The association of fungal infection in cervical cancer: a systematic review and meta-analysis

Abstract

Background

Emerging evidence suggests a potential link between the fungal mycobiome and carcinogenesis. However, the relationship between fungal infection and cervical cancer progression remains inconclusive. This meta-analysis aimed to determine the pooled prevalence of fungal infection across different cervical pathological stages and assess its association with cervical lesions and HPV infection. PROSPERO CRD42024588513 is the registration number for the systematic review.

Methods

We conducted a systematic review and meta-analysis in accordance with PRISMA guidelines. We systematically searched PubMed, Web of Science, and Embase for studies published between January 1, 2004, and December 1, 2025. Observational studies (cross-sectional, case–control, cohort) with sample sizes ≥ 100 were included. Two reviewers independently performed study selection, data extraction, and quality assessment using the Newcastle–Ottawa Scale. Pooled prevalence and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using random-effects models in STATA.

Results

Seventeen studies involving 97,382 patients with cervical lesions and 450,724 controls were included. The overall pooled prevalence of fungal infection was significantly higher in cervical cancer patients (19%, 95% CI: 0.07–0.30) compared to non-cancer patients (8%, 95% CI: 0.06–0.10). In patients with precancerous lesions (CIN1, CIN2/3), the fungal infection rate was similar to that in normal controls (NILM). Subgroup analysis among non-NILM patients revealed a higher risk of fungal infection in HPV-positive individuals than in HPV-negative individuals (OR = 1.39, 95% CI: 1.07–1.81). Meta-regression identified geographic region (Europe) as a significant source of heterogeneity.

Conclusions

Fungal infection was more prevalent in invasive cervical cancer than in precancerous lesions or normal cervix, suggesting a potential association with late-stage carcinogenesis rather than early initiation. The observed association between fungal infection and HPV-positive status warrants further investigation. These findings highlight the need for prospective studies to clarify the role of the cervical mycobiome in cervical cancer progression.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12905-026-04451-w.

Introduction

Cervical cancer is the fourth leading cause of cancer incidence and mortality among women worldwide, with more than 500,000 women diagnosed with it each year, and this number continues to rise in recent years [1, 2]. Currently, cervical cancer screening mainly relies on cytology examinations and HPV DNA testing. Cytology, as a traditional method, has relatively high specificity and lower cost, but its sensitivity is limited, and results can be influenced by the quality of sample preparation and the subjective experience of the evaluators [2, 3]. It has a low detection rate for precancerous lesions, which may lead to false negatives [4]. HPV testing, due to its high sensitivity, has become a frontline primary screening method and can effectively identify individuals at high risk of infection; however, its specificity is insufficient. Therefore, there are significant gaps in the existing screening system [5]. This drives research toward finding novel biomarkers that can supplement or optimize current strategies. In recent years, the cervical microbiome, particularly its fungal components, has received increasing attention for its role in regulating local immune microenvironments and affecting the progression of HPV infections, providing a potential direction for developing new adjunctive screening tools for cervical cancer.

The occurrence of cervical lesions is typically associated with cervical microecological dysbiosis, including a reduction in Lactobacillus, an overgrowth of anaerobic bacteria, as well as the presence of pathogens such as Chlamydia trachomatis and fungi. Among these, fungal overgrowth, particularly Candida albicans, emerges as a critical factor disrupting the homeostasis of the cervical microenvironment [6]. Although persistent high-risk HPV infection is the primary cause of cervical cancer, accumulating evidence suggests that the cervical microenvironment, including the local microbiome, may modulate HPV persistence and disease progression. Fungi, particularly Candida species, are common constituents of the cervical microbiota and can induce chronic inflammation and immune dysregulation, potentially creating a permissive environment for HPV oncogenesis. Recent studies suggest that fungal infections such as Candida albicans may create a favorable microenvironment for persistent HPV infection and viral gene integration by causing chronic inflammation of the cervical mucosa [7]. Among fungal infections in women, vaginal candidiasis is the second leading cause of vaginitis worldwide. A study indicates that approximately 1.1 billion people worldwide suffer from this condition [8]. Based on current trends, the annual number of individuals affected by recurrent vulvovaginal candidiasis is estimated to reach nearly 158 million by 2030, which would represent a significant challenge to global women’s health [9]. Recent in-depth research has revealed that direct stimulation of fungal structures (such as cell walls, cell membranes, and biofilms) and indirect effects of their metabolites can contribute to the development of numerous diseases, including tumors [10, 11]. Existing studies have not only revealed the presence of fungi in various cancers but have also systematically summarized new mechanisms and functional characterizations that affect cancer metastasis, providing new directions and potential strategies for clinical cancer treatment through microbiome regulation [12, 13]. However, there is currently no evidence to suggest the presence of fungi in cervical cancer, and existing literature on the relationship between cervical cancer and fungal infection is limited and inconsistent. A review of existing literature reveals conflicting viewpoints; some studies suggest a correlation between fungal biota and high-risk HPV infection and cervical intraepithelial neoplasia (CIN) severity, while others suggest that Candida albicans within the fungus does not appear to be a contributing factor to cervical cancer development. Previous studies investigating the association between fungal infection and cervical cancer have yielded heterogeneous findings, partly due to variations in study populations, diagnostic methods, and adjustment for confounders such as HPV status. Beyond etiological factors, the impact of cervical cancer on patients’ quality of life is a critical aspect of oncology research. Interventions such as physical activity have been shown to improve quality of life in cancer patients [14, 15]. Understanding the role of fungal infection in cervical cancer may also contribute to comprehensive patient management, though the current evidence base remains limited. A systematic synthesis of available evidence is therefore needed to clarify this relationship and inform future research directions.

This study conducted a systematic review of the literature published between 2004 and 2025 to identify studies investigating the association between fungal infections and cervical lesions (including cervical cancer and precancerous lesions). Data from eligible studies were extracted and analyzed to explore this potential correlation. This study aims to clarify the relationship between fungi and cervical cancer, thereby establishing a foundation for subsequent research on the mechanism by which fungi induce cervical epithelial carcinogenesis.

Materials and methods

Search strategy and information sources

This study was conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement. The study protocol was pre-filed in the PROSPERO database prior to study initiation. Included studies were retrieved from January 1, 2004 to December 1, 2025 from three databases (PubMed, Web of Science, and Embase). The search combined MeSH terms and keywords related to cervical cancer and fungi using Boolean operators. The complete search strings for each database are provided in Supplementary Table. Additionally, the reference lists and relevant review articles were manually screened to identify other relevant research.

Eligibility criteria and study selection

Studies reporting the prevalence of fungal infections in cervical cancer patients were included. Study types included observational studies with cross-sectional, case–control, and cohort designs. Studies were excluded if they were reviews, animal studies, case reports, letters, conference or poster abstracts, studies of precancerous lesions and malignancies with sample sizes less than 100, or studies that did not contain raw data. Studies were included regardless of whether they statistically adjusted for HPV status. However, to address HPV as a key confounder, we planned subgroup analyses restricted to studies that provided HPV-stratified data on fungal infection rates. Both cervical cancer patients and fungal infections must be diagnosed using effective and objective methods. Two reviewers independently screened the eligibility of all relevant study titles and abstracts, reaching consensus through discussion when disagreements arose. These two reviewers then reviewed the full text, extracted and analyzed the data, compared the results, and consulted a third reviewer when disagreements could not be resolved through discussion. We also contacted authors of several publications for more information on methods and data.

Data extraction

Two reviewers independently extracted data using a standardized form. Disagreements were resolved through discussion or consultation with a third reviewer. From studies meeting the inclusion criteria, the following data were extracted into a predefined Microsoft Excel spreadsheet, including author name, publication year, country, primary research method, sample size, number of fungal infections, HPV infection status. Potential disagreements were resolved through discussion with a third reviewer.

All pathological results were jointly discussed and agreed upon by two experienced pathologists; if consensus could not be reached after discussion, a third more senior pathologist provided the final diagnostic opinion. The text and all figures and tables strictly follow this definition system.

To clarify terminology definitions, this study defines population categories as follows: (Table 1).

Table 1.

Classification of study groups based on cervical cytology and histology

Group	Description
NILM	Negative for Intraepithelial Lesion or Malignancy and healthy individuals
< CC	include all individuals without invasive cervical cancer, encompassing those with NILM, ASCUS, and CIN/SIL lesions
> NILM	comprising individuals with abnormal cytology or histology, including ASCUS, CIN/SIL, and cervical cancer
CIN1	Low-grade Squamous Intraepithelial Lesions (LSIL), Cervical Intraepithelial Neoplasia Grade 1(CIN1)
CIN2/3	High-grade Squamous Intraepithelial Lesions (HSIL), Cervical Intraepithelial Neoplasia Grade 2/3(CIN2/3)
unclear	unclear pathological grading of cervical lesions or mixed pathological types

Abbreviations: CIN Cervical Intraepithelial Neoplasia, ASC Atypical Squamous Cells

Assessment of risk of bias

For studies, two reviewers independently assessed the quality of the final included articles using the Newcastle–Ottawa Scale (NOS). The NOS tool focuses on three domains: selection of study participants, comparability between groups, and outcome/exposure factors, which are used to assess key confounding factors such as age, HPV infection status, sexual history, and exposure to antibiotics. The Newcastle–Ottawa Scale (NOS) was used, with scores of 0–3, 4–6, and 7–9 indicating low, moderate, and high quality, respectively. There were four items in selection of study participants, with a maximum score of 4; one item in comparability between groups, with a maximum score of 2; and three items in outcome/exposure factors, with a maximum score of 3. When two reviewers give inconsistent quality scores, they each reassess the article’s quality score and then discuss the differences. If they cannot reach an agreement, they consult a third reviewer to resolve the issue.

Statistical analysis

Meta-analyses were performed using STATA software (version 18, STATA Inc, College Station, TX, United States). A p-value less than 0.05 was considered statistically significant. For each study, event numbers associated with the occurrence of cervical cancer were collected. The metan method was used to calculate the prevalence of fungal infections in the studies. The logit transformation and random-effects model were used to account for both within-trial sampling variability and potential between-study heterogeneity. Given the heterogeneity across studies, a random-effects model was chosen a priori to provide more conservative pooled estimates. Heterogeneity was determined using the I² statistic, where outcomes range from 0 to 100%. No heterogeneity was observed when I² = 0%, and a p-value less than 0.05 indicated significant heterogeneity when I² was greater than 50%. If heterogeneity was present, meta-regression was applied to explore the sources of heterogeneity. Sensitivity analyses were performed on the included studies to identify any studies that had a significant impact on the outcomes. Publication bias was assessed using the Begg’s test.

Result

Search and select

The study initially identified 1490 studies by searching literature from January 1, 2004 to December 1, 2025 (747 from PubMed, 563 from Web of Science, and 180 from Embase). Due to duplication, 219 studies were excluded. 1271 studies were initially screened as irrelevant to our objectives based on title and abstract. After reviewing the titles and abstracts, 1249 studies were excluded for failing to meeting the inclusion criteria. Upon reviewing the remaining articles, 5 studies were excluded for the following reasons: (1) 2 studies were Summary of meetings; (2) 3 studies did not specify the diagnostic method used. Ultimately, 17 studies were included in the final meta-analysis. The detailed process of study selection is shown in Fig. 1.

Fig. 1.

PRISMA research flow chart

Evaluation of methodological quality

During the literature screening process, two reviewers independently evaluated the included studies and used Cohen’s Kappa coefficient to assess their agreement. The results demonstrated an overall concordance of 86.4% between the two researchers, with a Kappa value of 0.637 (95% confidence interval: 0.265–1.000), indicating good agreement. According to the Kappa evaluation criteria proposed by Landis and Koch (0.61–0.80 for good agreement), the Kappa value in this study falls within this range, suggesting high reproducibility and reliability in the researchers’ judgments regarding the inclusion and exclusion of literature. Additionally, the p-value of the Kappa test was 0.003, indicating statistically significant agreement (p < 0.05), further supporting the reliability of the screening results.

Basic characteristics of included studies

Detailed characteristics of the 17 included studies are summarized in the characteristics table (Table 2). We assessed study quality using our Critical Appraisal Checklist for studies. The distribution of judgments is presented in Fig. 2 and Table 3. The included studies were published between 2004 and 2025, with 6 studies collected between 2004 and 2014, and 11 studies collected between 2015 and 2025. Among the eligible articles (in terms of study type), there were 11 cross-sectional studies, 2 case–control studies, and 4 cohort studies. Of these, 16 were rated as high quality, and 1 as medium quality, indicating a very good overall quality. Geographically, 6 studies were from East Asia, 5 from Europe, 2 from the Americas, 3 from South Asia, and 1 study was from an unknown region. To further explore the impact of different socioeconomic backgrounds on the results, we analyzed the included studies based on the Human Development Index (HDI). In terms of demographic data, our meta-analysis selected 17 studies that included 97,382 different cases of cervical cancer and precancerous lesions, of which 6,844 were fungal infections; And 450,724 patients with negative for intraepithelial lesion or malignancy (NILM), of which 6,259 were fungal infections.

Table 2.

Characteristics of studies included

Study ID	Participants	Country	Study type	Diagnostic method	Research characteristics
Jordan Hall 2009 [16]	> NILM = 7334	America	case–control study	cytological method	Hospital-based sample
Filip Jansaker 2022[32]	> NILM = 68,030	Switzerland	cohort study	morphological method	Population-based sample
Chris Yick-Kwong Lee, MMedSc 2007 [17]	> NILM = 1943	China	cohort study	cytological method	Hospital-based sample
M.K. Engberts 2006 [18]	NILM = 435,644,  > NILM = 10,010	Netherlands	cohort study	cytological method	Population-based sample
Rachel Masch 2024 [19]	> NILM = 864	El Salvador, China	cohort study	PCR	Hospital-based sample
Jata S. Misra 2006 [20]	> NILM = 1604	India	cross-sectional study	cytological method	Hospital-based sample
Tengfei Long 2022 [30]	NILM = 13,697, > NILM = 982	China	cross-sectional study	cytological method	Hospital-based sample
Geilson Gomes de Oliveira 2017 [29]	> NILM = 1357	Brazil	cross-sectional study	cytological method	Hospital-based sample
Subhojit Dey 2016 [21]	> NILM = 747	India	cross-sectional study	cytological method	Hospital-based sample
Yulong Zhang 2024 [22]	> NILM = 2517	China	cross-sectional study	cytological method	Hospital-based sample
Craciun Aurora 2021 [24]	NILM = 81,  > NILM = 169	Romania	cross-sectional study	cytological method	Hospital-based sample
Jing-Jing Zheng 2020 [25]	NILM = 100,  > NILM = 399	China	cross-sectional study	morphological method	Hospital-based sample
Ishita Ghosh 2017 [26]	NILM = 209,  > NILM = 274	India	case–control study	culture method	Hospital-based sample
Zhilian Wang 2017 [27]	NILM = 626,  > NILM = 410	China	cross-sectional study	morphological method	Population-based sample
P. Vieira-Baptista 2016 [28]	NILM = 367,  > NILM = 255	Portugal, Belgium	cross-sectional study	morphological method	Hospital-based sample
Xu Caiyan 2011 [31]	> NILM = 374	China	cross-sectional study	morphological method	Population-based sample
Elena Bernad 2010 [33]	> NILM = 113	Romania	cross-sectional study	culture method	Hospital-based sample

Fig. 2.

Forest plot of fungal infection rates by cervical pathological type (normal, ASC, CIN1, CIN2/3, carcinoma, CIN123, unclear) using a random-effects model

Table 3.

Newcastle–Ottawa scale

Author	Selection (4 stars max)	Comparability (2 stars max)	Outcome/Exposure (3 stars max)	Total NOS Score (9 max)	Risk of Bias
Filip Jansaker 2022 [32]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Chris Yick-Kwong Lee, MMedSc 2007 [17]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
M.K. Engberts 2006 [18]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Tengfei Long 2022 [30]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Geilson Gomes de- Oliveira 2017 [29]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Jing-Jing Zheng 2020 [25]	⋆⋆⋆	⋆⋆	⋆⋆⋆	8	Low
Zhilian Wang 2017 [27]	⋆⋆	⋆⋆	⋆⋆⋆	7	Low
P. Vieira-Baptista 2016 [28]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Craciun Aurora 2021 [24]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Subhojit Dey 2016 [21]	⋆⋆⋆	⋆⋆	⋆⋆⋆	8	Low
Yulong Zhang 2024 [22]	⋆⋆⋆	⋆⋆	⋆⋆	7	Low
Rachel Masch 2024 [19]	⋆⋆	⋆	⋆⋆⋆	6	Medium
Jata S. Misra 2006 [20]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Elena Bernad 2010 [33]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Xu Caiyan 2011 [31]	⋆⋆⋆	⋆	⋆⋆⋆	7	Low
Ishita Ghosh 2017 [26]	⋆⋆	⋆⋆	⋆⋆⋆	7	Low
Jordan Hall 2009 [16]	⋆⋆⋆	⋆⋆	⋆⋆	7	Low

Fungal infection rate in cervical cancer patients

Seventeen studies were analyzed, including 13 studies with a confirmed cervical cancer (6204 cases), HSIL (CIN2/3) (64,470 cases), LSIL (CIN1) (3436 cases), and normal cervix (450,724 cases). In four studies, the lesion grade was not clearly defined or the classification differed. These cases were placed in a subgroup labeled “unclear”. Figure 2 shows the 95% confidence interval for fungal detection in each subgroup (cancer, CIN1, CIN2/3, normal cervix, ASC, CIN123, unclear) using a random-effects model. The cancer subgroup had the highest overall prevalence at 19% (95% CI: 0.07–0.30, p < 0.001, I² = 96.3%) [16–22, 24–33]. Furthermore, our meta-analysis revealed an overall prevalence of 8% (95% CI: 0.06–0.09, p < 0.001, I² = 99.4%), 8% for NILM (95% CI: 0.06–0.09, p < 0.001, I² = 98.4%), 4% for ASC (95% CI: 0.02–0.06, p < 0.001, I² = 97.5%), 8% for CIN1 (95% CI: 0.02–0.15, p < 0.001, I² = 98.6%), and 7% for CIN2/3 (95% CI: 0.04–0.11, p < 0.001, I² = 97.0%). Figure 3 shows that the fungal infection rate in cervical cancer patients was 19% (95% CI: 0.07–0.30, p < 0.001, I² = 96.3%), significantly higher than the 8% infection rate in non-cervical cancer patients (95% CI: 0.06–0.10, p < 0.001, I² = 99.7%) [16–22, 24–33]. Figure 4 shows that the incidence of fungal infection was significantly higher in HPV-positive patients compared with HPV-negative patients (OR = 1.39, 95% CI: 1.07–1.81, p < 0.01, I² = 16.6%) [16, 18, 25]. Figure 5 shows that the incidence of fungal infection was higher in CIN1 patients compared with healthy individuals (OR = 1.13, 95% CI: 0.83–1.55, I² = 0.0%) [18, 26, 27]. Figure 6 shows that the incidence of fungal infection was higher in cervical cancer patients compared with healthy individuals (OR = 1.78, 95% CI: 0.93–3.42, I² = 50.9%) [24, 26]. Figure 7 shows that the fungal infection rate was higher in cervical cancer and CIN2/3 patients than in normal, ASC, and CIN1 individuals (OR = 1.11, 95% CI: 0.53–2.32, I² = 63.2%) [24–27].

Fig. 3.

Forest plot comparing fungal infection rates between cervical cancer patients and non-cancer controls (NILM, ASC, CIN)

Fig. 4.

Forest plot of the association between HPV status and fungal infection in non-NILM patients

Fig. 5.

Comparison of fungal infection rates between CIN1 patients and NILM

Fig. 6.

The rate of fungal infection between cervical cancer patients and NILM

Fig. 7.

Fungal infection rates in patients with cervical cancer and CIN2/3 compared to NILM, ASC, and CIN1

Meta-regression

Due to significant heterogeneity among the included studies, a meta-regression analysis was performed to explore the sources of heterogeneity. Meta-regression analyses were conducted on HDI, publication year, region, and study design. Table 4 shows the results of the meta-regression analysis, which indicates that the geographic distribution of patients (Europe) (p < 0.05) was significantly associated with the heterogeneity of the included studies.

Table 4.

Results of meta-regression analysis

Variable		Coefficient	SE	P value	I2
HDI	very high	−0.0361175	0.037706	0.353	99.7%
	high				98.4%
	medium				97.7%
Study design	case–control study	0.0137221	0.0405512	0.740	97.3%
	cohort study				99.8%
	cross-sectional study				98.6%
Year	2006–2016	0.0428154	0.0555749	0.453	97.3%
	2017–2024				98.6%
Region(Europe)	Yes	−0.1179127	0.054349	0.047	99.8%
	No				98.1%

Publication bias and sensitivity analysis

The Begg’s test in Fig. 8 (p = 0.758 > 0.05) indicates that publication bias had no significant effect. Begg’s test showed no significant publication bias, but the power of this test is limited given the relatively small number of included studies; thus, publication bias cannot be entirely ruled out. The sensitivity analysis in Fig. 9 indicates that no single study fundamentally altered the overall prevalence of all outcomes.

Fig. 8.

Begg’s test

Fig. 9.

Sensitivity Analysis

Discussion

Cervical cancer is a major public health problem threatening women’s health worldwide [34]. This study systematically depicted the fungal infection rate at various stages of cervical progression, including no intraepithelial lesion and malignant changes, CIN, and cervical cancer. Evidence confirms that fungal infection is often associated with an increased risk of common cancers such as pancreatic cancer, colon cancer, and skin cancer [35–37]. Accumulating evidence indicates that fungal overgrowth can contribute to tumor risk, highlighting the pivotal role of fungi in cancer prevention. Previous studies have reported a wide variation in fungal infection rates among cervical cancer patients. For instance, a large population-based study by Jansaker et al. in Switzerland reported a rate of 3.8%, whereas an investigation by Craciun et al. found a prevalence of 38%. Consistent with the notion that fungal infections are associated with cervical pathology, our meta-analysis of 17 articles revealed a significantly higher fungal infection rate in cervical cancer patients (19%) compared to non-cervical cancer patients.

However, this elevated fungal prevalence was not observed in precancerous stages. Our analysis revealed that the fungal infection rate in patients with cervical intraepithelial neoplasia (CIN, approximately 8%) was comparable to that in individuals without intraepithelial lesions or malignancy. This finding suggests that fungal infection may play a more prominent role in the advanced progression of cervical lesions rather than acting as an early driver of carcinogenesis.

Notably, subgroup analysis restricted to non-NILM patients showed that HPV-positive individuals had a significantly higher risk of fungal infection than in HPV-negative counterparts (OR = 1.39, 95% CI: 1.07–1.81, p < 0.01, I² = 16.6%). This interaction implies that HPV status may facilitate fungal colonization or proliferation, suggesting a potential synergistic role for both pathogens in reshaping the cervical microenvironment during disease progression.

Given the significant heterogeneity observed among the included studies, we performed a meta-regression analysis, which identified geographical region (Europe) as an important contributing factor. This finding underscores the need for future prospective studies to include diverse geographical populations to mitigate potential bias.

Fungal colonization in the vagina is common [38]. While typically existing as commensal yeasts, under conditions of microbiota imbalance, mucosal barrier dysfunction, or immunosuppression, they can transition to a pathogenic state [23, 39–43]. This transition, often fueled by hormonal changes that increase local glycogen availability, triggers a persistent inflammatory response [44]. Such chronic inflammation is a well-established driver of carcinogenesis [45]. Although experimental studies have proposed that Candida may promote carcinogenesis through chronic inflammation, toxin secretion, and immune modulation, these mechanisms have not been directly demonstrated in cervical cancer and remain hypothetical in this context. Mechanistically, C. albicans can promote tumor progression through multiple pathways: it induces pro-inflammatory cytokines like IL-6 via the Akt/PI3K signaling pathway [46]; it secretes the toxin candidalysin, which upregulates COX-2 through MAPK/p53 pathways and activates VEGF signaling to promote angiogenesis and progression [47–49]; and it activates the Th17/IL-17 immune axis via Dectin-1 and TLR receptors [50], leading to chronic inflammation that activates STAT3 and NF-κB pathways [51], thereby promoting epithelial cell proliferation and tumorigenesis. This persistent inflammatory state further disrupts cervical epithelial barrier function by downregulating tight junction proteins and may facilitate extracellular matrix remodeling through upregulation of matrix metalloproteinases (MMPs), ultimately creating conditions conducive to tumor cell invasion and metastasis [52].

Our subgroup analysis revealed a significant positive association between HPV and fungal infection in non-NILM populations, suggesting a potential interplay between these two pathogens. Current evidence points to a possible bidirectional relationship. On one hand, HPV infection may create a permissive environment for fungal colonization by disrupting the cervical epithelial barrier and inducing local immune dysregulation [53]. On the other hand, as detailed above, Candida-induced chronic inflammation and immunosuppression may impair the host’s ability to clear high-risk HPV, thereby promoting viral persistence and lesion progression [54]. This has led to the hypothesis of a mutually reinforcing “vicious cycle” between HPV and fungi in cervical carcinogenesis [49, 55, 56]. However, the exact molecular mechanisms remain poorly understood. While some studies suggest that Candida co-infection may affect HPV-related biomarkers such as p16 expression, data on direct molecular interactions between fungal virulence factors and HPV oncoproteins E6/E7 are still lacking [57–59].

Notably, due to insufficient data on HPV and fungal co-infection in the cervical cancer studies included in our meta-analysis, we were unable to directly assess the cancer risk associated with their co-occurrence. Therefore, while our findings reveal a significant association between HPV and fungal infection in the non-NILM population, this result should be interpreted with caution regarding its generalizability to cervical cancer patients. It primarily reflects the co-occurrence pattern in precancerous stages, and whether this synergy directly translates to an elevated risk of invasive cancer remains to be validated in future studies specifically designed for cancer cohorts.

Fungal infections do not occur in isolation; they are often both a contributor to and a consequence of broader cervical microbiota dysbiosis. This dysbiosis is typically characterized by reduced protective Lactobacillus species and an overgrowth of anaerobic bacteria, sometimes accompanied by other pathogens such as Chlamydia trachomatis. Such a “high-risk” microbial environment may synergize with HPV and fungi to exacerbate local immune dysregulation and chronic inflammation, further increasing the risk of persistent HPV infection and lesion progression [60, 61].

Interestingly, parallels can be drawn with other cancer types where microenvironmental modulation has proven critical. In oropharyngeal cancer, for instance, targeted surgical approaches like transoral robotic surgery (TORS) have demonstrated superior functional outcomes by preserving local tissue architecture compared to conventional treatments. This parallel suggests a broader principle that targeted interventions aimed at preserving function and modulating the local microenvironment, whether these approaches are surgical or microbiological, it could represent a promising strategy for optimizing cancer care. In the context of cervical cancer, this raises the possibility that restoring a healthy microbial ecosystem could serve as an adjunct strategy to enhance HPV clearance or slow disease progression.

However, fundamental knowledge gaps remain. Systematic studies are still lacking regarding whether Candida and its metabolites (e.g., candidalysin, cell wall components) directly regulate cell cycle, apoptosis, or oncogenic signaling pathways in cervical epithelial cells, and whether they synergize with HPV oncogenes E6/E7 at the molecular level. Elucidating these key questions will contribute to a deeper understanding of the inflammation-driven mechanisms of cervical cancer and provide a theoretical foundation for developing intervention strategies based on microecological regulation.

Despite our efforts to explore sources of heterogeneity through subgroup and meta-regression analyses, significant between-study heterogeneity persisted. Beyond the geographical region (Europe) identified in our analysis, other clinical and methodological factors may contribute to this variability. These include differences in fungal detection methods (e.g., culture-based vs molecular techniques), variations in diagnostic criteria for fungal infections, and inconsistent adjustment for key confounders such as antibiotic or immunosuppressive medication use across the primary studies.

Several limitations of this meta-analysis should be acknowledged when interpreting our findings. First, the observational nature of the included studies (cross-sectional, cohort, and case–control designs) precludes any causal inferences. The associations observed between fungal infection and cervical lesions should therefore be interpreted as correlations rather than causal relationships. Second, from a strict methodological perspective, the inclusion of case–control studies and the handling of multiple subgroups from the same study as independent data points may potentially inflate effect estimates and underestimate confidence intervals. This decision was necessitated by sample size constraints but represents a methodological compromise. Third, our review may be subject to language bias, as only studies published in English were included. Additionally, some subgroup analyses were based on a small number of studies, which may be prone to small-study effects and limit the generalizability of findings; thus, these findings should be considered exploratory and warrant confirmation in larger prospective cohorts. Fourth, data on key clinical outcomes, including lymph node involvement, distant metastasis, treatment response, and long-term survival, were not available in the primary studies, precluding analysis of the prognostic impact of fungal infections in cervical cancer.

Residual confounding is a concern, particularly regarding HPV status, sexual behavior, antibiotic use, and immunosuppression, which were not consistently adjusted for in the primary studies. Our subgroup analysis partially addressed HPV, but unmeasured or poorly measured confounders may still influence the observed associations.

In view of these limitations, our findings should be interpreted with caution. Large-scale, well-designed prospective studies with standardized microbiological assessments and comprehensive data collection are urgently needed to clarify the temporal relationship between fungal infection and cervical carcinogenesis and to determine its potential as a modifiable risk factor. If confirmed, mechanistic investigations, including appropriate in vivo models, will be essential to elucidate the underlying biological pathways.

Conclusion

This meta-analysis demonstrates a high prevalence of fungal infection in cervical cancer, and reveals a positive association between fungal colonization and HPV positivity in non-NILM women. These findings support the need for longitudinal studies to establish temporal relationships and clarify whether fungal infection plays a causal role in cervical carcinogenesis. Future research should also investigate the interaction between fungi and HPV using stratified analyses and explore underlying mechanisms in experimental models.

Abbreviations

CIN

Cervical Intraepithelial Neoplasia

ASC

Atypical Squamous Cells

NILM

Negative for Intraepithelial Lesion or Malignancy

HPV

Human Papillomavirus

NOS

Newcastle–Ottawa Scale

PRISMA

Preferred Reporting Items for Systematic Review and Meta-Analysis

HDI

Human Development Index

Authors’ contributions

Conceptualization, Yunlai Wu; methodology, Yunlai Wu, Yu Li and Yidan Zhang; software, Yunlai Wu; validation, Yunlai Wu, Yu Li and Xinxin Qian; formal analysis, Yunlai Wu; investigation, Yunlai Wu, Yu Li and Yidan Zhang; resources, Yunlai Wu, Yu Li and Xinxin Qian; data curation, Yunlai Wu, Yu Li and Yidan Zhang;writing—original draft preparation, Yunlai Wu, Yu Li and Xinxin Qian; writing—review and editing, Yunlai Wu, Yu Li and Yidan Zhang; visualization, Xinxin Qian; supervision, Yu Li; project administration, Yunlai Wu. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Basic Research Funding for Provincial Higher Education Institutions in Heilongjiang Province(2022-KYYWF-0825), and Qiqihar Medical University Young Doctor Special Research Fund Project(QMSI2024B-02).

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files. Further inquiries can be directed to the corresponding author.

Declarations

Ethics approval and consent to participate

This study is a systematic review and meta-analysis of previously published literature and did not involve the collection of primary data from human participants or animals.

Consent for publication

This manuscript does not contain any individual person’s data in any form.

Competing interests

The authors declare no competing interests.