Sex and risk of hyperprogressive disease in cancer patients receiving immune checkpoint inhibitors: a systematic review and meta-analysis

Abstract

Background

Hyperprogressive disease (HPD) refers to a phenomenon characterized by a significant acceleration in tumor growth during treatment with immune-checkpoint inhibitors (ICIs), far exceeding the pre-treatment growth rate, leading to rapid deterioration in the patient’s condition. HPD typically occurs within the first two months of treatment and often results in early mortality. While sex differences in immune responses have been extensively studied, the relationship between sex and HPD risk remains unclear.

Methods

We systematically searched MEDLINE (PubMed), Embase, and Scopus databases to identify studies reporting HPD in patients treated with ICIs, with available gender-stratified data. The search encompassed all records from database inception until 30 June 2025. Meta-analyses were performed using RevMan 5.4.1 software, with pooled odds ratios and 95% confidence intervals calculated under either fixed-effects or random-effects model, depending on heterogeneity. The primary objective was to evaluate the association between biological sex and the risk of HPD. The research protocol has been registered on the PROSPERO platform (CRD420261324007).

Results

This meta-analysis included 20 studies involving 2,291 patients (1,612 male patients and 679 female patients), all of which were retrospective. The overall incidence of HPD was 18.67% in male patients and 18.26% in female patients. Pooled analysis revealed no statistically significant association between biological sex and the risk of HPD development (I² = 27%; p = 0.13).

Conclusion

Our findings demonstrate that sex should not be considered a determinative factor for HPD risk in patients receiving ICIs.

1. Introduction

The emergence of immune checkpoint inhibitors (ICIs) has completely transformed the treatment landscape of advanced cancer, significantly improving the prognosis of some patients (1, 2). However, the efficacy of immunotherapy varies considerably across patient populations, and among the factors that may influence treatment outcomes, sex has emerged as a potential biological determinant (3). Sex differences in immune responses are well-documented. Women generally exhibit stronger innate and adaptive immune responses, while men are more susceptible to infections and malignancies (3, 4). In oncology, men not only have a higher cancer incidence but also a higher cancer mortality rate across multiple tumor types, including liver, bladder, esophageal, and head and neck cancers (5, 6). These disparities are largely attributable to genetic and hormonal factors. Notably, women possess two X chromosomes, which harbor numerous immune-related genes—including FOXP3, TLR7, CD40L, and CXCR3—that can escape X-chromosome inactivation, leading to higher expression levels and more robust immune activation in women (7). Concurrently, sex hormones such as estrogen and testosterone differentially modulate immune cell function, with estrogen generally promoting humoral immunity and testosterone exerting immunosuppressive effects (3). These molecular differences collectively underpin the observed sex disparities in immune competence and may have significant implications for the efficacy and safety of cancer immunotherapy.

Hyperprogressive disease (HPD) is a rare but highly aggressive pattern of tumor progression following ICIs, characterized by accelerated tumor growth and rapid clinical deterioration, often occurring within the first 2 months of treatment (8). The mechanisms underlying HPD remain incompletely understood, but emerging evidence suggests that abnormal remodeling of the immune microenvironment plays a central role (9, 10). Several potential pathways have been implicated: accelerated T-cell exhaustion kinetics driven by dysfunctional reinvigoration of exhausted T cells, expansion and recruitment of myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs) in response to programmed cell death protein 1 (PD-1)/programmed death-ligand 1 (PD-L1) blockade, and the upregulation of alternative immune checkpoints that suppress antitumor immunity (9, 11, 12). Given the established sex differences in immune cell composition, cytokine profiles, and the expression of X chromosome-linked immunoregulatory genes (e.g., FOXP3 and TLR7), it is biologically plausible that sex may influence the risk of HPD by modulating these microenvironmental processes.

Clinical evidence regarding the association between sex and HPD risk, however, remains conflicting. Kanjanapan et al. (13) reported a higher incidence of HPD among female patients (p = 0.01) in early-phase immunotherapy trials. Similarly, Saâda-Bouzid et al. (14) observed that gender was associated with HPD occurrence in patients with recurrent/metastatic head and neck squamous cell carcinoma. In contrast, several recent studies, including those by Petrova et al. (15), Aoki et al. (16), and Kim et al. (17), have failed to identify a significant sex difference. These divergent findings highlight the need for a systematic synthesis of the available evidence to clarify whether sex truly represents an independent risk factor for HPD.

2. Methods

This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines and registered in the International Prospective Register of Systematic Reviews, with the registration number CRD420261324007. This systematic review adheres to the PRISMA 2020 statement. The completed PRISMA checklist is provided as Supplementary Material S1.

2.1. Retrieval methods

For the current analysis, two researchers (ZJ and TCL) extracted data from MEDLINE (PubMed), Embase, and Scopus up to 30 June 2025, using the following search terms: immunotherapy, programmed death-ligand 1, PD-L1, cytotoxic T-lymphocyte-associated protein 4, CTLA-4, ipilimumab, tremelimumab, nivolumab, pembrolizumab, immune checkpoint inhibitor, hyperprogression, and hyperprogressive disease, and HPD. All duplicate records were excluded. We reviewed and excluded references from reviews, commentaries, editorials, included studies, and conference publications to ensure the accuracy of the included studies.

2.2. Inclusion and exclusion criteria

2.2.1. Inclusion criteria

The inclusion criteria were as follows: (1) prospective or retrospective studies reporting the characteristics of patients who developed HPD during immunotherapy, regardless of tumor type; (2) all patients were diagnosed with malignancy via biopsy; and (3) numerical data on sex and HPD incidence were available.

2.2.2. Exclusion

The exclusion criteria included: (1) duplicate studies, reviews, case reports, letters, reference abstracts, or studies for which full-text English articles were unavailable; (2) nonhuman studies, such as in vitro or animal studies; and (3) studies that did not provide numerical data on gender and HPD incidence.

2.3. Study selection and quality assessment

The literature retrieval was conducted independently by two authors. Any discrepancies were resolved through discussion with a third author until a consensus was reached.

Quality assessment was conducted using the Newcastle–Ottawa Scale (NOS), which evaluates study design based on eight criteria: selection of the study population, comparability of groups, and exposure. Only those with a score of ≥ 7 were included in this analysis.

2.4. Data extraction

Study characteristics were extracted, including the first author, publication year, study design, treatment type (PD-1, PD-L1, or cytotoxic T-lymphocyte antigen 4 (CTLA-4)), total number of HPD cases, and the number of male and female participants.

2.5. Statistical analysis

The pooled odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using either random- or fixed-effects models to assess the association between sex and HPD risk. Heterogeneity across studies was evaluated using the χ²-based Q test, and the I² statistic was computed to quantify the percentage of total variability attributable to study heterogeneity. When heterogeneity is significant (I² ≥ 50% or p < 0.10), a random-effects model is selected. Publication bias was assessed through funnel plot visualization supplemented by Begg’s test. In addition, each study was excluded from the meta-analysis for sensitivity analyses. To further reduce heterogeneity among studies and enhance the reliability of the results, stratified analyses were conducted. When stratified analysis failed to eliminate heterogeneity in the pooled effect size, individual studies were sequentially excluded, and the effect size was recalculated to identify studies contributing significantly to heterogeneity. All statistical tests were two-tailed, with a significance threshold of p < 0.05, and were implemented using RevMan 5.4.1.

3. Results

3.1. Characteristics of included studies

3.1.1. General characteristics of the studies

The search was updated to 30 June 2025, and a total of 835 articles were included. After full-text screening, 39 potentially relevant articles were identified. Ultimately, six studies (13, 15–19) met the inclusion criteria for both abstract and full-text review.

Additionally, 14 trials included in the meta-analysis by Zhao et al. (17) were also included. Therefore, a total of 20 retrospective studies were included in this meta-analysis (Figure 1), with publication dates ranging from 2019 to 2025. The sample sizes per trial ranged from 34 to 263. Among all the included patients, 15 studies were from Asian countries, including five from China, five from South Korea, and five from Japan; the other five studies were from non-Asian countries, including one from Canada, one from the Netherlands, one from Australia, one from Italy, and one from Bulgaria. All included studies were in English and had an NOS score of 7 or higher (Table 1).

Figure 1.

Details of the literature search and the study-selection processes.

Table 1.

Fundamental characteristics of the included studies.

Author	Year	N (male/female)	Study design	Country	Underlying malignancy	Treatment	Definition of HPD	HPD male	HPD female	Quality assessment of NOS	Incidence of HPD
Aoki et al. (18)	2019	34 (26/8)	R	Japan	Gastric cancer	Nivolumab	TGRpost/TGRpre ≥ 2	8	2	High	10/34 (29.4%)
Aoki et al. (16)	2023	147 (108/39)	R	Japan	Gastric cancer	Nivolumab	A more or less two fold increase in the tumor growth rate of measurable lesions	33	8	High	41/147(27.9%)
Castello et al. (19)	2020	46 (32/14)	R	Italy	NSCLC	Nivolumab, pembrolizumab, atezolizumab	PD on the first CT scan according to RECIST 1.1 (TGRpost-TGRpre)/TGRpre ≥ 50%	8	6	High	14/46 (30.4%)
Chen et al. (20)	2023	107 (73/34)	R	China	Pancreatic cancer	PD(L)1	TGRpost−TGRpreb > 50%	8	2	High	10/107 (9.3%)
Hagi et al. (21)	2020	136 (107/29)	R	Japan	Gastric cancer	Nivolumab	> 50% increase in the sum of the longest diameters of the target lesions TGKpost/TGKpre ≥ 2	25	5	High	30/136 (22.1%)
Jin et al. (22)	2021	41 (28/13)	R	China	NPC	Anti-PD-1 mAb	(1) The acceleration of tumor cell proliferation exceeding two fold based on three time points (2) TTF ≤ 2 months	13	3	High	16/41 (39%)
Kanjanapan et al. (19)	2019	182 (99/83)	R	Canada	Multiple cancers	Anti-PD-1 mAb, anti-PD-L1 mAb, anti-CTLA-4 mAb	RECIST 1.1 progression at the first on treatment scan with a 2-fold or greater increase in TGR between the reference and experimental periods.	2	10	High	12/182 (6.6%)
Kanjanapan et al. (23)	2023	87 (51/36)	R	Australia	Lung cancer and melanoma	PD(L)1, CTLA-4	An increase of ≥ 40% in the sum of the target lesions compared with baseline and/or an increase of ≥ 20% in the sum of the target lesions compared with baseline plus new lesions in at least two different organs.	5	6	High	11/87 (12.6%)
Kim CG et al. (24)	2021	263 (191/72)	R	Korea	NSCLC	PD(L)1	Defined as TGRpost/TGRpreb ≥2 and TGKpost/TGKprec ≥ 2 according to the RECIST 1.1 and TTF <2 mont	34	20	High	54/263 (20.5%)
Kim J, et al. (17)	2022	219 (169/50)	R	Korea	NSCLC	Nivolumab, pembrolizumab, atezolizumab, durvalumab	TGKpost = (Vafter – Vbase/Tafter − Tbase) TGKpost/TGKpre ≥ 2 TTF ≤ 2 months	24	11	High	35/219 (16.0%)
Kim S. H, et al. (25)	2020	40 (24/16)	R	Korea	NSCLC	Nivolumab	HPD according to RECIST 1.1 at the first evaluation TGRpost/TGRpre ≥ 2	8	8	high	16/40 (40%)
Kim Y. et al. (26)	2019	137 (94/41)	R	Korea	NSCLC	Nivolumab, pembrolizumab, atezolizumab, durvalumab, avelumab	TTF < 2 months TGKpost/TGKpre ≥ 2 > 50% increase in the sum of the total measured tumor burden	34	14	High	48/154 (35.6%)
Lu et al. (27)	2019	56 (39/17)	R	China	GI tract cancer	Anti-PD-1 mAb, anti-PD-L1 mAb, anti-CTLA-4 mAb	NA	3	2	High	5/56 (8.9%)
Li Y, et al. (28)	2024	105 (22/83)	R	China	NSCLC	PD(L)1, CTLA-4	> 50% increase in the sum of the longest diameters of the target lesions TGKpost/TGKpre ≥ 2	33	11	High	44/105 (41.0%)
Petrova et al. (15)	2020	154 (99/55)	R	Bulgaria	NSCLC	Pembrolizumab	(1) time to treatment failure < 3 months; (2) increase ≥ 50% in the sum of target lesion major diameters between baseline and first radiological evaluation; (3) appearance of at least two new lesions in an organ already involved between baseline and first radiological evaluation; (4) spread of the disease to a new organ between baseline and first radiological evaluation; and (5) clinical deterioration with ECOG PS ≥ 2 during the first 3 months of treatment. Criteria 1 and 5 were mandatory.	8	8	High	16/154 (10.4%)
Park et al. (29)	2020	125 (103/22)	R	Korea	HNSCC	Anti-PD-1 mAb, anti-PD-L1 mAb, anti-CTLA-4 mAb	TGKpost/TGKpre ≥ 2	15	3	High	18/125 (14.4%)
Sasaki A, et al. (30)	2019	62 (45/17)	R	Japan	Gastric cancer	Nivolumab	TGKpost/TGKpre ≥ 2 > 50% increase in the sum of the total measured tumor burden	12	1	High	13/62 (21%)
Schuiveling M, et al. (31)	2021	162 (99/63)	R	Netherlands	Melanoma	Anti-PD1, anti-CTLA-4	a time to treatment failure less than 2 months with doubling of tumor burden and a two fold increase in tumor growth rate during treatment	2	0	High	2/162 (1.2%)
Yildirim et al. (32)	2022	121 (85/36)	R	Turkey	Multiple cancers	Nivolumab, avelumab, pembrolizumab, atezolizumab	Defined by RECIST progression and at least three of the following symptoms: time to treatment failure <2 months (time to treatment failure is defined as the time from the start of treatment with ICI to ICI discontinuation for any reason); increase of ≥50% in the sum of target lesion major diameters between baseline and first radiologic evaluation; the appearance of at least two new lesions in an organ already involved between baseline and first radiologic evaluation; spread of the disease to a new organ between baseline and first radiologic evaluation; and clinical deterioration with a decrease in ECOG performance status ≥2 during the first 2 months of treatmenta time to treatment failure less than 2 months with doubling of tumor burden and a twofold increase in tumor growth rate during treatment	17	3	High	20/121 (16.5%)
Zhang et al. (33)	2021	69 (57/12)	R	China	HCC	Nivolumab, pembrolizumab, camrelizumab	Between the baseline and the first radiological evaluation; the appearance of at least two new lesions in an organ already involved between baseline and the first radiological evaluation; spread of the disease to a new organ between baseline and the first radiological evaluation; and clinical deterioration with a decrease in the ECOG performance status of ≥ 2 during the first 2 months of treatment (1) TTF < 2 months; (2) disease progression at the first evaluation and a > 50% increase in TGR; (3) TGRpost/TGRpre ≥ 2	9	1	High	10/69 (14.5%)

HPD, hyperprogressive disease; NSCLC, nonsmall cell lung cancer; LC, lung cancer; HCC, hepatocellular carcinoma; GC, gastrointestinal adenocarcinoma; R/M HNSCC, recurrent and/or metastatic head and neck squamous carcinoma; PD-1, programmed cell death protein 1; PD-L1, programmed death 1 ligand 1; CTLA-4, cytotoxic T-lymphocyte antigen 4.

3.1.2. Characteristics of the patients studied

A total of 2,291 cases were included in this study, among which 425 cases (14.5%) were HPD and 1,866 cases were non-HPD. The incidence of HPD in each study ranged from 1.2% to 41.0%. There were 301 male cases and 124 female cases. There were 229 cases of NSCLC, 109 cases of digestive system malignancies, 2 cases of melanoma, 18 cases of head and neck tumors, and the rest were other cancers (Table 1).

3.1.3. Characteristics of the design types included in the study

All 20 included studies were retrospective studies (13, 15–29, 32–35). The cases were divided into HPD and non-HPD groups. In assessing tumor growth kinetics indicators, because there is no standard, the definition of HPD varied across the included studies. The criteria reported in the included studies in this analysis were adopted (Table 1).

In these studies, Petrova et al. (15) employed criteria that integrated both clinical and radiological parameters. HPD was defined not only as a significant increase in tumor burden but also as the emergence of new lesions in distant organs. Notably, the definition of tumor growth kinetics (TGK) differed in the studies conducted by Kim et al. (24, 26). Furthermore, Zhang et al. (33) introduced two additional requirements in the study by Champiat et al. (11): (1) time to treatment failure (TTF) < 2 months, and (2) disease progression at the first assessment with a tumor growth rate (TGR) increase exceeding 50%.

3.2. Association between sex and hyperprogressive disease in cancer patients undergoing immunotherapy

A total of 20 retrospective studies involving 2,291 patients receiving immunotherapy provided numerical data on sex and the incidence of HPD. ORs and 95% CIs from these studies were pooled. In the meta-analysis of gender and HPD (Figure 2), no statistically significant difference was observed between men and women (OR = 0.90; 95% CI: 0.75–1.15; p = 0.41; I² = 27%, p = 0.13). Subgroup analyses were conducted based on tumor type, geographic region of included patients, and gender ratio.

Figure 2.

Forest plot of the association between sex and HPD in patients treated with immunotherapy.

3.3. Sensitivity analysis and publication bias

The analysis revealed low heterogeneity among the studies (I² = 27%; p = 0.13). Consequently, a sensitivity analysis was conducted on all included studies. Each study was sequentially excluded from the meta-analysis to evaluate its impact on the pooled OR. The results of the sensitivity analysis demonstrated that all included studies showed no evidence of potential publication bias (Figure 3). Furthermore, publication bias was assessed using Begg’s test. Funnel plot assessment (Supplementary Figure 1) showed that studies were symmetrically distributed on both sides of the funnel plot, indicating no significant publication bias (p = 0.230).

Figure 3.

Funnel plot for the association between sex and HPD risk.

3.4. Subgroup analysis

The overall analysis demonstrated no significant association between sex and HPD risk (OR = 0.90; 95% CI: 0.71–1. 15; I² = 27%; p = 0. 13), which supports our core hypothesis that gender should not be considered a determinative factor for HPD risk in ICI therapy. However, the prespecified subgroup analyses revealed several important effect modifiers, including geographic region, tumor type, gender ratio, and treatment modality, all of which warrant further investigation (Figure 4). To further investigate the potential impact of heterogeneous HPD diagnostic criteria on the pooled results (Figure 4), we performed a subgroup analysis stratifying the included studies according to their HPD definitions. The 20 included studies were categorized into three groups: (1) pure TGR/TGK-based definitions (n = 6), (2) pure RECIST 1.1-based definitions (n = 2), and (3) composite definitions combining TGR/TGK with TTF and/or new lesions/clinical deterioration (n = 12).

Figure 4.

Sex-based subgroup analysis of HPD in immunotherapy patients.

4. Discussion

Patients exhibit accelerated disease progression following immunotherapy, a phenomenon characterized by rapid tumor growth rather than control, which has been identified as HPD (23, 34). The incidence of HPD across various studies to date ranges from 4% to 29%, while our meta-analysis reveals an overall HPD incidence spanning from 1.2% to 41.0% (36). The lack of a standardized definition for HPD and its predictive biomarkers remains a challenge. Reported occurrences of HPD during immunotherapy have been shown to significantly reduce patients’ overall survival (37). Therefore, early identification of HPD is of paramount importance. Numerous studies have reported clinical factors that may contribute to a high incidence of HPD, including low PD-L1 expression levels, more than three metastatic sites, elevated NLR, and reduced hemoglobin (Hb) levels (8, 12, 33). In addition to these factors, gender represents a common clinical variable. Kanjanapan’s research identified a higher incidence of HPD among female patients (13). However, our analytical findings revealed no statistically significant association between sex and HPD. Furthermore, across geographical regions and diverse tumor types, we were unable to demonstrate significant gender-based differences in HPD. Multiple factors may account for these observations.

In the context of HPD, sex differences in immune responses may partly explain why men have a higher risk of developing malignant tumors, while women are more prone to autoimmune diseases. Sex disparities may be associated with genetic factors, as female cells contain two X chromosomes, whereas male cells contain one X chromosome and one Y chromosome (7). The latest research by Chen et al. revealed, for the first time, that the loss of the Y chromosome in tumor cells, which is uniquely present in male cells, can facilitate immune evasion by impairing antigen presentation, thereby enabling cancer cells to escape recognition and elimination by the immune system (38). This study elucidates the underlying mechanisms of male predisposition to cancer development and poorer prognosis from the perspective of Y chromosome loss. The research findings demonstrate that the molecular origins of sex disparities in cancer incidence extend beyond hormonal factors, encompassing significant genetic determinants (39). We updated the previous search conducted by Zhao et al. (36) by exclusively including trials that involved immunotherapy and excluding those with insufficient sex-specific data, thereby enabling a focused evaluation of the relationship between sex and HPD in immunotherapy. Although the overall analysis revealed no significant association between sex and HPD risk, significant subgroup differences were observed across tumor types (p = 0.005), suggesting that tumor type may modulate the effect of sex on HPD. Our subgroup analysis revealed varying incidence rates of HPD across different sex ratios. Specifically, within a sex ratio range of 1–2, a statistically significant association was observed between patient sex and the occurrence of HPD OR = 0.45 (95% CI: 0.25 - 0.81) p < 0.32), with a higher incidence of HPD observed in female patients. In the study by Kanjanapan et al. (13), the sex ratio was between one and two. Notably, the number of studies in the subgroup with a sex ratio greater than two was much larger than that in the subgroup with a sex ratio between one and two, which contributed substantially to the combined HR effect. Furthermore, three studies conducted by Petrova et al. (15), Aoki et al. (16), and Kim et al. (17) employed PD-1/PD-L1 inhibitor monotherapy, while Kanjanapan’s research utilized PD-1/PD-L1 inhibitor monotherapy or combination therapy with CTLA-4 or other ICIs (13). Some studies suggest that the efficacy of combining ICIs with chemotherapy is superior in female patients compared with male patients. This is attributed to chemotherapy’s ability to enhance the tumor mutational burden in women, thereby increasing the antigenicity of cancer cells. Compared with men, women exhibit a more efficient capacity to eliminate tumor cells with higher antigenicity (40). Tumor mutational burden in the study by Kanjanapan et al., as an unadjusted potential confounding factor, may account for the higher incidence of HPD observed in female patients (13). Subgroup analysis according to treatment regimen revealed no significant association between sex and HPD risk in either the monotherapy subgroup (OR = 0.86; 95% CI: 0.65 -1.15; p = 0.31; I² = 17%) or the combination therapy subgroup (OR = 1.01; 95% CI: 0.64 -1.60; p = 0.95; I² = 43%). The test for subgroup differences was not significant (p = 0.56, I² = 0%), indicating that treatment type did not significantly modify the association between sex and HPD risk. Although combination therapy has been hypothesized to potentially amplify sex specific differences in immune responses --particularly through chemotherapy induced immunomodulation or the addition of CTLA-4 inhibitors our findings did not support this hypothesis in the context of HPD risk. Previous meta analyses have shown that immunotherapy plus chemotherapy provides equivalent survival benefits for both sexes across cancer types (HR for OS: 0.76 in males vs. 0.78 in females) (41), suggesting that the addition of chemotherapy does not substantially alter sex-based treatment effects. However, the combination therapy subgroup in our analysis included heterogeneous regimens, such as ICIs combined with chemotherapy or with other ICIs (e.g., CTLA-4 inhibitors), which may have contributed to the wide confidence intervals observed in this subgroup. Furthermore, the combination therapy subgroup had a relatively small sample size and was enriched with specific tumor types (e.g., colorectal cancer, esophageal cancer) that may themselves influence HPD risk. Therefore, the apparent lack of association in the combination therapy subgroup should be interpreted with caution, and future studies with larger sample sizes and homogeneous treatment regimens are needed to further evaluate the potential interaction between sex, treatment type, and HPD risk.

5. Limitations

This study has several limitations. First, all 20 included studies were retrospective, which limits causal inference and introduces potential selection and information biases. Second, the marked variability in HPD assessment criteria, tumor growth kinetic indicators, and definitions across studies may impede the pooling of results and potentially introduce selection and reporting biases. This underscores the urgent need for a standardized HPD definition. Third, our analysis relied on unadjusted odds ratios rather than individual patient-level data. We were unable to control for key confounders, including tumor stage, line of immunotherapy, PD-L1 expression, TMB, and menopausal status. The observed effect size may therefore be subject to residual confounding, and our findings should be considered hypothesis-generating rather than confirmatory. Fourth, critical factors such as the menopausal status of female patients, the pivotal role of gender hormones in immune system regulation, and their potential therapeutic implications warrant further consideration in the context of HPD. Finally, future studies should incorporate comprehensive data on gender hormone status and immune-related factors to further elucidate the relationship between sex and HPD risk. Further high-quality, large-scale, multicenter studies are necessary.

6. Conclusion

The study concludes that the present meta-analysis does not provide sufficient evidence to support a significant association between patient sex and HPD risk. Subgroup analyses according to geographic region, tumor type, gender ratio, definitions and treatment modality recapitulated these findings.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. This work was supported by a special research grant from the Wu Jieping Medical Foundation (320.6750.2025-10-3).

Data availability statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Author contributions

JZ: Conceptualization, Data curation, Software, Writing – original draft. CT: Data curation, Writing – review & editing.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1860831/full#supplementary-material