Risk factors for severe Muckle-Wells syndrome

Authors


Abstract

Objective

Muckle-Wells syndrome (MWS) is an inherited autoinflammatory disease resulting in excessive interleukin-1 release. It is unknown whether demographic, clinical, or laboratory characteristics at the time of diagnosis may identify patients who are at high risk for severe disease activity. This study was undertaken to analyze clinical and laboratory features of MWS, compare genetically defined subcohorts, and identify risk factors for severe MWS.

Methods

A multicenter cohort study of consecutive MWS patients was performed. Parameters assessed included clinical features, MWS Disease Activity Score (MWS-DAS), inflammation markers, and cytokine levels. E311K mutation–positive patients were compared with E311K mutation–negative patients. Putative risk factors for severe MWS (defined as an MWS-DAS score of ≥10) were assessed in univariate analyses, and significant predictors were entered into a multivariate model.

Results

Thirty-two patients (15 male and 17 female) were studied. The most frequent organ manifestations were musculoskeletal symptoms and eye and skin disorders. Renal disease and hearing loss were seen in >50% of the patients. Genetically defined subcohorts had distinct phenotypes. Severe disease activity was documented in 19 patients (59%). Predictors of severe MWS identified at the time of diagnosis were female sex, hearing loss, musculoskeletal disease, increased erythrocyte sedimentation rate, and low hemoglobin level. Female sex and hearing loss remained significant after adjustment for age in a multivariate model (relative risk 1.8 and 2.6, respectively).

Conclusion

MWS patients at high risk for severe disease can be identified at the time of diagnosis. Female patients presenting with hearing loss have the highest likelihood of manifesting severe MWS and should be considered a high-risk group.

Muckle-Wells syndrome (MWS) is one of the inherited cryopyrin-associated periodic fever syndromes (CAPS), which are caused by mutations in the NLRP3 gene, leading to alterations of the gene product cryopyrin (also known as NALP3). The protein is part of the inflammasome, a multiprotein complex crucial for intracellular host defense (1–3). CAPS encompass the mildest form, familial cold autoinflammatory syndrome (FCAS), MWS, and the most severe form, neonatal-onset multisystem inflammatory disease/chronic infantile neurologic, cutaneous, articular syndrome (NOMID/CINCA syndrome) (1). All 3 diseases are characterized by seemingly unprovoked inflammation as a consequence of excessive production of the proinflammatory cytokine interleukin-1β (IL-1β) (2, 3).

Patients with MWS typically present with acute attacks of fever, rash, musculoskeletal symptoms, and conjunctivitis. These characteristic features occur episodically and may last between 1 day and 2 weeks. The MWS-associated rash may be cold induced, urticarial, or erythematous in nature. Musculoskeletal symptoms include arthralgia, arthritis, and significant myalgia. Conjunctivitis together with progressive sensorineural hearing loss are the pathognomonic clinical findings that support the diagnosis of MWS (4). Constitutional features of severe, persistent fatigue are commonly found and have a significant impact on the quality of life of MWS patients. Sequelae of MWS include deafness and amyloidosis, causing renal failure. Levels of inflammation markers are commonly elevated in patients with MWS, particularly during fever episodes (5–7). Levels of proinflammatory cytokines were recently shown to also be elevated in active MWS (8).

Mutations in the NLRP3 gene on chromosome 1q44 are found in most patients with CAPS (1, 9, 10). Inheritance of CAPS is generally autosomal-dominant, although sporadic cases of NOMID/CINCA syndrome, MWS (6), and FCAS (9) due to de novo mutations have been described. To date, >115 NLRP3 mutations and polymorphisms have been identified. The majority are missense mutations localized in exon 3 encoding the NACHT domain (11). In MWS, NLRP3 mutations are associated with a broad clinical phenotype (12). Demographic, clinical, or laboratory characteristics at diagnosis that might be associated with high risk for a severe MWS phenotype have not been identified to date. The aims of this study were therefore to systematically analyze the clinical and laboratory characteristics of a multicenter cohort of MWS patients, to compare distinct, genetically defined subcohorts with specific MWS features, and to identify risk factors for a severe MWS phenotype.

PATIENTS AND METHODS

Patients.

A multicenter cohort study of consecutive patients diagnosed as having MWS between February 1999 and January 2007 was performed. MWS patients at the pediatric and rheumatology departments of 5 centers (University Hospital [Tuebingen, Germany], University Hospital [Wuerzburg, Germany], Center for Pediatrics and Neonatology [Sankt Augustin, Germany], German Center for Rheumatology in Children and Adolescents [Garmisch-Partenkirchen, Germany], and Bregenz National Hospital [Bregenz, Austria]) were included. Informed consent was obtained from all patients (or from the legal guardian if the patient was a minor) prior to the DNA sequence analysis of NLRP3 exons 3, 4, and 6. The study was approved by the institutional ethics committees (325/2007B01; 326/207B01).

Patients were eligible for the study if they had a diagnosis of active MWS and genetic testing for NLRP3 mutations had been performed. Patients were excluded if they had any clinical features characteristic of NOMID/CINCA syndrome, such as disease onset at an early age, bony inflammatory changes, and neurologic manifestations or clinical features characteristic of other autoinflammatory syndromes. All MWS patients were assessed according to a standardized protocol developed at the interdisciplinary Autoinflammatory Diseases Clinic at University Hospital Tuebingen. All patients were followed up by experienced rheumatologists at the respective centers.

Data collection.

Information on each patient's ethnicity, sex, age at diagnosis, duration of symptoms/illness, duration of followup, and types and duration of treatment was collected. Identified NLRP3 gene mutations were documented. Family history determined parental consanguinity and identified family members diagnosed as having autoinflammatory syndromes and a history of possible organ involvement such as hearing loss. Patients were assessed at diagnosis prior to initiation of therapy and at standardized intervals as reported previously (13). All data were recorded in a designated Web-based documentation system incorporated in the Arthritis and Rheumatism Database and Information System (ARDIS), a previously described Web-based prospective data collection tool capturing all variables of the standardized assessment (14). For the present study, data were extracted from the ARDIS and reviewed retrospectively.

Subcohorts.

Patients were assigned to the E311K subcohort if they had confirmatory genetic evidence of this NLRP3 mutation. Patients were assigned to the non-E311K subcohort if they had an NLRP3 mutation other than E311K or had classic MWS in the absence of a confirmatory NLRP3 mutation.

Assessment of clinical features.

Clinical assessment included evaluation for constitutional symptoms such as fatigue and fever pattern and duration, as well as organ-specific clinical characteristics such as headache, conjunctivitis, uveitis, papilledema, sensorineural hearing loss, oral ulcers, abdominal pain, arthralgia, arthritis, myalgia, and skin symptoms including erythematous and cold-induced rash. Data on associated conditions, disease complications, and sequelae including hearing impairment or loss, renal impairment, or delayed puberty were also recorded.

Audiologic assessment.

Clinical examination and audiography were performed to evaluate the status of hearing prior to treatment and every 6 months thereafter. The audiologic examination included bone and air conduction pure tone threshold for frequencies of 250–8,000 Hz, and tympanometry. According to a standardized 4-frequency pure tone average score, hearing was classified as normal (0–20% loss; grade 0), mildly impaired (20–30% loss; grade 1), moderately impaired (40–60% loss; grade 2), severely impaired (60–80% loss [or up to 90 dB]; grade 3), or profoundly lost (>80%; grade 4) (15).

Ophthalmologic evaluation.

A standardized ophthalmologic evaluation was performed by a consultant ophthalmologist prior to treatment, at 2 weeks, and then every 6 months. The examination focused on visual changes that are sensitive indicators of optic nerve function, conjunctivitis, and uveitis.

MWS Disease Activity Score (MWS-DAS).

The previously developed MWS-DAS (13) captures data on active MWS in 10 domains, 9 of which reflect organ involvement of MWS (fever, headache, eye involvement, hearing impairment, oral ulcers, abdominal pain, renal disease, musculoskeletal disease, and rash). The tenth domain is the Patient Global Assessment Score. The MWS-DAS attributes 0, 1, or 2 points to each level of disease activity. Two points are given for severe symptoms, 1 point for mild symptoms, and 0 points for the absence of symptoms in a domain. As previously determined, an MWS-DAS score of <10 reflects overall mild MWS disease, whereas a score of ≥10 is an indicator of severe MWS disease activity (13).

Global measures of disease activity.

Three patient-derived measures of health status were determined at each visit: 1) the Patient Global Assessment Score, recorded on a 10-cm visual analog scale (VAS) with 0 representing no disease activity/perfect health and 10 representing maximum disease activity; 2) the Patient Mood Score, recorded on a 3-point Likert scale with 1 representing excellent mood and 3 representing the lowest mood possible; and 3) the Patient Performance Score, recorded on a 0–10 VAS with 0 representing inability to perform and 10 representing the best possible performance level. The Physician Global Assessment Score (0–10 VAS, with 0 representing no disease activity and 10 representing maximum disease activity) was also recorded at each visit.

Inflammation markers and cytokine profiles.

The following laboratory parameters were assessed/measured at the time of diagnosis: 1) classic inflammation markers and hematologic parameters, i.e., erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), total white blood cells, neutrophils, hemoglobin, and platelets; 2) MWS biomarkers, i.e., serum amyloid A (SAA), IL-1β, IL-6, and tumor necrosis factor α (TNFα), determined by enzyme-linked immunosorbent assay; 3) renal function parameters, including serum creatinine, urea, and uric acid, urine blood and albumin measured by urinalysis with dipstick test, urine protein (α1-microglobulin and β2-microglobulin) measured by spot urine test, urine creatinine, protein, and creatinine clearance measured 24-hour urine collection, and calculation of the protein:creatinine ratio.

Study outcome measure and variables investigated.

The primary outcome measure of the study was evidence of severe MWS disease activity, defined as an MWS-DAS score of ≥10 at the time of diagnosis (13). Putative clinical and laboratory risk factors associated with severe MWS disease activity were selected for analysis. These included 1) demographic and baseline disease characteristics (sex, NLRP3 mutation); 2) clinical features (fever, MWS eye disease [conjunctivitis], hearing loss, MWS skin disease [oral ulcers], MWS musculoskeletal disease [arthralgia, arthritis, myalgia], and MWS renal disease [proteinuria, hematuria, renal failure]); 3) global measures of disease activity (Physician Global Assessment Score, Patient Mood Score, Patient Performance Score); and 4) laboratory features (leukocyte, neutrophil, hemoglobin, and CRP levels and ESR).

Predictive model.

Putative risk factors were assessed separately by univariate analysis. Selection of variables was based on statistically significant association with a severe disease phenotype. Significant predictors were then entered into a multivariate modified Poisson regression analysis in order to appropriately estimate an outcome of high prevalence (59% had the outcome).

Statistical analysis.

Baseline demographic findings were analyzed using descriptive statistics. Characteristics of subgroups were compared using Student's t-test for continuous data and chi-square analysis or Fisher's exact test for categorical variables. All analyses were performed using SAS for Windows, version 9.2.

RESULTS

Patient characteristics (Table 1).

A total of 32 MWS patients (15 male, 17 female) were included in the study. The median age at diagnosis was 15.4 years (range 3.7–75.0). All patients were white. Twenty-five of the patients (78%) had a confirmed NLRP3 mutation. Thirteen (41%) of the patients (5 male and 8 female; median age at diagnosis 36.6 years [range 6.7–75.0]) had an E311K mutation. The remaining 19 patients (59%) (10 male and 9 female; median age at diagnosis 10.6 years [range 3.7–51.5]) did not have an E311K mutation.

Table 1. Demographic features, NLRP3 mutations, family history, treatment, and followup of the patients with Muckle-Wells syndrome (MWS)*
 Total MWS group (n = 32)E311K subcohort (n = 13)Non-E311K subcohort (n = 19)P
  • *

    Except where indicated otherwise, values are the number (%) of patients.

Demographics    
 Age at diagnosis, median (range) years15.4 (3.7–75.0)36.6 (6.7–5.0)10.6 (3.7–51.5)0.7
 Males/females, no. (ratio)15/17 (0.9:1)5/8 (0.6:1)10/9 (1.1:1)0.4
NLRP3 mutations    
 Total25 (78)13 (100)12 (63)
 Mutation type    
  E311K13 (41)13 (100)
  V198M5 (16)5 (26) 
  T348M4 (13)4 (21) 
  E525K1 (3)1 (5) 
  D303N1 (3)1 (5) 
  R135H1 (3)1 (5) 
Family history    
 Overall positive27 (84)13 (100)14 (74)0.06
 Fever23 (72)13 (100)10 (53)0.003
 Renal failure17 (53)13 (100)4 (21)<0.0001
 Hypertension15 (47)13 (100)2 (11)<0.0001
 Hearing loss19 (59)13 (100)6 (32)0.0001
 Arthralgia27 (84)13 (100)14 (74)0.06
 Arthritis20 (63)13 (100)7 (37)<0.0003
 Myalgia17 (53)13 (100)4 (21)<0.0001
 Eye disease23 (72)13 (100)10 (53)0.003
 Rash24 (75)13 (100)11 (58)0.007
Anakinra treatment18 (56)7 (54)11 (58)0.8
 Duration, mean months21.718.823.6
Followup, mean (range) months37.5 (5.0–96.5)33.5 (33.3–33.7)40.3 (5.0–96.5)

The following mutations were identified among patients in the non-E311K subcohort: V198M in 5 patients (26%) from 2 unrelated families (2 patients and 3 patients, respectively), T348M in 4 patients (21%) from 2 unrelated families (3 patients and 1 patient, respectively), and R135H, D303N, and E525K in 1 patient each (5%). In 7 MWS patients, no mutation was found by sequence analysis of NLRP3 exons 3, 4, and 6. (The Q703K substitution, which was identified in 1 of these 7 patients, should be regarded as a polymorphism rather than a mutation [12].)

A positive family history of MWS-related clinical features was reported in 27 individuals (84%). In the E311K subcohort, 100% of patients had a positive family history, including fever, renal failure, hypertension, hearing loss, arthritis, myalgia, eye disease, and rash. In the non-E311K subcohort, 74% of the patients had family members with MWS features, most commonly arthralgia (74%), rash (58%), fever (53%), eye disease (53%), arthritis (37%), and hearing loss (32%). There was no consanguinity in either subcohort. Eighteen patients (56%) received anakinra therapy. The mean followup time was 37.5 months (range 5.0–96.5) in the total cohort, 33.5 months (range 33.3–33.7) in the E311K subcohort, and 40.3 months (range 5.0–96.5) in the non-E311K subcohort. Treatment prior to diagnosis of MWS included nonsteroidal antiinflammatory drugs, corticosteroids and TNF inhibitors. None of the patients had been treated with IL-1 blocking agents prior to diagnosis of MWS.

Clinical findings (Table 2).

Fever.

Fever episodes were reported by 47% of the patients (31% in the E311K subcohort, compared with 58% in the non-E311K subcohort). The development of fever episodes tended to occur later in the E311K subcohort, at a mean age 21.5 years, versus 4.3 years in the non-E311K subcohort. Resolution of fever episodes occurred at an average age of 41.3 years in the E311K subcohort, compared with 11.1 years in the non-E311K subcohort. Fever episodes lasted an average of 1.6 days in the 2 subcohorts combined, ranging from 0.3 days to 4.0 days.

Table 2. Clinical features, disease activity, scores on global measures of health, inflammation markers, and cytokine profiles in the patients with MWS*
 Total MWS group (n = 32)E311K subcohort (n = 13)Non-E311K subcohort (n = 19)P
  • *

    Except where indicated otherwise, values are the number (%) of patients. MWS = Muckle-Wells syndrome; ENT = ear, nose, throat; MWS-DAS = MWS Disease Activity Score; VAS = visual analog scale; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; SAA = serum amyloid A; IL-1β = interleukin-1β; TNFα = tumor necrosis factor α.

  • Data available on 23 patients (13 in the E311K subcohort and 10 in the non-E311K subcohort).

  • Data available on 13 patients (8 in the E311K subcohort and 5 in the non-E311K subcohort).

  • §

    Data available on 21 patients (13 in the E311K subcohort and 8 in the non-E311K subcohort).

Clinical features    
 Fever    
  Patients with fever episodes15 (47)4 (31)11 (58)0.2
  Age at onset of fevers, mean (range) years8.9 (0.5−57.0)21.5 (1.0−57.0)4.3 (0.5−12.0)
  Age at resolution of fevers, mean (range) years21.1 (8.0−67.0)41.3 (15.0−67.0)11.1 (8.0−13.5)
  Fever duration, mean (range) days1.6 (0.3−4.0)1.75 (1.0−3.0)1.52 (0.3−4.0)
 MWS eye disease25 (78)11 (85)14 (74)0.5
  Conjunctivitis22 (69)10 (77)12 (63)0.4
  Uveitis4 (13)2 (15)2 (11)1.0
  Papilledema3 (9)0 (0)3 (16)
 MWS ENT disease    
  Hearing loss18 (56)12 (92)6 (32)0.0007
  Age at hearing loss, mean (range) years20.4 (5.5−41.0)21.9 (5.5−41.0)16.9 (7.0−39.0)
  Oral ulcers13 (41)6 (46)7 (37)0.59
 MWS skin disease23 (72)7 (54)16 (84)0.06
  Cold-induced rash5 (16)0 (0)5 (26)0.04
 MWS musculoskeletal disease29 (91)11 (85)18 (95)0.33
  Arthralgia29 (91)11 (85)18 (95)0.33
  Arthritis20 (63)9 (69)11 (58)0.51
  Myalgia10 (31)7 (54)3 (16)0.02
 MWS-associated pain    
  Headache18 (50)7 (54)11 (58)0.82
  Abdominal pain14 (44)4 (31)10 (53)0.22
 MWS renal disease19 (59)10 (77)9 (47)0.09
  Proteinuria17 (53)10 (77)7 (37)0.03
  Hematuria4 (13)1 (8)3 (16)0.45
  Renal failure2 (6)1 (8)1 (5)1.00
MWS disease activity    
 Severe MWS (MWS-DAS ≥10)19 (59)8 (62)11 (58)0.8
 Mild MWS (MWS-DAS <10)13 (41)5 (38)8 (42)0.8
Global measures of health    
 Physician Global Assessment Score, 0−10 VAS, mean ± SD5.7 ± 2.25.9 ± 1.55.6 ± 2.60.8
 Patient Global Assessment Score, 0−10 VAS, mean ± SD5.2 ± 2.44.8 ± 2.25.6 ± 2.60.3
 Patient Mood Score, 1−3, mean ± SD2.2 ± 0.62.1 ± 0.52.2 ± 0.70.6
 Patient Performance Score, 0−10 VAS, mean ± SD5.5 ± 2.45.9 ± 2.25.3 ± 2.60.6
Laboratory features    
 Inflammation markers    
  Leukocytes, mean (range)/μl (normal 6,000−17,500)12,868 (5,840−25,000)10,507 (6,530−21,200)14,484 (5,840−25,000)0.04
  Neutrophils, mean (range)/μl (normal 1,500−8,000)8,730 (2,321−22,250)7,694 (3,434−18,211)9,439 (2,321−22,250)0.4
  Hemoglobin, mean (range) gm/dl (normal 10.1−18.0)12.7 (7.4−17.2)13.7 (10.6−16.9)12.0 (7.4−17.2)0.03
  Platelets, mean (range) ×103/μl (normal 150−450)417 (206−881)378 (319−474)445 (206−881)0.3
  ESR, mean (range) mm/hour (normal >20)36 (2−111)31 (3−55)39 (2−111)0.4
  ESR abnormal23 (72)9 (69)14 (74)
  CRP, mean (range) mg/dl (normal >0.5)13.9 (0.01−113.0)2.3 (0.66−5.60)21.9 (0.01−113.0)0.05
  CRP abnormal27 (84)13 (100)14 (74)
  SAA, mean (range) mg/liter (normal >0.9)46.2 (2.0−388.0)40.7 (5.2−254.0)53.3 (2.0−388.0)
  SAA abnormal14 (61)10 (80)4 (40)
 Cytokines    
  IL-1β, mean (range) pg/ml (normal <0.5)0.8 (0.2−2.1)0.5 (0.2−1.5)1.3 (0.5−2.1)
  IL-1β abnormal9 (69)4 (50)5 (100)
  TNFα, mean (range) pg/ml (normal <8)§8.9 (4.0−22.4)9.4 (4.0−16.5)8.2 (4.0−22.4)
  TNFα abnormal§9 (43)7 (54)2 (25)
  IL-6, mean (range) pg/ml (normal <8)§10.5 (2.0−46.6)6.4 (2.3−15.7)17.2 (2.0−46.6)
  IL-6 abnormal§10 (48)6 (46)4 (50)

MWS organ manifestations.

The most commonly reported organ manifestations were musculoskeletal symptoms, including arthralgia (91% of the total group), arthritis (63%), and myalgia (31%). Eye disease was found in 78% of the patients, including conjunctivitis in 69%, uveitis in 13%, and papilledema in 9%. MWS rash was seen in 23 patients (72%), of whom 5 had cold-induced urticaria. Renal disease was present in 59%, with proteinuria occurring in 53% and renal failure in 6% (2 patients). Hearing loss was documented in 56% of the patients, occurring at an average age of 20.4 years.

The E311K subcohort and the non-E311K subcohort presented with distinctly different clinical phenotypes. The E311K subcohort included a significantly higher proportion of patients with hearing loss, myalgia, and proteinuria. In contrast, the non-E311K patients were found to have a significantly higher frequency of cold-induced urticaria. There was no statistically significant difference for any of the other MWS organ manifestations.

Disease activity and global measures of disease activity.

Severe MWS disease activity at diagnosis (MWS-DAS score ≥10) was noted in 19 patients (59%), while 13 patients (41%) had mild MWS disease activity. No significant difference based on E311K mutation was observed. In the E311K subcohort, severe MWS disease activity was documented in 8 patients (62%) and mild MWS disease activity in 5 (38%); in the non-E311K subcohort, 11 patients (58%) presented with severe disease activity and 8 (42%) with mild disease activity. Similarly, the mean Physician Global Assessment Score, Patient Mood Score, and Patient Performance Score were comparable. Patient-derived measures consistently revealed moderately elevated scores in both subcohorts.

Inflammation markers and cytokine profiles.

Classic inflammation markers, including leukocyte and neutrophil counts, hemoglobin level and platelet count, ESR, and CRP and SAA levels were abnormal in the majority of MWS patients tested. On average, the levels were moderately elevated. Comparison of the E311K and non-E311K subcohorts revealed statistically significant differences in mean leukocyte count, hemoglobin level, and CRP level. Patients with the E311K mutation had significantly lower levels of inflammation markers overall, with lower leukocyte counts (P = 0.04), higher hemoglobin levels (P = 0.03), and lower CRP levels (P = 0.05). There were no significant differences in neutrophil or platelet counts or ESR. SAA levels tended to be lower in the E311K subcohort (mean 40.7 mg/liter, compared with 53.3 mg/liter in the non-E311K subcohort [10 of 19 non-E311K patients tested]).

Cytokine levels at diagnosis were determined in 21 patients, including all 13 patients in the E311K subcohort and 8 of the 19 patients in the non-E311K subcohort. Abnormal levels were identified in approximately half of the patients, including elevations of IL-1β in 69%, IL-6 in 48%, and TNFα in 43%. Patients with the E311K mutation tended to have elevated levels of IL-1β (50% of the patients), TNFα (54%), and IL-6 (46%). In contrast, whereas patients in the non-E311K subcohort had increased levels of IL-1β (100% of the patients) and IL-6 (50%), TNFα levels were less commonly elevated in this subgroup (25%).

Achievement of and predictors of the outcome measure (severe MWS).

As noted above, severe MWS disease activity at diagnosis, as determined by an MWS-DAS score of ≥10, was documented in 19 patients (59%); 13 patients (41%) had mild disease activity. Univariate analyses performed to identify putative predictors of the severe MWS phenotype identified several significant variables (Table 3). These included 1) the demographic characteristic female sex (P = 0.04); 2) the clinical features hearing loss (P = 0.02), MWS musculoskeletal disease (P = 0.03), arthralgia (P = 0.03), and arthritis (0.02); 3) the global disease activity measures Physician Global Assessment Score (P < 0.001), Patient Mood Score (P = 0.002), and Patient Performance Score (P < 0.001); and 4) the laboratory features hemoglobin level (P = 0.008) and ESR (P = 0.04).

Table 3. Variables associated with a severe MWS phenotype in the univariate analysis*
 Mild MWS (MWS-DAS <10) (n = 13)Severe MWS (MWS-DAS ≥10) (n = 19)P
  • *

    Except where indicated otherwise, values are the number (%) of patients. See Table 2 for definitions and for normal ranges of laboratory measures.

Baseline characteristics   
 Female sex4 (31)13 (69)0.04
 E311K mutation5 (38)8 (42)0.8
Clinical features   
 Fever3 (23)12 (63)0.1
 MWS eye disease8 (62)17 (90)0.06
 Conjunctivitis7 (54)15 (79)0.1
 Hearing loss4 (31)14 (74)0.02
 MWS skin disease8 (62)15 (79)0.3
 Oral ulcers3 (23)10 (53)0.1
 MWS musculoskeletal disease10 (77)19 (100)0.03
  Arthralgia10 (77)19 (100)0.03
  Arthritis5 (38)15 (79)0.02
  Myalgia2 (15)8 (42)0.1
 MWS renal disease6 (46)13 (68)0.2
  Proteinuria6 (46)11 (58)0.5
  Hematuria0 (0)4 (21)0.07
  Renal failure0 (0)2 (11)0.22
Global measures   
 Physician Global Assessment Score, 0−10 VAS, mean ± SD3.53 ± 1.507.21 ± 1.0<0.001
 Patient Mood Score, 1–3, mean ± SD1.77 ± 0.592.42 ± 0.500.002
 Patient Performance Score, 0–10 VAS, mean ± SD7.53 ± 1.454.15 ± 1.95<0.001
Laboratory features   
 Leukocytes, mean (range)/μl12,414 (6,530−23,100)13,179 (5,840−25,000)0.7
 Neutrophils, mean (range)/μl7,662 (2,321.3−18,211)9,462 (2,628−22,250)0.3
 Hemoglobin, mean (range) gm/dl14.0 (11.2−17.2)11.9 (7.4−16.9)0.008
 ESR, mean (range) mm/hour24 (2−89)44 (6−111)0.04
 CRP, mean (range) mg/dl7.3 (0−56.0)18.5 (0−113.0)0.3

Results of predictive model application.

The features shown in the univariate analysis to be significantly associated with severe disease were subjected to multivariate model analysis with adjustment for age. In this analysis, female sex and evidence of hearing loss at diagnosis were significantly and independently associated with the severe MWS phenotype, even after adjustment for age (for female sex, relative risk [RR] 1.8, 95% confidence interval [95% CI] 1.3–5.1, P = 0.006; for hearing loss at diagnosis, RR 2.6, 95% CI 1.0–3.3, P = 0.048).

DISCUSSION

This study evaluated a multicenter cohort of 32 MWS patients ages 3–75 years. The majority presented with musculoskeletal symptoms, conjunctivitis, and skin disease. Renal disease and hearing loss were seen in >50%. We identified 2 genetically distinct subgroups of MWS patients, an E311K mutation–positive group and an E311K mutation–negative group. These subcohorts had characteristic clinical and laboratory features. Patients with the E311K mutation presented significantly more often with hearing loss, myalgia, and proteinuria. The non-E311K group had a significantly higher frequency of cold-induced urticaria, elevated levels of inflammation markers including leukocytes and CRP, and reduced levels of hemoglobin. The subcohorts did not differ significantly in their cytokine profiles at diagnosis. Severe MWS disease as defined by an MWS-DAS score of ≥10 was found in 59% of the patients studied. The percentage of patients presenting with severe MWS disease activity did not differ significantly between the E311K and non-E311K subcohorts. The study identified a predictive model for high risk for a severe MWS phenotype, which included female sex and hearing loss at diagnosis.

The collaborative, multicenter study design provided a unique opportunity to analyze the impact of one distinct genetic mutation on the clinical phenotype of MWS. The E311K mutation was previously reported as being associated with a “pure,” non-overlapping MWS phenotype (16). We were able to confirm this observation in the 13 E311K-positive patients included in the present study. All of these individuals had a positive family history of MWS; 92% presented with hearing loss, 77% had proteinuria, and 54% had severe myalgia. None of these patients had cold-induced urticaria. Interestingly, the E311K subcohort overall had lower levels of inflammation markers than the non-E311K patients, who had other NLRP3 mutations or, in 7 cases, no NLRP3 mutation at all. In this non-E311K group, MWS skin disease was found in 84% of the patients and cold-induced urticaria in 26%. As noted above, inflammation markers were more commonly elevated in this latter subcohort.

The E311K genotype was associated with a specific MWS phenotype. Such distinct clinical patterns have been described for other mutations as well, in particular, the T348M mutation (17, 18), identified in 4 patients in the present study. In contrast, the clinical presentation of patients with the V198M mutation, found in 5 patients in the present study, is extremely variable and ranges from asymptomatic to both FCAS and MWS phenotypes (17), supporting the concept of a mutation with reduced penetrance that requires additional genetic factors in order to produce a disease phenotype (12). The V262 substitution has been shown to be associated with a clinical overlap of FCAS and MWS (18, 19), while the D303N mutation (12), identified in 1 patient in the present study, and the I480F mutation (20) have been found in patients with MWS–NOMID overlap. The E525K mutation, in contrast, found in 1 patient in the present study, has in the past been described exclusively in patients with FCAS (21). The novel R135H mutation also encoded by NLRP3 exon 3 has not been reported previously. The patient in the present study who had this amino acid substitution had the typical clinical picture of MWS with fever, sensorineural hearing loss, and musculoskeletal symptoms. In addition, this patient had a cold-induced rash as seen in FCAS.

Seven of the 32 study patients (22%) exhibited typical symptoms of MWS even though no NLRP3 mutation could be identified. In an extensive study of patients with CAPS, Aksentijevich et al identified novel mutations and reported a NLRP3 mutation rate of 75% in typical MWS patients (12).

No previously published studies have determined MWS risk factors beyond possible genetic associations with disease activity. However, recognition of potential risk factors for a severe MWS disease phenotype may enable physicians to tailor monitoring and treatment approaches. This collaborative study utilized the MWS-DAS to identify patients with severe disease, which was present in 19 of 32 patients (59%). Baseline predictors of severe MWS included female sex, evidence of hearing loss, MWS musculoskeletal disease, low hemoglobin level, and increased ESR at the time of diagnosis. When significant predictors were tested in a regression model, female sex and evident hearing loss were strongly and independently associated with severe MWS disease. This analysis supports the clinical observation that overall, female patients appear to have more severe disease.

Global measures of disease activity, i.e., the Physician Global Assessment Score, Patient Mood Score, and Patient Performance Score, were significantly associated with severe MWS disease activity as measured by the MWS-DAS. The finding that disease severity strongly corresponded with fatigue and mood scores and physician global assessments demonstrates that disease activity in MWS can be estimated using physician- and patient-derived measures. NLRP3 mutations, in contrast, were not associated with disease activity.

There are several limitations to this study. Although it represents one of the largest series of MWS patients reported to date, the total number of patients in the cohort was only 32. State-of-the-art MWS genetic testing, standardized assessments, and measurements of classic inflammation markers were performed, but cytokine profiles were obtained in only a subset of the cohort, and their predictive value remains to be determined in future studies.

In conclusion, MWS patients at high risk for a severe disease activity can be identified at the time of diagnosis. Female patients presenting with hearing loss, musculoskeletal symptoms, elevated ESR, and/or low hemoglobin levels have a higher likelihood of severe MWS. Women with hearing loss (22) at MWS diagnosis should be considered a high-risk group for severe disease.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Benseler had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Kümmerle-Deschner, Tyrrell, Reess, Kötter, Lohse, Girschick, Huemer, Horneff, Haas, Koitschev, Deuter, Benseler.

Acquisition of data. Kümmerle-Deschner, Tyrrell, Reess, Kötter, Lohse, Girschick, Huemer, Horneff, Haas, Koitschev, Deuter, Benseler.

Analysis and interpretation of data. Kümmerle-Deschner, Tyrrell, Reess, Kötter, Lohse, Girschick, Huemer, Horneff, Haas, Koitschev, Deuter, Benseler.

Ancillary