Abstract

Objective

Familial Mediterranean fever (FMF) is characterized by episodic inflammation, but inflammatory activity may continue between clinically apparent attacks and contribute to complications such as amyloidosis. We aimed to identify clinical and laboratory variables associated with serum amyloid A (SAA) concentrations during attack-free periods and to examine the relationship between SAA and the systemic immune-inflammation index (SII).

Methods

We retrospectively analyzed data from 710 patients with FMF followed at a tertiary rheumatology center between January 2016 and December 2020. Demographic, clinical, genetic, treatment, and laboratory data were evaluated. SII was calculated as platelet count × neutrophil/lymphocyte ratio. Correlation and multivariable regression analyses were used to examine factors associated with SAA concentrations.

Results

SAA levels showed significant positive correlations with disease duration, leukocyte and neutrophil counts, proteinuria, C-reactive protein (CRP), erythrocyte sedimentation rate, fibrinogen levels, and SII (all p<0.001), while a negative correlation was observed with lymphocyte count. Higher SAA levels were detected in patients with amyloidosis and/or kidney disease and in those carrying the M694V mutation. In multivariable analysis, longer disease duration, higher CRP, and higher fibrinogen levels remained independently associated with SAA, whereas SII did not.

Conclusion

SAA remains a useful indicator of subclinical inflammatory activity in patients with FMF during attack-free periods. Although SII was positively correlated with SAA, it was not independently associated with SAA after multivariable adjustment. As a readily available and inexpensive index derived from routine blood tests, SII may provide complementary information; however, its clinical utility requires prospective validation.

Introduction

Familial Mediterranean fever (FMF) is a hereditary autoinflammatory disorder characterized by recurrent, self-limited inflammatory attacks involving serosal, articular, and cutaneous sites. During these episodes, systemic inflammation is commonly accompanied by neutrophilic leukocytosis and elevations in acute-phase reactants (APRs), including C-reactive protein (CRP), fibrinogen, serum amyloid A (SAA), and erythrocyte sedimentation rate (ESR). Although these markers are not specific to FMF, SAA may be particularly sensitive to inflammatory activity and has been proposed as a useful parameter for guiding colchicine treatment and reducing the risk of AA amyloidosis.^[1] Importantly, inflammatory activity may not resolve completely between clinically apparent attacks. Persistent SAA elevation during these periods is clinically relevant because SAA is the hepatic precursor of AA amyloid fibrils, and sustained SAA elevation has been associated with the development of AA amyloidosis.^[1]

Several scoring systems are available for assessing disease activity and cumulative damage in FMF^[2-5], while efforts continue to identify laboratory markers that may complement clinical assessment.^[6] Systemic immune-inflammation index (SII), which integrates peripheral neutrophil, lymphocyte, and platelet counts, represents one such readily obtainable inflammatory index. Its prognostic value has been investigated predominantly in malignancies^[7-10], but associations with clinical outcomes have also been reported in several non-malignant disorders, including cardiovascular and cerebrovascular diseases, osteoporosis, infective endocarditis, and interstitial lung disease.^[11-15] Evidence regarding the clinical relevance of SII in rheumatological diseases, however, remains relatively limited.^[16-18]

Accordingly, this study sought to characterize FMF patients during attack-free periods and to determine the clinical and laboratory factors associated with SAA concentrations, with particular emphasis on the relationship between SAA and SII.

Materials and Methods

Study Design and Population

We reviewed the records of 710 adult patients with FMF who were followed at our center from January 2016 through December 2020. Patients were eligible if they were older than 18 years and met the Tel Hashomer diagnostic criteria. Those with an active infection at the time of evaluation or a history of solid or hematological malignancy were excluded.

Data Collection

Data extracted from the medical records included age, sex, age at diagnosis, genetic mutations, FMF-related treatments, comorbidities, and amyloidosis. Laboratory data included CRP, ESR, SAA, fibrinogen, complete blood count parameters, and proteinuria was assessed using the spot urine protein-to-creatinine ratio and/or 24-hour urinary protein excretion. Genetic data were available for 478 patients (67.32%); mutation results for the remaining patients were unavailable because of incomplete retrospective records.

Laboratory Measurements

Blood samples were obtained during routine outpatient visits while patients were clinically attack-free. The attack-free state was defined by the absence of symptoms consistent with an FMF attack and of clinical evidence of acute infection at the time of evaluation. SAA was measured by nephelometry in the hospital laboratory using 1 mL of serum (minimum volume, 0.5 mL), with a reference range of 0-10 mg/L.

Calculation of SII

SII was derived from complete blood count parameters and calculated as platelet count × neutrophil/lymphocyte ratio.

Ethical Approval

Ethical approval for the study was obtained from the Ankara University Human Research Ethics Committee (approval no: İ5-364-21, date: 25.06.2021). All study procedures were performed in compliance with the principles of the 1964 Declaration of Helsinki and subsequent amendments.

Statistical Analysis

Statistical analyses were conducted using IBM SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA). Distributional characteristics of continuous variables were examined visually using histograms and Q-Q plots. Normally distributed continuous data are reported as mean ± standard deviation, whereas variables with non-normal distributions are summarized as median and interquartile range (Q1-Q3). Categorical data are presented as numbers and percentages.

Associations between SAA concentrations and continuous variables were examined using Spearman’s rank correlation analysis, while between-group differences in SAA were assessed with the Mann-Whitney U test. Stepwise multiple linear regression was used to examine variables independently associated with SAA. Because of its skewed distribution, SAA was natural-log transformed before regression analysis. Variables significant in univariate analyses were considered for inclusion in the multivariable model. A two-sided p-value <0.05 was considered statistically significant.

Results

The study included 710 patients with FMF (461 women and 249 men), with a mean age of 41.87±12.70 years and a mean age at diagnosis of 25.73±13.46 years. Amyloidosis and/or kidney disease was present in 29 patients (4.08%), while 48 (6.76%) had a concomitant rheumatological disease. Genetic data were available for 478 patients (67.32%), and M694V was the most frequently identified mutation. Treatment data were available for 685 patients (96.48%). Median SAA was 3.8 (interquartile range, 1.5-10.8) (Table 1).

SAA concentrations correlated positively with disease duration (r=0.138; p<0.001), leukocyte count (r=0.200; p<0.001), neutrophil count (r=0.273; p<0.001), SII (r=0.288; p<0.001), proteinuria (r=0.219; p<0.001), CRP (r=0.918; p<0.001), ESR (r=0.411; p<0.001), and fibrinogen (r=0.526; p<0.001). Conversely, an inverse correlation was observed between SAA and lymphocyte count (r=-0.122; p=0.001). SAA was not significantly correlated with patient age, age at diagnosis, or platelet count (Table 2).

Receiver operating characteristic (ROC) analysis evaluated the ability of SII to discriminate patients with elevated SAA (≥5 mg/L), which was used as an indicator of subclinical inflammation. The area under the curve was 0.672 [95% confidence interval (CI): 0.632-0.713; p<0.001], indicating modest discrimination. The Youden index identified an SII threshold of 575, with 55.1% sensitivity, 70.8% specificity, and a positive likelihood ratio of 1.89 (Figure 1).

Patients with amyloidosis and/or kidney disease had significantly higher SAA concentrations (p=0.007). Similarly, SAA was higher among patients carrying the M694V mutation (p=0.001). In contrast, the presence of E148Q and other exon 2 mutations was associated with lower SAA concentrations compared with their respective counterparts (p=0.001 and p<0.001, respectively).

In the multiple linear regression model, longer disease duration (β=0.008; 95% CI: 0.000-0.016; p=0.045), higher CRP (β=0.152; 95% CI: 0.137-0.167; p<0.001), and higher fibrinogen levels (β=0.257; 95% CI: 0.164-0.350; p<0.001) remained independently associated with increased SAA concentrations (Table 3). No independent association with SAA was identified for sex (p=0.188), amyloidosis and/or kidney disease (p=0.952), M694V (p=0.141), E148Q (p=0.263), other exon 2 mutations (p=0.123), anakinra treatment (p=0.094), leukocyte count (p=0.056), neutrophil count (p=0.094), lymphocyte count (p=0.181), SII (p=0.102), proteinuria (p=0.639), or ESR (p=0.180).

Discussion

Subclinical inflammatory activity may persist in patients with FMF even during clinically attack-free periods and may contribute to long-term complications such as amyloidosis. Reliable detection of this ongoing inflammatory burden is therefore important for long-term disease management. In the present study, we evaluated factors associated with SAA during attack-free periods, with particular emphasis on its relationship with SII. SAA was associated with disease duration, conventional inflammatory markers, leukocyte parameters, and proteinuria. Higher SAA levels were observed in patients with amyloidosis and in M694V carriers, whereas lower levels were found in patients with E148Q and other exon 2 mutations. Although SII correlated positively with SAA, this association did not remain significant in the multivariable analysis.

The biological characteristics of SAA support its value as a sensitive marker of inflammatory activity. Circulating SAA levels are normally very low in healthy individuals but may increase by up to 1000-fold within 24-48 hours after the onset of an acute-phase response.^[19] This rapid increase is largely driven by inflammatory stimuli that promote de novo hepatic synthesis^{[20, 21]}, with concentrations declining promptly as the inflammatory response resolves. Beyond reflecting acute-phase activity, SAA may also participate in the inflammatory process by promoting complement activation and pro-inflammatory cytokine responses.^[22-24] It can enhance the expression of interleukin-6 (IL-6) and pro-IL-1β, while some SAA1-mediated effects involve signaling through TLR2 and TLR4.^{[19, 25]} SAA may also contribute to autoinflammatory pathways through direct activation of the NLRP3 inflammasome and subsequent caspase-1 activation.^[23]

Clinical evidence further supports the value of SAA as a marker of inflammatory activity. A meta-analysis showed that SAA concentrations were higher in patients with rheumatological diseases than in healthy controls and were also higher during active disease than during remission.^[26] In FMF, SAA increases together with other APRs during attacks. Its potential value in detecting residual inflammation has also been reported in other inflammatory conditions; for example, SAA has been proposed as an additional monitoring marker in patients with clinically quiescent inflammatory bowel disease, particularly when CRP concentrations are low.^[27] In FMF, periodic assessment of SAA may be particularly relevant because it has been reported to be more sensitive than CRP and ESR for detecting residual inflammation during attack-free periods.^[28]

Our findings are consistent with the potential role of SAA in identifying ongoing inflammation during clinically quiescent periods. SAA correlated positively with disease duration, leukocyte and neutrophil counts, proteinuria, CRP, ESR, and fibrinogen, whereas an inverse correlation was observed with lymphocyte count. Mutation status was also associated with SAA levels: patients carrying M694V had higher levels than non-carriers, while E148Q and other exon 2 mutations were associated with lower levels. These findings suggest that variation in SAA during attack-free periods may reflect not only the current inflammatory burden but also differences associated with the underlying mutation profile.

Laboratory-derived inflammatory indices have gained increasing attention as practical tools for assessing systemic inflammatory activity.^[29] Although much of the available evidence concerns autoimmune diseases, SII has also been reported to be elevated in adult-onset Still’s disease, an autoinflammatory disorder.^[16] As SII is derived solely from routine complete blood count parameters, it offers a simple and inexpensive measure that can be readily obtained in clinical practice. In our cohort, SII correlated positively with SAA during attack-free periods. This association is biologically plausible given the contribution of its cellular components to inflammatory responses: SAA correlated positively with neutrophil counts and negatively with lymphocyte counts, while neutrophils are key early participants in the inflammatory response.

Beyond their established roles in thrombosis and hemostasis, platelets also participate in inflammatory and immune responses.^[30] These functions are mediated partly through interactions with innate immune cells, including neutrophils, monocytes, and macrophages. Platelet-expressed TLRs contribute to thromboinflammatory signaling^{[30, 31]}, while TLR2 and TLR4 have also been implicated in the pathogenesis of FMF.^[32-34] These mechanisms provide a biological basis for the contribution of platelet count to SII.

Platelet-neutrophil interactions represent another important link between hemostatic and inflammatory pathways and contribute to the regulation of immune responses.^[35] Their reciprocal activation under both physiological and pathological conditions may sustain thromboinflammatory responses, creating a self-amplifying cycle.^[36] Platelet activation is also associated with the release of pro-inflammatory cytokines, including tumor necrosis factor-α, IL-1β, IL-6, and IL-8.^[37] Among these mediators, IL-1β is particularly relevant to FMF because of its central role in autoinflammatory pathways. Taken together, these mechanisms provide a biological rationale for investigating SII as a marker reflecting inflammatory activity in FMF.

Despite its positive correlation with SAA, SII did not remain independently associated with SAA in the multivariable analysis. The very strong correlation between SAA and CRP may have introduced multicollinearity into the regression model, reflecting the substantial overlap between these inflammatory markers. This may have attenuated the independent contribution of SII and should be considered when interpreting the regression findings. Further studies are needed to clarify the relationship of SAA and SII with clinical characteristics, including mutation status, during attack-free periods.

Study Limitations

Several limitations should be considered when interpreting our findings. The retrospective, single-center design may restrict their generalizability, while the cross-sectional assessment of SAA and SII during attack-free periods precluded evaluation of their longitudinal relationship with clinical outcomes. In addition, the absence of a control group limits conclusions regarding the specificity of these markers for FMF-related inflammation, as both may be influenced by other inflammatory conditions. Residual confounding related to the retrospective design also cannot be excluded.

The SII cut-off identified by ROC analysis should be interpreted cautiously, as the cross-sectional design does not allow a definitive threshold for clinical use to be established. Incomplete genetic and treatment data may also have introduced selection bias and affected the subgroup analyses and regression modeling. In addition, the attack-free state was determined clinically without a standardized interval from the preceding attack or a requirement for complete normalization of APRs, which may have influenced the assessment of subclinical inflammation. Prospective studies with well-defined clinical outcomes are therefore needed to validate clinically meaningful cut-off values for SII and SAA. Despite these limitations, the large sample size and comprehensive assessment of clinical, laboratory, and genetic characteristics represent important strengths of the study.

From a clinical perspective, our findings support the potential utility of SAA for identifying subclinical inflammatory activity during attack-free periods. SII, as a readily available and inexpensive index derived from routine laboratory parameters, may provide complementary information. However, the SII cut-off identified in this study should be interpreted cautiously and should not be directly applied to clinical decision-making without prospective validation. Prospective studies are needed to establish clinically relevant cut-off values and determine the optimal frequency of SAA and SII assessment in routine FMF follow-up.

Conclusion

In conclusion, SAA remains a useful marker of subclinical inflammatory activity during attack-free periods in FMF. Although SII was positively correlated with SAA, it was not independently associated with SAA in the multivariable analysis, suggesting that its potential role may be complementary rather than independent. As a readily available and inexpensive index derived from routine laboratory parameters, SII may provide additional information alongside established inflammatory markers; however, its clinical utility requires further prospective validation.