Divergent neuroimmune signatures in the cerebrospinal fluid predict differential gender-specific survival among patients with HIV-associated cryptococcal meningitis

Introduction Survival among people with HIV-associated cryptococcal meningitis (CM) remains low, particularly among women, despite the currently optimal use of antifungal drugs. Cryptococcus dissemination into the central nervous system [brain, spinal cord, and cerebrospinal fluid (CSF)] elicits the local production of cytokines, chemokines, and other biomarkers. However, no consistent diagnostic or prognostic neuroimmune signature is reported to underpin the risk of death or to identify mechanisms to improve treatment and survival. We hypothesized that distinct neuroimmune signatures in the CSF would distinguish survivors from people who died on antifungal treatment and who may benefit from tailored therapy. Methods We considered baseline clinical features, CSF cryptococcal fungal burden, and CSF neuroimmune signatures with survival at 18 weeks among 419 consenting adults by “gender” (168 women and 251 men by biological sex defined at birth). Results Survival at 18 weeks was significantly lower among women than among men {47% vs. 59%, respectively; hazard ratio (HR) = 1.4 [95% confidence interval (CI), 1.0 to 1.9; p = 0.023]}. Unsupervised principal component analysis (PCA) demonstrated divergent neuroimmune signatures by gender, survival, and intragender-specific survival. Overall, women had lower levels of programmed death ligand 1, Interleukin (IL) (IL-11RA/IL-1F30, and IL-15 (IL-15) than men (all p < 0.028). Female survivors compared with those who died expressed significant elevations in levels of CCL11 and CXCL10 chemokines (both p = 0.001), as well as increased T helper 1, regulatory, and T helper 17 cytokines (all p < 0.041). In contrast, male survivors expressed lower levels of IL-15 and IL-8 compared with men who died (p < 0.044). Conclusions Survivors of both genders demonstrated a significant increase in the levels of immune regulatory IL-10. In conclusion, the lower survival among women with CM was accompanied by distinct differential gender-specific neuroimmune signatures. These female and male intragender-specific survival–associated neuroimmune signatures provide potential targets for interventions to advance therapy to improve the low survival among people with HIV-associated CM.

parameters in females compared with males.Nevertheless, in cryptococcal meningitis, the cryptococcal fungal burden, white cell counts, and protein in CSF, and the white cells, and CD4 + T cell numbers in blood were similar by gender 7 .The cryptococcal host immune evasion mediating factors, e.g., Cryptococcus fungal vomocytosis and the cryptococcal intracellular fungal proliferation in the cryptococcal infected macrophages, were similar by gender 47 .
We evaluated CSF-specific soluble cytokine, chemokine, and immunoregulatory responses to cryptococcal meningitis between females and males overall, and the differences in these neuroimmune responses in relation to survival by gender (Biological sex assigned at birth).We hypothesized that CSF neuroimmune cytokine and chemokine signatures would differ between females and males and across survival.In the CSF at baseline, we characterized representative Th1, Th2, Th17, Tfh cytokine, and innate myeloid-neutrophils activation regulating cytokines, and immune checkpoint markers (PD-L1) among people who survived or died during the 18 weeks of follow-up.We identified discrete differences in patterns of neuroimmune mediators by gender and by intragender-specific survival at the site of infection in the CSF.

Participants, Sites, and Setting
We enrolled 419 participants from a cohort of adults with CM enrolled to receive meningitis treatment in the Adjunctive Sertraline for the Treatment of HIV-Associated Cryptococcal Meningitis trial (ClinicalTrials.gov:NCT01802385).The trial was conducted between 2015 -2017 at the Infectious Diseases Institute (Administrative site) and at Mulago and Kiruddu National Referral Hospitals (patient catchment sites) in Kampala, Uganda 2,48 .The participants were selected based on gender and available survival-specific data for cross-sectional analyses with reference to 18 weeks of survival after cryptococcal meningitis diagnosis.The majority (94%; 168 females and 251 males) had complete demographic, clinical, and CSF datasets.Participants or their surrogates gave informed written and signed consent under protocols approved by the institutional review board of Makerere University and the University of Minnesota Medical School.Enrolled in the study were people ≥18 years of age, with a confirmed diagnosis of HIV-associated cryptococcal meningitis co-infection as previously described 49,50 .Only participants whose survival status was known at censoring or at study termination at 18 weeks of follow-up (379 of 419; 90.5%) were included in the survival sensitivity analyses.Of note, parent cohort survival did not differ statistically between either sertraline randomized participants or those on standard therapy alone 2 .

Specimen Preparation
CSF was drawn from lumbar punctures at diagnosis of CM prior to antifungal therapy initiation.The CSF specimens were spun to pellet out cells.The CSF supernatant was stored in a -80 0 C freezer prior to thawing for testing using Luminex.At enrolment, all participants in the current study were antifungal treatment naïve.

Statistical Analysis
Data were analyzed using GraphPad Prism version 9.3.0,GraphPad Software, LLC for Macintosh (San Diego, California, USA).The databases were compiled using Microsoft Excel.The data variability was visualized using unsupervised Principal Component Analysis, (PCA) using Eigenvector covariates projection on biplots as described elsewhere [58][59][60][61] .Further interrogation of the individual or independent principal component clustering and variability, factor differences, and interactions or associations of independent predictor variables by gender and by survival were performed using univariate and multivariate analyses.
Univariate analytic methods comprised pairwise comparisons using Mann-Whitney nonparametric U test that compared arithmetic medians and unpaired parametric t-test that compared arithmetic means.In this context of non-normally distributed population variables, statistical differences were reported based on the difference in the sample medians with interquartile ranges (IQR).The univariant difference in the survival outcome or survival risk was determined using Kaplan-Meier/Log-rank test (Mantel-Cox Chi-Square test) or Mann-Whitney U-test.The difference in binary outcome was determined using Fisher's exact Chi-square test or Mann-Whitney U-test.
In multivariate models, data were Log2 transformed to normalize the variable prior to the statistical interrogation using multivariate factor analysis and or survival-adjusted logistic regression least square models that measure the risk of likelihood (proportional hazard ratios).In all models, missingness was not imputed.In this context, statistical significance in both the univariate and the multivariate models were based on difference among variables in the original available dataset among participants.The statistical level of the significantly different covariates was reported at a p-value <0.05 and at a 95% level of confidence.The rigor of our interrogations included a large sample size (N=419), controlled comparative covariables at nearly a 1:1 ratio, and minimal missingness with approximately 94% data completeness in gender analyses and approximately 90% data completeness in the 18 weeks of survival analyses.

Baseline Demographics, by Gender
Consistent with published reports (Figure 1), among 419 participants studied with HIVassociated CM, the majority were males, who were older and heavier than females (Table 1).About half of these participants were on ART for a median of 1.6 months (IQR; 0 -22 months).Neurologic abnormalities predominated among clinical signs and symptoms, with almost all describing headache for one to four weeks, a third reporting changes in mental status, and a half with abnormal Glasgow Coma Score (GCS<15), with each variable reported at comparable frequencies in males and females.
As anticipated, CD4 + T cell numbers in circulation were low, but were marginally higher in females than males, as were platelet counts and hemoglobin levels (Table 1).CSF protein was not consistently elevated, and WBC counts were low, despite a high burden of yeast.Each result was generally comparable by gender, except CSF glucose which tended to be lower among females than males (Table 1).

The 18 Weeks Survival on Antifungal Treatment
In the parent trial for this analysis, males and females from Uganda and South Africa were randomized 1:1 to receive sertraline (an antidepressant with putative antifungal activity) or standard treatment but showed no differences in survival between treatment groups at 18 weeks 2 .An initial analysis by gender of an expanded data set, from which this current report is derived, showed significantly lower survival among 400 females vs. 577 males (at 50% vs. 57% survival, respectively) at 10 weeks, unadjusted (Hazard Ratio (HR) = 1.20; 95% CI, 1.00-1.45;p= 0.050) 7 .These differences were greater yet in this current subset extended to 18 weeks of observation.Survival among females was 47% (71/150 females) vs. 59% among males (136/230 males) (Mantel-Cox proportional hazard ratio (HR)=1.4(95% confidence interval, CI: 1.0-1.9);p=0.023) (Figure 2).Of note, survival was similar among males and females in two earlier studies reported prior to ART optimization (before 2015) but lower among females in two studies reported after ART optimization (after 2015), which include the current report (Figure 1-2).In the current study, survival overall, and by gender did not differ by ART experience (data not shown).
Survival was similar by gender in the first two weeks of antifungal therapy (Figure 2).However, survival diverged thereafter, remaining consistently lower among females throughout the 18 weeks of observation.As noted, baseline demographics, signs and symptoms, blood and CSF analytes, and cryptococcal fungal burden were relatively similar among participants by gender and by survival (Table 1).Thus, we considered whether the concentrations and the composition of neuroimmuneinduced factors at the site of severe cryptococcal disease in the CSF could underlie subsequent differences in survival by gender over time of observation.

Significant Differences in Neuroimmune Signatures in Cerebrospinal Fluid by Survival,
Gender, and Intragender-Specific Survival.
At baseline, we performed unsupervised Principal Component Analysis (PCA) as a primary approach to visualize the data variability and to explore potential unbiased differences in data clustering by gender, survival, and gender-specific survival (Figure 3).The PCA identified individual clusters that offered opportunities to structure downstream data analyses of the indicated model outcomes.The members in the cluster aggregated based on common attributes of the datasets showing high variability and distinct distribution of cytokines and chemokines between females and males (Figure 3A), between survivors and participants who died (Figure 3B), and within gender survival, among females (Figure 3C) and among males (Figure 3D).Eigenvector projections on principal components 1 and 2 (PC1 and PC2 respectively) indicated a high probability of neuroimmune variables predicting association with gender and survival outcome (Figure 3 A-D).High Eigenvalues >5 indicate the high capability of the selected covariables in predicting model-associated outcomes (Table below Figure 3).For each of the comparative groups, almost all showed such high Eigenvalues and, thereby, the separation between determinants in each group.These distinct patterns for PC1 and PC2 were closely correlated among participants by gender, subject by survival, and females by survival but less so for male survival in which 3 clusters were identified (correlation data not shown).Due to the striking data clustering observed in these groups, we next determined the specific neuroimmune factors contributing to the observed patterns using supervised univariate and multivariate data interrogation approaches.

Innate Neuroinflammatory Cytokines in the Cerebrospinal Fluid Differ by Gender
The CSF of females with CM at baseline had significantly lower levels of selected innate cytokines than that of their male counterparts, particularly IL-1RA and IL-15, TNF-α and immune checkpoint, PD-L1 (all p<0.050) (Table 2).The remainder of the cytokines and chemokines interrogated did not differ significantly by gender but tended to be lower still among females than males (Table 2).We next considered whether such gender-specific differences were associated with the differences in survival by gender.

(i). Female Gender-Specific Survival Attributes
A number of relevant factors differed between females who survived or died during the 18 weeks of observation.The circulating CD4 + T cell numbers were generally very low, but median CD4 + T cells were 31 cells/µL vs. 14 cells/µL among females who survived vs. those who died (p=0.009)but did not differ by survival in males (data not shown).Using several alternative models, and after adjusting for cytokines (Table 4 model 1), CXCL10 and CCL11 consistently predicted female survival but not in males.These soluble immune factors were consistently higher in magnitude among female survivors compared to those who died (Table 3).The levels of CXCL10 were significantly higher among females who survived than those who died, (p=0.013)(Table 3).By survival, CXCL10 levels in females who survived correlated with the number of CSF white cell counts (r=0.292,95% CI: 0.067-0.488;p=0.010), but not in males (Supplementary Figure 1).The fungal burden (CFUs) did not correlate with CXCL10 levels by either gender or gender-specific survival (data not shown).
The level of CCL11 (a chemoattractant produced by activated astrocytes, lymphocytes, and macrophages), was significantly higher in female survivors than in those who died, but not among males (Table 3).After adjusting for other cytokines, higher CCL11 expression still predicted female but not male survival (Table 4, Model 1).T cell-related factors (e.g., IL-12p70, IL-17A, and IFN-g) in the univariate models were also consistently increased in females who survived compared to those who died, but not in males (Table 3).The levels of other cytokines and the neutrophils chemokine IL-8 showed no difference among females by survival (Table 3).Only the IL-10 level was increased in both genders among those who survived in univariable analysis (Table 3).However, after adjusting for other factors, the levels of IL-10 expression were correlated with only female survival, but not with male survival (Table 4, Models 2 and 3), as was IL-2 (Table 4, Model 3).
In summary, although immune parameters were lower among all females vs. males, the females who survived showed consistently elevated levels of both myeloid-derived chemokines (CCL11), lymphoid-derived chemokine (CXCL10), T cell-derived cytokines, and the regulatory molecule PD-L1, the differences that were mostly distinct to survival in females, but not in males (Supplementary Figure 1).

(ii). Male Gender-Specific Survival Attributes
Males who survived had significantly higher hemoglobin compared with those who died, even after adjusting for the platelet counts and CD4+ T cell counts (Table 4, Model 3).Males, but not females, who survived had significantly lower IL-15 than those who died (Table 3), even after adjusting for other cytokines (Table 4, all Models).The IL-15 levels were independent of fungal burden (CFUs) and CSF white cell counts (data not shown).After adjusting for other factors, lower levels of the neutrophil chemoattractant IL-8 also consistently predicted male survival but not among females (Table 4, Models 1, 3-4).Unlike in males, neither low levels of IL-15 nor IL-8 were associated with survival in females.As noted above, elevated levels of regulatory IL-10 were associated with survival in both genders but did not correlate with male survival (Supplementary Figure 1).

DISCUSSION
In a large cohort of adults with CM in Uganda on anti-fungal therapy, survival among females was significantly lower compared with that of their male counterparts.Evaluating gender-specific mortality is uncommon in this context (only 5 including the current 21 studies; Supplemental Table S1), Unlike two previous studies which showed no difference in mortality by gender reported 26,44 , the two studies showing such a differences 7,42 were performed in people treated with antifungal therapy, which helps control fungal burden.Combining antifungal drugs with immunomodulatory interventions could limit the fungal assault on the CNS, illustrated by brain lesions from magnetic resonance imaging and from autopsy examinations among cases with neurological focal infections [62][63][64][65] .With appreciable control of fungal burden, immune responses may be the more prominent determinant of clinical outcomes.
We have identified distinct immune-biological differences in CSF by gender and genderspecific survival.The signatures associated with survival in females are distinctive from those in males.These differences were independent of baseline clinical features and cryptococcal fungal burden (CFUs).Most consistent in both the univariate and in multivariate gender-associated survival predictive models were the lower levels of CSF CXCL10/IP-10 and CCL11/Eotaxin that distinguished females who died from survivors, but not males.In contrast, high levels of IL-15 and IL-8 differentiated males who died matched to those who survived, but not females.Several T cell cytokines (IFN-γ, TNFα, IL-13, and IL-17A) similarly exhibited diminished levels of expression in females who died matched survivors, but not in males.Lower levels of immune regulatory IL-10 expression were linked with an increased threat of death in both genders.Thus, the biological significance of gender-specific CSF immune signatures suggested among participants with CM may underlie important immune-based mechanisms indispensable to improving antifungal treatment and survival predominantly in women with demonstrated distinct survival-associated cytokine and chemokine patterns.
Consistent with the univariate model observations, improved survival at 2-10 weeks was associated with elevation in the levels of soluble CSF cytokines Th1 IL-2, IFN-γ, TNF-α, Th2 IL-4, and Th17 IL-17A, together with innate IL-6 cytokine 66,67 .Soluble T-lymphokines levels in CSF are proposed to derive from CXCL10 68 and CCL11 69 mediated stimulation of CNS resident immune cells.Moreover, the use of cryptococcal antigens (such as GXM) to stimulate peripheral immunocytes demonstrated, that patients with improved 10 weeks survival had increased expression of these cytokines among polyfunctional differentiated CD4 T cells 67 .Other chemokines demonstrated to predict improved survival included the elevated levels of monocytes chemotactic protein-1 and macrophages inflammatory protein-1 66 .Despite demonstrated correlates of survival in the peripheral circulation 70,71 , the importance of characterizing immune responses in the CNS, at the foci of infection, cannot be underestimated since CNS responses in larger studies do not often correlate with those in the peripheral circulation 70,71 .The disparity between peripheral and CNS observed responses is a characteristic indicator of local mechanisms that influences responses.Compartment-specific responses have been noted previously 19,20 and argue that we should be investigating the response in CSF.Thus, the current study is unique in characterizing the distinctive neuroimmune signatures defining survival by gender in CNS, at the foci of disease.
In theory, the microbial invasion of the CNS activates resident immune cells to produce neuroimmune cellular activating cytokines, chemoattractant chemokines, and surrounding tissue basal cellular immune mediators (prostaglandins, leukotrienes, et cetera).Diverse effector cells include resident microglial, astrocytes, oligodendrocytes, CNS surveillance phagocytes (monocytes, macrophages, neutrophils), dendritic cells, adaptive T and B cells, innate (natural killer cells), and basal barrier epithelial cells 72 .Other cells may be drawn by extravasation and diapedesis through the protective vascular barriers to infiltrate the CNS in response to infection and/or to ensuing meningoencephalitis 19,20 .The consistent use of antifungal therapy in this study helps to control fungal replication and the burden of fungi 2,9,26,73 .With relatively effective antifungal therapy modeling immune factors highlights the role of the immune system in determining immune-mediated damage vs. protection in the CSF, and ultimately disease outcome 67,74,75 .
Although cryptococcal fungal burden and other clinical parameters in blood or in CSF did not differ by gender or gender-associated survival, increased production of IL-1RA, IL-15, and PD-L1 in the CSF of males overall may have influenced their poor survival than in their female counterparts.Lower male but not female survival was associated with high levels of myeloid cytokines, including the pleiotropic IL-15 and neutrophil-activating factors IL-8, which may influence poor survival in males by the distinctive mechanisms.In literature, IL-8 expression derived from microglia in CNS in response to neuroinflammation 76,77 , from monocytes and autocrine IL-8 production by neutrophils enhances immune activation and cellular proliferation 18 .The elevated intrathecal IL-15 and IL-8 in the CSF can enhance neutrophils-induced NETosis (Neutrophils traps).The increased neutrophil trap formation may propagate thrombotic vasculitis, cryptococcomas, and fungal occlusions in subarachnoid ventricular and arterial spaces in those with poor survival.Together IL-15 and IL-8 elevation can increase local inflammation in the brain, leading to vasculitis in subarachnoid spaces, interfering with blood supply, promoting ischemia, necrosis, and infarction of the brain tissue, and potentially leading to increased risk of death.Thus, the relatively high levels of IL-15 expression among males who died and the inverse relationship of IL-15 levels with those of immune regulatory IL-10 and immune checkpoint PD-L1 among males who died highlights the dynamic functional interrelationship of the potential loss of IL-15 immune regulatory function in influencing the greater risk of death among patients presenting with this neuroimmunological phenotype.Low levels of IL-15 may limit the mitotic division of especially fungal-infected macrophages which may attenuate the propagation of endogenous pathogens in quiescent cells 18 .Thus, the observed low levels of IL-15 and IL-8 among male survivors may limit intracellular cryptococcal fungal replication potentially leading to low fungal burden and improved survival.
Among neutrophils activated factors that shape neuroimmune response and outcome in CNS infection and in sequelae include observations that neutrophil induced responses exacerbate tissue injury through self-directed host and pathogen-mediated mechanisms 78,79 .That neutrophil activation is associated with the propagation of tissue necrosis, hypoxia, and nutritional supply impairment especially among people with sepsis and in those with multiple organ failure, that underlie brain injury 78,79 .In this context, activation of neutrophil recruiting cytokines and chemokine, at the foci of infection, in the CNS is detrimental to the host 76,78 , leading to the release of toxic granules, that are destructive to connective tissues, leading to vasculitis with tissue necrosis, infarction, hypoxia, and potentially death.Indeed, in cryptococcal meningitis, neutrophilia in circulation was associated with poor survival outcome 80 , brain hypoxia 15,81,82 , subarachnoid blood vessel occlusion 83 , brain tissue necrosis, and cerebellar infarction 64 .Subsequently, survivors of cryptococcal meningitis suffered longterm CNS abnormalities associated with impaired faculty observed in sequelae among people without HIV infection 62,84 .The persistent CNS abnormalities in sequelae noted in humans with fungal infections were similar to those observed among mouse models of Alzheimer's disease 85 .In other fatal meningitis focal infections including tuberculosis meningitis and bacterial meningitis, neutrophils account for the majority of cells in the CSF 86,87 .However, neither we nor others have characterized in detail the neutrophil myeloid lineages and role among CSF white cells in the setting of CM-associated survival.
Among key differences in female intragender survival with cryptococcal meningitis, survivors tended to shade into CSF relatively high amount of neuroimmune mediating soluble cytokine and chemokine responses across the panel.The higher levels of Th1 IL-2, IFN-γ, TNF-α, IL-12p70, Th17 IL-17A, and immune regulatory IL-10 may influence survival in females by distinct CXCL10 and CCL11 chemokine-stimulated mechanisms.In contrast to females who died, females with higher levels of the myeloid chemokine CCL11 and the lymphoid chemoattractant CXCL10 were more likely to survive, but not men.CXCL10 produced by astrocytes and microglia cells or by CNS resident monocytes, macrophages, and dendritic cells stimulates Th1 cytotoxic CD8 T cells, immune modulatory CD4 T cells, regulatory T cells, NK and regulatory B cells, and regulatory T cells via its CXCR3 receptor-mediated stimulation 88 .The cytotoxic response to infected macrophages is potentially enhanced in the presence of elevated chemokine in the central nervous, because infected cells respond to chemoattractant proteins at the site of infection 89,90 , whereas immune regulatory cells can modulate inflammation and cellular activation to stimulate tissue recovery 91 .Thus, independent of the baseline CSF cryptococcal CFUs and the baseline clinical features, the CCL11 and CXCL10 balance is critical in immune regulation of the Th1/Th2/17 balance, modulation of CNS cellular influx that is important in better survival of host with HIV-associated cryptococcal meningitis.
Interleukin (IL)-12 is observed as an important factor in the immune control of Cryptococcus meningitis as it is observed to influence mechanisms that inhibited cryptococcal fungal replication among Cryptococcus-infected macrophages in vitro 92 .In mouse models, early preemptive use of IL-12 as an adjuvant promoted survival of mice in early use after cryptococcal fungal 92 .Whereas the late treatment of mice with IL-12 after the establishment of cryptococcal infection resulted in more mice dying from fungal infection-attributable deaths 92 .In the current study, IL-12p70 was elevated among female survivors but not males.In the CNS IL-12 is produced by dendritic cells, and is observed to regulate T cell, and NK cell function, and influence the differentiation of the Th1 cellular phenotype 93 .With the prominence of T cells and B cells, in the CSF with cryptococcal meningitis 19,20 , increased IL-12p70 shading into the CSF can enhance fungal-specific immunocytes activation to mediate fungal neuroimmune mechanisms for control and for improved host survival.
The Th17 cytokine, IL17A, modulated through RORγt regulatory pathway was observed to work in tandem with Th1-modulated cytokines to regulate neuroimmune responses induced to CNS infections 94 .In particular, IL-17A activates Th17 immunocytes in response to tissue-resident infection.In cryptococcal model infection, IL-17 promoted the development of majority giants cryptococcal cells that have limited ability to cross the blood-brain barrier, leading to localized cryptococcal fungal infection with limited spread to the brain parenchyma where it would cause lethal disease 65,95 .Among females, IL-17A was upregulated among those with better survival outcomes.Other Th1 cytokines upregulated among females survivors included IFN-γ, TNF-α and their immune activating factors, CD40 ligand that was demonstrated to facilitate the formation of reactive phagosomes, activation of reactive oxygen and nitrogen species that mediate endogenous killing mechanisms of Cryptococci ingesting macrophages 65,95 .Collectively, these immune activating complementary factors, were each increased with increased survival among females.And in the aggregate, these novel findings (reported for the very first in the setting of CM gender-specific survival), drive the question as to how these results may be harnessed for targeted gender-specific therapy to optimize survival where fatal outcomes seem to be propagated by immune-mediated mechanisms that appear to operate independent of the fungal burden, clinical presentation, and current antifungal therapy.

CONCLUSION
Survival from HIV-associated CM remains significantly lower in females.We found novel distinctive and divergent intragender-associated CSF cytokine signatures at the time of clinical presentation that was independent of the baseline fungal burden and clinical features.Among females, increased concentration of CCL11, CXCL10, and T cell subset associated cytokines mediating the Th1 and th17 fungal immunopathogenesis pathways to be divergently augmented with female survival but not in males.However, among males, low levels of the neutrophil-activating chemokines IL-8 and the pleiotropic cytokine, IL-15 potentially fashion antifungal regulation neuroimmune milieu, distinctively expressed with improved male survival but not in females.Increased expression of IL-10 and PD-L1 regulatory mechanisms accompany improved survival in both genders identifying common features of a successful CSF profile.The elevated induction of CCL11, CXCL10, IL-17A, and IL-12p70 neuroimmune cytokine signature by female intragender-specific survival in the CSF and their strong positive correlation to IL-10 and PD-L1 highlight the interdependence of these responses in shaping immune homeostasis and survival associated-neuroimmune modulating milieu.These key observations reveal promising signals to consider the impact of more robust responses in females, particularly T cell-dependent (CXCL10, Th1/Th17 cytokines), and lower myeloid responses in males (e.g., IL-15 and IL-8) and their regulators in controlling outcomes.This data point to the potential neuroimmune mechanistic targets which could be modified to improve treatment and clinical outcomes in those with reversible deaths with immune-based combination antifungal therapy.

LIMITATIONS
One limitation is that the cross-sectional, exploratory study design generated data could not explain marker and clinical variations after the baseline and prior to death outcome or censor, which might be key to explaining the factors that underlie the observed gender-survival subjective mechanisms.The differential of myeloid vs. lymphoid white cells in the CSF compartment was largely unavailable.The limitations of our interpretations of the baseline cross-sectional findings could be enhanced with longitudinal data to appropriately capture and model changes in the immune profile from diagnosis to eventual survival or death.

CONFLICT OF INTEREST
The authors conformed to the International Committee of Medical Journal Editors (ICMJE) article publication criteria.The authors declared no conflict of interest.Study participants or funders had no role in the design, data curation, or intention to publish.Part of this work was presented in an accepted abstract and poster at the Conference on Retroviruses and Opportunistic Infections (CROI) at Seattle Convention Center, Seattle, Washington, 19-22 February 2023 and was published online on CROI website 96 .SO was a doctoral fellow and the examined thesis abstract was published online by the Makerere University Online Library as a requirement degree award 97 .

AUTHOR CONTRIBUTION
SO, EO, DBM, ENJ, YCM, DRB, JR, OOJ, HK, BB, and FB, conceptualized the study, framed the hypothesis, and edited the manuscript.SO, YCM, DRB, JR, DBM, and ENJ sourced for funding, designed experiments, guided data analysis, and drafted the manuscript.SO, performed statistical data analysis, compiled the results, put together the first drafted manuscript, and managed the review and publication process.The authors read and approved the final version of the manuscript for online publication.

CONTRIBUTIONS TO THE FIELD
The study involved a large sample size with equal ratios between comparative variables that give strong confidence to weighted treatment outcome predictors of HIV-associated cryptococcal meningitis.We discovered the critical role of biological gender in the assessment of survival among patients with HIV-associated cryptococcal meningitis with discrepant bias in women's poor survival.And that immune response differs within gender-specific survival.Together we highlighted the potential role of neuroimmune mechanisms in contributing to the differences in biological gender in survival.The observed immune mechanism illustrates the potential ability of the underlying stimulus in advancing immune-based treatment to work in synergy to improve the efficacy of antifungal drugs to improve survival among preventable deaths.
Interim, we propose close immune monitoring of CCL11/Eotaxin, CXCL10/IP-10, IL-12p70, and IL-17A as key indicators of the beneficial Th1 neuroinflammatory cytokine milieu among HIV and cryptococcal meningitis diagnosed women.We recommended mechanisms to augment CCL11/Eotaxin, CXCL10/IP-10, IL-12p70, and IL-17A cytokine deficient among women in addition to interventions to holistically refurbish the dysregulated or exhausted immune response among those co-infected in an attempt to significantly improve women's survival outcome.We propose close monitoring of IL-15, IL-8/CXCL8 together with IL-10 and PD-L1/B7-H1 dysregulation as key indicators of men's poor survival leading pathway.We recommend mechanisms to restore IL-15 and IL-8/CXCL8, IL-10, and PD-L1/B7-H1 checkpoint responses to improve men's survival outcomes.It is plausible that switching the CCL11/Eotaxin and CXCL10/IP-10 chemokine balance in later stages of disease progression to a potentially myeloid pathway can lead to a significant self-reliant neuroinjury inflicting pathway leading to poor survival outcomes.1).Odds ratios represent the independent likelihood of survival based on levels of the indicated factor with an adjusted multivariate logistic regression model.Statistic: intragender (male or female) multiple logistic regression.Not statistically significant variables had adjusted p-value ≥0.05 at a 95% confidence interval.The main effect of variables on gender survival.3A).The CSF secreted cytokines clustering by gender (Figure 3B).Cytokine clustering by female survival (Figure 3C).Cytokine clustering by men's survival (Figure 3D).

FIGURES
work was supported in part by funding from the National Institutes of Health and the National Institute of Allergy and Infectious Diseases R01 AI078934, U01 AI089244, R21 NS065713, R01 AI108479, and T32 AI055433 grants to DBM and DRB; R01 AI108479 grant to JR and R01 AI108479 funding to ENJ.National Institute of Neurologic Diseases and Stroke R01 NS086312, R25 TW009345, and K24 AI096925 grant to JR. Fogarty International Center D43 TW009771 grant to YCM and R01 NS086312 grant to RJ. GlaxoSmithKline Trust in Science Africa grant COL 100044928 grant to SO. DELTAS Africa Initiative DEL-15-011 grant to THRiVE-2, DBM.Wellcome Trust grants 107742/Z/15/Z grant to DBM.Veterans Affairs Research Service I01CX001464 grant to ENJ.The Wellcome Trust Training Health Researchers into Vocational Excellence (THRiVE) in East Africa, 087540 grant to DBM.United Kingdom Medical Research Council/Wellcome Trust/Department for International Development (MRC MR/M007413/1 grant to JR and the Grand Challenges Canada S4-0296-01 grant to JR. Funding agencies had no role in study design, data collection, data analysis plan, preparation of the manuscript, or in the decision to publish.Participants provided prospective informed written and signed consent and storage consent for specimen use in future studies.Makerere University Higher Degrees Research and Ethics Committee of the School of Biomedical Sciences granted a waiver of informed consent and ethical approvals to analyze the data.The protocol was registered and approved by the National Council of Science and Technology.The parent clinical trial was approved by the National Drug Authority and by the Uganda National Council for Science and Technology.The patient's identifiers were alphanumeric coded and could not identify those patients during the data search.

Figure 3 .
Figure 3. Projection of cytokine responses on Eigenvector correlation covariates on PC1 and PC 2 (axes) using unsupervised principal component analysis.Baseline cerebrospinal fluid (CSF) immune signature among participants who were diagnosed with HIV-associated CM showed distinct clusters by gender and by survival.PC -Principal component.Dots -show individual cytokines produced among participants projected on Eigenvector correlation covariates.Variables near the center (0; zero) are uncorrelated to the model and variables further away from the center (0; zero) are strongly correlated to the model.Variables among all cytokine projected on the variety of CSF cytokine patterns among participants by host survival (Figure3A).The CSF secreted cytokines clustering by gender (Figure3B).Cytokine clustering by female survival (Figure3C).Cytokine clustering by men's survival (Figure3D).

9 ) 897 *Supplementary Figure 1 .
-hospital deaths.= -12 weeks deaths.# -10 weeks deaths.^ -24 weeks.a -all deaths.Spearman's correlation of regulatory elements IL-10 and PD-L1 by gender among multivariate predictors of gender-specific survival among females and males who died on antifungal treatment.A-C correlation of IL-10 with A -CCL11/Eotaxin, B -CXCL10/IP-10, and C -IL-15 among females and males who died on antifungal treatment, (i -ii, respectively).D -correlation of IL-8 with IL-15 among females and males who died on antifungal treatment.E-H -correlation of PD-L1 with E -IL12p70, F -IL-15, G -CXCL10, and H -CCL11 among females and males who

Table 1 , Baseline Demographics of People With HIV-Associated cryptococcal meningitis by
805

Table 4 , Baseline Independent Factors Predicting Intragender 18 Weeks Survival Among Patients with HIV-Associated Cryptococcal Meningitis on Antifungal Therapy.
Multivariate adjusted models of gender-specific associated factors (univariate gender-specific survivalassociated factors in Table3), were interrogated with statistically different gender-associated demographic variables (Table

.
The proportion of cases with HIV-associated CM by gender and of related survival by gender in 21 published case series.These studies include 38,485 cases with 5,834 reported deaths, which accounted for a 15.2% case fatality rate (see Supplementary Table1for details and references7,9,19,26,28-  44.Bars show median values.5 studies report survival by gender (4 references and this report; Supplemental Table1).