechanism responsible for homologous recombination is compromised. This can lead to an accumulation of DNA damage and genomic instability. Other studies have shown that activation of the STING pathway as a result of cytosolic DNA fragment accumulation is an emerging potential mechanism that can foster potent antitumor immune response. 39 , 40 (3) Combination therapy, especially those that include agents that target DDR pathways, can be highly effective in patients with HRD‐positive tumors. The HRD assay is highly associated with sensitivity to platinum‐based chemotherapy. 41 The mechanism behind this is that HRD‐positive cells are already defective in their ability to repair DNA damage, and the addition of DDR‐targeting agents can further exacerbate this defect, leading to increased cell death and tumor regression.

Consistent with the existing knowledge of the blocking mechanism of PD‐1/PD‐L1, PD‐L1 expression was the first response biomarker identified, and so far it is also the most successful guiding therapy choice in locally advanced and metastatic nononcogene addicted NSCLC. 7 , 8 However, there are some limitations in using immunohistochemistry (IHC) to evaluate PD‐L1 TPS of tumor specimens. Mechanically speaking, the expression of PD‐L1 on epithelial cells is a sign of inflammatory microenvironment, which is mainly induced by IFN‐γ produced by surrounding dendritic cells and activated T lymphocytes. 42 , 43 Therefore, in essence, its expression in tissue varies greatly over time and space, even under the same tumor lesion. Furthermore, tumor histology may influence the predictive ability of PD‐L1. Phase III trials of nivolumab after the failure of chemotherapy showed that there was a correlation between the expression of PD‐L 1 in lung adenocarcinoma and the clinical benefit, but in squamous NSCLC, it could not predict the prognosis or curative effect. 44 , 45 In addition to this inherent limitation, some technical problems compromise the reliability of PD‐L1 as a single biomarker. Potential differences in staining qualities and the adoption of different expression cutoff points have long been a matter of debate. 46 On the whole, the expression of PD‐L1 is a weak biomarker when it is used alone, and we are looking for a substitute biomarker for comprehensive evaluation. There was no statistical difference in PFS between patients with high and low PD‐L1 expression in our cohort. We also found co‐status of PD‐L1 expression and HRD score was a good prognostic factor, and the prognostic effect of co‐status was better than that of HRD score alone, which provide alternative biomarkers for use in a composite evaluation.

We acknowledge several limitations relevant to this study. First, the study was exploratory in nature, meaning that it was designed to investigate a potential association between HRD score and PD‐L1 expression with the outcome of combination therapy. The results should therefore be considered preliminary until further validation in larger, randomized controlled trials is available. Second, the sample size of the study was limited, which could have affected the statistical power of the analysis and limited the ability to detect significant associations between HRD score, PD‐L1 expression, and the outcome of combination therapy. Larger studies are needed to confirm the findings and to investigate the prognostic value of these factors in different populations. Finally, the population studied was heterogeneous, with patients receiving different types of combination therapy and with varying baseline characteristics. This could have introduced confounding factors that may have influenced the outcome of the study and limited the ability to generalize the results to other populations. In conclusion, while the study suggests that HRD score and PD‐L1 expression may be useful prognostic factors in patients with advanced NSCLC receiving first‐line combination therapy, the findings should be interpreted with caution given the limitations of the study. Future research is needed to confirm these results and to investigate the role of HRD score and PD‐L1 expression in different populations and treatment settings.

In conclusion, our study associated the resulting HRD event with the improved outcomes of first‐line chemotherapy combined with ICI therapy in NSCLC, offering an unprecedented guide prior to the armamentarium of immuno‐treatments for NSCLC patients.

AUTHOR CONTRIBUTIONS

Ai Gao: Designed and implemented the study, analyzed the data and drafted the manuscript. Xin Wang and Jing Wang: Responsible for sample collection and data interpretation. Linlin Zhang and Diansheng Zhong: Conceived the research, interpreted the results, revised the manuscript and managed the research process.

FUNDING INFORMATION

This study was funded by the National Nature Science Foundation of China (grant nos. 82103045, 82000113), Beijing Xisike Clinical Oncology Research Foundation (grant nos. Y‐XD202001‐0332, Y‐zai2021/ms‐0247), and Tianjin Pharmaceutical Young and Middle aged Research Project (grant no. TJYX2023‐12).

CONFLICT OF INTEREST STATEMENT

The authors confirm there are no conflicts of interest.

Supporting information

Figure S1. Study flow chart of the 22 patients included in this study.

Figure S2. Prognostic analysis between BRCA1/2 mutation and wildtype groups. (a) ORR distribution between BRCA1/2 mutation and wild‐type groups. (b) DCR distribution between BRCA1/2 mutation and wildtype groups. (c) Progression‐free survival by BRCA1/2 status.

Table S1. Patient clinical and demographic data.

ACKNOWLEDGMENTS

All the patients who participated in this study are appreciated. For technical assistance, we also acknowledge Tianjin Medical Laboratory, Beijing Genomics Institution (BGI).

DATA AVAILABILITY STATEMENT

The raw sequence data reported in this study have been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2021) in the National Genomics Data Center (Nucleic Acids Res 2022), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA‐Human: HRA003385) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa-human.
==== Refs
REFERENCES

1 Siegel RL , Miller KD , Jemal A . Cancer statistics, 2020. CA Cancer J Clin. 2020;70 (1 ):7–30.31912902
2 Reck M , Rodríguez‐Abreu D , Robinson AG , Hui R , Csőszi T , Fülöp A , et al. Pembrolizumab versus chemotherapy for PD‐L1‐positive non‐small‐cell lung cancer. N Engl J Med. 2016;375 (19 ):1823–1833.27718847
3 Herbst RS , Baas P , Kim DW , Felip E , Pérez‐Gracia JL , Han JY , et al. Pembrolizumab versus docetaxel for previously treated, PD‐L1‐positive, advanced non‐small‐cell lung cancer (KEYNOTE‐010): a randomised controlled trial. Lancet (London, England). 2016;387 (10027 ):1540–1550.26712084
4 Brueckl WM , Ficker JH , Zeitler G . Clinically relevant prognostic and predictive markers for immune‐checkpoint‐inhibitor (ICI) therapy in non‐small cell lung cancer (NSCLC). BMC Cancer. 2020;20 (1 ):1185.33272262
5 Wang Q , Wu X . Primary and acquired resistance to PD‐1/PD‐L1 blockade in cancer treatment. Int Immunopharmacol. 2017;46 :210–219.28324831
6 Topalian SL , Hodi FS , Brahmer JR , Gettinger SN , Smith DC , McDermott DF , et al. Safety, activity, and immune correlates of anti‐PD‐1 antibody in cancer. N Engl J Med. 2012;366 (26 ):2443–2454.22658127
7 Planchard D , Popat S , Kerr K , Novello S , Smit EF , Faivre‐Finn C , et al. Metastatic non‐small cell lung cancer: ESMO clinical practice guidelines for diagnosis, treatment and follow‐up. Annals of oncology: official journal of the European society for. Med Oncol. 2018;29 (4 ):iv192–iv237.
8 Hanna NH , Schneider BJ , Temin S , Baker S Jr , Brahmer J , Ellis PM , et al. Therapy for stage IV non‐small‐cell lung cancer without driver alterations: ASCO and OH (CCO) joint guideline update. J Clin Oncol. 2020;38 (14 ):1608–1632.31990617
9 Man J , Millican J , Mulvey A , Gebski V , Hui R . Response rate and survival at key timepoints with PD‐1 blockade vs chemotherapy in PD‐L1 subgroups: meta‐analysis of metastatic NSCLC trials. JNCI Cancer Spectr. 2021;5 (3 ):pkab012.34084999
10 Xu Y , Wan B , Chen X , et al. The association of PD‐L1 expression with the efficacy of anti‐PD‐1/PD‐L1 immunotherapy and survival of non‐small cell lung cancer patients: a meta‐analysis of randomized controlled trials. Transl Lung Cancer Res. 2019;8 (4 ):413–428.31555516
11 Cortes J , Cescon DW , Rugo HS , Nowecki Z , Im SA , Yusof MM , et al. Pembrolizumab plus chemotherapy versus placebo plus chemotherapy for previously untreated locally recurrent inoperable or metastatic triple‐negative breast cancer (KEYNOTE‐355): a randomised, placebo‐controlled, double‐blind, phase 3 clinical trial. Lancet. 2020;396 (10265 ):1817–1828.33278935
12 Jiang M , Jia K , Wang L , Li W , Chen B , Liu Y , et al. Alterations of DNA damage response pathway: biomarker and therapeutic strategy for cancer immunotherapy. Acta Pharm Sin B. 2021;11 (10 ):2983–2994.34729299
13 Burrell RA , McGranahan N , Bartek J , Swanton C . The causes and consequences of genetic heterogeneity in cancer evolution. Nature. 2013;501 (7467 ):338–345.24048066
14 Chae YK , Davis AA , Raparia K , Agte S , Pan A , Mohindra N , et al. Association of tumor mutational burden with DNA repair mutations and response to anti‐PD‐1/PD‐L1 therapy in non‐small‐cell lung cancer. Clin Lung Cancer. 2019;20 (2 ):88–96.e86.30425022
15 Mouw KW , Goldberg MS , Konstantinopoulos PA , D'Andrea AD . DNA damage and repair biomarkers of immunotherapy response. Cancer Discov. 2017;7 (7 ):675–693.28630051
16 Asaoka Y , Ijichi H , Koike K . PD‐1 blockade in tumors with mismatch‐repair deficiency. N Engl J Med. 2015;373 (20 ):1979.
17 Germano G , Lamba S , Rospo G , Barault L , Magrì A , Maione F , et al. Inactivation of DNA repair triggers neoantigen generation and impairs tumour growth. Nature. 2017;552 (7683 ):116–120.29186113
18 Ricciuti B , Recondo G , Spurr LF , Li YY , Lamberti G , Venkatraman D , et al. Impact of DNA damage response and repair (DDR) gene mutations on efficacy of PD‐(L)1 immune checkpoint inhibition in non‐small cell lung cancer. Clin Cancer Res. 2020;26 (15 ):4135–4142.32332016
19 Strickland KC , Howitt BE , Shukla SA , Rodig S , Ritterhouse LL , Liu JF , et al. Association and prognostic significance of BRCA1/2‐mutation status with neoantigen load, number of tumor‐infiltrating lymphocytes and expression of PD‐1/PD‐L1 in high grade serous ovarian cancer. Oncotarget. 2016;7 (12 ):13587–13598.26871470
20 Bono JSD , Goh JC , Ojamaa K , Rodriguez JMP , Antonarakis ES . KEYNOTE‐199: Pembrolizumab (pembro) for docetaxel‐refractory metastatic castration‐resistant prostate cancer (mCRPC). J Clin Oncol. 2018;36 (15 ):5007.
21 Chalmers ZR , Connelly CF , Fabrizio D , Gay L , Ali SM , Ennis R , et al. Analysis of 100,000 human cancer genomes reveals the landscape of tumor mutational burden. Genome Med. 2017;9 (1 ):34.28420421
22 Boudadi K , Suzman DL , Luber B , Wang H , Silberstein J , Sullivan R , et al. Phase 2 biomarker‐driven study of ipilimumab plus nivolumab (Ipi/Nivo) for ARV7‐positive metastatic castrate‐resistant prostate cancer (mCRPC). J Clin Oncol. 2017;35 (15 ):5035.
23 Wang Z , Zhao J , Wang G , Zhang F , Zhang Z , Zhang F , et al. Comutations in DNA damage response pathways serve as potential biomarkers for immune checkpoint blockade. Cancer Res. 2018;78 (22 ):6486–6496.30171052
24 Telli ML , Timms KM , Reid J , Hennessy B , Mills GB , Jensen KC , et al. Homologous recombination deficiency (HRD) score predicts response to platinum‐containing neoadjuvant chemotherapy in patients with triple‐negative breast cancer. Clin Cancer Res. 2016;22 (15 ):3764–3773.26957554
25 Flaum N , Crosbie EJ , Edmondson RJ , Smith MJ , Evans DG . Epithelial ovarian cancer risk: a review of the current genetic landscape. Clin Genet. 2020;97 (1 ):54–63.31099061
26 Zhou Z , Ding Z , Yuan J , Shen S , Jian H , Tan Q , et al. Homologous recombination deficiency (HRD) can predict the therapeutic outcomes of immuno‐neoadjuvant therapy in NSCLC patients. J Hematol Oncol. 2022;15 (1 ):62.35585646
27 Mills GB , Timms KM , Reid JE , Gutin AS , Krivak TC , Hennessy B , et al. Homologous recombination deficiency score shows superior association with outcome compared with its individual score components in platinum‐treated serous ovarian cancer. Gynecol Oncol. 2016;141 :2–3.27016222
28 Marquard AM , Eklund AC , Joshi T , Krzystanek M , Favero F , Wang ZC , et al. Pan‐cancer analysis of genomic scar signatures associated with homologous recombination deficiency suggests novel indications for existing cancer drugs. Biomark Res. 2015;3 (9 ):1–10.
29 Feng Z , Shao D , Cai Y , Bi R , Ju X , Chen D , et al. Homologous recombination deficiency status predicts response to platinum‐based chemotherapy in Chinese patients with high‐grade serous ovarian carcinoma. J Ovarian Res. 2023;16 (1 ):53.36922847
30 Yang C , Zhang Z , Tang X , Zhang X , Chen Y , Hu T , et al. Pan‐cancer analysis reveals homologous recombination deficiency score as a predictive marker for immunotherapy responders. Hum Cell. 2022;35 (1 ):199–213.34628623
31 Golshan M , Loibl S , Wong SM , Huober JB , O'Shaughnessy J , Rugo HS , et al. Breast conservation after neoadjuvant chemotherapy for triple‐negative breast cancer. JAMA Surg. 2020;155 (3 ):e195410.31913413
32 Evron E , Ben‐David AM , Goldberg H , Fried G , Kaufman B , Catane R , et al. Prophylactic irradiation to the contralateral breast for BRCA mutation carriers with early‐stage breast cancer. Ann Oncol. 2019;30 (3 ):412–417.30475942
33 Poggio F , Bruzzone M , Ceppi M , Pondé NF , la Valle G , del Mastro L , et al. Platinum‐based neoadjuvant chemotherapy in triple‐negative breast cancer: a systematic review and meta‐analysis. Ann Oncol. 2018;29 (7 ):1497–1508.29873695
34 Robson M , Im SA , Senkus E , Xu B , Domchek SM , Masuda N , et al. Olaparib for metastatic breast cancer in patients with a germline BRCA mutation. N Engl J Med. 2017;377 (6 ):523–533.28578601
35 Loibl S , Weber KE , Timms KM , Elkin EP , Hahnen E , Fasching PA , et al. Survival analysis of carboplatin added to an anthracycline/taxane‐based neoadjuvant chemotherapy and HRD score as predictor of response‐final results from GeparSixto. Ann Oncol. 2018;29 (12 ):2341–2347.30335131
36 Mayer EL , Abramson V , Jankowitz R , Falkson C , Marcom PK , Traina T , et al. TBCRC 030: a phase II study of preoperative cisplatin versus paclitaxel in triple‐negative breast cancer: evaluating the homologous recombination deficiency (HRD) biomarker. Ann Oncol. 2020;31 (11 ):1518–1525.32798689
37 Fasching PA , Link T , Hauke J , Seither F , Jackisch C , Klare P , et al. Neoadjuvant paclitaxel/olaparib in comparison to paclitaxel/carboplatinum in patients with HER2‐negative breast cancer and homologous recombination deficiency (GeparOLA study). Ann Oncol. 2021;32 (1 ):49–57.33098995
38 Samstein RM , Krishna C , Ma X , Pei X , Lee KW , Makarov V , et al. Mutations in BRCA1 and BRCA2 differentially affect the tumor microenvironment and response to checkpoint blockade immunotherapy. Nat Cancer. 2021;1 (12 ):1188–1203.33834176
39 Le DT , Durham JN , Smith KN , et al. Mismatch repair deficiency predicts response of solid tumors to PD‐1 blockade. Science. 2017;357 (6349 ):409–413.28596308
40 Unterholzner L , Dunphy G . cGAS‐independent STING activation in response to DNA damage. Mol Cell Oncol. 2019;6 (4 ):1558682.31211228
41 Telli ML , Hellyer J , Audeh W , Jensen KC , Bose S , Timms KM , et al. Homologous recombination deficiency (HRD) status predicts response to standard neoadjuvant chemotherapy in patients with triple‐negative or BRCA1/2 mutation‐associated breast cancer. Breast Cancer Res Treat. 2018;168 (3 ):625–630.29275435
42 Chen DS , Mellman I . Elements of cancer immunity and the cancer‐immune set point. Nature. 2017;541 (7637 ):321–330.28102259
43 Hashimoto M , Kamphorst AO , Im SJ , Kissick HT , Pillai RN , Ramalingam SS , et al. CD8 T cell exhaustion in chronic infection and cancer: opportunities for interventions. Annu Rev Med. 2018;69 :301–318.29414259
44 Borghaei H , Paz‐Ares L , Horn L , Spigel DR , Steins M , Ready NE , et al. Nivolumab versus docetaxel in advanced nonsquamous non‐small‐cell lung cancer. N Engl J Med. 2015;373 (17 ):1627–1639.26412456
45 Brahmer J , Reckamp KL , Baas P , Crinò L , Eberhardt WEE , Poddubskaya E , et al. Nivolumab versus docetaxel in advanced squamous‐cell non‐small‐cell lung cancer. N Engl J Med. 2015;373 (2 ):123–135.26028407
46 Liberini V , Mariniello A , Righi L , Capozza M , Delcuratolo MD , Terreno E , et al. NSCLC biomarkers to predict response to immunotherapy with checkpoint inhibitors (ICI): from the cells to in vivo images. Cancer. 2021;13 (18 ):4543.
