Aneurysmal subarachnoid hemorrhage (SAH) results in significant morbidity (1). Delayed cerebral ischemia (DCI) is a late complication occurring typically 4–14 days after onset in one-third of SAH caused by a combination of angiographic vasospasm, arterial constriction and thrombosis, and cortical spreading ischemia (2, 3) and is a major contributor to clinical outcomes (4, 5). Early brain injury (EBI) occurring within 72 h after aneurysmal rupture has been shown to be predictive of clinical outcomes (6). Global cerebral edema (GCE) quantifies the effacement of hemispheric sulci, and is a radiographic marker of early brain injury (7).

Uncontrolled inflammation occurs during EBI due to the reaction to extravascular blood (8), impaired cerebral autoregulation (9), release of products from injured brain tissue, and ischemia-reperfusion injury (7). Pro-inflammatory cytokines such as interleukin-1β (IL-1β) triggers the release of astrocyte derived extracellular vesicles that enter the peripheral circulation and promote the transmigration of leukocytes to the central nervous system (CNS) (10). The subsequent robust systemic inflammatory response occurring after SAH peaks at 24–48 h and contributes to delayed neurological deterioration (11, 12).

Peripheral leukocytes have been shown to migrate to the cerebrospinal fluid and brain after SAH (13) with activated neutrophils damaging brain micro-vessels (14). Preclinical models have shown that neutrophil and monocyte levels peaked by day 5 post SAH and that the leukocytes found in the brain originated from systemic blood circulation (5, 15). Early elevation in total peripheral leukocytes has been linked with the occurrence of DCI and poor functional outcomes (16, 17). However, leukocyte subtypes likely play distinct roles. After intracerebral hemorrhage (ICH), increased peripheral monocyte counts as well as neutrophil to lymphocyte ratios (NLR) are associated with worse outcomes (18–21). Similarly, after SAH, numerous studies have demonstrated that increased NLR is associated with worse outcomes (22–26). Peripheral monocytosis has also been linked to development of DCI (27). However, the biological mechanism underpinning the associations between leukocytes and outcomes and the link between inflammatory cells and EBI remain poorly understood.

The objective of this study is to investigate trends in leukocyte counts after SAH and the associations between leukocytes and outcomes. Total WBC count, monocytes, leukocytes, neutrophils, and basophils were collected at admission and each day for the next 8 days. We hypothesized that increased peripheral inflammatory cells will be associated with poor clinical outcomes and DCI.

This study demonstrates that peripheral leukocytes robustly distinguish outcomes after SAH, with persistently elevated WBCs, neutrophils, and NLR in those with DCI and poor outcomes. Although leukocytes were affected by clinical severity, there was an independent effect of cell counts on outcomes. Monocyte count (both on admission and medians from day 1–5) robustly predicted DCI and outcomes, in a sex dependent fashion. We found an association between infection and outcome and importantly demonstrated a link between neutrophils and NLR with the occurrence of infection after SAH.

Monocytes remained significantly elevated across all study days in subjects with DCI and poor functional outcomes. Peripheral monocyte count has been shown to increase early after brain injury (33). Monocytes play a beneficial role by removing debris in the subarachnoid space (34), however increased monocyte counts have also been associated with worse outcome after ICH (18). The pathophysiology underlying this association is unclear. However, monocytes have been linked to hematoma expansion in patients with ICH (35). Monocytes are coated with physiological anticoagulants such as thrombomodulin and tissue factor pathway inhibitor 2 (36, 37). A mouse model also found decreased extent of bleeding in an ICH model with impaired CCL2-CCR2 chemoattractant system (38). Peripheral monocytes play an important role in cerebral inflammation after injury in stroke, ICH, Alzheimer's and status epilepticus (34, 35, 39–42). Mice deficient in Toll-like receptor 4 (TLR4) demonstrated decreased recruitment of monocytes after ICH associated with decreased perihematomal edema (39). Therefore, monocytes may play a role both in hemorrhage expansion and inflammation after SAH. Indeed, the innate immune system has been shown to be activated early after SAH, with cells trafficking to the brain from the periphery (43). CD14⁺CD16⁻ monocytes (22) as well as monocytes expressing programmed death-1 (PD-1) are thought to play a role in the development of vasospasm, as blocking their ingress to the CNS after SAH attenuated vasospasm (44).

The early (day 0 and 1) increase among subjects with DCI and poor functional outcomes distinguished monocytes and NLR. Indeed, increased monocytes were robustly associated with DCI and poor functional outcome both considering median cell counts over days 1–5 (Tables 2, 3) and day 0 (Supplementary Table 3). The ability of monocytes to predict delayed cerebral infraction and poor outcomes after SAH has recently been demonstrated in a cohort of 204 patients (27). These authors found an optimal cutoff of 0.60 to discriminate development of cerebral infarction and poor outcomes after SAH, with AUC of 0.622. We found an improved AUCs for discrimination of DCI (0.640) and poor functional outcome (0.719). Our results also suggest a higher monocyte count may improve discrimination of DCI (0.683) and functional outcomes (0.810). When adding day 0 monocyte count to a baseline model including HHS, age, mFS, infection, IVH, and GCE, the AUCs further improved to 0.754 to DCI and 0.877 for functional outcome (Supplementary Figure 3).

Monocyte counts were significantly higher in males (Figure 4). Of note, monocytes had a better ability to discriminate DCI (AUC 0.703 vs. 0.611) and outcomes (AUC 751 vs. 655) in males compared with females. Monocyte function is sex specific and absolute numbers of monocytes have been shown to be higher in males than females (45). Several studies have also reported more robust cytokine production in male compared with female monocytes (46–48). Sex hormones impact immune cell function. Estrogen has been found to promote M2 polarization, which was dependent on the function of estrogen receptor α (49). We theorize that male monocytes are more likely to undergo M1 polarization with increased pro-inflammatory cytokine production after SAH. Furthermore, unlike neutrophils and NLR, monocyte count was not significantly associated with occurrence of infection (Table 4). Interestingly, an association between infection and monocyte count was only found in women and not seen in males. To summarize, monocytes increased steadily in both sexes. Monocyte counts were predictive of the occurrence of DCI in men. Monocyte counts were predictive of the occurrence of infections in women. This suggests that there are sex-specific effects of monocytes after SAH. Further examination of the underlying pathophysiology of these differences should be pursued.

An elevated NLR likely represents dysregulation of the immune system reflecting an increased innate and attenuated adaptive immune response (8, 50). In addition to playing a role in ICH and stroke (26, 51) an elevated NLR has been linked to outcomes in patients with cancer, myocardial infarction, and sepsis (31, 52–56). Neutrophils infiltrate the cerebral vasculature within 10 min after SAH, and reduction of neutrophil levels or activity improves vascular integrity and outcomes in animal models (14, 57). In addition to producing pro-inflammatory cytokines, neutrophils generate substantial oxidative stress by releasing factors such as myeloperoxidase, which can generate hypochlorous acid and consume nitric oxide resulting in impaired vasodilation (58). Isoprostanes generated by neutrophil mediated oxidative stress can also result in vasoconstriction (59, 60). Indeed, neutrophils may also contribute to early cerebral hypoperfusion after SAH (61).

This study builds upon results from prior studies, which have demonstrated the role of peripheral leukocytes after SAH. While several important studies have shown that WBC count has an association with DCI (62, 63), herein we have provided a comprehensive assessment of the role of differential cell counts on DCI and functional outcomes. Our assessments of the role of sex differences and associations with infection have also not been addressed in prior studies. We also support findings from previous studies indicating an association between elevated NLR and DCI (23, 25) and outcomes (24, 26). Our study has key differences with these reports, with a main difference being that most studies have exclusively assessed the role of early leukocyte counts. The largest study to date, which evaluated 1067 SAH patients, found an admission NLR cutoff of 5.9 to be highly predictive of the development of DCI with no association between NLR and functional outcomes (25). Our results demonstrate that WBCs, neutrophils, and NLR remain elevated in patients with DCI and poor functional outcome for a week after the initial bleed and these differences increase after the first two post-bleed days (Figure 1). This is in line with granulocytes being persistently elevated in the brain for at least a week after SAH in a rat model (64). We did not find associations with outcomes using early (day 0) WBCs, neutrophils, and NLR but found robust associations using median values over 5 days (Tables 2, 3). We suggest that the earliest time point for WBCs and NLR may not be as tightly linked to ultimate clinical outcome.

There was an association between neutrophils and NLR and the occurrence of infection after SAH, while lymphocytes were associated with fewer infections (Table 4). Elevated NLR plays a role in post SAH immunosuppression (26) and contributes to the risk of pneumonia (65). Neutrophils may play a role in the suppression of the adaptive immune system (50). The immunosuppressive capacity of some neutrophils is due to the ability to inhibit T-cell activation, proliferation, and effector functions (66). However, it is unclear from our results whether neutrophilia precedes infection, or whether early infections drive persistently elevated neutrophils and NLR. The role of lymphocytes in the injured brain is complex and incompletely understood. T cells infiltrate 48–96 h after injury and peak around 5 days (67, 68) in mice but have been found in the CSF after ICH in humans after 6 h (69). T cells comprise a heterogenous population and have been shown to have both protective and deleterious roles (70). Both pro-inflammatory (γδT cells) and immunosuppressive regular T cells (Tregs) traffic to the area of hemorrhage (32). γδT cells have been shown to contribute to injury after ischemic stroke mainly due to the production of IL-17 (71). Tregs may provide benefits against the late tissue damage (72), while in the earlier time frame they have been shown to modulate the activity of invading pro-inflammatory cells (73). The role of γδT cells may be primarily in the periphery rather than brain parenchyma as they accumulate in the leptomeninges and control trafficking of other inflammatory cells (74). While our study is not able to distinguish between lymphocyte subtypes, it is possible that the serum lymphocytes and decreased NLR associated with good outcomes and decreased risk of infection outcomes are primarily Tregs. We suspect that increased neutrophils and decreased lymphocytes are key factor leading to risk of infection after SAH, with this serving as a major determinant of clinical outcomes.

A major caveat to the interpretation of the results presented herein is the prevalent use of corticosteroids. At our institution, patients with aneurysmal SAH are typically treated with a standard regimen of corticosteroids during the first week of hospitalization, affecting most subjects in this study. Corticosteroids have an important effect on circulating leukocytes. Corticosteroid treatment increases neutrophils due to increased entrance from bone marrow and decreased vascular removal. However, monocytes, lymphocytes, basophils, and eosinophils are decreased after steroid treatment. This effect is thought to be due to redistribution of these cells, although lymphocytes are known to undergo steroid induced apoptosis (75). In addition, corticosteroids have potent anti-inflammatory effects. Corticosteroids inhibit recruitment of neutrophils and monocytes (76) and inhibit neutrophil adherence to capillary endothelial cells by decreasing expression of adhesion molecules and the synthesis and release of prostaglandins (77, 78). It is therefore very likely that corticosteroid treatment has affected the cell counts used in this study. As nearly all patient received corticosteroid treatment, it is not possible to control for this effect in our multivariate regression models. Additional studies will be needed to confirm the findings described herein in SAH subjects not treated with corticosteroids.

In addition to the effect of corticosteroids, this study has several other important limitations. This is a single center study, which limits its generalizability. This study is also retrospective, and it is therefore prone to selection bias. Given that multiple comparisons among cell types and across different days were made, there is a risk of type I error. We have adjusted P-values considered to be significant in univariate analyses for cell counts for multiple comparisons using a Bonferroni correction to mitigate this risk. We attempt to explore the pathophysiological basis for the connection between leukocytes and outcomes, however our study is limited to clinically available differentials. Subtypes of each leukocyte may play distinct roles (e.g., lymphocyte subtypes may be beneficial or harmful). Future flow cytometry studies will be used to answer questions about specific cell types. Several variables of interest in the study require qualitative assessment such as GCE, mFS, and DCI. We attempt to control for any variability and subjectivity by having all variables adjudicated by multiple reviewers.

This study determined the time course of peripheral leukocyte responses after SAH. We found that monocytes were associated with DCI and poor functional outcome, in a sex dependent fashion. Monocytes were elevated in men compared to women. In men, monocyte counts predicted the occurrence of DCI, while in women monocyte counts were associated with infection. Both neutrophil count and NLR were associated with worse outcomes. We suspect increased NLR (reflective of increased neutrophils and decreased lymphocytes) results in an immunosuppressive environment that contributes to infection risk, and this plays a role in outcomes. We suggest that monocytes (especially in males) play a key role in driving systemic and central inflammation after SAH, while neutrophils and NLR affect outcomes by modulating infection risk. Monocyte count may help to predict DCI and outcomes after DCI and may serve as a target for therapeutic intervention.

The data supporting the conclusions of this article will be made available by the authors upon reasonable request.

The studies involving human participants were reviewed and approved by University of Texas McGovern School of Medicine Institutional Review Board. The patients/participants provided their written informed consent to participate in this study.

AG designed the study, acquired and analyzed data, and drafted the article. JS analyzed data and helped to draft the article. ES acquired and analyzed data and helped to draft the article. AP acquired data and revised the article. AA acquired data and revised the article. S-BK contributed to study design and revised the article. LM contributed to study design and revised the article. HC contributed to study designed, analyzed data, and critically revised the article. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

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