Browse the corpus

Mpox is a zoonotic infectious disease caused by the mpox virus (MPXV). The WHO declared the mpox outbreak a Public Health Emergency of International Concern (PHEIC) in 2022 and again in 2024, issuing a global alert []. Since 2022, nearly 180,000 confirmed mpox cases have been reported from 144 countries []. MPXV belongs to the orthopoxvirus genus of the family Poxviridae ; is closely related to vaccinia virus, variola virus, and cowpox virus; and can be divided primarily into Central African and West African clades []. The incubation period for MPXV infection is 7–21 days. Nonspecific clinical symptoms include headache, fever, and lymphadenopathy. Characteristic skin lesions typically progress through stages of macules, papules, vesicles, pustules, and crusts, which can be present simultaneously. In men infected via sexual contact, the rash is predominantly located in the anogenital region []. Although MPXV infection is often self-limiting, severe immunosuppression is a significant risk factor for more severe disease and even death [,]. In mpox patients with immune dysfunction, infection can still trigger a strong nonspecific immune response at the local lesion site, but the ability of the body to control viral replication and clear infected cells is markedly reduced [].

Notably, HIV and MPXV exhibit a high rate of coinfection, a pattern that is particularly prominent among men who have sex with men (MSM) []. HIV primarily infects CD4 + T cells, leading to their progressive decline and severe functional impairment, resulting in defective immune responses and dysregulation of immune coordination mechanisms []. Antiretroviral therapy (ART) cannot fully reverse the resulting immunopathological damage, and immune function recovery is often incomplete []. Severe immune suppression is a significant risk factor for prolonged disease duration, more complex and severe clinical manifestations, and higher mortality in mpox patients []. However, the precise effects of HIV coinfection on key clinical outcomes in mpox patients, including hospitalization rates, mortality, sexually transmitted coinfections, clinical manifestations, skin lesion characteristics, and biochemical markers, remain incompletely characterized. In particular, whether these outcomes differ according to the level of HIV viral control is poorly defined. In the absence of specific anti-MPXV drugs, summarizing and analysing the impact of HIV coinfection on mpox patients is crucial for informing the development of clinical guidelines [].

This meta-analysis synthesizes available observational studies published since the onset of the mpox outbreak, combining data on hospitalization, death, coinfection with sexually transmitted pathogens, clinical features, skin lesions, and blood test results among mpox patients. HIV infection was used as the exposure factor to compare differences between the exposed and control groups. When the data were available, exploratory subgroup analyses were performed stratifying the data by economic development level, HIV viral suppression status, and the proportion of participants with advanced HIV disease (AHD). This work aims to clarify the effects of HIV coinfection on mpox patients and provide a reference for guiding clinical management.

This meta-analysis was registered with PROSPERO ( https://www.crd.york.ac.uk/PROSPERO , ID: CRD420251021372). It was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [].

Searches were performed in PubMed, Cochrane, EMBASE, and Web of Science for original studies investigating the impact of HIV on MPXV infection. The search time frame ranged from database inception to March 2025. The search strategy included terms for HIV (Acquired Immune Deficiency Syndrome Virus; Human Immunodeficiency Virus) and MPXV (Monkey Orthopox Virus; Monkey Pox Virus). The detailed search strategy is provided in Table S1.

On the basis of the patient, exposure, comparison, outcome, and study design (PECOS) framework, the inclusion criteria were as follows: (1) participants: mpox patients; (2) exposure and comparison: exposed group: mpox patients coinfected with HIV; control group: mpox patients who were HIV-negative; (3) outcomes: the outcomes of interest were hospitalization rate, mortality, proportion of patients coinfected with other sexually transmitted pathogens, clinical features, skin lesion characteristics (location, morphology, number), and blood test results; and (4) study design: observational studies. The exclusion criteria were as follows: (1) case reports, animal studies, reviews, guidelines, expert consensus opinions, preprints, editorials, and conference abstracts were excluded to maintain evidence quality; (2) incomplete data: studies for which the full text was unavailable or those that did not report any of our specified outcomes; and (3) for the hospitalization outcome specifically, studies that enrolled only hospitalized patients or no patients with hospitalization events were excluded. This meta-analysis included only literature published in English.

Two researchers independently screened the retrieved records using predefined inclusion and exclusion criteria. The initial search results were imported. Two authors independently screened the titles and abstracts to identify potentially eligible records. The full texts of these potentially eligible publications were then reviewed against the inclusion and exclusion criteria. The references of all included studies were also screened for additional eligible studies. Two reviewers independently extracted data from each study using a prespecified data extraction form. Disagreements between reviewers were resolved through consultation with a third reviewer. For each publication, the following information was extracted: first author’s name, cohort year, study location and period, sample size with HIV coinfection, and outcome measures (hospitalization rate, mortality, proportion with other STIs, clinical features, skin lesion characteristics, and blood test results). If the data were presented only graphically, the graphic digitization software Engauge Digitizer (version 12.1; https://markummitchell.github.io/engauge-digitizer/ ) was used to digitize and extract the numerical values.

Two researchers independently assessed study quality by reading the full texts. Case–control and cohort studies were evaluated using the Newcastle-Ottawa Scale (NOS) []. The NOS assesses three domains: selection of study groups, comparability of groups, and ascertainment of either the exposure or the outcome. The quality is classified on the basis of scores: high quality (≥ 7), moderate quality (4–6) and low quality (< 4). Cross-sectional studies were assessed using the Joanna Briggs Institute (JBI) critical appraisal tool to determine eligibility for inclusion [].

All subgroup analyses in this study were post hoc and exploratory, and the results were intended only for hypothesis generation and not for confirmatory conclusions. (1) Economic development level: After the literature search and study selection were completed, we observed that the included studies originated from regions with substantial differences in economic development and hypothesized that this variation might be a major source of between-study heterogeneity. To explore this possibility, we conducted a post hoc, exploratory subgroup analysis based on the economic development level of the country in which each study was conducted (developing vs. advanced economies) []. (2) HIV viral suppression and CD4 + T-cell count: Two separate post hoc subgroup analyses were conducted using study-level aggregate data. Viral suppression: We extracted the proportion of participants per study who had achieved viral suppression (HIV RNA < 200 copies/mL or undetectable) and grouped studies with proportions of < 95% or ≥ 95% [,]. Advanced HIV disease (AHD): We extracted the proportion of participants per study with a CD4 + T-cell count < 200 cells/μL and grouped studies with proportions ≤ 5% or > 5% []. Only studies with ≥ 90% completeness for the relevant variable were included. The two indicators were analysed in separate univariate subgroup analyses, without cross-stratification.

For outcomes ≥ 2 subgroup analyses, the Benjamini-Hochberg false discovery rate (FDR) procedure was applied within each outcome family to account for multiplicity. Subgroup interaction P values were extracted from meta-regression models (or from subgroup comparisons using the Q-test for interaction), and adjusted P values (q values) were calculated to control the expected proportion of false positives among all significant results. Raw and adjusted P values are reported for all subgroup analyses subject to correction. Subgroup signals with an adjusted P < 0.05 were considered robust to multiplicity; otherwise, they were interpreted as exploratory. For outcomes with only a single subgroup analysis, no multiplicity correction was applied.

The approach to assessing publication bias was determined by the number of available studies per outcome, which is consistent with the recommendation of the Cochrane Handbook that funnel-plot-asymmetry tests should be performed only when at least ten studies are included []. For outcomes with fewer than ten studies, we did not perform formal statistical tests; instead, we qualitatively assessed potential small-study effects by visually inspecting the distribution of effect estimates in relation to study size. Given these limitations, we did not assert the “absence of publication bias” for any analysis. Sensitivity analysis was performed by sequentially excluding each study to check the robustness of the pooled results.

This meta-analysis across four databases yielded a total of 1,879 articles. After screening according to the inclusion and exclusion criteria, 29 articles were ultimately included (). The studies were conducted in 12 different countries: 14 in advanced economies and 14 in developing economies, and one was a multicentre cohort study involving 15 sites. A total of 15,522 patients were included, 94.54% of whom were male at birth, and 52.43% of whom were coinfected with HIV (). A total of 20 studies reported the proportion of participants with AHD in the exposed group, of which 10 had ≥ 90% data completeness. Among these, the AHD proportion was > 5% in 4 studies and was ≤ 5% in 6 studies. Twenty studies reported the viral suppression rates in the exposed group, of which 15 had ≥ 90% data completeness; viral suppression was < 95% in 12 studies and ≥ 95% in 3 studies. Eight studies reported the number of individuals newly diagnosed with HIV during the mpox episode ( and Table S4). Across the overall study population, prior smallpox vaccination was documented in 14 studies (with > 10% vaccination coverage in 4 studies), and mpox vaccination was reported in 5 studies, predominantly with JYNNEOS (Table S5).

The hospitalization status of patients was reported in 27 articles. In 7 of these articles, all the patients were hospitalized or none were hospitalized. The specific number of hospitalized patients could not be obtained in 1 article. Finally, the results from 19 articles were included in this meta-analysis. The overall hospitalization rate was 6.43%. Among the exposed group (n = 7,101), the hospitalization rate was 7.82%, whereas it was 4.84% in the control group (n = 6,201). The random-effects model revealed that the hospitalization rate was significantly greater in the exposed group [RR = 1.50, 95% CI (1.17, 1.93)] (). Between-study heterogeneity was present (I 2 = 76.2%).

(1) Economic development level: In advanced economies, the hospitalization rate was greater in the exposed group [RR = 1.81, 95% CI (1.31, 2.51)]. In developing economies, the difference between groups was not statistically significant [RR = 1.32, 95% CI (0.90, 1.95)] (). (2) AHD: Seven articles were included in the subgroup analysis. No differences were found in the subgroup with AHD > 5% [RR = 1.29, 95% CI (0.80, 2.09)] or AHD ≤ 5% [RR = 1.06, 95% CI (0.50, 2.55)] (). Egger’s test indicated no significant publication bias for this outcome (P Egger = 0.177). The funnel plot is shown in Figure S1 (A). No single study had a decisive influence on the overall conclusion, indicating that the findings are robust [Figure S2(A)]. After Benjamini-Hochberg correction for multiple testing, none of the subgroup interactions for hospitalization rate retained statistical significance (Table S7). These findings remain exploratory.

Cases of patient death were reported in 20 articles. No patient deaths were reported in 11 articles, and the results from 9 articles were included in this meta-analysis. The overall mortality rate was 0.58%. The mortality rate was 0.88% in the exposed group (n = 5,328) and 0.16% in the control group (n = 3,858). The random-effects model revealed that the hospitalization rate was significantly greater in the exposed group [RR = 4.54, 95% CI (2.11, 9.77)] (). Between-study heterogeneity was present (I 2 = 0.0%).

In advanced economies, the difference between groups was not statistically significant in the exposed group [RR = 2.55, 95% CI (0.18, 35.69)]. In developing economies, mortality was statistically greater in the exposed group [RR = 4.90, 95% CI (2.16, 11.13)] (). Fewer than ten studies were available; therefore, publication bias was assessed qualitatively only. The funnel plot showed no obvious asymmetry, although bias could not be ruled out [Figure S1 (B)]. No single study had a decisive influence on the overall conclusion, indicating that the findings are robust [Figure S2 (B)].

Eight studies reported syphilis coinfection. The prevalence of syphilis coinfection was significantly greater in the exposed group [RR = 2.08, 95% CI (1.56, 2.77)]. Five studies reported hepatitis C virus (HCV) coinfection, which was significantly more common in the exposed group [RR = 5.48, 95% CI (3.60, 8.34)]. Four studies reported hepatitis B virus (HBV) coinfection, with a significantly greater prevalence in the exposed group [RR = 2.57, 95% CI (1.47, 4.49)]. Six studies reported overall STIs rates, with no significant difference between the exposed and control groups [RR = 1.18, 95% CI (0.80, 1.76)]. The proportions of patients with gonorrhoea, chlamydia, and herpes simplex virus (HSV) coinfections did not significantly differ between the exposed and control groups.

In advanced economies, the difference in syphilis coinfection between groups was not statistically significant [RR = 1.62, 95% CI (0.79, 3.32)], whereas in developing economies, the exposed group had a higher syphilis rate [RR = 2.24, 95% CI (1.60, 3.12)]. For HCV, the exposed group had higher coinfection rates in both advanced [RR = 7.39, 95% CI (3.68, 14.82)] and developing [RR = 4.63, 95% CI (2.73, 7.83)] economies. For HBV, the exposed group in advanced economies had a higher coinfection rate [RR = 3.53, 95% CI (1.44, 8.69)], but no statistically significant difference was found in developing economies [RR = 1.44, 95% CI (0.55, 3.79)]. The detailed results mentioned above are presented in . Fewer than ten studies were available in all of these subgroups; therefore, publication bias was assessed qualitatively only. The funnel plot showed no obvious asymmetry, although bias could be ruled out [Figure S1 (C–H)]. No single study had a decisive influence on the overall conclusion, indicating that the findings are robust [Figure S2 (C-G)].

In this meta-analysis, the differences in 15 clinical features between the exposed and control groups were pooled. The random-effects model revealed that the exposed group had significantly greater incidences of fever [RR = 1.06, 95% CI (1.02, 1.10)], proctitis [RR = 1.50, 95% CI (1.16, 1.96)], malaise [RR = 1.08, 95% CI (1.02, 1.15)], rectal irritation syndrome (RIS) [RR = 1.29, 95% CI (1.11, 1.50)], diarrhoea [RR = 1.78, 95% CI (1.20, 2.63)], skin and soft tissue infection [RR = 2.65, 95% CI (1.33, 5.31)], and pneumonia [RR = 11.00, 95% CI (2.55, 47.35)]; no significant differences were found for the incidence of lymphadenopathy [RR = 0.94, 95% CI (0.86, 1.04)], headache [RR = 1.04, 95% CI (0.99, 1.09)], myalgia [RR = 1.04, 95% CI (0.94, 1.16)], pharyngitis [RR = 1.05, 95% CI (0.83, 1.32)], lower urinary tract symptoms (LUTS) [RR = 0.66, 95% CI (0.36, 1.23)], genital oedema [RR = 0.76, 95% CI (0.40, 1.47)], rhinitis [RR = 0.88, 95% CI (0.21, 3.60)], or tonsillitis [RR = 1.27, 95% CI (0.36, 4.53)].

(1) Economic development level : In advanced economies, the exposed group had higher rates of rectal irritation syndrome [RR = 1.35, 95% CI (1.04, 1.75)] and skin and soft tissue infection [RR = 2.45, 95% CI (1.17, 5.10)], but these differences were not statistically significant in developing economies. In developing economies, the exposed group had higher rates of fever [RR = 1.10, 95% CI (1.03, 1.17)], proctitis [RR = 1.65, 95% CI (1.23, 2.20)], and malaise [RR = 1.14, 95% CI (1.04, 1.24)], but these differences were not statistically significant in advanced economies (). (2) AHD : The exposed group had higher rates of fever in the subgroup with AHD > 5% [RR = 1.26, 95% CI (1.09, 1.46)] but this finding was not observed in the AHD ≤ 5% subgroup [RR = 1.08, 95% CI (0.98, 1.20)]. The exposed group had higher rates of proctitis in the subgroup with AHD ≤ 5% [RR = 2.01, 95% CI (1.54, 2.63)] but this finding was not observed in the AHD > 5% subgroup [RR = 1.11, 95% CI (0.65, 1.88)]. No statistically significant differences were found for pharyngitis between the two subgroups (Table S6).

(3) Viral suppression: The exposed group had higher rates of fever in the subgroup with viral suppression < 95% [RR = 1.11, 95% CI (1.04, 1.18)], but this finding was not observed in the viral suppression ≥ 95% subgroup [RR = 1.00, 95% CI (0.87, 1.17)]. The exposed group had significantly lower rates of lymphadenopathy [RR = 0.90, 95% CI (0.84, 0.97)] and LUTS [RR = 0.49, 95% CI (0.24, 0.99)] in the subgroup with viral suppression < 95%, but this finding was not observed in the viral suppression ≥ 95% subgroup. No statistically significant differences were found for headache, proctitis or genital oedema between the two subgroups (Table S6). Egger’s tests and qualitative assessment revealed no obvious asymmetry. Funnel plots are shown in Figure S1 (I-T). No single study had a decisive influence on the overall conclusions, indicating that the findings are robust [Figure S2 (H-S)]. After Benjamini-Hochberg correction for multiple testing, none of the subgroup interactions for fever, lymphadenopathy, headache, proctitis, pharyngitis, LUTS and genital oedema retained statistical significance (Table S7). These findings remain exploratory.

In this meta-analysis, differences in the location, morphology, and number of skin lesions between the exposed and control groups were pooled by a random-effects model.

The prevalence of skin lesions located on the genitals did not differ significantly between groups overall [RR = 0.91, 95% CI (0.76, 1.10)]. Subgroup analysis revealed that the differences were lower in the advanced economies [RR = 0.79, 95% CI (0.68, 0.92)] and the viral suppression ≥ 95% [RR = 0.80, 95% CI (0.66, 0.95)] subgroups. No statistically significant differences were found in the developing economies and viral suppression < 95% subgroups. The prevalence of skin lesions located in the perianal region was greater in the exposed group [RR = 1.25, 95% CI (1.09, 1.45)]. Subgroup analysis revealed that this difference was statistically greater in the developing economies [RR = 1.46, 95% CI (1.11, 1.92)] and the viral suppression < 95% [RR = 1.54, 95% CI (1.20, 2.00)] subgroups. No statistically significant differences were found in the advanced economies and viral suppression ≥ 95% subgroups. The overall prevalence of skin lesions located in the anogenital region did not differ significantly between the groups [RR = 1.00, 95% CI (0.83, 1.22)]. The detailed results mentioned above are presented in and Table S6. After Benjamini-Hochberg correction for multiple testing, none of the subgroup interactions for the prevalence of skin lesions located in the genitals and perianal region retained statistical significance (Table S7). These findings remain exploratory.

The proportion of patients with ≥ 10 skin lesions was not significantly different between the two groups [RR = 1.06, 95% CI (0.82, 1.36)]. Subgroup analysis revealed that this difference was greater in the developing economies subgroup [RR = 1.91, 95% CI (1.16, 3.15)]. The morphology of the skin lesions (vesicle, pustule, ulceration, papule, macula, scab) did not significantly differ between the two groups (). Egger’s tests and qualitative assessment revealed no obvious asymmetry. Funnel plots are shown in Figure S1 (U-Z). No single study had a decisive influence on the overall conclusion, indicating that the findings are robust [Figure S2 (T-W)].

In this meta-analysis, differences in 16 blood parameters between the exposed and control groups were pooled. The exposed group had significantly lower levels of albumin [SMD = −0.52, 95% CI (−0.94, −0.11)], calcium [SMD = −0.56, 95% CI (−0.90, −0.21)] and haemoglobin [SMD = −0.61, 95% CI (−0.96, −0.27)] and significantly higher levels of creatine kinase isoenzyme-MB (CK-MB) [SMD = 0.52, 95% CI (0.10, 0.93)]. No significant differences in white blood cell counts, C-reactive protein (CRP) levels, lymphocyte counts, platelet counts, alanine aminotransferase levels, lactate dehydrogenase levels, neutrophil counts, monocyte counts, aspartate aminotransferase levels, creatinine levels, creatine kinase levels or D-dimer levels were detected. Owing to the limited number of studies, subgroup analysis was not performed for blood test results ().