Association between neutrophil-to-lymphocyte ratio and hematoma expansion in spontaneous intracerebral hemorrhage: A systematic review and meta-analysis

Abstract

BACKGROUND

Hematoma expansion (HE) typically portends a poor prognosis in spontaneous intracerebral hemorrhage (ICH). Several radiographic and laboratory values have been proposed as predictive markers of HE.

AIM

To perform a systematic review and meta-analysis on the association of neutrophil-to-lymphocyte ratio (NLR) and HE in ICH. A secondary outcome examined was the association of NLR and perihematomal (PHE) growth.

METHODS

Three databases were searched (PubMed, EMBASE, and Cochrane) for studies evaluating the effect of NLR on HE and PHE growth. The inverse variance method was applied to estimate an overall effect for each specific outcome by combining weighted averages of the individual studies’ estimates of the logarithm odds ratio (OR). Given heterogeneity of the studies, a random effect was applied. Risk of bias was analyzed using the Newcastle-Ottawa Scale. The study was conducted following the Preferred Reporting Items for Systematic Review and Meta-analysis guidelines. The protocol was registered in PROSPERO (No. CRD42024549924).

RESULTS

Eleven retrospective cohort studies involving 2953 patients were included in the meta-analysis. Among those, HE was investigated in eight studies, whereas PHE growth was evaluated in three. Blood sample was obtained on admission in ten studies, and at 24 hours in one study. There was no consensus on cut-off value among the studies. NLR was found to be significantly associated with higher odds of HE (OR = 1.09, 95%CI: 1.04-1.15, I² = 86%, P < 0.01), and PHE growth (OR = 1.28, 95%CI: 1.19-1.38, I² = 0%, P < 0.01). Qualitative analysis of each outcome revealed overall moderate risk of bias mainly due to lack of control for systemic confounders.

CONCLUSION

The available literature suggests that a possible association may exist between NLR on admission and HE, and PHE growth. Future studies controlled for systemic confounders should be designed to consolidate this finding. If confirmed, NLR could be added as a readily available and inexpensive biomarker to identify a subgroup of patients at higher risk of developing HE.

Key Words: Neutrophil-to-lymphocyte ratio; Hematoma expansion; Perihematomal growth; Intracerebral hemorrhage; Hemorrhagic stroke

Core Tip: In this work, we provide an updated and concise quantitative synthesis of the literature addressing the association between neutrophil-to-lymphocyte ratio and hematoma expansion (HE) in intracerebral hemorrhage. We found an independent association between neutrophil-to-lymphocyte ratio (NLR) on admission and HE, and perihematomal growth. Our findings support the idea that NLR could be added as a readily available and inexpensive biomarker to identify a subgroup of patients at higher risk of developing HE.

INTRODUCTION

Spontaneous intracerebral hemorrhage (ICH) is the most devastating type of stroke, and it is associated with high morbidity and mortality[1,2]. Initial hematoma expansion (HE) has been shown to be associated with poor outcome[3]. As such, early predictors of HE, including clinical and radiographic markers, have been extensively investigated[4,5]. Prediction scores for HE have been proposed and compared to each other[6] in an effort to best predict which patients would benefit from enrollment in trials targeted to reducing HE[7,8].

Detection of abnormal systemic inflammatory markers is a frequent encounter in the critically ill patient[9]. In acute ischemic stroke, elevated inflammatory markers have been demonstrated to have a negative prognostic significance given their role in developing secondary neurological injury[10,11]. Neutrophil-to-lymphocyte ratio (NLR) has been reported to be an independent prognostic factor in ICH, being associated with neurological deterioration, short-term mortality, and major disability at 90 days[12–16]. A meta-analysis from 2022 has failed to observe a significant association between NLR and HE[17]. However, the bulk of literature on the topic has significantly grown since then. Here we presented an updated systematic review and meta-analysis aimed at assessing the association between NLR and HE in ICH. The secondary objective was to assess the role of NLR in perihematomal (PHE) growth in ICH.

MATERIALS AND METHODS

Protocol

This systematic review was registered via PROSPERO (No. CRD42024549924) and conducted and presented in accordance with the Preferred Reporting Items for Systematic Review and Meta-analysis (PRISMA) statement[18]. Ethical approval was not required because data was synthetized from previously published studies.

Study design

This is a systematic review and meta-analysis aimed at assessing the association of NLR and HE in ICH. Additional outcome investigated was the association between NLR and PHE growth.

Search strategy

Three databases (PubMed, EMBASE, and Cochrane) were searched from inception to May 22, 2024, without linguistic or geographical limitation using the following keywords "neutrophil-to-lymphocyte ratio" or "neutrophil lymphocyte ratio" or "neutrophil-lymphocyte ratio" in conjunction with "hematoma expansion" or "hematoma growth" or "intracerebral hemorrhage" or "ICH". The search strategy is available in supplementary material. Reference lists were searched manually for additional sources.

Eligibility criteria

All articles identified from the literature search were screened by two authors (Loggini A and Hornik J), independently, as per the following inclusion criteria: (1) Adult population with spontaneous ICH; (2) Laboratory value indicating NLR; and (3) Clinical outcome focused on HE and PHE edema growth. Randomized clinical trials and controlled observational cohort studies were included in the review. Detailed information on participants, interventions, comparisons, outcomes and types of studies are provided in Supplementary Table 1. The relevance of the studies was assessed using a hierarchical approach based on title, abstract, and full manuscript.

Assessment of study quality

Risk of bias was computed using the Newcastle-Ottawa Scale (NOS) by two authors (Loggini A and Hornik J), independently[19]. Disagreement between quality of the study was resolved by consensus.

Data extraction

For each included study, the following variables were extracted independently by two authors (Loggini A and Hornik J): (1) Lead author; (2) Publication date; (3) Study design; (4) Number of participants; (5) NLR and its cut-off value; (6) Time of sample collection; (7) Markers of stroke severity; (8) Clinical outcomes of HE and PHE edema; and (9) Corresponding odds ratio (OR) and 95%CI values from multivariable analyses. If any of the above variables were not available in the full text publication, further information was sought by correspondence with the lead author of the study.

Statistical analysis

The inverse variance method was applied to estimate an overall, unconfounded, effect estimate of NLR for each specific outcome by combining weighted averages of the individual studies’ estimates of the logarithm OR. A Z test was carried out to assess the significance of the OR. The I² was calculated by χ² test to assess variability due to heterogeneity rather than chance. A substantial heterogeneity was assumed with I² > 50%. 95%CI for HE and PHE edema were calculated with the Wilson method and placed in forest plots. A P value < 0.05 was considered statistically significant. Statistical analysis was performed using Review Manager 5.3[20,21].

RESULTS

Study selection

A total of 241 studies were identified through database searching, including PubMed (n = 103), EMBASE (n = 134), and Cochrane (n = 4). After removing 97 duplicates, the remaining studies (n = 144) were screened for eligibility. A total of 120 records were excluded by title and by abstract. As a result, 24 studies were assessed with full-text review. Among those, 13 articles were excluded due to not investigating the desired population (n = 4), protocol only (n = 3), reviews (n = 3), lack of control group (n = 2), and pilot study (n = 1). Finally, this meta-analysis included 11 articles that met the inclusion criteria[22–32]. The detailed PRISMA flow diagram is shown in Figure 1.

Figure 1 Preferred Reporting Items for Systematic Review and Meta-analysis flow diagram.

Study characteristics

The characteristics of the included studies are summarized in Table 1[22-32]. No randomized controlled trials were identified; all studies were single center retrospective cohort analyses. HE was evaluated in eight studies[23,25–29,31,32]. HE was defined as an increase of 33% from baseline hematoma volume or absolute growth of more than 6 mL in five study[25,26,28,31,32], increase of 33% or absolute increase of 12.5 mL from baseline hematoma volume in two study[22,23], whereas it was not defined in one study[27]. PHE growth was evaluated in three studies[24,29,30]. All of the included studies defined PHE growth as the absolute difference in the PHE volume between follow-up and first scan. Blood sample was obtained on admission in ten studies[23–32], and at 24 hours in one study[22]. There was no consensus on cut-off value of NLR among the studies.

Table 1 Characteristics of included studies.

Ref.	Design	N	Sample time	Cut-off value	Outcome	Outcome definition	Factors controlled for
Mao et al[30], 2024	Retrospective	117	Admission	-	PHE	Difference between follow up scan and admission PHE	ICH volume, GCS score
Pisco et al[29], 2023	Retrospective	215	Admission	-	PHE	Difference between follow up scan and admission PHE	NIHSS, glucose, neurosurgery procedure, ICH volume, ICH location, time to first HCT
Kim et al[31], 2023	Retrospective	520	Admission	5.63	HE	ICH growth > 6 mL or > 33% of initial volume	Age, hypertension, diabetes, prior anticoagulation, GCS, NIHSS, WBC, glucose, SBP
Chu et al[27], 2023	Retrospective	301	Admission	5	HE	-	Sex, diabetes, GCS, time to first HCT, ICH volume, WBC
Alimohammadi et al[28], 2022	Retrospective	221	Admission	-	HE	ICH growth > 6 mL or > 33% of initial volume	ICH volume, SBP, GCS, WBC
Zhang et al[32], 2022	Retrospective	506	Admission	-	HE	ICH growth > 6 mL or > 33% of initial volume	GCS, blend sign, swirl sign, hypodensities on CT
Fonseca et al[25], 2019	Retrospective	135	Admission	7.8	HE	ICH growth > 6 mL or > 33% of initial volume	Systemic infection, time to fist CT, ICH volume, use of anticoagulant
Pektezel et al[22], 2019	Retrospective	383	At 24 hours	-	HE	ICH growth > 12.5 mL or > 33% of initial volume	ICH location, use of anticoagulant
Wang et al[26], 2019	Retrospective	123	Admission	6.49	HE	ICH growth > 6 mL or > 33% of initial volume	Age, sex, GCS, WBC, ICH volume, midline shift
Zhang et al[23], 2018	Retrospective	279	Admission	14.53	HE	ICH growth > 12.5 mL or > 33% of initial volume	Time to fist CT, hydrocephalus, IVH, GCS, ICH volume, island sign
Gusdon et al[24], 2017	Retrospective	153	Admission	-	PHE	Difference between follow up scan and admission PHE	ICH volume, HE, IVH, external ventricular drain placement, GCS, time to first CT

CT: Computed tomography; HCT: Head computed tomography; HE: Hematoma expansion; ICH: Intracerebral hemorrhage; IVH: Intraventricular hemorrhage; GCS: Glasgow Coma Scale; NIHSS: National Institute of Health Stroke Scale; PHE: Perihematomal edema; SBP: Systolic blood pressure; WBC: white blood cells.

Study quality

The summary of the risk of bias for each of the outcomes is listed in Supplementary Tables 1, 2 and 3[22-32]. Their overall quality was acceptable, ranging on the NOS from 8/9 to 9/9 (9/9 being the highest study quality). Specifically, ten studies had moderate risk of bias for lack of control for systemic factors (systemic infections).

The presence of publication bias was not able to be assessed through visual assessment of asymmetry of the funnel plot as there were less than 10 studies included for each outcome.

Study outcomes

Results of study outcomes are presented in Figure 2[22-32]. NLR was associated with significant higher odds of HE (OR = 1.09, 95%CI: 1.04-1.15, I² = 86%, P < 0.01). A sensitivity analysis on the HE outcome was conducted by excluding studies one by one. After removing the study of Kim et al[30], the heterogeneity was decreased (I² = 83%) without affecting the result (OR = 1.07, 95%CI: 1.02-1.12, P < 0.01). A subgroup analysis for the HE outcome was performed, including only the studies that uniformly considered HE as growth > 6 mL or > 33% of initial ICH volume. The result of the meta-analysis held true (OR = 1.13, 95%CI: 1.05-1.22, I² = 86%, P < 0.01). NLR was also found to be associated with PHE growth (OR = 1.28, 95%CI: 1.19-1.38, I² = 0%, P < 0.01).

Figure 2 Meta-analysis of the association. A: Between NLR and hematoma expansion; B: Between NLR and perihematomal edema.

DISCUSSION

Summary of evidence

This systematic review and meta-analysis included eleven retrospective cohort studies involving 2953 patients with ICH. The increasing volume of published data in recent years testifies to the rapidly developing interest in this topic. Our study showed that in patients with ICH, the NLR is associated with a higher trend of HE and PHE growth.

NLR is a marker of systemic inflammation[33]. In ICH, the local inflammatory response that occurs immediately after hematoma development involves the recruitment of peripheral leukocytes, with neutrophils being the first centrally recruited[34,35]. Additionally, the systemic stress response after ICH promotes the demargination of neutrophils and suppression of lymphocytes, further imbalancing the peripheral NLR[36,37]. Elevation of neutrophils in the brain parenchyma promotes neurotoxicity and alters the permeability of the blood-brain barrier[30,38–40]. The former appears to be the leading mechanism underlying PHE growth, while the latter drives HE.

The association between NLR and HE appears robust. Among the eight studies quantitatively summarized, most were homogeneous in the definition of HE, and a subgroup analysis restricted to a uniformed definition of HE confirmed the results. An association between NLR and PHE growth was also found. However, this association needs to be considered with more caution due to the low number of studies on the topic.

HE is an independent predictor of poor prognosis in ICH patients[3]. Several radiographic and laboratory values have been investigated as predictors of HE[4,5]. Among these, inflammatory markers are inexpensive laboratory tests frequently evaluated on admission. This makes NLR an exciting, inexpensive, and reliable potential early biomarker of HE in ICH. Developing sensitive early biomarkers of HE holds potential for transforming prognostic modeling, precision therapeutic targeting, and tailoring patient-specific treatment plans in ICH.

The association between high NLR and poor outcome has been well-documented in the literature[17]. Several studies have proven a correlation between NLR and mortality, and disability at 30 days and 90 days after ICH[12–16]. Furthermore, NLR has also been found to be associated with poor outcomes in other acute neurological conditions, such as severe traumatic brain injury[41,42]. The results of this meta-analysis, combined with the proven association with outcomes, suggest that NLR may not simply be a bystander marker of poor outcome but could play an active role in the early development of secondary brain injuries, such as HE and PHE growth, that affect long-term outcomes.

As a nonspecific peripheral inflammatory marker, NLR has been evaluated in several diseases, such as cancer, heart failure, and infectious disease[43–45]. In the hyperacute phase, systemic infections can cause a sudden rise in the innate response alongside suppression of the adaptive response, leading to a rise in NLR[21]. While all included studies were appropriately controlled for known markers of ICH severity, only one study considered systemic infections as a covariable of interest[45].

Finally, it is important to note that the studies included in this meta-analysis were designed as retrospective cohort studies. The association between NLR and HE and PHE appears robust. However, no assumptions can be made about the effect that modulating the NLR could have on ICH outcomes.

A future large multi-center prospective study should be designed to enroll subjects from different geographical areas worldwide, sampling blood of participants at presentation, controlling for markers of ICH severity and systemic confounders, and following the temporal behavior of NLR. This would help to consolidate the available literature and clarify the association and relationship between NLR, HE, PHE, and outcomes in ICH.

Limitations

The main limitation of this meta-analysis is that all included studies were retrospective cohort analyses, which are inherently subject to selection bias and confounding factors. Additionally, the definition of HE varied among the studies, and the cut-off values for NLR were not consistent. Finally, the relatively small number of studies included in this meta-analysis limited the ability to perform further subgroup analyses and assess consistency across different populations, aside from a subgroup analysis on a homogeneous definition of HE.

CONCLUSION

The available literature suggests that a possible association may exist between NLR on admission and HE, and PHE growth. Future studies controlled for systemic confounders should be designed to consolidate this finding. If confirmed, NLR could be added as a readily available and inexpensive biomarker to identify a subgroup of patients at higher risk of developing HE.