The Prognostic Significance of BRCA1-Associated Protein 1(BAP1) in Malignant Pleural Mesothelioma: A Systematic Review and Meta-Analysis

Abstract

Background

Malignant pleural mesothelioma (MPM) originates from pleural mesothelial cells and represents a highly aggressive malignancy. Characterized by a prolonged latency period, poor survival prognosis, and non-specific clinical manifestations, MPM poses significant diagnostic and therapeutic challenges. BRCA-1-associated protein 1 (BAP1) is a nuclear deubiquitinating enzyme involved in chromatin regulation, homologous recombination, and programmed cell death. Although multiple studies have suggested that BAP1 loss or mutation holds prognostic significance in malignant pleural mesothelioma, this remains a matter of debate.

Methods

A meta-analysis was conducted using data from the Web of Science, PubMed, Embase, Cochrane, and CNKI up to October 21, 2025. We included studies involving patients with malignant pleural mesothelioma that assessed the prognostic significance of BAP1. These studies reported hazard ratios (HR) and 95% confidence intervals (CI) for median overall survival (mOS) and median progression-free survival (mPFS).

Results

A total of 14 articles meeting the inclusion criteria were identified, encompassing 1,835 patients. We conducted both univariate and multivariate analyses of mOS, revealing that BAP1 deletion or mutation consistently correlated with improved prognosis, with hazard ratios of 0.51 (95% CI, 0.45–0.58, P < 0.0001) and 0.62 (95% CI, 0.54–0.71, P < 0.0001), respectively. The values were statistically significant (P < 0.0001). Furthermore, analyses of mPFS continued to demonstrate that BAP1 deletion or mutation consistently correlated with improved prognosis (U: HR = 0.62, 95% CI = 0.39–0.99, P = 0.04; M: HR = 0.79, 95% CI = 0.51–1.24, P = 0.31), though this did not reach statistical significance. Sensitivity analyses confirmed the robustness of these findings, with minimal heterogeneity between studies.

Conclusion

BAP1 serves as a potential prognostic marker for MPM, with its deletion or mutation indicating a favourable prognosis for patients. However, whether it provides valuable insights for various treatment strategies, including immunotherapy, requires further clinical investigation to determine.

Introduction

Malignant pleural mesothelioma (MPM) originates from pleural mesothelial cells and is a rare, highly aggressive malignancy closely associated with asbestos exposure, with approximately 80% of patients having a history of asbestos contact. ¹ Malignant pleural mesothelioma has a prolonged latency period, with symptoms potentially emerging 20 to 60 years after asbestos exposure.^2,3 It presents no specific clinical symptoms, with the most common manifestations being chest discomfort and respiratory distress. The majority of patients are diagnosed at an advanced stage. The prognosis for malignant pleural mesothelioma is extremely poor, with an overall survival (OS) period of 9 to 17 months.^4,5 The current guideline recommends pemetrexed in combination with platinum-based chemotherapy as the first-line treatment regimen. ⁶ However, the survival benefit for MPM patients was not significant. The study indicated a response rate of 26.3%, a median time to progression (TTPD) of 7 months, and a one-year survival rate of 63.1%. ⁷ There are three main histological subtypes of MPM: epithelioid, biphasic, and sarcomatoid. Among these, epithelioid MPM is the most common, typically indicating a better prognosis and superior response to chemotherapy. Sarcomatoid MPM is the rarest histological type and carries the poorest prognosis.^8,9 Early studies indicate that the median overall survival for epithelial-like tumours is significantly longer than for non-epithelial-like tumours (12.2 months versus 4.4 months). ¹⁰

In malignant pleural mesothelioma, BRCA-1-associated protein 1 (BAP1) is one of the most frequently mutated genes. ¹¹ BAP1 is a nuclear deubiquitinating enzyme involved in chromatin regulation, homologous recombination, and programmed cell death. ¹² Germline and somatic alterations in BAP1 have been described in multiple cancers, including uveal and cutaneous melanoma, clear cell renal cell carcinoma, and approximately 54% of malignant pleural mesothelioma. ¹³ Over the past several years, numerous studies have demonstrated the prognostic role of BAP1 expression in malignant pleural mesothelioma (MPM) and the potential association between BAP1 status and tumour response to chemotherapy. Nevertheless, findings remain controversial. Although some investigations suggest that loss of BAP1 expression correlates with a more favourable prognosis in MPM,^14,15 other studies indicate that there is no correlation between BAP1 status and survival. ¹⁶ The study indicates that BAP1 deletion or mutation is more prevalent in the epithelioid subtype. For epithelioid malignant pleural mesothelioma, patients with BAP1 deletion or mutation exhibit longer overall survival. ¹⁷ Additionally, a study indicates that BAP1 deletion or mutation enriches immune-related pathways in MPM, leading to increased mRNA signatures of the interferon-α/γ response, activation of dendritic cells, immune checkpoint receptors, and T-cell inflammation. Consequently, MPM with BAP1 deletion or mutation may benefit from immunotherapy. ¹⁸ However, there remains a lack of relevant clinical research data to substantiate this view.

In recent years, prospective and retrospective studies have progressively clarified the relationship between BAP1 expression and prognosis. However, no meta-analysis synthesising the prognostic role of BAP1 based on existing research has been conducted. Therefore, we summarised the current evidence through a systematic review and meta-analysis to determine the prognostic significance of BAP1 expression in MPM.

Materials and Methods

Search Strategy

The systematic literature search was conducted using the following search terms:

• (1) ‘malignant pleural mesothelioma’ OR ‘pleural mesothelioma’ OR ‘MPM’

• (2) ‘BRCA1-associated protein 1’ OR “BAP1” OR ‘BRCA1-associated protein-1’ OR ‘BRCA1 associated protein 1’

The search query was constructed as: (1) AND (2).

The search was performed in the following databases: Web of Science, PubMed, Embase, and Cochrane Library, up to 21 October 2025. Additionally, the China National Knowledge Infrastructure (CNKI) was searched using equivalent Chinese search terms.

Two authors independently screened the titles and abstracts of primary studies identified through electronic searches. Duplicate studies were excluded, and any discrepancies were resolved through discussion with a third author. Reference lists of all retrieved articles meeting the inclusion criteria were also cross-checked. The final list of included studies was determined through a process of discussion, integration, and synthesis. The protocol for this systematic review and meta-analysis was registered on the International Registry of Systematic Review and Meta-Analysis Protocols (PROSPERO: CRD420251175905). Furthermore, this research report complies with the PRISMA 2020 guidelines ¹⁹ and follows the REMARK (Reporting Recommendations for Tumor Marker Prognostic Studies) guidelines where applicable. ²⁰

Inclusion and Exclusion Criteria

We evaluated only human studies meeting the following inclusion criteria: (1) full-text observational cohort studies (prospective and retrospective) concerning MPM; (2) assessable BAP1 status determined via Immunohistochemistry (IHC) or molecular methods (e.g., mutation, deletion); (3) sufficient data to calculate the correlation between BAP1 expression and median overall survival (mOS) or median progression-free survival (mPFS).

Exclusion criteria: (1) Studies concerning cancers other than MPM (e.g., malignant peritoneal mesothelioma); (2) Studies lacking reported patient survival outcomes or failing to assess BAP1 status; (3) Where multiple studies from the same institution or authors were reported, only the most recent or highest-quality study (with the largest reported patient cohort) was included.

Data Collection and Evaluation

Data extraction from the selected articles was conducted independently by two researchers, with a third researcher overseeing the synthesis. The collected data comprised fundamental details, including authors, publication year, study region, sample size, study type, cut-off values, treatment strategies, and outcomes (including hazard ratio estimates). To assess the quality of included studies, the Newcastle-Ottawa Scale (NOS) was employed. Studies scoring no less than 6 points on the NOS were categorised as high-quality research.

Statistical Analysis and Data Analysis

We extracted the following information from each study: author names, publication year, number of patients, histological type, mOS, mPFS, hazard ratio (HR), 95% confidence interval (95%CI), p-value(P), and other relevant details. The primary objective of this meta-analysis was to determine the hazard ratio for death in patients whose tumours retained BAP1 expression (BAP1-preserved) compared to those whose tumours did not express BAP1 (BAP1-lost or mutated). An HR > 1 indicates a higher risk of death in MPM patients with BAP1 loss or mutation. An HR < 1 signifies a lower risk. When considering HR values obtained from univariate and multivariate analyses, priority was given to the HR value derived from the multivariate analysis.

We conducted both univariate and multivariate meta-analyses for the following reasons. Univariate analyses were performed to evaluate the crude association between BAP1 status and survival outcomes without adjusting for confounding factors, providing an overall estimate of the prognostic effect. Multivariate analyses were performed to account for potential confounders (such as age, sex, histological subtype, and treatment modality) that may influence survival outcomes, thereby providing a more accurate estimate of the independent prognostic value of BAP1. By presenting both analyses, we aimed to allow readers to assess the impact of confounding factors on the observed association and to enhance the robustness of our conclusions. When considering HR values obtained from both univariate and multivariate analyses, priority was given to the HR value derived from the multivariate analysis, as it provides a more conservative and reliable estimate.

We employed R 4.5.0 (R Foundation for Statistical Computing) and STATA 18.0 software (STATA Corporation, College Station, Texas, USA) for the statistical analysis and forest plot generation of this meta-analysis. To assess the prognostic significance of BAP1 in patients with malignant pleural mesothelioma, we combined the HR and its 95% CI for each outcome. Unless otherwise specified, the HR from multivariate analyses was prioritised. HR and its 95% CI were pooled to evaluate the association between BAP1 mutation or deletion and survival prognosis in patients with malignant pleural mesothelioma. The DerSimonian-Laird random-effects model was employed to calculate pooled HRs with corresponding 95% CIs comparing BAP1-deficient or -mutated patients with BAP1-preserved MPM patients. This method was selected a priori due to the heterogeneity of observational studies. Inconsistencies between study results were assessed using the Higgins I² index and chi-square statistic. Publication bias was investigated through visual inspection of funnel plots. Sensitivity analyses were conducted by sequentially excluding individual studies to evaluate the robustness. Assess publication bias using Begg’s test, Egger’s test, and the trim-and-fill method. Statistical significance was defined as a two-sided P value < 0.05.

Results

Search Results and Baseline Characteristics

Following the predefined retrieval strategy, we initially identified 1,346 articles from Web of Science (n=552), PubMed (n=232), Embase (n=498), Cochrane (n=22) and CNKI(n=42). After removing duplicates, 819 articles remained. During the literature review process, one additional relevant article was included, bringing the total to 820 articles. Following review of titles and abstracts, 777 articles were excluded for reasons including being reviews, animal or cell studies, unrelated to disease or BAP1, addressing the prognostic significance of BAP1, Chinese articles or lack of full-text availability. Full-text reviews were then conducted on the remaining 43 articles, leading to the exclusion of 29 articles for reasons including absence of BAP1 subgroup analysis (n = 18), and lack of survival data (n = 11). Ultimately, 14 articles were selected for this meta-analysis (Figure 1).

Figure 1.

PRISMA 2020 flow diagram illustrating the study selection process for the meta-analysis on the prognostic role of BRCA1-associated protein 1 (BAP1) in malignant pleural mesothelioma (MPM). The diagram presents the number of records identified from each database (Web of Science, PubMed, Embase, Cochrane, and CNKI), the number of duplicates removed, records screened, full-text articles assessed for eligibility, and the final number of studies included in the qualitative synthesis and quantitative meta-analysis. CNKI, China National Knowledge Infrastructure

These retrospective or prospective studies were conducted in Turkey, ²¹ Australia,^14,15,22,23 Denmark, ²² Italy,^24-26 Israel, ²⁷ the United States,^28-31 Egypt, ³² and other regions, with publications up to 21 October 2025. All included patients were diagnosed with malignant pleural mesothelioma, totalling 1,835 individuals. These patients may have undergone treatment strategies including surgery, chemoradiotherapy, targeted therapy, or immunotherapy. Patient numbers per study ranged from 17 to 369, with groups stratified according to BAP1 gene mutation status. Twelve studies reported mOS, whilst five studies provided mPFS. Among the twelve studies reporting mOS, five studies used HR estimated via univariate and multivariate analyses, two studies reported HR based solely on univariate analysis, and three studies reported HR based solely on multivariate analysis. Among the five studies reporting mPFS, one study estimated HR using both univariate and multivariate analyses, one reported HR based solely on univariate analysis, and one reported HR based solely on multivariate analysis. All selected studies were deemed high quality according to their NOS scores. Further details are presented in Table 1.

Table 1.

Main Characteristics of Studies Included in the Meta-Analysis

Author(year)	Study region	Research Type	Sample size	mOS	Treatment strategy	Testing method	Epithelioid (BAP1 deletion or mutation)	Biphasic(BAP1 deletion or mutation)	Sarcomatoid(BAP1 deletion or mutation)	BAP1 deletion or mutation(n)	BAP1 deletion or mutation	BAP1 positive	HR(95%CI)	P-value	Quality score
Univariate analysis
Yuce(2024)	Turkey	prospective	107	12	pemetrexed-cisplatin or pemetrexed-carboplatin	IHC	74(48)	25(18)	8(3)	69	mOS 14	mOS 10	0.71(0.47-1.06)	0.095	7
Louw(2022)D	Danish	retrospective	114	11.3	Platinum and Pemetrexed or BSC	IHC	55(37)	43(27)	12(4)	71	mOS 14.2	mOS 6.3	0.44(0.29-0.65)	< 0.001	8
Louw(2022)A	Australia	retrospective	234	14.3	Platinum and Pemetrexed or BSC	IHC	85(58)	23(6)	30(4)	142	mOS 18.9	mOS 10.8	0.53(0.40-0.69)	< 0.001	8
Fuchs(2022)	Australia	retrospective	369	8.2	NS	IHC	220	NS	NS	178	mOS 15.98	mOS 6.99	0.50(0.40-0.64)	<0.0001	8
Dudnik(2021)	Israel	retrospective	45	NS	Platinum-based drugs and pemetrexed, either combined with or without anti-angiogenic drugs	NGS	(7)	(1)	(0)	8	mOSDx 98.3	mOSDx 19.4	0.69(0.22-2.17)	0.53	8
											mPFS 9.1	mPFS 9.2	1.01(0.41-2.50)	0.97
Signorelli(2020)	Italy	retrospective	56	NS	MTO	IHC, NGS	NS	NS	NS	14	mOS NS	mOS NS	0.88(0.43-1.73)	0.73	8
Righi(2016)	Italy	retrospective	143	15	NS	IHC, FISH	107(74)	13(10)	23(5)	89	mOS NS	mOS NS	0.49(0.34-0.69)	<0.0001	9
Farzin(2015)	Australia	retrospective	229	9	NS	IHC	120(75)	42(19)	67(12)	106	mOS 16.11	mOS 6.34	0.39(0.28-0.55)	<0.01	8
McGregor(2015)	US	retrospective	111	NS	NS	IHC	58(47)	43(21)	10(0)	68	mPFS NS	mPFS NS	0.53(0.31-0.92)	0.0068	8
Zauderer(2013)	US	retrospective	121	NS	MTO	Sanger sequencing	NS	NS	NS	24	mPFS 14.3	mPFS 14.8	NS	NS	8
Multivariate analysis
Yuce(2024)	Turkey	prospective	107	12	pemetrexed-cisplatin or pemetrexed-carboplatin	IHC	74(48)	25(18)	8(3)	69	mOS 14	mOS 10	0.54(0.35-0.83)	0.004	7
Illini(2024)	Austria	retrospective	52	16.9	MTO	IHC	48(33)	3(3)	1(0)	36	mOS 20	mOS 11.3	0.80(0.28-2.28)	0.683	8
Louw(2022)D	Danish	retrospective	114	11.3	Platinum and Pemetrexed or BSC	IHC	55(37)	43(27)	12(4)	71	mOS 14.2	mOS 6.3	0.40(0.25-0.65)	< 0.001	8
Louw(2022)A	Australia	retrospective	234	14.3	Platinum and Pemetrexed or BSC	IHC	85(58)	23(6)	30(4)	142	mOS 18.9	mOS 10.8	0.68(0.50-0.91)	0.01	8
Fuchs(2022)	Australia	retrospective	369	8.2	NS	IHC	220	NS	NS	178	mOS 15.98	mOS 6.99	0.67(0.52-0.85)	0.001	8
Murrone(2021)	Italy	retrospective	55	14.6	platinum-based chemotherapy	IHC	41(29)	7(1)	7(3)	33	mOS 14.9	mOS 11.3	0.75(0.35-1.61)	0.456	8
											mPFS 3.8	mPFS 5.5	0.89(0.43-1.86)	0.754
Cantini(2020)	US	retrospective	85(60)	14.76	platinum-based chemotherapy	IHC	56(32)	10(3)	13(2)	37	mOS 14.8	mOS 18.1	1.09(0.51-2.31)	0.81	8
Righi(2016)	Italy	retrospective	143	15	NS	IHC, FISH	107(74)	13(10)	23(5)	89	mOS NS	mOS NS	0.67(0.45-1.01)	0.055	9
Farzin(2015)	Australia	retrospective	229	9	NS	IHC	120(75)	42(19)	67(12)	106	mOS 16.11	mOS 6.34	0.52(0.36-0.75)	<0.01	8
McGregor(2015)	US	retrospective	111	NS	NS	IHC	58(47)	43(21)	10(0)	68	mPFS NS	mPFS NS	0.74(0.42-1.30)	0.2931	8
Data were not shown
Markowski(2018)	US	retrospective	17	10.8	MTO	NGS	10(3)	2(0)	3(2)	7	mOS 14.5	mOS 7.2	NS	NS	8
Bahnasy(2018)	Egypt	prospective	122	NS	MTO	PCR, sequencing, IHC	92(37)	16	10	47	mOS 2.5	mOS 6.3	NS	NS	7

US, United States; mOS, median overall survival; mPFS, median progression-free survival; BAP1, BRCA1 associated protein 1; HR, hazard ration; n, number; m, month; MTO,multiple treatment options; NS, data were not shown; U, univariate analysis; M, multivariate analysis; PCR, polymerase chain reaction; IHC, Immunohistochemistry; NGS, Next-generation sequencing technology.

Overall, ten studies involving 1,157 individuals underwent BAP1 testing with confirmed pathological diagnoses. Patients exhibiting BAP1 deletion or mutation outnumbered those with BAP1 positivity, totalling 626 cases (approximately 54%). Among these, the majority received an epithelial subtype diagnosis, comprising 473 cases (approximately 41%). Additionally, 153 cases were non-epithelial (approximately 13%). Detailed information is presented in Figure 2A. Furthermore, we pooled data from nine studies involving 1,039 individuals, revealing 654 epithelial, 211 biphasic, and 174 sarcomatoid subtypes. Within these nine studies, consistent with prior findings, the epithelial subtype exhibited the highest prevalence of BAP1 mutations or deletions at 66.7%. This was followed by the biphasic subtype at 51.2%, with the sarcomatoid subtype being the least common at 20.1%. See Figure 2B for specifics.

Figure 2.

Distribution of BAP1 status and histological subtypes in patients with malignant pleural mesothelioma (MPM). (A) Pie chart showing the proportion of patients with BAP1 deletion or mutation versus BAP1-preserved (wild-type) status among 1,157 patients with confirmed pathological diagnoses from 10 studies. (B) Bar chart showing the distribution of BAP1 deletion or mutation across different histological subtypes (epithelioid, biphasic, and sarcomatoid) among 1,039 patients from 9 studies. BAP1, BRCA1-associated protein 1

Association of BAP1 With Overall Survival in MPM

To investigate whether BAP1 deletion or mutation correlates with survival in patients with malignant pleural mesothelioma. Across these 12 studies evaluating mOS in malignant pleural mesothelioma patients, we pooled eight univariate and nine multivariate HR, all indicating that BAP1 deletion or mutation significantly predicts patient survival. Our univariate meta-analysis confirmed that patients with BAP1 deletion or mutation exhibited markedly prolonged mOS compared to BAP1-positive patients (HR = 0.51, 95% CI = 0.45-0.58, P < 0.0001; Figure 3A). Our multivariate meta-analysis similarly confirmed that patients with BAP1 deletion or mutation exhibited markedly prolonged mOS compared with BAP1-positive patients (HR = 0.66, 95% CI = 0.59-0.77, P < 0.0001, Figure 3B).

Figure 3.

Forest plots of the association between BAP1 deletion or mutation and overall survival (OS) in patients with malignant pleural mesothelioma (MPM). (A) Univariate analysis; (B) Multivariate analysis. Hazard ratios (HRs) and 95% confidence intervals (CIs) comparing BAP1-deleted or -mutated patients with BAP1-preserved patients were pooled using the DerSimonian-Laird random-effects model. An HR < 1 indicates improved survival in patients with BAP1 deletion or mutation. The size of each square is proportional to the weight of the corresponding study. Horizontal lines represent 95% CIs. The diamond represents the pooled estimate. BAP1, BRCA1-associated protein 1; CI, confidence interval; HR, hazard ratio; mOS, median overall survival

Association of BAP1 With Progression-Free Survival in MPM

In the PFS analysis, we observed that BAP1 deletion or mutation influenced patient survival prolongation. Univariate and multivariate analyses yielded respective results of (HR=0.62, 95% CI=0.39-0.99, P=0.04) and (HR=0.79, 95% CI=0.51-1.24, P=0.31), respectively, though these differences did not reach statistical significance (Figure 4). This may be attributable to the limited scope of PFS studies and insufficient sample sizes. Expanding sample sizes, incorporating additional studies, or extending follow-up periods could yield more meaningful results.

Figure 4.

Forest plots of the association between BAP1 deletion or mutation and progression-free survival (PFS) in patients with malignant pleural mesothelioma (MPM). (A) Univariate analysis; (B) Multivariate analysis. Hazard ratios (HRs) and 95% confidence intervals (CIs) comparing BAP1-deleted or -mutated patients with BAP1-preserved patients were pooled using the DerSimonian-Laird random-effects model. An HR < 1 indicates prolonged progression-free survival in patients with BAP1 deletion or mutation. The size of each square is proportional to the weight of the corresponding study. Horizontal lines represent 95% CIs. The diamond represents the pooled estimate. BAP1, BRCA1-associated protein 1; CI, confidence interval; HR, hazard ratio; mPFS, median progression-free survival

Analyses of Heterogeneity, Stability, and Publication Bias

In the HR analyses for both univariate and multivariate mOS, sensitivity analyses confirmed the robustness of our findings, with all predicted estimates falling within the corresponding 95% confidence intervals for malignant pleural mesothelioma patients (Figure 5). Moreover, no significant publication bias was detected according to the Egger test (U: P=0.275, Figure 6A; M: P=0.757, Figure 6B) and Begg test (U: P=0.386, Figure 6C; M: P=0.754, Figure 6D).

Figure 5.

Sensitivity analysis of the association between BAP1 deletion or mutation and overall survival (OS) in patients with malignant pleural mesothelioma (MPM). (A) Univariate analysis; (B) Multivariate analysis. Sensitivity analysis was performed using the leave-one-out method, whereby each study was sequentially excluded to evaluate the robustness of the pooled hazard ratio (HR). The x-axis represents the pooled HR with the corresponding study removed, and the y-axis lists the excluded study. The vertical line indicates the pooled HR from the full analysis. Error bars represent 95% confidence intervals (CIs). BAP1, BRCA1-associated protein 1; CI, confidence interval; HR, hazard ratio; mOS, median overall survival

Figure 6.

Publication bias assessment and stability evaluation of BAP1 loss on median overall survival (mOS) in malignant pleural mesothelioma. (A and B) Egger’s linear regression test for publication bias. (A) Univariate analysis; (B) Multivariate analysis. The x-axis represents precision (inverse of standard error), and the y-axis represents the standardized effect. The solid diagonal line indicates the regression line, and the horizontal solid line represents the null hypothesis of no effect. Asymmetry of the scatter points around the regression line suggests potential publication bias. (C and D) Begg’s rank correlation test for publication bias. (C) Univariate analysis; (D) Multivariate analysis. The x-axis represents the standard error of log(HR), and the y-axis represents the log(HR). The solid horizontal line indicates the pooled effect estimate, and the diagonal dashed lines represent the pseudo 95% confidence limits. The Kendall’s tau and p-value were calculated to statistically assess funnel plot asymmetry. (E and F) Trim and fill funnel plots. (E) Univariate analysis; (F) Multivariate analysis. Blue circles represent observed studies (n=8 in e; n=9 in f), and orange circles represent imputed studies estimated by the trim and fill method to account for potential publication bias (n=2 in e; n=1 in f). The vertical solid line indicates the pooled effect estimate, and the diagonal dashed lines represent the pseudo 95% confidence limits

We used multiple complementary methods to assess publication bias. First, we visually inspected the funnel plot for asymmetry (Supplementary Figure 1). Second, for the OS analysis (8-9 studies), we performed no significant publication bias was detected according to the Egger test (U: P=0.275, Figure 6A; M: P=0.757, Figure 6B) and Begg test (U: P=0.386, Figure 6C; M: P=0.754, Figure 6D). Third, considering the limited statistical power of Egger and Begg tests when there are fewer than 10 studies, we performed the trim-and-fill method to complement the publication bias assessment. For the univariate OS analysis (8 studies), the trim-and-fill method estimated that 2 studies were missing from the left side of the funnel plot (Figure 6). After imputing these hypothetical studies, the adjusted pooled HR was 0.63 (95% CI: 0.54–0.73), which remained statistically significant (P < 0.0001), suggesting that the observed association between BAP1 loss/mutation and improved OS was robust to potential publication bias. For the multivariate OS analysis (9 studies), only 1 study was estimated as missing. After imputation, the adjusted pooled HR was 0.57 (95% CI: 0.47–0.69), which also remained statistically significant. The smaller number of imputed studies and the minimal change in the effect estimate (ΔHR = 0.02) further confirm the robustness of our primary multivariate analysis.

Tissue Subtypes and Detection Methods

We extracted histology-specific information and attempted exploratory subgroup analyses on epithelial and non-epithelial MPM, but only two studies analysed whether BAP1 is an independent prognostic factor within each histological subtype, and neither provided specific values. Both studies indicated that BAP1 loss had a positive predictive effect on the survival of patients with epithelial disease, but there is controversy regarding its effect on the survival of patients with non-epithelial disease. One study concluded that BAP1 loss had a positive predictive effect on the survival of patients with non-epithelial disease, and that the survival of both groups diagnosed with epithelial and non-epithelial disease showed a positive predictive effect, ²² but another study did not observe this effect when analysing biphasic cases. ³⁰

To address potential measurement bias introduced by different BAP1 detection methods, we excluded studies that used only NGS detection to assess whether the results remained consistent when limited to IHC studies, and performed new forest plots and sensitivity analyses. In the multivariate analysis of OS, there were no studies using only NGS detection, so the HR remained consistent, and the sensitivity analysis results of the forest plot were the same as Figures 3 and 5. In the univariate analysis of OS, after excluding one study that used only NGS detection (Dudnik2021), the remaining seven studies (mainly using immunohistochemistry IHC) had an HR of (HR=0.51, 95% CI=0.45-0.58, P<0.0001). The forest plot and sensitivity analyses are shown in Supplementary Figure 2, indicating that different detection methods did not affect the accuracy of our study.

Discussion

IHC is the most widely used method for BAP1 assessment in clinical practice. Most of the articles included in this meta-analysis detected BAP1 solely through IHC, but there has been recent controversy regarding prognostic immunohistochemical markers. The results of BAP1 immunohistochemistry may be affected by inter-observer variability and technical factors. ³³ Moreover, the emergence of next-generation sequencing and other molecular techniques provides enhanced diagnostic and prognostic capabilities. ³⁴ Different detection methods may have different sensitivities and specificities, which can introduce measurement bias. Future meta-analyses should consider stratified analysis based on detection methods to account for these technical advances. However, the forest plots and sensitivity analyses conducted in this meta-analysis according to detection methods indicate that different detection techniques did not affect the accuracy of our study.

This is the first meta-analysis confirming that BAP1 is a potential prognostic marker of malignant pleural mesothelioma, and it can effectively predict the overall survival (OS) of patients with malignant pleural mesothelioma, with statistical significance. In 2020, Luca Cantini and colleagues published the first meta-analysis studying the prognostic correlation between BAP1 and malignant pleural mesothelioma. However, their study suggested that the prognostic and predictive value of BAP1 IHC in malignant pleural mesothelioma cannot be assessed independently, which may be due to the influence of technical factors on BAP1 immunohistochemistry results. ²⁸ However, in recent years, the advent of multiple detection methods, including next-generation sequencing, has improved the detection rate of BAP1, yielding more meaningful conclusions for BAP1 analysis.

Precision treatment for malignant tumours remains one of the enduring challenges facing the medical community. However, the complexity and variability inherent in malignant tumours pose significant obstacles to achieving such targeted therapies. To address this challenge, numerous researchers are exploring predictive biomarkers for malignant tumours. Predictive biomarkers not only guide risk stratification but also improve treatment outcomes, reduce treatment-related toxicity, and lower overall healthcare costs. This meta-analysis indicates that BAP1, derived from both univariate and multivariate analyses of overall survival, serves as a potential biomarker for predicting prognosis in patients with malignant pleural mesothelioma. However, including univariate and multivariate HRs may introduce heterogeneity due to differences in the degree of adjustment for confounding factors between studies. Furthermore, the study simultaneously evaluating BAP1 status and miR-31 levels has demonstrated that this combination can also enhance prognostic stratification in MPM patients. ²⁴

The primary first-line treatment for malignant pleural mesothelioma currently remains pemetrexed combined with platinum-based chemotherapy. Although nivolumab combined with ipilimumab has also been incorporated into first-line treatment regimens, delivering some survival benefit for MPM, this advantage is not significantly greater than that of chemotherapy, with median overall survival (mOS) of 18.1 months versus 14.1 months.^18,35 To address the current poor survival prognosis for MPM patients, numerous researchers have commenced the search for predictive biomarkers for immunotherapy and novel therapeutic targets. MPM is characterised by frequent inactivating mutations in tumour suppressor genes (TSGs) – such as homozygous deletions (HD) and point mutations – including cyclin-dependent kinase inhibitor 2A/2B (CDKN2A/2B), BRCA1-associated protein 1 (BAP1), neurofibromin 2 (NF2), large tumour suppressor kinase 2 (LAST2), and tumour protein p53 (TP53). ³⁶ Among these, BAP1 mutations are the most prevalent and currently hold the greatest significance in terms of diagnosis, prognosis, and treatment prediction, being present in 47–67% of MPM tumours. ³⁷ Multiple studies have identified the potential role of BAP1 in immunotherapy for malignant pleural mesothelioma (MPM). They observed that BAP1-deficient MPM exhibits increased expression of several inhibitory immune checkpoints, such as PD-L1, PD-1, and LAG3. ¹⁸ Deficiency of BAP1 in MPM upregulates immune checkpoints and T cells depleted of precursor cells, a key mediator of the superior efficacy of cancer immunotherapy, suggesting BAP1’s potential role in predicting response to immunotherapy.

Interestingly, this meta-analysis included 1,835 patients with malignant pleural mesothelioma from 14 studies. The pooled analysis confirmed the prognostic value of BAP1 in lung cancer patients. Patients with BAP1-positive expression demonstrated poorer survival outcomes regardless of whether overall survival (OS) or progression-free survival (PFS) was used as the endpoint. Furthermore, the low heterogeneity observed enhances the credibility of the conclusion that BAP1 constitutes an independent prognostic factor for MPM, underscoring its robustness and reliability. However, the prevalence of BAP1 mutations is higher in epithelioid MPM, which itself has a better prognosis, potentially confounding the observed survival benefit. Future studies with histological stratification data are needed to determine whether BAP1 is an independent prognostic factor within each histological subtype.

This meta-analysis provides important findings, though several limitations warrant attention. Firstly, the relatively small number of included studies, most of which were retrospective, limits the overall strength of the findings. Secondly, the lack of geographical diversity, with the majority of studies conducted in Italy and Australia, may affect the generalisability to broader populations. In addition, due to the lack of reported data in the included studies, we were unable to analyse disease-free survival or disease-specific survival. Future studies that include these endpoints will more comprehensively assess the prognostic value of BAP1 in MPM. Finally, although no apparent publication bias was evident in our analysis, the exclusion of non-English studies and potential unpublished negative results may have influenced the overall conclusions.

Given these limitations, future research should focus on validating the prognostic value of BAP1 through large-scale, prospective, and multicentre studies. Efforts should be made to combine BAP1 with other predictive biomarkers (such as PD-L1, PD-1, and LAG3) to develop more robust predictive models for accurately forecasting treatment response. Furthermore, elucidating the underlying molecular mechanisms linking systemic inflammation and immune responses to MPM progression is crucial for fully realizing BAP1’s potential as a component of personalised therapeutic strategies, ultimately improving patient prognosis.

Conclusion

This meta-analysis indicates that BAP1 serves as a potential prognostic marker for malignant pleural mesothelioma, effectively predicting overall survival (OS) in patients with malignant pleural mesothelioma with statistical significance. However, across these 14 studies, data on progression-free survival (PFS) were largely unavailable, rendering assessments of PFS statistically nonsignificant. Nevertheless, analyses of OS in MPM patients indicate that BAP1 serves as a practical and cost-effective biomarker for guiding treatment decisions and optimising patient outcomes.

ORCID iDs

Jiaojiao Hong https://orcid.org/0009-0001-9831-9938

Chengyu Chen https://orcid.org/0009-0008-8281-8338

Ethical Considerations

This study is a systematic review and meta-analysis of previously published data. It does not involve new patient data, human tissues/samples, or direct contact with patients. Therefore, this study does not require ethical approval or informed consent.

Consent for Publication

All authors have read and approved the content and agree to submit it for publication in your journal.

Data Availability Statement

The data sources are authentic and reliable, and can all be verified on platforms such as Web of Science, PubMed, Embase, Cochrane, and CNKI.*