Key words
albumin - C-reactive protein - Familial Mediterranean Fever - subclinical inflammation
Schlüsselwörter
Albumin - C-reaktives Protein - Familiäres Mittelmeerfieber - subklinische Entzündung
Introduction
Familial Mediterranean Fever (FMF) is an autoinflammatory condition inherited
autosomal recessively and described as recurrent, self-limiting attacks of fever and
synovitis, peritonitis and pleuritis [1]
[2]. FMF varies according to different ethnic
groups, being frequently seen in Jews, Turks, Arabs, and Armenians [3]. The gene encoding the
pyrin/marenostrin proteins located on chromosome 16p, regulates fibroblasts
and leukocytes inflammation and apoptosis cascade and encodes the Mediterranean
fever gene (MEFV) [4]. The rate of MEFV gene
mutation is reported as 1/5 in Turks, Jews and Arabs, and 1/7 in
Armenians [5]. Mutations in the MEFV gene
initiate an uninterrupted inflammatory cascade [6]. Inflammatory mediators, such as tumour necrosis factor alpha (TNF-
α), interleukin-6 (IL-6), IL-8, and IL-10 have been evaluated to be elevated
during attacks and attack-free intervals of FMF [7]
[8]. Acute phase reactants, such
as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), fibrinogen, and
serum amyloid A (SAA) also increase during attack periods but return normal during
attack-free intervals [9]. In some patients,
subclinical inflammation (SI) may progress during attack-free intervals in FMF and
triggers amyloidosis despite colchicine therapy [10]
[11]. Recently, inflammation
during the symptom-free periods of FMF has been reported to correlate with simple
inflammatory-based scores, including the mean platelet volume (MPV), red cell width
distribution (RDW), neutrophil/lymphocyte ratio (NLR), and
platelet/lymphocyte ratio (PLR) [12]
[13]
[14]
[15]
[16]. The CRP/serum
albumin ratio (CAR) is a new inflammation score used as a marker of chronic
inflammation and prognosis in solid tumours [17], critically ill patients [18],
rheumatoid arthritis [19], and ANCA-associated
vasculitis [20]. The current study aimed to
assess CAR levels as a marker of SI during the attack-free period in FMF patients.
We also compared the CAR levels with the NLR, PLR, RDW and MPV values between the
patients with and without SI.
Material and Methods
We retrospectively evaluated 100 patients with FMF that presented to the Paediatric
Nephrology and Rheumatology Department of Manisa Celal Bayar University School of
Medicine between June 2010 and June 2020. All patients were diagnosed with FMF
according to the Tel-Hashomer criteria [21]
and included in regular colchicine therapy. As the control group, we obtained the
medical records of 70 healthy age- and sex-matched children referred to our
outpatient clinic without any inflammatory manifestation. Patients with any
infection, acute or chronic inflammatory or other chronic disease, malnutrition, or
failure to thrive that might cause hypoalbuminemia and those without drug compliance
were excluded from the study. The study was approved by the local ethics committee.
Written informed consent was obtained from all participants.
The demographic, clinical and laboratory parameters including a complete blood cell
count (CBC) analysis, biochemical data, and genetic findings were retrospectively
obtained from the hospital records of the patients during the attack-free period
when the patients were symptom-free. SI was defined as the elevation of at least one
acute phase reactant during an attack-free period [22]. NLR, PLR, MPV, RDW and CAR were derived from the CBC analysis.
Genetic
A genetic analysis was routinely performed when FMF was diagnosed. DNA was
obtained using the QIAamp DNA blood isolation kit (Qiagen GmbH) following the
manufacturer’s instructions using 2 mL of peripheral blood that
had been gathered into ethylenediaminetetraacetic acid-anticoagulated tubes with
the regular venepuncture method. A Thermo Scientific NanoDrop spectro-photometer
was used to determine the DNA concentration. The profiles of the 22 most common
pathogenic MEFV gene variants (E148Q, P369S, H478Y, F479L, S675N, G678E, M680I
(G>C), M680I (G>A), M680L, T681I, I692del, M694V, M694I, M694L,
K695R, K695M, R717S, I720M, V722M, V726A, A744S, and R761H) were genotyped by
pyrosequencing (Qiagen, Germany).
Statistical Analysis
The Statistical Package for the Social Sciences (SPSS) for Windows, v. 22 (IBM
SPSS Inc. Chicago, USA) was used for statistical analyses. Quantitative
variables were expressed as mean±standard deviation (SD) or median as
appropriate while categorical variables were obtained as percentages.
Student’s t-test or the Mann-Whitney U test was conducted to evaluate
the differences between the groups in terms of continuous variables. The
χ2-test was used to evaluate the categorical variables. The Spearman
correlation was performed to evaluate the relationship between the variables. A
two-tailed p value of<0.05 was considered as statistically significant.
Pairwise comparison of area under ROC curves was conducted using
DeLong’s test by MedCalc statistical software ((version 11.3.8.0,
Mariakerke, Belgium).
Results
The demographic characteristics and laboratory data of the FMF and control groups
are
presented in [Table 1]. The mean age of the
FMF patients was 12.7±4 years and that of the control group was
12.09±3.41 years. The female/male ratio was 42/58 for the
patients with FMF and 37/33 for the control group. There was no significant
difference between the FMF patients and the control group in terms of age and gender
(p=0.31 and p=0.164, respectively). The mean follow-up duration was
5.29±2.9 years. The CRP levels were 0.8±1.4 mg/dl
for the FMF patients during the attack-free period and
0.28±0.17 mg/dl in the control group, indicating
significantly higher values in the patient group (p=0.002). The mean value
of albumin, a negative acute phase reactant, was significantly lower in FMF patients
(4.11±0.58 g/dL) compared to the control group
(4.34±0.26) (p=0.003). The CAR levels were 0.21±0.41 for the
FMF group and 0.06±0.05 for the control group, being significantly higher in
the former (p=0.003). The NLR levels were also significantly higher in the
patients with FMF (2.53±3.8) compared to the controls (1.5±1)
(p=0.038). The PLR levels were 134.6±91.3 in the FMF patients and
110.6±48.3 in the control group, indicating significantly higher values in
the FMF patients (p=0.046). No significant difference was found between the
attack-free FMF patients and the control group in terms of the lymphocytes,
platelets, MPV and RDW (p=0.06, p=0.62, p=0.97,
p=0.48, respectively).
Table 1 Demographic characteristics and laboratory findings of the
FMF and control groups.
|
Characteristics
|
FMF group (n=100)
|
Control group (n=70)
|
P-value
|
|
Age (years)
|
12.7±4
|
12.09±3.41
|
0.313
|
|
Sex (F/M)
|
42/58
|
37/33
|
0.164
|
|
Family history of FMF (%)
|
61 (%35.9)
|
10 (%14.2)
|
<0.001
|
|
CRP (mg/dL)
|
0.8±1.4
|
0.28±0.17
|
0.002
|
|
Albumin (g/dL)
|
4.11±0.58
|
4.34±0.26
|
0.003
|
|
Lymphocytes (10
3
/µL)
|
3.10±2.1
|
3.762±2.23
|
0.06
|
|
Platelets (10
3
/µL)
|
332±97
|
339±84
|
0.62
|
|
MPV (fL)
|
8.5±1.2
|
8.5±0.9
|
0.97
|
|
RDW (%)
|
14.57±2
|
15.6±14.4
|
0.48
|
|
CAR
|
0.21±0.41
|
0.06±0.05
|
0.003
|
|
NLR
|
2.53±38
|
1.5±1
|
0.038
|
|
PLR
|
134.6±91.3
|
110.6±48.3
|
0.046
|
F: female, M: male, FMF: familial Mediterranean fever, CRP: C-reactive
protein, MPV: mean platelet volume, RDW: red cell width distribution, CAR:
C-reactive protein/serum albumin ratio, NLR:
neutrophil/lymphocyte ratio, PLR: platelet/lymphocyte
ratio.
The FMF patients were categorised into two groups according to the association of
SI
defined as the elevation of at least one acute phase reactant during an attack-free
period. There were 29 patients with high CRP (>0.5 mg/dl),
25 with high ESR (>8 mm/hr), and 13 with high fibrinogen
(>393 mg/dl). Thirty-nine patients had SI. The demographic
characteristics, laboratory findings and PRAS disease severity scores of the
patients with and without SI are shown in [Table
2]. There was no significant difference between the FMF patients with and
without SI according to age, gender, consanguinity, symptoms (e. g., fever,
abdominal pain, chest pain, and arthritis), PRAS disease severity score,
lymphocytes, NLR, PLT, and PLR (p=0.52, p=0.57, p=0.11,
p=0.84, p= 0.22, p=0.75, p=0.51, p=0.28,
p=0.75, p=0.067, p=0.95, p=0.24, respectively). A
family history of FMF, ESR, fibrinogen, CRP, albumin, CAR, and neutrophil count were
significantly higher in FMF patients with SI than those without SI (p=0.004,
p<0.001, p=0.001, p<0.001, p<0.001, p<0.001,
p=0.015, respectively).
Table 2 Comparison of the participants with and without SI.
|
Characteristics
|
FMF with SI (n : 39)
|
FMF without SI (n : 61)
|
P-value
|
|
Age onset of disease (years)
|
5.94±3
|
6.31±3.2
|
0.57
|
|
Sex (F/M)
|
15/24
|
27/34
|
0.57
|
|
Consanguinity (%)
|
4/39(10.25%)
|
14/61(23%)
|
0.11
|
|
Fever (%)
|
38/39
|
59/61
|
0.84
|
|
Abdominal pain (%)
|
31/39
|
54/61
|
0.221
|
|
Chest pain (%)
|
4/39
|
6/61
|
0.75
|
|
Arthritis (%)
|
16/39
|
21/61
|
0.51
|
|
Family history of FMF (%)
|
17/39
|
44/61
|
0.004
|
|
PRAS
|
8.1±1.9
|
7.61±1.9
|
0.24
|
|
ESR (mm/hr)
|
15.23±10.47
|
5.46±1.65
|
<0.001
|
|
Fibrinogen (mg/dL)
|
299±95
|
245.6±68.8
|
0.001
|
|
CRP (mg/dL)
|
1.64±2.04
|
0.25±0.16
|
<0.001
|
|
Albumin (g/dL)
|
3.8±0.6
|
4.23±0.45
|
<0.001
|
|
CAR
|
0.45±0.59
|
0.06±0.056
|
<0.001
|
|
Neutrophil (×10
3
)
|
6.3±3.37
|
4.9±2.5
|
0.015
|
|
Lymphocytes (×10
3
)
|
3.2±3
|
3.06±1.5
|
0.75
|
|
NLR
|
3.4±5.6
|
1.97±1.7
|
0.067
|
|
Platelets (×10
3
)
|
333±93
|
332±100
|
0.95
|
|
PLR
|
147±130
|
126±56
|
0.245
|
SI: subclinical inflammation, F: female, M: male, FMF: familial Mediterranean
fever, PRAS: disease severity score, ESR: erythrocyte sedimentation rate,
CRP: C-reactive protein, RDW: red cell width distribution, CAR: C-reactive
protein/serum albumin ratio, NLR: neutrophil/lymphocyte
ratio, PLR: platelet/lymphocyte ratio.
The FMF patients were also categorised according to their MEFV gene mutations. The
distribution of the MEFV mutations of the patients was as follows: 37 cases
(37%) had two mutations in the MEFV gene [23 homozygous (M694V,
n=19) and 14 (14%) compound heterozygous], 48 patients (48%)
were heterozygous [18 had M694V, 13 had E148Q, 8 had V726A, 3 had M680I ,3 had
R761H, 2 had R202Q, 1 had A744S] and 12 patients (12%) had no mutation. The
M694V homozygous (n=19) and M694V heterozygous (n=33) cases and the
patients with no M694V mutation (n=48) were compared with regard to their
demographic characteristics and laboratory findings ([Table 3]). SI was significantly higher in
patients with M694V mutations (p=0.025). However, there was no significant
difference between these three groups in terms of age, gender, consanguinity, FMF
symptoms, family history of FMF, PRAS score, and laboratory parameters (ESR,
fibrinogen, CRP, albumin, CAR, and neutrophil, lymphocyte and platelet counts).
Table 3 Comparison of M694V homozygous, M694Vheterozygous and no
M694V mutation patients.
|
Characteristics
|
M694Vhomozygous (n : 20)
|
M694Vheterozygous (n : 32)
|
No M694V mutation (n : 48)
|
P-value
|
|
Age onset of disease (Years)
|
5.9±3.5
|
6.23±3
|
6.25±3.1
|
0.89
|
|
Sex (F/M)
|
12/8
|
11/21
|
19/29
|
0.174
|
|
Consanguinity (%)
|
5/20 (25%)
|
4/32 (12.5%)
|
9/48(18.75%)
|
0.51
|
|
Fever (%)
|
20/20(100%)
|
31/32(96.9%)
|
46/48(95.8%)
|
0.66
|
|
Abdominal pain (%)
|
16/20(80%)
|
29/32(90%)
|
40/48(83%)
|
0.53
|
|
Chest pain (%)
|
1/20(5.5%)
|
4/32(12.5%)
|
9/48(18.8%)
|
0.32
|
|
Arthritis (%)
|
7/20(33%)
|
11/32(34%)
|
19/48(39.6%)
|
0.88
|
|
Family history of FMF (%)
|
14/20(66%)
|
18/32(56%)
|
29/48(60%)
|
0.62
|
|
PRAS
|
8.1±1.92
|
7.47±1.9
|
7.6±2.14
|
0.48
|
|
SI (%)
|
12 (60%)
|
16 (50%)
|
11 (23%)
|
0.005
|
|
ESR (mm/hr)
|
9.65±6.6
|
11.78±10.1
|
7.44±6.8
|
0.063
|
|
Fibrinogen (mg/dl)
|
285±81.6
|
264±82
|
260±86
|
0.521
|
|
CRP (mg/dL)
|
1.5±2.7
|
0.78±0.81
|
0.51±0.8
|
0.035
|
|
Albumin (g/dL)
|
3.97±0.65
|
4±0.55
|
4.2±0.56
|
0.24
|
|
CAR
|
0.41±0.8
|
0.2±0.21
|
0.14±0.24
|
0.047
|
|
Neutrophil (×10
3
)
|
6.4±2.9
|
4.8±2.3
|
4.8±2.0
|
0.027
|
|
Lymphocytes (×10
3
)
|
2.63±1.2
|
3.6±3.4
|
2.94±1.2
|
0.199
|
|
NLR
|
2.9±1.7
|
1.9±2
|
1.9±1.24
|
0.063
|
|
Platelets (×10
3
)
|
326±72
|
318±103
|
344±102
|
0.493
|
|
PLR
|
168±167
|
116±61
|
133±55
|
0.135
|
F: Female, M: Male, FMF: Familial Mediterranean Fever, PRAS: disease severity
score, SI: subclinical inflammation, ESR: erythrocyte sedimentation rate,
CRP: C-reactive protein, CAR: C-reactive protein/serum albumin
ratio, NLR: neutrophil/lymphocyte ratio, PLR:
platelet/lymphocyte ratio.
According to the receiver operating characteristics (ROC) analysis, CAR had an area
under the curve value of 0.928 (95%CI, 0.879–0.977) and was
determined to be a reliable marker for SI in FMF patients. CAR had a slightly higher
AUC value than CRP (0.916, 95%CI, 0.863–0.972). CAR had a higher AUC
value than the other systemic inflammation parameters, including albumin, ESR,
fibrinogen, RDW, NLR, PLR with AUC values of 0.727, 0.822, 0.664, 0.629, 0.598,
0.541, respectively. After pairwise comparison of AUC values, we found no
significant difference between AUC of CAR and CRP (AUC:0.928; 95%CI:
0.879–0.977, z=0.75, p=0.45), there was significant
difference between CAR and albumin levels (AUC: 0.928; 95%CI:
0.879–0.977vs AUC: 0.727; %95CI: 0,629–0,811; z=
3.6, p= 0,003).
Discussion
Although FMF is characterized by self-limiting, remitting relapsing 24- to 72-hour
episodes of fever with serositis involving the pleura, peritoneum and joints with
elevations in serum inflammatory markers, acute phase reactants can remain high in
nearly half of the patients even during the attack-free period [11]
[22].
Our results indicate that 39% of our patients had SI defined as an increased
level of at least one acute phase reactant. In previous studies, the rate of SI was
determined as 63% by Korkmaz et al. [23] and 31.4% by Çakmak et al [24]. SI may lead to impairment in quality of
life, growth retardation, normocytic normochromic anaemia, predisposition to
atherosclerosis and coronary heart disease, female infertility, depression, and
amyloidosis [10]. Amyloidosis is the most
important and devastating complication of FMF [25]. Our patients did not have amyloidosis or severe proteinuria. The
development of SI despite colchicine therapy may work as a contributor to the main
pathogenesis of amyloidosis [11]
[14]. To date, some inflammatory markers have
been studied to determine their ability to predict SI in FMF [11]
[12]
[13]
[14]
[15]
[16]. Duzova et al. defined SAA
as a useful marker to assess SI [11]. Ahsen et
al. found that the NLR values were significantly higher in patients with SI and in
FMF patients with M694V mutations [12].
Sakallı et al. suggested that higher MPV values, especially in with FMF
patients with proteinuria might be a critical determinant of SI and amyloidosis
[13]. Ozer et al. found that the PLR, NLR,
MPV and RDW levels could be used to determine SI, and NLR had the strongest
correlation with SI and amyloidosis [14].
Although we determined higher AUC values for NLR in the ROC analysis compared to the
PLR and MPV levels, NLR in our hands failed to significantly discriminate between
SI
and Non-SI, in accordance with Basaran et al. who reported that neither NLR nor MPV
was useful in detecting SI but they were significant in comparing inflammation
between the FMF group and the control group [15]. Marzouk et al. suggested that RDW and ESR could be used as
indicators of SI in children with FMF [16]. In
our study, we aimed to demonstrate the relationship between CAR and SI and
identified the best marker for SI in this patient group by comparing their CAR, RDW,
NLR, PLR and MPV values. Our results revealed that CRP and CAR had the strongest
association with SI in children with FMF. But replacing CRP by CAR in the assessment
of SI gives no additional benefit. We also observed that ESR and fibrinogen could
be
useful to determine SI but they have no better contribution than CRP and CAR ([Fig. 1])([Table 4]).
We also investigated SI in patients with M694V homozygous and heterozygous mutations
and those with no M694V mutation, considering that M694V mutations are correlated
with SI and amyloidosis. We found that SI was significantly higher in homozygous and
heterozygous patients with M694V mutations in accordance with previous findings
[12]
[26]. However, we did not obtain any significant difference in SI between
the M694V homozygous or M694V heterozygous cases (p=0.5).
CRP is an acute phase reactant produced by hepatocytes during inflammation however,
many studies have shown that CRP is also associated with chronic inflammation [27]. Albumin, the most abundant protein of
blood plasma, is a negative acute phase reactant showing inflammation and oxidative
stress [28]. The measurement of CRP and
albumin levels is an easy, inexpensive and simple method compared to the analysis
of
inflammatory cytokines, including IL-1β, IL-6, IL-8, and TNF-α or
SAA for FMF patients. CAR is widely used to predict the patient’s outcome in
many types of cancer [17], critically ill
patients [18], vasculitis [20], Crohn’s disease [29], sepsis [30], and moreover it was reported to perform better than CRP alone [8]
[30].
In one of the first studies investigating the role of CAR in rheumatologic disease,
Seringec Akkececi et al. found that CAR was correlated with disease activity in
Takayasu arteritis [31]. Yang et al. reported
that the CAR and the albumin/fibrinogen ratio could serve as inflammatory
markers for monitoring disease activity in rheumatoid arthritis [19].
The strengths of our study are that we had a high number of patients with SI
(39%), which allowed us to compare their data to those with no subclinical
inflammation. We also compared the findings according to the presence of M694V
mutations in the FMF attack-free group. Neither analysis revealed statistical
differences in markers, namely CAR, CRP, NLR, and PLR. Furthermore, considering that
lymphocyte and neutrophil counts vary in children with age, we included patients of
similar age in the sample. Nevertheless, there were also certain limitations to our
study; e. g., the retrospective single-centre design and the small sample
size. We also did not study the IL-1β, IL-6, IL-8, TNF-α or SAA
levels in our patients.
Conclusion
To our best knowledge, this is the first study showing the relationship between CAR
and SI in FMF patients. A higher CAR level indicates a higher inflammatory state and
seems to be as accurate as CRP in detecting SI based on the ROC curve analysis.
However, the prognostic value of CAR is not superior to CRP. Thus, merging CRP and
albumin in a single index provides no additional benefit in detecting SI in FMF.
Fig. 1 Receiver- operating characteristics (ROC) curve for CAR for
subclinical inflammation in attack free FMF patients.
Table 4 Receiver operating characteristics analysis for subclinical
inflammation.
|
Test Result Variable(s)
|
Area Under the Curve
|
P-value
|
95% Confidence Interval
|
|
Lower Bound
|
Upper Bound
|
|
CARAF
|
0.928
|
<0.001
|
0.879
|
0.977
|
|
CRPAF
|
0.917
|
<0.001
|
0.863
|
0.972
|
|
Albumin
|
0.727
|
<0.001
|
0.629
|
0.811
|
|
ESR
|
0.822
|
<0.001
|
0.725
|
0.918
|
|
Fibrinogen
|
0.664
|
0.006
|
0.549
|
0.779
|
|
RDWAF
|
0.629
|
0.031
|
0.515
|
0.742
|
|
NLRAF
|
0.598
|
0.102
|
0.483
|
0.712
|
|
PLRAF
|
0.541
|
0.492
|
0.424
|
0.658
|
|
MPVAF
|
0.517
|
0.772
|
0.402
|
0.633
|
CRP: C-reactive protein, AF: Attack-free, CAR: C-reactive
protein/serum albumin ratio, NLR: neutrophil/lymphocyte
ratio, PLR: platelet/lymphocyte ratio, MPV: mean platelet volume,
ESR: erythrocyte sedimentation rate.
