Prognostic Role of Neutrophil to Lymphocyte Ratio in Nonalcoholic Fatty Liver Disease: A Systematic Review and Meta-Analysis

Introduction Nonalcoholic steatohepatitis (NASH) and liver fibrosis are the most common complications of nonalcoholic fatty liver disease (NAFLD). In this systematic review and meta-analysis, we aim to analyze the current literature to evaluate the association of neutrophil to lymphocyte ratio (NLR) with NASH and fibrosis in patients with NAFLD. Methods PubMed, Web of Science, and Scopus were used to conduct a systematic search for relevant publications published before May 24, 2022. The Newcastle–Ottawa scale was used for quality assessment. Results Thirteen studies were included in our study. The pooled results showed that NAFLD patients with significant NASH had elevated levels of NLR compared to those with nonsignificant or without NASH (SMD = 0.97, 95% CI = 0.59–1.39, p < 0.001). The pooled sensitivity and specificity of NLR were 78.16% (95% CI = 73.70%–82.04%), and 76.93% (95% CI = 70.22%–82.50%), respectively. In addition, NAFLD patients with significant liver fibrosis had elevated levels of NLR compared to those with nonsignificant or without fibrosis (SMD = 1.59, 95% CI = 0.76–2.43, p < 0.001). The pooled sensitivity and specificity of NLR were 82.62% (95% CI = 70.235%–90.55%) and 81.22% (95% CI = 75.62%–85.78%), respectively. Conclusion Our findings support NLR to be a promising biomarker that can be readily integrated into clinical settings to aid in the prediction and prevention of NASH and fibrosis among patients with NAFLD.


Introduction
Nonalcoholic fatty liver disease (NAFLD) is an increasingly prevalent (approximately 25% global prevalence) [1] clinical disease that is associated with obesity, type II diabetes, and other metabolic comorbidities that are increasingly common in modern society. [2] Nonalcoholic steatohepatitis (NASH) and liver fbrosis are the most common complications of NAFLD. In fact, NASH is a histological phenotype of NAFLD that represents a signifcant infammatory progression from simple steatosis that may subsequently progress to cirrhosis and hepatocellular carcinoma and is becoming an increasingly common indication for liver transplantation. [2] Te neutrophil to lymphocyte ratio (NLR) is an easily obtained serum measure that corresponds to systemic infammation and has been demonstrated to be a useful prognostic measure in a variety of pathologies, including stroke and colorectal cancer. [3,4] In several studies, NLR has been positively associated with NASH and fbrosis stage in patients with NAFLD [5][6][7][8][9][10][11][12][13][14][15][16][17]. Tus, NLR might serve as an easily obtainable predictive tool to guide clinical decision-making, intervene earlier and improve patient outcomes. Te goal of this systematic review and meta-analysis is to analyze existing retrospective and prospective studies to establish the potential utility of NLR in the prediction of NASH and fbrosis among patients with NAFLD. To the best of our knowledge, this is the frst meta-analysis in this context.

Study Design and Eligibility Criteria.
Tis study is conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2020 reporting guideline [18]. We searched databases of PubMed, Web of Science, and Scopus up to May 24, 2022. In our literature search, we used the following search strategy: ('Neutrophil to lymphocyte ratio' or NLR) and ('nonalcoholic fatty liver' or 'nonalcoholic fatty liver' or NAFLD) and (steatohepatitis or NASH or fbrosis).
Additionally, we reviewed the reference lists of included and relevant studies to identify further eligible studies. Our inclusion criteria were based on the following PICO terms: (e) Study Design: We expected papers to be case-control or cross-sectional. However, we did not limit our search to any particular research design.
Our exclusion criteria were as follows: (1) review articles, editorials/letters, case series, case reports, abstracts, and randomized controlled trials; (2) duplicate studies; (3) not peer-reviewed publications. Tere were not any limitations on language or date of publication.

Data Extraction and Quality Assessment.
Te frst author, year of publication, study design, study location, total sample size, number of cases and controls, mean and SD of NLR level, or any data for estimating the mean and SD (median and IQR or/and range), a cut-of value of NLR and its false/true positive and false/true negative from 2 × 2 table were all extracted. When the number of patients in false/true positive and false/true groups was not reported, we calculated it using sensitivity, and specifcity. Two authors conducted the quality assessment of included studies, based on the Newcastle-Ottawa scale (NOS), including three components selection of the cohort, comparability of cohorts based on the design or analysis, how the exposure was ascertained, and how the outcomes of interest were assessed [19]. Disagreements between the authors were fnally resolved via consensus. Tose studies with six or more points were deemed to have good quality.

Data Synthesis and Analysis.
We performed the metaanalysis by using Stata 11.2 software (Stata Corp, College Station, TX). We used standardized mean diference (SMD) with a 95% confdence interval (CI) to compare the NLR level between cases and controls. Te I 2 and Cochran's Q tests were adopted to determine the heterogeneity of the included studies. Signifcant heterogeneity between studies was conceived as I 2 > 50% and p value of the Q test < 0.05. Finally, because a signifcant level of heterogeneity was found, we applied the random-efects model to calculate pooled efects. In order to determine the diagnostic value of NLR for NASH or fbrosis, we used the "metandi" command which estimated pooled sensitivity, specifcity, diagnostic odds ratio (DOR), negative likelihood ratio, and positive likelihood ratio. In addition, a summary receiver operating characteristic (SROC) curve was drawn. In order to determine the publication bias, we used the funnel plot and Egger test.

Characteristics of Included Studies.
Te characteristics and methodological qualities of these studies were shown in Table 1. Te overall study quality ranged from 6 to 8 stars. Tirteen studies were included in our systematic review and meta-analysis. Tree studies were retrospective and others were prospective. All of them were written in English.

Discussion
In the presence of infammatory disease, circulating neutrophils often increase, and circulating lymphocytes often decrease. [20] As infammatory markers, neutrophils and lymphocytes may play multiple roles in the progression of chronic infammatory diseases, including NAFLD. Currently, the progression of NAFLD is described as a 'two-hit hypothesis.' [21] Initially, the 'frst hit' is defned as triglycerides accumulate in hepatocytes and insulin resistance develops as a hepatic manifestation. [21] Tis steatosis essentially desensitizes the liver to further infammation, allowing the progression to NASH. [22] In 2011, Kamari and colleagues illustrated signifcant associations between increased NLR and insulin resistance pathologies using mouse models. [23] Notably, an overproduction of IL-1α and IL-1β from resident liver cells is observed in the development of NASH, as a defciency in either interleukin was observed to sufciently protect against NASH development. [23] Additionally, the lipotoxicity leading to the development of NASH kills hepatocytes through apoptosis and necrosis. [24] In turn, necrosis activates macrophages, neutrophils, and proinfammatory pathways, resulting in an elevated NLR. [24,25].       Figure 3: Subgroup analysis of diferences in NLR level between NAFLD patients with signifcant NASH compared to those with nonsignifcant or without NASH, according to study design.

Canadian Journal of Gastroenterology and Hepatology
Te triglyceride accumulation and lipotoxicity are followed by the 'second hit' activation of systemic proinfammatory pathways. Specifcally, infammatory cytokines (notably, IL-1, IL-6, and TNF-α), chemokines, chemokine receptors, adhesion molecules, and signal molecules are increased in both NASH animal models and clinical studies of NASH patients. [25] Farrell and colleagues describe nuclear factor-kappa B (NF-κB) and c-Jun N-terminal kinase (JNK) as the key proinfammatory signal molecules increased in NASH, as these signaling pathways provide a link between hepatic infammation and insulin resistance. [25].
It has been recently described the roles of interferon regulatory factors (IRFs), a family of transcription factors that regulate IFN expression, that play important roles in both innate and adaptive immune responses and the potential of IRF regulators in NAFLD treatment, as recently well described in a comprehensive review conducted by Zhang et al. [26].     Additionally, the NLR may also increase through a hormonal mechanism. In 2012, Ahmed and colleagues demonstrated induction of hepatic 11β-HSD1 expression and activity following NAFLD progression of worsening hepatic infammation and injury. [27] 11β-HSD1 serves as a primary regulator that catalyzes the reduction reaction of inactive cortisone to active cortisol. [27] As patients progress from steatosis to NASH along the NAFLD spectrum, hepatic glucocorticoid levels are activated and relative hypercortisolemia develops. [27] In response to the relative hypercortisolemia, leukocytosis and lymphopenia are also observed; thus, resulting in a markedly elevated NLR. [17] Terefore, while the current literature is unsure of the primary etiology of NASH, it is clear infammation plays a central role.
In 2022, Lesmana and colleagues investigated the difference in NLR values among the varying degrees of steatosis and fbrosis against transient elastography (TE) with controlled attenuation parameter (CAP), a gold standard diagnostic tool in steatosis and fbrosis. Patients with mild steatosis had a mean NLR of 1.492 (p < 0.001), compared to patients with moderate-severe steatosis with a mean NLR of 2.198 (p < 0.001). [14] Patients with nonsignifcant fbrosis had a mean NLR of 1.744 (p < 0.001), compared to patients with signifcant fbrosis with a mean NLR of 2.617 (p < 0.001). [14] Teir data suggest NLR can accurately predict the condition of liver steatosis. [14] As discussed earlier, patients with insulin pathologies including obesity and type 2 diabetes mellitus are at risk of increased NLR alone, without steatosis. Te data from Lesmana and colleagues found these comorbidities were not confounding factors in comparing the NLR to CAP as diagnostic tools. [14].   In 2015, Yilmaz and colleagues compared NLR and C-reactive protein (CRP) as variables in understanding liver histopathology and fbrosis. Tey found NLR to increase signifcantly with steatosis and fbrosis, whereas CRP did not. [17] Terefore, they concluded NLR to be a noninvasive clinical diagnostic tool for NASH and fbrosis compared to CRP. [17] Yilmaz and colleagues also suggest plasma fasting glucose coupled with NLR to independently predict the severity of the NAFLD activity score, as diabetes can accelerate the pathology of NASH in experimental mouse models. [17,28].
Some studies have suggested hs-CRP levels to be signifcantly elevated in patients with NASH versus simple steatosis, as well as in patients with advanced fbrosis. [29,30] In contrast, Hui and colleagues suggest hs-CRP did not accurately predict the severity of NAFLD from a histological standpoint, as they found no correlation between hs-CRP levels and grades of steatosis, fbrosis, or necroinfammation. [31] Yilmaz and colleagues did fnd NLR to be associated with both fbrosis and necroinfammation. [17] Tese data further suggest NLR to be a more powerful predictor of NASH and fbrosis severity.
Our results indicate a signifcant diference in the predictive value of NLR in NAFLD in retrospective versus prospective studies. We hypothesize this is due to the diference in sample size between the two groups of studies, as we included two retrospective studies versus ten prospective studies. Evaluating more retrospective studies may help clarify the signifcance of the diference in the predictive value of NLR in the settings of NASH, fbrosis, and NAFLD.
An interesting application may be to evaluate the utility of the NLR applied to alcoholic and pregnancyrelated fatty liver disease. One study noted NLR to be signifcantly higher in both alcoholic liver cirrhosis (ALC) and NAFLD compared to controls. [32] Since infammatory cell infltration is the common feature of steatohepatitis in alcoholic liver disease and NAFLD, we expect NLR to also predict histological grade and fbrosis stage in alcoholic liver disease as it seems to do in NAFLD. [33] Furthermore, there do not appear to be any studies evaluating the role of NLR in acute fatty liver of pregnancy (AFLP). Histology of AFLP does not feature infammation [34,35], so we predict that NLR may not be able to serve as a marker in this pathology, in contrast to NAFLD and alcoholic liver disease.

Biomarker Usage and Pharmacologic Insights.
New biomarkers are important to guide potential treatments as recently reported. In this regard, in the last years. It has been proposed the potential efcacy of sodium-glucose cotransporter 2 (SGLT2) inhibitors on NAFLD and "metabolic associated fatty liver disease (MAFLD) as recently reported by Goya al. [36]. Discussing the potential treatments, also Silymarin can be considered benefcial in treating NAFLD and should be initiated as early as possible and continued as long as necessary as recently suggested by Hashem et al. [37]. Given the results of our study, medications aimed at reducing NLR levels may prove efcacious for treating and even preventing such complications.

Limitations.
Our study has a few limitations that are important to address. Te main limitation of this study is the small number of papers that were included in the metaanalysis of the association of NLR with liver fbrosis. As such, our results may be limited in power and additional studies would be warranted to further strengthen the results of our study. Furthermore, the studies included in our analysis exhibited high heterogeneity. Although this was accounted for with the random-efect model, such measures may not entirely eliminate the issue of heterogeneity. Nonetheless, our systematic search-in conjunction with a manual review of references from resulting articles-has ensured a thorough and reliable search of the literature and serves as a notable strength of this study.

Conclusion
In conclusion, the data regarding cirrhotic patients suggest that NLR may be useful as an independent prognostic marker of NASH and liver fbrosis among NAFLD patients. Further studies need to be conducted to determine precise cut-of guidelines in which to utilize NLR.

Data Availability
Te dataset supporting the conclusions of this article is included within the article.