Introduction
Multiple sclerosis (MS) is an autoimmune inflammatory demyelinating disease of the central nervous system and is the second leading cause of disability among adults after trauma [1] . Because appropriate treatment can prevent or mitigate disability, disease-modifying therapies play a critical role in the management of MS. Fingolimod is the first highly effective oral therapy approved for MS [2] . Its primary mechanism of action involves modulation of sphingosine-1-phosphate receptor 1, resulting in the sequestration of lymphocytes within secondary lymphoid tissues, thereby preventing their release into the peripheral circulation and reducing their passage through the blood–brain barrier.
Discontinuation of fingolimod, whether controlled or uncontrolled, can lead to severe neurological symptoms within 2–4 months and the development of new or enlarging focal lesions (T1 Gadolonium contrast-enhanced lesions or T2 lesion) exceeding pretreatment radiological activity on Magnetic Resonance Imaging. This phenomenon has been described as rebound activity (RA) [3] , [4] . Several risk factors have been associated with RA, including a high relapse frequency before treatment, younger age, lower disability, a rapid increase in lymphocyte counts, and a prolonged interval before initiation of another MS therapy. The development of RA has also been associated with substantial accumulation of disability [5] . However, some studies suggested that, when RA is closely monitored and appropriately managed, it may not affect long-term disability outcomes over periods of up to 5 years [6] . Therefore, patients discontinuing therapies such as fingolimod, which regulate lymphocyte trafficking into the central nervous system, require close monitoring. Although several risk factors for RA have been identified, the ability to predict this complication using readily available haematological parameters would be extremely valuable.
Studies have demonstrated the importance of blood inflammatory parameters in vascular diseases, cancers, infections, and demyelinating diseases, particularly for assessing prognosis and treatment responses [7] . Similar approaches have also been investigated in RA. The haemoglobin, albumin, lymphocyte, and platelet (HALP) score is a composite biomarker that incorporates haematological and nutritional parameters to assess systemic inflammatory status. It has been investigated as a prognostic marker in many conditions, particularly cancers, as well as vascular diseases and hyperlipidaemia [8] .
In this study, we investigated the predictive potential of HALP score in RA development following fingolimod discontinuation for various reasons in patients with MS. We aimed to elucidate whether the inflammatory and nutritional status before fingolimod initiation was associated with subsequent disease activity following treatment discontinuation. We also included parameters such as monocyte-to-lymphocyte ratio (MLR) and neutrophil-to-lymphocyte ratio (NLR), which can be used as biomarkers indicating systemic inflammation in several diseases, including MS.
Method
Study design
This retrospective study investigated the association between pre-treatment haematological indices and clinical and radiological disease activity following fingolimod treatment and discontinuation in patients with MS.
Patients were identified from the MS outpatient clinic of the Department of Neurology, Faculty of Medicine, Sivas Cumhuriyet University. Patients who were followed between January 2016 and January 2025, met the McDonald 2010 diagnostic criteria for MS, had received fingolimod treatment, and subsequently discontinued for various reasons were screened. A total of 68 patients met the initial eligibility criteria. A control group without statistically significant differences in age or sex compared with the patient group was also included.
Participants
The same general exclusion criteria were applied to both patients and healthy controls. Individuals younger than 18 years; those with active infections, diabetes, hypertension, acute or chronic liver or kidney failure, acute or chronic inflammatory diseases (e.g., inflammatory bowel disease or Sjögren's syndrome), other autoimmune diseases, malignancy, or a history of surgery, acute myocardial infarction, or trauma within the previous 3 months were excluded.
The control group consisted of a population who presented to our outpatient clinic for a report stating that they were healthy, had no systemic diseases, were not taking any medications, and had normal blood values. The absence of chronic pulmonary disease was also required. Participants were required to be nonsmokers or to have stopped smoking for at least 1 year.
For patients, haematological and biochemical measurements obtained during the period before initiation of fingolimod treatment were used. Specifically, samples obtained within 0–168 h before fingolimod initiation were included in the analysis. HALP score, MLR, and NLR were evaluated. Patients were also excluded if haemoglobin or albumin levels changed by at least one unit (1 g/dL for haemoglobin or 1 g/L for albumin) relative to baseline during routine complete blood count and biochemical follow-up performed for at least 3–6 months. Seven patients were excluded on this basis. The observed fluctuations in these patients were considered potentially attributable to factors such as laboratory measurement variability, hydration status, or malnutrition, and these patients were excluded to minimize potential confounding.
The following additional criteria were required for the patient group: relapsing-remitting MS with an Expanded Disability Status Scale (EDSS) score of 0–1.5; newly diagnosed MS; or no prior disease-modifying therapy, or if receiving treatment, no treatment other than interferon, glatiramer acetate, teriflunomide, or dimethyl fumarate; absence of lymphopenia; and no prior immunosuppressive therapy.
Blood samples taken before initiating fingolimod treatment had to be taken at least 8 weeks after methylprednisolone treatment, if given. Fingolimod may impact lymphocyte subsets for several years, patients were required to have received fingolimod for at least 18 months before radiological and clinical activity assessment. Participants had a body mass index between 18.5 and 24.9. Pregnancy and breastfeeding were exclusion criteria. Patients were also required to have had no substantial changes in dietary or body mass index (BMI) changes during the follow-up period and to have received no iron (Fe) or vitamin B12 supplementation during the 3 months preceding fingolimod initiation.
Patients were followed for at least 24 months after initiation of fingolimod. Following discontinuation, patients receive did not methylprednisolone or another intervention specifically to prevent disease activity. Patients who initiated another disease-modifying therapy were included only if the treatment had not reached expected onset of clinical effect during the period used to assess disease activity.
Reasons for discontinuation of fingolimod treatment were: 17 patients due to clinical activity, 14 patients due to radiological activity, 19 patients due to both clinical and radiological activity, 4 patients due to side effects, 4 patients due to pregnancy, and 3 patients voluntarily. The EDSS was used to determine neurological disability. To assess the presence and number of active lesions, Magnetic Resonance Imaging (MRI) scans obtained using the same MRI system were jointly reviewed by an experienced radiologist and a neurologist specialising in neuroradiology.
Variables
The criteria used in this study were established a priori by the investigators to distinguish severe activity from mild-to-moderate clinical and radiological activity. Rebound activity (RA) was defined as meeting at least two of the following criteria: (1) at least two T1 and/or one tumefactive gadolinium-enhancing lesion exceeding pre-fingolimod radiological activity; (2) neurological deterioration of 1.5 points or more in the EDSS score; (3) lack of response to steroids and receiving acute treatment progressing to plasmapheresis.
Attack activity (AA) was defined as meeting at least two of the following criteria: (1) radiological activity comparable to that observed during the pre-fingolimod period; (2) an EDSS increase of <1.5 points; and (3) regression of neurological symptoms following steroid treatment.
Patients meeting the criteria for either RA or AA were classified as having disease activity (DA) group. Patients who did not develop radiological or clinical activity following fingolimod discontinuation were classified as the non-disease activity (NA) group.
The discriminative capacity of haematological indices, primarily the HALP score as a marker reflecting systemic inflammation and nutritional status, as well as MLR and NLR values, at diagnostic levels and across clinical subgroups in patients with MS was evaluated.
Data measurement
Blood samples for biochemical analyses were collected into dry tubes, whereas samples for haematological analyses were collected into EDTA-containing tubes. Complete blood counts were performed using Diagon kits on a Mindray BC-6800 device. Biochemical analyses were performed using a fully automated nephelometric method (Beckman Coulter AU5800; Beckman Coulter Inc., Hialeah, Florida) with reagents from the same manufacturer.
The HALP score developed by Chen et al. was calculated using the following formula: HALP Score = [haemoglobin (g/L) × albumin (g/L) × lymphocyte count (/L)] / platelet count (/L). The monocyte-to-lymphocyte ratio (MLR) was calculated by dividing the absolute monocyte count by lymphocyte count, and whereas the neutrophil-to-lymphocyte ratio (NLR) was calculated by dividing the absolute neutrophil count by lymphocyte count.
Statistical analysis
Statistical analyses were performed using SPSS version 22.0 (IBM Corp., Armonk, NY, USA). The distribution of continuous variables was assessed using the Kolmogorov–Smirnov test. Continuous variables with normal distributions were expressed as mean ± standard deviation (SD), whereas non-normally distributed variables were expressed as median and interquartile range (IQR). Categorical variables were presented as frequency and percentage.
The chi-square test was used, where appropriate, to compare proportions, and Student's t -test was used to compare continuous variables. The Kruskal–Wallis H test was used to compare values among three or more independent groups. Diagnostic screening tests (sensitivity, specificity, AUC) and receiver operating characteristic (ROC) curve analysis were used to determine the threshold value. P -value <0.05 was considered statistically significant.
Ethical considerations
The study was conducted in accordance with the Declaration of Helsinki. The rights and confidentiality of all participants were protected. Ethical approval was obtained from the Sivas Cumhuriyet University Ethics Committee (approval no. 2025-06/55).
Results
Among the 61 participants included in the patient group, 43 (70.5%) were female and 18 (29.5%) were male. In the healthy control group, 30 (52.6%) were female and 27 (47.4%) were male. No statistically significant difference in sex distribution ( p = 0.09) between the two groups. In the patient group, 10 (16.4%) patients developed AA, 7 (11.5%) patients developed RA, and 17 (27.9%) patients formed the DA group, defined as a combination of AA and RA groups. The mean time to develop AA was 64.80 ± 17.87 days, whereas the mean time to RA was 68.86 ± 10.12 days.
Compared with healthy controls, the patient group had significantly lower HALP scores ( p < 0.001) and significantly higher MLR ( p < 0.001). Although NLR values were also higher in patients with MS than in controls, the difference did not reach statistical significance ( p = 0.057) ( Fig. 1 ). Fig. 1 Comparison of haematological indices and clinical parameters between healthy controls and MS patients. Data are presented as boxplots with individual data points, where the central line represents the median, the box bounds the interquartile range (IQR), and whiskers extend to 1.5 times the IQR. Statistical comparisons were performed using the Mann-Whitney U test. Significant p -values are displayed above each boxplot panel. Abbreviations: HALP, Haemoglobin, Albumin, Lymphocyte, and Platelet score; MLR, Monocyte-to-Lymphocyte Ratio; NLR, Neutrophil-to-Lymphocyte Ratio. Fig. 1
Patients were divided into two groups according to the presence or absence of AA. Ten patients developed AA, whereas 51 did not. HALP scores and MLR values were significantly higher in patients with AA than in those without AA ( p < 0.01 and p = 0.01, respectively). No significant difference was observed in NLR values ( p = 0.28) ( Fig. 2 ). Fig. 2 Comparison of inflammatory scores according to attack activity status in patients with multiple sclerosis. Abbreviations: AA, Acute Attack; Non-AA, Non-Acute Attack; HALP, Haemoglobin, Albumin, Lymphocyte, and Platelet score; MLR, Monocyte-to-Lymphocyte Ratio; NLR, Neutrophil-to-Lymphocyte Ratio. Statistical evaluation was performed using the Mann-Whitney U test. Fig. 2
Patients were similarly categorized according to the presence or absence of RA. Seven patients developed RA, whereas 54 did not. Patients with RA had significantly higher HALP scores and MLR values ( p = 0.001 and p = 0.005, respectively). NLR values did not differ significantly between the two groups ( p = 0.367) ( Fig. 3 ). Patients were required to have used fingolimod for at least 18 months, and although RA is expected to appear 4 months after treatment, they were monitored for up to 6 months. Fig. 3 Comparison of inflammatory scores according to rebound activity status following Fingolimod discontinuation. Abbreviations: RA, Rebound Activity; Non-RA, Non-Rebound Activity; HALP, Haemoglobin, Albumin, Lymphocyte, and Platelet score; MLR, Monocyte-to-Lymphocyte Ratio; NLR, Neutrophil-to-Lymphocyte Ratio. Statistical evaluation was performed using the Mann-Whitney U test. Fig. 3
Patients with DA had significantly higher HALP scores and MLR than inactive patients (NA group) (both p < 0.001). In contrast, NLR values did not differ significantly according to disease activity status ( p > 0.05) ( Fig. 4 ). Fig. 4 Comparison of inflammatory scores according to overall disease activity status. Abbreviations: DA, Disease Activity; NA, No Activity; HALP, Haemoglobin, Albumin, Lymphocyte, and Platelet score; MLR, Monocyte-to-Lymphocyte Ratio; NLR, Neutrophil-to-Lymphocyte Ratio. Statistical evaluation was performed using the Mann-Whitney U test. Fig. 4
The association between the reason for fingolimod discontinuation and the development of RA is presented in Table 1 . Patients were categorized into four groups according to the reason for discontinuation: no response, side effect, left by himself, and pregnancy. The reasons for discontinuation were significantly associated with the development of RA ( p = 0.04). Among patients who developed RA, 42.9% had discontinued fingolimod because of pregnancy, whereas 14.3% discontinued because of side effects and left by himself. In contrast, 87.0% of patients who did not develop RA had discontinued fingolimod because of treatment inefficacy. These findings suggest that discontinuation due to treatment inefficacy occurs in a more controlled manner, and that the rebound risk may be relatively lower for this group ( Table 1 ). Table 1 Demographic features, clinical parameters, and reasons for treatment discontinuation in patient with MS and control groups. Control n (%) Patient n (%) Total p Female 30 (52.6) 43 (70.5) 73 (61.9) 0.046 ⁎ ,a Male 27 (47.4) 18 (29.5) 45 (38.1) Total 57 (100.0) 61 (100.0) 118 (100.0) Age 37 (31–42) 37 (33–45.5) 37 (32–43) 0.255 EDDS – 1 (0.5–2.5) – – Non-Rebound Activity n (%) Rebound Activity n (%) Total p No Response 45 (83.3) 4 (57.1) 49 (80.3) 0.126 b Side Effect 4 (7,4) 0 (0.0) 4 (6.6) Patient Preference 2 (3.7) 2 (28.6) 4 (6.6) Pregnancy 3 (6.6) 1 (14.3) 4 (6.6) Total 54 (100.0) 7 (100.0) 61 (100.0) Data are presented as $n$ ($\%$) for categorical variables, median (interquartile range, Q1–Q3) for non-normally distributed continuous variables. Categorical comparisons were performed using Pearson's chi-square test ( χ 2 ) ( a ) or Likelihood Ratio test ( b ) where expected cell counts were below standard thresholds. Continuous parameters between groups were compared using the Mann-Whitney U test. ⁎ p < 0.05 indicates statistical significance.
Table 1
Demographic features, clinical parameters, and reasons for treatment discontinuation in patient with MS and control groups.
Data are presented as $n$ ($\%$) for categorical variables, median (interquartile range, Q1–Q3) for non-normally distributed continuous variables. Categorical comparisons were performed using Pearson's chi-square test ( χ 2 ) ( a ) or Likelihood Ratio test ( b ) where expected cell counts were below standard thresholds. Continuous parameters between groups were compared using the Mann-Whitney U test.
⁎ p < 0.05 indicates statistical significance.
| Control n (%) | Patient n (%) | Total | p | |
| Female | 30 (52.6) | 43 (70.5) | 73 (61.9) | 0.046 ⁎ ,a |
| Male | 27 (47.4) | 18 (29.5) | 45 (38.1) | |
| Total | 57 (100.0) | 61 (100.0) | 118 (100.0) | |
| Age | 37 (31–42) | 37 (33–45.5) | 37 (32–43) | 0.255 |
| EDDS | – | 1 (0.5–2.5) | – | – |
| Non-Rebound Activity n (%) | Rebound Activity n (%) | Total | p | |
| No Response | 45 (83.3) | 4 (57.1) | 49 (80.3) | 0.126 b |
| Side Effect | 4 (7,4) | 0 (0.0) | 4 (6.6) | |
| Patient Preference | 2 (3.7) | 2 (28.6) | 4 (6.6) | |
| Pregnancy | 3 (6.6) | 1 (14.3) | 4 (6.6) | |
| Total | 54 (100.0) | 7 (100.0) | 61 (100.0) |
The discriminative ability of HALP, MLR, and NLR was evaluated using ROC curve analysis for MS status, AA, RA, and DA. Due to the limited number of outcome events, these analyses were considered exploratory and hypothesis-generating. The area under the ROC curve (AUC), sensitivity, specificity, and corresponding 95% confidence intervals (CIs) were calculated ( Table 2 ). Table 2 ROC analysis of HALP, MLR, PLR and NLR Score results in MS. Cut point Sensitivity Specificity PPV NPV Youden AUC 95% CI (Lower-Upper) MS disease HALP 4,26 78,95 60,66 65,22 75,51 ,396 ,708* ,613-,802 MLR ,21 81,97 59,65 68,49 75,56 ,416 ,750* ,661-,838 PLR 138,05 63,93 70,18 69,64 64,52 ,341 ,667* ,567-,766 NLR – – – – – – ,602 ,498-,706 Attack Prediction HALP 3,98 90 70,45 40,91 96,88 ,605 ,845* ,732-,956 MLR 0,25 100 52,27 32,26 100 ,523 ,750* ,619-,881 PLR – – – – – – ,677 ,493-,862 NLR – – – – – – ,611 ,369-,854 Rebound Prediction HALP 4,62 85,71 77,27 37,5 97,14 ,630 ,857* ,725-,989 MLR 0,31 85,71 72,73 33,33 96,97 ,584 ,825* ,679-,971 PLR 103,98 90,91 57,14 93,02 50 ,481 ,766* ,583-,950 NLR – – – – – – ,610 ,366-,855 Activity Prediction HALP 3,98 88,24 70,45 53,57 93,94 ,586 ,850* ,753-,948 MLR ,29 82,35 68,18 50,0 90,91 ,505 ,781* ,666-,895 PLR 138,05 75,0 64,71 84,62 50,0 ,397 ,714* ,570-,858 NLR – – – – – – ,611 ,431-,791
Table 2
ROC analysis of HALP, MLR, PLR and NLR Score results in MS.
| Cut point | Sensitivity | Specificity | PPV | NPV | Youden | AUC | 95% CI (Lower-Upper) | ||
| MS disease | HALP | 4,26 | 78,95 | 60,66 | 65,22 | 75,51 | ,396 | ,708* | ,613-,802 |
| MLR | ,21 | 81,97 | 59,65 | 68,49 | 75,56 | ,416 | ,750* | ,661-,838 | |
| PLR | 138,05 | 63,93 | 70,18 | 69,64 | 64,52 | ,341 | ,667* | ,567-,766 | |
| NLR | – | – | – | – | – | – | ,602 | ,498-,706 | |
| Attack Prediction | HALP | 3,98 | 90 | 70,45 | 40,91 | 96,88 | ,605 | ,845* | ,732-,956 |
| MLR | 0,25 | 100 | 52,27 | 32,26 | 100 | ,523 | ,750* | ,619-,881 | |
| PLR | – | – | – | – | – | – | ,677 | ,493-,862 | |
| NLR | – | – | – | – | – | – | ,611 | ,369-,854 | |
| Rebound Prediction | HALP | 4,62 | 85,71 | 77,27 | 37,5 | 97,14 | ,630 | ,857* | ,725-,989 |
| MLR | 0,31 | 85,71 | 72,73 | 33,33 | 96,97 | ,584 | ,825* | ,679-,971 | |
| PLR | 103,98 | 90,91 | 57,14 | 93,02 | 50 | ,481 | ,766* | ,583-,950 | |
| NLR | – | – | – | – | – | – | ,610 | ,366-,855 | |
| Activity Prediction | HALP | 3,98 | 88,24 | 70,45 | 53,57 | 93,94 | ,586 | ,850* | ,753-,948 |
| MLR | ,29 | 82,35 | 68,18 | 50,0 | 90,91 | ,505 | ,781* | ,666-,895 | |
| PLR | 138,05 | 75,0 | 64,71 | 84,62 | 50,0 | ,397 | ,714* | ,570-,858 | |
| NLR | – | – | – | – | – | – | ,611 | ,431-,791 |
For discrimination between patients with MS and healthy controls, the optimal HALP cut-off value was 4.26, corresponding to a sensitivity of 78.95% (moderate–high) and specificity of 60.66% (moderate–high). The AUC was 0.708 (95% CI, 0.613–0.802), with a Youden index of 0.396. For MLR, the optimal cut-off value was 0.21, with a sensitivity of 81.97% (high) and specificity of 59.65% (moderate). The AUC was 0.750 (95% CI, 0.661–0.838), with Youden index of 0.416. The AUC for NLR was 0.602 (95% CI, 0.498–0.706), indicating limited discriminative ability and no statistically significant discrimination between patients with MS and healthy controls ( Fig. 1 ). Overall, these findings suggest that low HALP scores were associated with MS status in this cohort, predicting with moderate-high sensitivity and specificity. MLR levels may also be effective in identifying MS patients, although they may lead to false positivity in healthy individuals ( Fig. 5 ). Fig. 5 Receiver Operating Characteristic (ROC) curve analysis evaluating the diagnostic performance of HALP, MLR, and NLR scores across different clinical contexts: (A) Multiple Sclerosis prediction vs. healthy controls, (B) Acute Attack prediction, (C) Rebound Activity prediction, and (D) Disease Activity prediction. Diagonal dashed lines indicate chance-level performance (AUC = 0.50). Abbreviations: AUC, Area Under the Curve; HALP, Haemoglobin, Albumin, Lymphocyte, and Platelet score; MLR, Monocyte-to-Lymphocyte Ratio; NLR, Neutrophil-to-Lymphocyte Ratio. Fig. 5
For prediction of AA among patients with MS, the HALP score showed the highest AUC among biomarkers. The optimal cut-off value was 3.98, with a sensitivity of 90% (high) and specificity of 70.45% (high). The AUC was 0.845 (95% CI, 0.732–0.956), with a Youden index of 0.605. For MLR, the optimal cut-off value was 0.25, corresponding to a sensitivity of 100% (high) and specificity of 52.27% (moderate). The AUC was 0.750 (95% CI, 0.619–0.881), with a Youden index of 0.523. Although NLR showed meaningful contributions, it was not a significant marker in this cohort ( p > 0.05) ( Fig. 5 ).
For RA, both HALP scores and MLR values demonstrated high discriminative performance. The optimal cut-off value for HALP scores was 4.62, with a sensitivity of 85.71% (high) and specificity of 77.27% (high). The AUC was 0.857 (95% CI, 0.725–0.989), with a Youden index of 0.630. The optimal cut-off value for MLR values was 0.31, with a sensitivity of 85.71% (high) and specificity of 72.73% (high). The AUC was 0.825 (95% CI, 0.679–0.971), with a Youden index of 0.584. NLR did not demonstrate statistically significant discriminative performance for RA ( p > 0.05) ( Fig. 5 ).
According to ROC analyses, the HALP score emerged as the strongest biomarker for predicting DA. The optimal cut-off value was 3.98, with a sensitivity of 88.24% (high) and specificity of 70.45% (high). The AUC was 0.850 (95% CI, 0.753–0.948), with a Youden index of 0.586. MLR also showed statistically significant discriminative ability in predicting DA. The optimal cut-off value was 0.29, with a sensitivity of 82.35% (high) and specificity of 68.18% (moderate–high). The AUC was 0.781 (95% CI, 0.666–0.895), with a Youden index of 0.505. AUC value = 0.714, Youden index = 0.397. NLR was not found to be a statistically significant marker in this sample ( Fig. 5 ).
Discussion
In this retrospective study, we evaluated the association of the HALP score, MLR, and NLR with clinical and radiological activity following fingolimod discontinuation in patients with MS. Several findings emerged. Patients with MS had lower HALP score, indicating systemic inflammatory state and the utility as a potential biomarker. In addition, subgroup analyses indicate that both HALP scores and MLR were significantly higher in those who developed AA, RA, and DA following discontinuation of fingolimod. We also hypothesize that our predictive power for RA could be enhanced with specific HALP cut-off value of 4.62 or higher.
MS is characterized by the interaction of inflammatory and neurodegenerative processes, resulting in demyelination and axonal damage [1] . The inflammatory phase clinically presents with relapses and focal radiological activity on MRI. With disease progression, neurodegenerative mechanisms contribute to a decrease in neuronal reserve, brain atrophy, and the development of disability [7] .
The HALP score was originally introduced by Chen et al. in 2015 as a composite index incorporating haemoglobin, albumin, lymphocyte count, and platelet count to assess systemic inflammation and nutritional status, initially in patients with gastric cancer [8] . In general, lower HALP values have been associated with poorer clinical outcomes prognosis, indicating strong relationship between nutrition and inflammation [9] . Nutritional biomarkers such as haemoglobin, albumin, and BMI are known to reflect inflammatory processes and can be used to predict disease activity and prognosis, especially cancers. On the contrary, malnutrition is associated with lower cytokine response and changes in the inflammatory profile. In a study of MS and neuromyelitis optica spectrum disorder (NMOSD), positive correlations were observed between disease activity, disability, and lesion activity and inflammatory markers such as platelet-to-lymphocyte ratio, MLR, and the Systemic Immune-Inflammation Index, whereas negative correlations were detected with haemoglobin and BMI [10] .
The HALP score has also been evaluated in patients with Primary Sjögren's Syndrome (PSS), where lower HALP score was observed and associated with disease severity [11] . Interestingly, we observed higher HALP score and identified it as a biomarker with high sensitivity and specificity in predicting DA, RA, and AA. One possible explanation is that higher HALP scores may indicate an intense inflammatory process and a strong response following drug discontinuation. Moreover, the inflammatory period during which disease-modifying therapies are effective (“window of opportunity”), may be more dominant in patients exhibiting RA, AA, or DA [7] . These findings demonstrate a close relationship between inflammation, nutrition, and other systems in MS.
The potential contribution of individual components of the HALP score also warrants consideration. Albumin is a major plasma protein with antioxidant properties and the ability to cross the blood–brain barrier (BBB) when disrupted. Albumin may serve as a substrate for reactive oxygen species (ROS) and reactive nitrogen species (RNS), thereby eliciting protective role during the pathogenesis of EAE and MS [12] . Microglia also secrete albumin during ischemia and are associated with lower inflammatory states in Alzheimer's disease [13] . During acute and chronic inflammation, due to factors such as regulation of hepatic synthesis and altered vascular permeability, albumin may act as a negative acute-phase reactant beyond its nutritional significance [14] . Previous studies have reported associations between albumin concentrations and disease-related outcomes in neurological disorders, including MS, NMOSD, and myasthenia gravis [15] , [16] . Higher albumin is linked to good prognosis in cerebrovascular diseases, although other studies indicate no association [17] , [18] . In our cohort, lower HALP scores may be attributed to the reduction in albumin level, whereas higher HALP score after Fingolimod discontinuation may indicate a strong immune response during the inflammatory period before the neurodegenerative phase.
Haemoglobin may also contribute to the HALP score through mechanisms extending beyond oxygen transport. Haemoglobin can also cross the BBB and contribute to the neurodegenerative process via oxidative stress, mitochondrial dysfunction, and its role in the production of immune system cells, particularly T and B cells [19] . Iron metabolism is closely associated with myelin synthesis and oligodendrocyte function, and alterations in iron homeostasis have been implicated in neuroinflammatory and neurodegenerative processes. Astrocytes facilitate iron transport through astrocyte endfoot and contribute to oxygenation and remyelination of oligodendrocytes. Previous studies have reported the associations between iron deficiency anaemia and relapse risk as well as disease severity in MS [20] , [21] . Thus, low haemoglobin can be considered a marker in chronic diseases due to its complex functions [22] .
Platelets are increasingly recognized as participants in neuroinflammatory processes. They interact with components of innate immunity such as leukocytes, neutrophils, and monocytes, and may influence the inflammatory processes. Platelets induce immunoglobulin formation in B cells and activate cytolytic and helper T cells. Because platelet count serves as the denominator of the HALP scores, lower platelet count is expected to increase HALP score values. Thus, the higher HALP scores observed among patients with RA, AA, and DA could theoretically reflect changes in platelet counts, although it may also be related to an increase in haemoglobin, albumin, or lymphocyte counts.
The MLR findings provide additional evidence for an association between peripheral immune-cell composition and disease activity. Monocytes and their CNS-derived counterparts, including macrophages and microglia, are important components of inflammatory responses in MS. Both microglia and macrophages have been observed around demyelinating plaque lesions [23] . Microglia are also present in the periplaque region and even in distant normal-appearing white matter. Disruption of the BBB facilitates the recruitment of antigen-presenting cells, such as B cells and myeloid cells, to plaque regions, and results in the differentiation of memory T cells into pro-inflammatory T helper lymphocytes [24] . Consistent with previous studies that showed associations between MLR and active lesions [8] , we found significantly higher MLR values in MS patients. MLR was also significantly higher among patients with AA, RA, and DA than among those without disease activity. Our finding supports the potential role of peripheral monocyte–lymphocyte balance as a marker of inflammatory activity with high specificity and sensitivity.
In contrast to HALP and MLR, NLR did not differ significantly between patients with RA, AA, or DA. The significant increase in MLR values in these groups may be associated with an increased monocyte count resulting from peripheral immune system activation. This finding suggest that DA, whether manifesting as RA or AA after abrupt drug discontinuation, may be associated with immune mechanisms. Furthermore, based on our results, MLR may be useful for predicting AA, RA, and their combination, DA, with high sensitivity and specificity. These findings suggest that MLR may serve as a potential marker for these three clinical conditions, for which specific indicators are currently lacking. MLR is also readily accessible and incurs no additional cost. However, because of its components, MLR may be insufficient to reflect the relationship between nutritional status and inflammation.
In MS pathophysiology, pro-inflammatory Th1 helper cells are known to contribute to the development of new lesions, whereas Th2 helper cells suppress local inflammation [25] . The elevation of NLR, calculated by dividing the neutrophil count by lymphocyte count and considered a useful indicator of systemic inflammation and an important biomarker in autoimmune diseases, has been demonstrated in many diseases, including MS [26] . Neutrophils may contribute to MS pathogenesis through cytokine release, disruption of BBB permeability, and contribution to parenchymal damage [27] . Neagle et al. demonstrated that neutrophils may increase their proportions by escaping apoptosis through changes in surface expression molecules [28] . Although we observed higher NLR in patients with MS, the absence of a significant associations may reflect the relatively small simple size and the potential effects of disease-modifying treatment on peripheral leukocyte populations.
Fingolimod was approved by the United States Food and Drug Administration (FDA) in late 2010 as the first oral MS drug. Fingolimod acts primarily by modulating sphingosine-1 phosphate (S1P) receptor and preventing lymphocyte egress from lymphoid tissues, thereby reducing the number of circulating lymphocytes available for migration into the CNS [6] . Abrupt discontinuation requires proactive measures (e.g., switching to an alternative treatment, increasing monitoring intensity) to prevent the emergence of RA risk. In our study, seven of 61 patients developed RA. This proportion is consistent within the range reported in previous studies and highlights the importance of close clinical supervision upon treatment discontinuation. Our study also revealed that discontinuation due to pregnancy planning or pregnancy showed the highest risk of rebound development. This finding underscores the need for careful treatment planning for women of reproductive age who receive fingolimod and require discontinuation.
The present study has some limitations. First, it was a retrospective, single-centre study with relatively small sample size, limiting generalizability of the findings. Second, the number of RA events was particularly small, substantially limiting statistical power. Although multivariable logistic regression analysis adjusting for age, sex, disease duration, and lymphocyte count could theoretically be used to assess whether HALP independently predicts disease activity, the low number of RA events would result in a high risk of model overfitting. Accordingly, the associations observed in this study should be regarded as exploratory and hypothesis-generating. Third, only NLR and MLR were evaluated as inflammatory indicators. Although whether patients received methylprednisolone treatment to prevent disease activity, or the efficacy durations of the most recently initiated molecules during transitions to other disease-modifying therapies, were taken into account, it is not definitively known whether some patients individually used fingolimod every other day or less frequently before discontinuing. It is also appropriate to mention here whether the definition of the diet has not been evaluated in detail, taking into account individual, cultural, religious, and economic factors.
Our study represents the first examining the correlation between HALP Score and nutrition and inflammatory responses in patients with MS. MS group exhibited lower HALP scores and higher MLR values. The absence of any clinical and radiological activity after fingolimod discontinuation was associated with lower HALP score, suggesting that the neurodegenerative process is more prominent. While high HALP scores may indicate negative consequences and exacerbating disability, it is also a sign of active neuroinflammatory conditions that neuroinflammation (current disease-modifying therapies used in MS treatment are particularly effective during this period) is still present in these patients and that the neurodegenerative process is not yet fully established. Considering the relationship between nutrition and inflammation, promoting a healthy lifestyle may delay neurodegenerative process.
Further prospective, multicentre studies with larger cohorts are required to determine whether HALP and MLR can reliably identify patients at increased risk of disease activity following fingolimod discontinuation.
CRediT authorship contribution statement
Seyda Figul Gokce: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition. Asli Bolayir: Writing – review & editing, Formal analysis, Data curation, Conceptualization. Ecem Demir: Writing – original draft, Visualization, Supervision, Software, Methodology, Investigation, Data curation. Burhanettin Cigdem: Writing – original draft, Visualization, Software, Funding acquisition, Conceptualization.
Ethical approval
The study was performed in accordance with ethical standards of Helsinki declaration and its later amendments and was approved by the institutional ethical board.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
1 Ebers G.C. Environmental factors and multiple sclerosis Lancet Neurol. 7 2008 268 277 10.1016/S1474-4422(08)70042-5 18275928
2 Thomas K. Proschmann U. Ziemssen T. Fingolimod hydrochloride for the treatment of relapsing remitting multiple sclerosis Expert. Opin. Pharmacother. 18 2017 1649 1660 10.1080/14656566.2017.1373093 28844164
3 Berger B. Baumgartner A. Rauer S. Mader I. Luetzen N. Farenkopf U. Stish O. Severe disease reactivation in four patients with relapsing-remitting multiple sclerosis after fingolimod cessation J. Neuroimmunol. 15 2015 118 122 10.1016/j.jneuroim.2015.03.022 25903738
4 Forci B. Mariottini A. Mechi C. Massacesi L. Repice A. Disease reactivation following fingolimod withdrawal in multiple sclerosis: two case reports Mult. Scler. Relat. Disord. 15 2017 24 26 10.1016/j.msard.2017.05.003 28641768
5 Fragoso Y.D. Adoni T. Gomes S. Goncalves M.V.M. Parolin L.F. Rosa G. Ruocco H.H. Severe exacerbation of multiple sclerosis following withdrawal of fingolimod Clin. Drug Investig. 39 2019 909 913 10.1007/s40261-019-00804-6 31152369
6 Gulec B. Everest E. Gorkey O.D. Koc M. Tutuncu M. Saip S. Siva A. Uygunoglu U. Comparison of multiple sclerosis patients with or without rebound activity after fingolimod cessation: five-year clinical outcomes Eur. J. Neurol. 30 2023 2745 2751 10.1111/ene.15913 37300847
7 Gokce S.F. Bolayır A. Cigdem B. Yildiz B. The role of systemic immune inflammatory index in showing active lesion in patients with multiple sclerosis: SII and other inflamatuar biomarker in radiological active multiple sclerosis patients BMC Neurol. 23 2023 64 10.1186/s12883-023-03101-0 36765289 PMC9912215
8 Chen X.L. Xue L. Wang W. Chen H.N. Zhang W.H. Liu K. Prognostic significance of the combination of preoperative hemoglobin, albumin, lymphocyte and platelet in patients with gastric carcinoma: a retrospective cohort study Oncotarget 6 2015 41370 41382 10.18632/oncotarget.5629 26497995 PMC4747412
9 Stumpf F. Keller B. Gressies C. Schuetz P. Inflammation and nutrition: friend or foe? Nutrients 25 2023 1159 10.3390/nu15051159 PMC10005147 36904164
10 Li X. Jiang W. Li G. Ding Y. Li H. Sun J. Inflammatory and nutritional markers as indicators for diagnosing and assessing disease activity in MS and NMOSD J. Inflamm. Res. 29 2024 10065 10078 10.2147/JIR.S489502 PMC11613729 39628706
11 Bircan A.Ö. Acemoğlu Ş.Ş.Z. Türk İ. Özer H.T.E. Immunonutritional biomarkers in primary Sjögren’s syndrome disease activity: CALLY index and HALP score Arch. Rheumatol. 1 2025 376 382 10.5152/ArchRheumatol.2025.25012 PMC12502840 40977111
12 Wang P. Xie K. Wang C. Bi J. Oxidative stress induced by lipid peroxidation is related with inflammation of demyelination and neurodegeneration in multiple sclerosis Eur. Neurol. 72 2014 249 254 10.1159/000363515 25277682
13 Eckart A. Struja T. Kutz A. Baumgartner A. Baumgartner T. Zurfluh S. Relationship of nutritional status, inflammation, and serum albumin levels during acute illness: a prospective study Am. J. Med. 133 2020 713 722 10.1016/j.amjmed.2019.10.031 31751531
14 Li Y. Chen L. Yang X. Cui H. Li Z. Chen W. Dynamic association of serum albumin changes with inflammation, nutritional status and clinical outcomes: a secondary analysis of a large prospective observational cohort study Eur. J. Med. Res. 28 2025 679 679 10.1186/s40001-025-02925-5 PMC12302892 40717075
15 Chen K. Li C. Zhao B. Shang H. Albumin and multiple sclerosis: a prospective study from UK Biobank Front. Immunol. 10 2024 1415160 10.3389/fimmu.2024.1415160 PMC11194376 38915402
16 Xiang W. Wu Y. Li H. Zhu D. Yao X. Ding J. Serum albumin level can predict immunotherapy response of neuromyelitis optica spectrum disorders in the acute phase J. Inflamm. Res. 12 2024 909 917 10.2147/JIR.S442532 PMC10873210 38370469
17 Chinnammanavar P.K.B. Somannavar V.G. Mohan P.B. Role of serum calcium, serum albumin, and serum uric acid as markers of initial neurological severity and short-term outcome indicators in acute ischemic stroke J. Assoc. Physicians India 2 2024 41 44 10.59556/japi.72.0496 38881109
18 Pantha B. Khadka M. Karki L. Koirala P. Correlation between serum albumin level and severity of acute ischemic stroke: a cross-sectional study J. Nepal Med. Assoc. 62 2024 669 674 10.31729/jnma.8792 PMC11619538 40655892
19 Kłodnicka K. Januszewski J. Forma A. Pająk W. Teresińska B. Baj J. Iron in multiple sclerosis – from pathophysiology to disease progression - a narrative literature review Acta Neurobiol. Exp. (Wars) 31 2025 75 93 10.55782/ane-2025-2680 40742010
20 Koudriavtseva T. Renna R. Plantone D. Mandoj C. Piattella M.C. Giannarelli D. Association between anemia and multiple sclerosis Eur. Neurol. 73 2015 233 237 10.1159/000381212 25823947
21 Goischke H.K. Comorbidities in multiple sclerosis-a plea for interdisciplinary collaboration to improve the quality of life of MS patients Degener. Neurol. Neuromuscul. Dis. 13 2019 39 53 10.2147/DNND.S204555 PMC6584285 31354380
22 Aringer M. Inflammatory markers in systemic lupus erythematosus J. Autoimmun. 110 2020 102374 10.1016/j.jaut.2019.102374 31812331
23 Prineas J.W. Kwon E.E. Cho E.S. Sharer L.R. Barnett M.H. Oleszak E.L. Hoffman B. Morgan B.P. Immunopathology of secondary-progressive multiple sclerosis Ann. Neurol. 50 2001 646 657 10.1002/ana.1255 11706971
24 Filippi M. Bar-Or A. Piehl F. Preziosa P. Solari A. Vukusic S. Rocca M.A. Multiple sclerosis Nat. Rev. Dis. Prim. 43 2018 10.1038/s41572-018-0041-4 30410033
25 Ziaber J. Pasnik J. Baj Z. Pokoca L. Chmielewski H. Tchórzewski H. The immunoregulatory abilities of polymorphonuclear neutrophils in the course of multiple sclerosis Mediat. Inflamm. 7 1998 335 338 10.1080/09629359890857 PMC1781867 9883968
26 Bolayir A. Cigdem B. Gokce S.F. Yilmaz D. The relationship between neutrophil/lymphocyte ratio and uric acid levels in multiple sclerosis patients Bratisl. Med. J. 122 2021 357 361 10.4149/BLL_2021_060 33848187
27 Pierson E.R. Wagner C.A. Goverman J.M. The contribution of neutrophils to CNS autoimmunity Clin. Immunol. 189 2018 23 28 10.1016/j.clim.2016.06.017 27377536 PMC5203971
28 Naegele M. Tillack K. Reinhardt S. Schippling S. Martin R. Sospedra M. Neutrophils in multiple sclerosis are characterized by a primed phenotype J. Neuroimmunol. 242 2012 60 71 10.1016/j.jneuroim.2011.11.009 22169406
Acknowledgments
We would like to thank Dr. Mehmet Arı from Sivas Medicana Hospital, Clinic of Radiology, for his contributions to the MRI evaluation.