Association of EPHA3 mutations with tumor mutational burden and efficacy of immune checkpoint inhibitors in cancer patients

doi:10.21203/rs.3.rs-1308057/v1

Download PDF

Research Article

Association of EPHA3 mutations with tumor mutational burden and efficacy of immune checkpoint inhibitors in cancer patients

https://doi.org/10.21203/rs.3.rs-1308057/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: The erythropoietin-producing human hepatocellular receptor A3 (EPHA3), which was identified as a tumor antigen targeted by a lytic T-cell response in melanoma patients, is relatively tumor specific. Moreover, EPHA3 is frequently mutated in human cancers, and the association of EPHA3 mutations with immunotherapy efficacy in cancer patients has not been reported to date.

Methods: We conducted a retrospective analysis of 1756 patients with 11 cancer types after immune checkpoint inhibitor (ICI) monotherapy identified from The Cancer Genome Atlas (TCGA) data portal in April 2021. The associations between EPHA3 mutations and tumor mutational burden (TMB), durable clinical benefit (DCB), and overall survival (OS) was assessed.

Results: Among all 1756 patients, the mutational frequency of EPHA3 was 6.7%. The TMB of patients with EPHA3 mutations was substantially higher than that in patients without the mutations (odds ratio [OR], 5.35, 95% confidence interval [CI], 3.58–8.04; P < .001). EPHA3 mutations were significantly associated with better DCB (median: 52.3% vs 32.8%; OR, 2.24; 95%CI, 1.15–4.41; P = .011) and OS (median: 40.0 vs 16.0 months; hazard ratio, 0.57; 95% CI, 0.42–0.77; P < .001) compared with EPHA3 wild-type. Multivariable and subgroup analyses showed that these associations were independent of all known covariates, including TMB.

Conclusions: These findings indicate that EPHA3 mutations may be associated with higher TMB, better immune response, and better survival than EPHA3 wild-type. Despite the need for validation, cancer patients with EPHA3 mutations might be appropriate candidates for ICI monotherapy.

EPHA3

Immune Checkpoint Inhibitor

Immunotherapy Efficacy

Tumor Mutational Burden

TMB

Immune checkpoint inhibitors (ICIs) are changing the standard of care for many cancer types. Positive predictive markers for ICI treatment include microsatellite instability, elevated tumor mutational burden (TMB), and programmed cell death ligand 1 (PD-L1) overexpression [1]. Recent studies have identified a number of gene mutations as possible ICI predictive markers, including KRAS [2], LRP1B [3], POLE, and POLD1 [4]

The erythropoietin-producing human hepatocellular (EPH) receptor tyrosine kinases control cell to cell adhesion, and contribute to cell migration and axon guidance during development and homeostasis [5, 6]. EPHA3, which encodes EPH receptor A3, is highly expressed in some types of cancer cells, but is almost absent in normal adult tissues, making it relatively tumor specific [7, 8]. Although its role in cancers is largely unknown, EPHA3 was identified as a tumor antigen targeted by a lytic T-cell response in melanoma patients [9], suggesting a potential role in immunotherapy. EPHA3 is frequently mutated in human cancers, including lung adenocarcinoma, melanoma, and colorectal carcinoma [10-13]. To our knowledge, the relationship between EPHA3 mutations and immunotherapy outcomes has not yet been reported. Here, we investigated whether EPHA3 mutations are associated with TMB and clinical outcomes in cancer patients after ICI monotherapy.

Patient selection

The somatic mutational and clinical data from The Cancer Genome Atlas (TCGA) datasets were downloaded using cBioPortal [6]. Patients treated with ICI monotherapy (atezolizumab, avelumab, durvalumab, ipilimumab, nivolumab, pembrolizumab, or tremelimumab) and tumor tissue sequenced by whole genome, whole exome, or a target gene panel including the EPHA3 gene were included. Patients were excluded if they were treated with ICI combination therapy (single ICI combined with chemotherapy, targeted therapy, or another ICI) or if treatment data was missing. This study was deemed exempt from institutional board approval and patient informed consent was waived due to the retrospective nature of the study and use of anonymized data.

TMB classification

TMB was converted into binary variables to minimize the influence of cancer types, sequencing scope (target gene panel, whole-exome, or whole-genome), and study conditions on the analyses regarding TMB. Briefly, subsets were defined as groups of samples of the same cancer type with the same dataset sequencing scope. In each subset, the TMB score was directly used, if provided; otherwise, it was defined as the total number of nonsynonymous alterations [8, 9]. Subsets were excluded if they comprised fewer than 10 samples. TMB in the subsets was classified by percentile: high (top 20%) and non-high TMB (bottom 80%) [1].

Statistical analyses

The primary endpoints were TMB, durable clinical benefit (DCB; complete, partial, or stable response for > 6 months), and overall survival (OS). Between-group comparisons were conducted using Fisher’s exact test for categorical variables and the Wilcoxon rank sum test for continuous variables. The association of EPHA3 mutations with TMB and DCB were assessed using logistic regression models; the results were presented as odds ratios (ORs) with 95% confidence intervals (CIs), using EPHA3 wild-type as the reference. The OS rates were compared using log-rank tests and plotted using the Kaplan–Meier method. The association between EPHA3 mutations and OS was assessed using Cox proportional hazards regression models; the results were presented as hazard ratios (HRs) with 95% CIs, using EPHA3 wild-type as the reference. The multivariate analyses of the associations between EPHA3 mutations and TMB incorporated age, sex, cancer type, line of treatment, sample type, sequencing scope, and tumor purity. TMB was incorporated in the multivariate analyses of associations between EPHA3 mutations and DCB and OS. Subgroup analyses were conducted to evaluate whether the associations between EPHA3 mutations and TMB, DCB, and OS were dependent on these covariates. To conduct a subgroup analysis, at least 10 patients with EPHA3 mutations were required. Two-sided P < 0.05 was considered statistically significant. Analyses were performed using R 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).

Eleven datasets (Additional file 1) including 1756 ICI monotherapy patients involving 11 cancer types were identified. The selection flowchart is shown in Additional file 2. The clinical characteristics of the patients according to EPHA3 status are summarized in Table 1. The main cancer types included melanoma (24.0%, n = 421) and non-small cell lung cancer (NSCLC, 22.7%, n = 398). All patients had locally advanced stage, recurrent or metastatic disease, and received palliative therapy. Among 1756 patients, the mutational frequency of EPHA3 was 6.7%, and the most EPHA3 mutations were found in patients with melanoma (10.0%) and NSCLC (10.3%).

TMB was substantially higher in patients with compared to patients without EPHA3 mutations (Fig. 1A, Table 2; OR, 5.35; 95%CI, 3.58–8.04; P < .001). The EPHA3 mutations were significantly associated with better DCB (Fig. 1B, Table 2; median: 52.3% vs 32.8%; OR, 2.24; 95%CI, 1.15–4.41; P = .011) and OS (Fig. 1C, Table 2; median: 40.0 [95%CI, 26.7–not reached] vs 16.0 [95%CI, 14.4–18.0] months; HR, 0.57; 95%CI, 0.42–0.77; P < .001) compared to EPHA3 wild-type. The associations of EPHA3 mutations and DCB and OS remained statistically significant in the multivariable analyses after controlling for all known covariates, including TMB (Table 2). In the subgroup analysis, all ORs showed that patients with EPHA3 mutations had a higher TMB, and all ORs for DCB and HRs for OS favored EPHA3 mutations except for the urinary tract OS subgroup (Fig. 2).

We investigated whether the association between EPHA3 mutations and clinical outcomes might be attributable to general prognostic benefits of EPHA3 mutations, unrelated to ICI monotherapy. In all 11 datasets included in the analysis, we identified the DCB of 166 patients who received ICI combination therapy and of 273 patients not treated with ICIs, there was no association between EPHA3 mutations and DCB, respectively (Fig. 3). We identified the OS of 306 patients treated with ICI combination therapy, and found no association between EPHA3 mutations and OS (Fig. 3).

To our knowledge, this is the first study to investigate the association of EPHA3 mutations and immunotherapy outcomes in cancer patients. EPHA3 mutations were found to be associated with higher TMB and better DCB and OS compared with EPHA3 wild-type. This association was independent of age, sex, cancer type, line of treatment, sample type, sequencing scope, and tumor purity.

It is well known that higher TMB is associated with better responses to ICI treatment in cancer patients [1]. To determine if these responses were primarily attributable to differences in TMB, we conducted multivariable and subgroup analyses according to TMB. The magnitude of the OS benefit associated with EPHA3 mutations decreased but was still significant after adjustments by TMB. Moreover, subgroup analyses showed OS benefits regardless of TMB. Similar results were indicated regarding DCB. These findings suggested that EPHA3 mutations were likely to be clinically meaningful, and underscored the robustness of EPHA3 mutations as a predictive biomarker of ICIs.

This study had several limitations. First, it was a retrospective analysis with potential selection bias. Second, different EPHA3 mutation subtypes were not compared due to the small number of patients per subgroup. Third, PD-L1 influence on the associations between EPHA3 mutations and TMB, DCB, and OS were not assessed because PD-L1 expression was only known in 108 of 1756 (6.2%, Table 1) patients included in the analysis, and the statistical power was insufficient for reasonable analysis. Finally, we did not assess the regulation of immune system signaling pathways, antigen processing and presentation, and cell cycle checkpoints, which may be related to immunotherapy responses.

This study showed that EPHA3 mutations were associated with higher TMB and better clinical efficacy in terms of DCB and OS in patients treated with ICI monotherapy compared to patients with EPHA3 wild-type. Mechanistic studies into the function of EPHA3 mutations are needed, as well as prospective validation of its predictive role in ICI responses.

CI, confidence interval; DCB, durable clinical benefit; EPH, erythropoietin-producing human hepatocellular; HR, hazard ratio; ICI, immune checkpoint inhibitor; OR, odds ratio; OS, overall survival; PD-L1, programmed cell death ligand 1; TCGA, The Cancer Genome Atlas; TMB, tumor mutational burden.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used during the current study are available from cBioPortal at https://www.cbioportal.org/datasets (Additional file 1).

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the Scientific and Technological Developing Scheme of Ji Lin Province [grant number 20200201603JC]; the growth foundation for Chinese National Natural Science Foundation of Jilin University [grant number 2020-CXM-08]; Jilin Province medical and health talents special project [JLSWSRCZX2021-110].

Authors' contributions

Siqi Zhang, Mengge Zheng, Deheng Nie, and Fujun Han wrote the manuscript; Huimin Tian and Wenjia Liu collected related literature and performed the data analysis; Fengli Pei, Wenhui Liu, helped in drafting figures; Fujun Han provided guidance and revised the manuscript. All authors read and approved the final manuscript.

Acknowledgements

Not applicable.

Authors' information

Not applicable.

Samstein RM, Lee CH, Shoushtari AN, Hellmann MD, Shen R, Janjigian YY, et al. Tumor mutational load predicts survival after immunotherapy across multiple cancer types. Nat Genet. 2019;51:202-6.
Jeanson A, Tomasini P, Souquet-Bressand M, Brandone N, Boucekine M, Grangeon M, et al. Efficacy of Immune Checkpoint Inhibitors in KRAS Mutant Non-Small Cell Lung Cancer (NSCLC). J Thorac Oncol. 2019;14:1095-101.
Chen H, Chong W, Wu Q, Yao YL, Mao M, Wang X. Association of LRP1B Mutation With Tumor Mutation Burden and Outcomes in Melanoma and Non-small Cell Lung Cancer Patients Treated With Immune Check-Point Blockades. Front Immunol. 2019;21:10-1113.
Wang F, Zhao Q, Wang YN, Jin Y, He MM, Liu ZX, et al. Evaluation of POLE and POLD1 Mutations as Biomarkers for Immunotherapy Outcomes Across Multiple Cancer Types. JAMA Oncol. 2019;5:1504-6.
London M, Gallo E. Critical role of EphA3 in cancer and current state of EphA3 drug therapeutics. Mol Biol Rep. 2020;47:5523-33.
Pasquale EB. Eph-ephrin bidirectional signaling in physiology and disease. Cell. 2008;133:38-52.
Day BW, Stringer BW, Al-Ejeh F, Ting MJ, Wilson J, Ensbey KS, et al. EphA3 maintains tumorigenicity and is a therapeutic target in glioblastoma multiforme. Cancer Cell. 2013;23:238-48.
Charmsaz S, Al-Ejeh F, Yeadon TM, Miller KJ, Smith FM, Stringer BW, et al. EphA3 as a target for antibody immunotherapy in acute lymphoblastic leukemia. Leukemia. 2017;31:1779-87.
Chiari R, Hames G, Stroobant V, Texier C, Maillère B, T Boon, et al. Identification of a tumor-specific shared antigen derived from an Eph receptor and presented to CD4 T cells on HLA class II molecules. Cancer Res. 2000;60:4855-63.
Ding L, Getz G, Wheeler DA, Mardis ER, McLellan MD, Cibulskis K, et al. Somatic mutations affect key pathways in lung adenocarcinoma. Nature. 2008;455:1069-75.
Sjöblom T, Jones S, Wood LD, Parsons DW, Lin J, Barber TD, et al. The consensus coding sequences of human breast and colorectal cancers. Science. 2006;314:268-74.
Zhuang GL, Song WQ, Amato K, Hwang Y, Lee K, Boothby M, et al. Effects of Cancer-Associated EPHA3 Mutations on Lung Cancer. J Natl Cancer Inst. 2012;104:1182-97.
Wood LD, Calhoun ES, Silliman N, Ptak J, Szabo S, Powell SM, et al. Somatic mutations of GUCY2F, EPHA3, and NTRK3 in human cancers. Hum Mutat. 2006;27:1060-1.

Table 1

Patient characteristics by EPHA3 status.

Characteristic	Total patients (n=1756)		EPHA3wt (n=1639)		EPHA3m (n=117)		P value
Characteristic	No.	%	No.	%	No.	%	P value
Age (years)							.061^a
< 65	896	50.0	848	51.7	48	41.0
≥ 65	748	42.6	689	42.0	59	50.4
Unknown	112	6.4	102	6.2	10	8.5
Median (range)	63 (15-92)		63 (15-92)		67 (25-87)		.012^a
Sex							.115
Male	1087	61.9	1023	62.4	64	54.7
Female	669	38.1	616	37.6	53	45.3
Cancer type							< .001
Melanoma	421	24.0	379	23.1	42	35.9
NSCLC	398	22.7	357	21.8	41	35.0
Urinary tract	214	12.2	202	12.3	12	10.3
Renal cell carcinoma	152	8.7	152	9.3	0	0
Head and neck	126	7.2	123	7.5	3	2.6
Brain glioma	116	6.6	113	6.9	3	2.6
Colorectal	97	5.5	93	5.7	4	3.4
Esophagogastric	90	5.1	87	5.3	3	2.6
Breast	37	2.1	34	2.1	3	2.6
Hepatocellular carcinoma	27	1.5	26	1.6	1	0.9
Unknown primary	78	4.4	73	4.5	5	4.3
Line of treatment							.021
1	343	19.5	312	19.0	31	26.5
≥ 2	253	14.4	231	14.1	22	18.8
Unknown	1160	66.1	1096	66.9	64	54.7
Sample type							.001
Primary	561	31.9	538	32.8	23	19.7
Metastatic	633	36.0	592	36.1	41	35.0
Unknown	562	32.0	509	31.1	53	45.3
Sequencing							.152
Whole-exome or genome	355	20.2	325	19.8	30	25.6
Target gene panel	1401	79.8	1314	80.2	87	74.4
Tumor purity							.204
<50%	592	33.7	561	34.2	31	26.5
≥50%	458	26.1	425	25.9	33	28.2
Unknown	706	40.2	653	39.8	53	45.3
TMB							<.001
Bottom 80%	1361	77.5	1311	80.0	50	42.7
Top 20%	395	22.5	328	20.0	67	57.3
PD-L1 (%)							.617^a
< 1	60	3.4	55	3.4	5	4.3
≥ 1	48	2.7	44	2.7	4	3.4
Unknown	1648	93.8	1540	94.0	108	92.3
^a: Patients with unknown characteristics was not included in the comparison. Abbreviation: m, mutations; wt, wild-type.

Table 2

Multivariable regression models for associations between EPHA3 status and TMB and immunotherapy outcomes

	TMB		OS		DCB
	OR(95% CI)	P value	HR (95% CI)	P value	OR (95% CI)	P value
Unadjusted	5.35 (3.58-8.04)	<.001	0.57 (0.42-0.77)	<.001	2.24 (1.15-4.41)	.011
Adjusted by other covariables^a	5.86 (4.19-8.19)	<.001	0.57 (0.42-0.77)	<.001	2.70 (1.56-4.67)	.003
Adjusted by TMB	N.A.	N.A.	0.68 (0.50-0.92)	.013	1.60 (0.92-2.79)	.164
Adjusted by TMB and other covariables^a	N.A.	N.A.	0.67 (0.49-0.92)	.012	1.99 (1.11-3.57)	.052
The top 20th percentile was used as the cutoff to dichotomize tumor mutational burden (TMB) into high TMB and non-high TMB. The association between EPHA3 status and TMB and durable clinical benefit (DCB) was assessed using logistic regression models. The results were presented as odds ratios (ORs) with 95% confidence intervals (CIs) using EPHA3 wild-type as the reference. The association between EPHA3 status and overall survival (OS) was assessed using Cox proportional hazards regression models. The results were presented as hazard ratios (HRs) with 95% CIs using EPHA3 wild-type as the reference. ^a: age, sex, cancer type, line of treatment, sample type, sequencing scope, and tumor purity. Abbreviations: N.A., not applicable.

No competing interests reported.

Additionalfile1.doc
Additional file 1.doc. Flow diagram of patient selection from cBioPortal
Additionalfile2.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Association of EPHA3 mutations with tumor mutational burden and efficacy of immune checkpoint inhibitors in cancer patients

Status:

Version 1

Abstract

Figures

Background

Methods

Patient selection

TMB classification

Statistical analyses

Results

Discussion

Conclusion

Abbreviations

Declarations

Ethics approval and consent to participate

Consent for publication

Availability of data and materials

Competing interests

Funding

Authors' contributions

Acknowledgements

Authors' information

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1