Dermatologic Manifestations of Relapsing Polychondritis

Introduction Relapsing polychondritis (RP) is an uncommon disease that may be associated with other rheumatologic or autoimmune conditions (30%–35%) and with myelodysplasia (5%–10%) (24,28,34). RP is characterized by recurrent inflammation and degeneration of cartilaginous tissue, especially auricular, nasal, and laryngeal/tracheal cartilages. In fact, many other tissues may be involved, such as the eye, heart, blood vessels, and inner ear (23,25,32,34). Dermatologic manifestations have been relatively poorly studied compared to other signs. In previous studies (23,34), a significant number of patients with RP were reported to have leukocytoclastic vasculitis. We conducted the present study to analyze the dermatologic manifestations of 200 patients with RP. Patients and Methods Between 1974 and 2000, 200 patients with RP were followed in our university referral center for connective tissue disorders as in-or outpatients. To qualify for the study, patients had to fulfill the diagnostic criteria defined by Michet et al (25). Indeed, all had a documented history of inflammatory episodes, involving at least 2 of 3 sites (auricular, nasal, laryngotracheal cartilages) or 1 of these sites together with 2 other manifestations, including ocular inflammation (conjunctivitis, keratitis, episcleritis, uveitis), hearing loss, vestibular dysfunction, or seronegative inflammatory arthritis (25). Histologic proof of cartilage involvement was not required for inclusion. The patients, 129 women and 71 men, had a mean age of 43 ± 18 years (range, 6–87 yr) at the time of the first chondritis. Patients were seen by 1 of us (JCP) and the presence of cutaneous or mucosal lesions was systematically checked. All those who displayed dermatologic involvement were examined at least once by an experienced dermatologist (CF) who recorded all previous and current dermatologic manifestations. Clinical and laboratory data were analyzed retrospectively. Oral aphthosis was diagnosed when recurrences occurred at least 3 times during a 12-month period according to O’Duffy criteria for aphthae in Behçet disease (22). Genital aphthae were always noted regardless of the number of recurrences (18). Complex aphthosis is defined as the association of oral and genital aphthoses. Raynaud phenomenon and frequent infectious skin lesions, such as herpes infections, were not taken into account. Associated diseases were diagnosed using standard criteria. Patients with an associated disease with possible dermatologic involvement or chronic dermatitis were analyzed separately; the clinical and laboratory data for the remaining patients, with or without dermatologic manifestations, were compared. Myelodysplasia was diagnosed in patients with chronic macrocytic refractory anemia (hemoglobin < 10 g/dL) not explained by vitamin or hormone deficiency or renal failure. Age at the time of the first chondritis, sex ratio, and frequency of dermatologic manifestations were compared between patients without associated disorders and those with myelodysplastic syndrome. A total of 82 skin or mucosal biopsy specimens available from 59 patients were examined by 1 of us (JLL) who had no knowledge of the clinical data. All specimens were stained with hematoxylin-eosin-saffron. Direct immunofluorescence was performed on 17 skin biopsy specimens clinically suggestive of vasculitis. Statistical comparisons were made using the chi-square test for qualitative data and the Wilcoxon test for quantitative data. Values of p less than 0.05 were considered significant. Results Seventy-three (27 male/46 female patients) of the 200 patients with RP had at least 1 associated disease with potential dermatologic involvement or chronic dermatitis (Table 1). The other 127 patients (44 male/83 female) had a mean age of 41.4 ± 17 years (range, 5–84 yr) at the time of the first chondritis. Forty-five (35.4%) of them had dermatologic manifestations (Table 2), which were the first manifestations of RP in 15 (12%) cases. The dermatologic manifestations preceded the first chondritis in 20 (15.7%) cases by 10 days to 20 years (mean, 45.6 ± 64 mo), appeared concomitantly with the first chondritis in 11 (8.6%) cases, and appeared secondarily in 14 (11%) cases with a mean time lapse from the first chondritis of 59 ± 56 months (range, 1–172 mo). They occurred concomitantly with chondritis attacks in 23 cases and independently in 22 cases. Oral aphthosis was the most common manifestation, present in 21 patients, 7 of whom also had genital involvement. No patient had genital without oral aphthae. Nodules on the limbs frequently appeared as erythema nodosum-like lesions. However, the histologic findings in these lesions were diverse since septal panniculitis was observed in only 3 of 8 biopsy specimens. Other histologic findings included thrombosis or vasculitis of skin vessels and deep dermal neutrophil infiltrates resembling those seen in some deep lesions of Sweet syndrome. Purpura corresponded to leukocytoclastic or, rarely, lymphocytic vasculitis. Papules were either urticarial corresponding to leukocytoclastic or lymphocytic vasculitis, or bluish-red corresponding to Sweet syndrome or leukocytoclastic vasculitis. Superficial phlebitis was confirmed by venous echoDöppler without biopsy. Skin ulcerations were observed predominantly on the limbs and were secondary to small-vessel vasculitis or thrombosis. In 1 case, the ulceration had the clinical and histologic features of pyoderma gangrenosum. Taken altogether regardless of the clinical features, histologic findings showed leukocytoclastic vasculitis (n = 17), lymphocytic vasculitis (n = 2), neutrophil infiltrates (n = 6), thrombosis of skin vessels (n = 4), septal panniculitis (n = 3), or minor changes (n = 2). Direct immunofluorescence performed on skin biopsy specimens with leukocytoclastic vasculitis (n = 9) was either negative (n = 5) or positive with vascular deposits of C3 (n = 4), IgG (n = 2), and/or IgM (n = 2). No clear preferential association was observed among the various clinical dermatologic lesions (Table 3). Patients with and without dermatologic manifestations did not differ with regard to male/female ratio (15/30 versus 29/53); age at RP onset (40.5 ± 18 yr versus 42 ± 17 yr); prevalence of auricular (93% versus 76%), nasal (71% versus 71%), or laryngotracheobronchial chondritis (62% versus 55%); and prevalence of articular (96% versus 79%), ocular (78% versus 60%), audiovestibular (51% versus 41%), large artery (26% versus 12%), or deep venous (6% versus 2%) involvement.TABLE 1: Associated diseases in 73 patients with relapsing polychondritisTABLE 2: Dermatologic manifestations and histologic findings in 45 of 127 patients with relapsing polychondritis without associated diseaseTABLE 3: Number of patients with associated dermatologic manifestations among the 45 patients without associated diseaseTwenty-two patients had myelodysplasia; some of these patients were reported briefly previously (26). One had dermatomyositis diagnosed 2 years before RP without other dermatologic lesions. In another patient, Crohn disease was diagnosed 1 month after RP. The dermatologic manifestations observed in 20 of these 22 patients and their corresponding clinicopathologic findings are listed in Table 4. Eight patients presented with dermatologic manifestations 10.3 ± 10 months (range, 0.5–24 mo) before the first chondritis episode. Dermatologic manifestations developed concomitantly with the first chondritis in 5 cases and secondarily in 7 (mean time lapse, 56 ± 20 mo; range, 3–153 mo). They occurred simultaneously with chondritis attacks in 12 cases and independently in 8 cases. One of the 2 patients with IgA myeloma had urticarial vasculitis with IgA skin vessel deposits. Direct immunofluorescence performed on skin biopsy specimens from patients with myelodysplastic syndrome with leukocytoclastic vasculitis (n = 7) was either negative (n = 3) or positive with vascular deposits of C3 (n = 4), IgG (n = 1), and/or IgM (n = 1).TABLE 4: Dermatologic manifestations and histologic findings in 22 patients with relapsing polychondritis and myelodysplastic syndromeDermatologic manifestations that may be related to RP in patients with other diseases are reported in Table 5. Among patients with systemic lupus erythematosus, 4 had acute cutaneous lupus erythematosus, 2 had subacute cutaneous lupus erythematosus, and 1 had discoid lupus erythematosus. All patients with mixed connective tissue disease had acrosclerosis. Cutaneous lesions of dermatomyositis were obvious in patients with this diagnosis.TABLE 5: Dermatologic manifestations that may be related to relapsing polychondritis in 51 patients with relapsing polychondritis and other diseases (except myelodysplastic syndrome)*Eight patients had autoimmune thyroid diseases. One of them also had myelodysplasia with several dermatologic manifestations: livedo reticularis, nodules, sterile pustules, nonurticarial papules, superficial phlebitis, and distal necrosis. Four skin biopsy specimens of different lesions showed dermal neutrophil infiltrates, leukocytoclastic vasculitis, lymphocytic vasculitis, or deep vessel thromboses. Among the other 7 patients, 3 had dermatologic manifestations: purpura (n = 2), nodules (n = 1), superficial phlebitis (n = 1), and limb ulcerations (n = 1). Nine patients had associated chronic dermatitis. One also had dermatomyositis, and another had aseptic visceral abscesses. In only 1 of the 2 patients with lichen planus, erythema nodosum-like lesions of the legs could be associated with RP and occurred concomitantly with attacks. Both patients with histologically proven Crohn disease presented with dermatologic manifestations, which could be attributed to either Crohn disease or RP. One patient had oral aphthosis; the other had purpura, urticarial vasculitis, and superficial phlebitis in association with myelodysplasia. The 4 patients with deep aseptic abscesses had dermatologic manifestations, including erythema nodosum-like lesions (n = 3), aphthosis (complex: n = 1; oral: n = 1), and aseptic skin abscesses with subsequent ulceration (n = 2). The frequency of dermatologic manifestations (91% versus 35.4%; p < 0.0001), sex ratio (18 male/4 female versus 44 male/83 female; p < 0.0001), and age at the time of the first chondritis (63.3 ± 14 yr versus 41.4 ± 17 yr; p < 0.0002) were significantly higher in the 22 patients with myelodysplasia than in the 127 patients without any associated disease. Discussion Dermatologic manifestations have been reported frequently in various series of patients with RP, but without details of the clinical features and corresponding pathologic findings. The frequency of skin lesions ranged from 17% to 36% in the largest series reported in the literature (23,25,32,34), but adjacent mucosal involvement was not specified. In contrast, dermatologic manifestations have been described accurately in isolated case reports (4,8,30). The lack of specificity of the observed dermatologic manifestations and the high frequency of associated diseases potentially featuring similar skin or mucosal involvement complicated the analysis of dermatologic manifestations in this series. All patients with such overlapping disorders, with the exception of Sjögren syndrome which was not systematically searched for, were analyzed separately. This approach meant that a minimum frequency of dermatologic manifestations, that is, 35.5% in the group of 127 patients with “pure” RP, was reported. Indeed, the frequency of dermatologic manifestations was extremely high in some groups of patients, especially those with RP-associated myelodysplasia or Crohn disease (91% and 2 of 2, respectively), confirming our preliminary report (28). These frequencies were much higher than those usually observed when these disorders are not associated with RP, suggesting that RP plays its own role in their occurrence. Cutaneous involvement in unselected patients with myelodysplasia can take the form of either a malignant infiltration or various benign lesions. Both types of lesions were observed in 14 (7.8%) patients in a recent series of 180 patients, with the latter present in 12 (2). Besides the cutaneous infections common in neutropenic patients, benign lesions included cutaneous vasculitis and various neutrophilic dermatoses (2). The frequency of dermatologic manifestations in patients with Crohn disease has been reported to be as high as 34%. These manifestations include granulomatous-specific lesions, reactive lesions, other miscellaneous disorders, and malnutrition-associated lesions (6,15). The most common reactive lesions are similar to those observed in RP, that is, aphthosis, neutrophilic dermatosis, vasculitis, and urticaria. Oral aphthosis was the most common dermatologic manifestation seen in our series. Aphthae in RP did not differ from those of simple recurrent aphthosis. Oral aphthae were associated with genital aphthae in 12 cases: 7 patients without associated disease (see Table 2), 2 patients with myelodysplastic syndrome (see Table 4), and 3 patients with other diseases (see Table 5). Complex aphthosis is not synonymous with Behçet disease; it has also been reported in patients with inflammatory bowel diseases (6) and chronic myeloid leukemia patients treated with interferon-alpha (5). Due to the lack of specificity of complex aphthosis, patients with oral and genital aphthae were not considered to have the association of Behçet disease and RP, termed “MAGIC syndrome” (Mouth And Genital ulcers with Inflamed Cartilage) by Firestein et al in 1985 (10). Indeed, patients with and without complex aphthosis did not differ with regard to other dermatologic (see Table 3) or systemic manifestations of RP. For example, ocular manifestations, present in 9 of 12 (75%) patients with complex aphthosis, were similar to those observed in RP (19) : scleritis (n = 5), conjunctivitis (n = 5), and/or keratitis (n = 3). One patient with associated spondylarthropathy had iridocyclitis. No patients suffered from retinal vasculitis, posterior uveitis, or anterior uveitis with hypopyon, which are common in patients with Behçet disease (3). Nodules on the limbs were the most common skin lesions; they have been described frequently as erythema nodosum-like lesions (30). Indeed, septal panniculitis was noted in many cases. However, other pathologic findings were also observed, such as thrombosis, vasculitis (29), and deep neutrophil infiltrates. Nodules corresponding to the latter were always associated with other neutrophilic dermatoses (see Tables 2, 4), such as Sweet syndrome (7 cases), pyoderma gangrenosum (1 case), and erythema elevatum diutinum (2 cases). These neutrophilic dermatoses may occur as diseases confined to the skin (idiopathic), but they also may be associated with hematologic disorders and arthritis. All these diverse neutrophilic dermatoses have been reported previously in isolated cases of patients with RP (19,21). Sweet syndrome-like lesions and pyoderma gangrenosum may be associated also with inflammatory bowel diseases (6) or Behçet disease (14). Pyoderma gangrenosum-like lesions have been observed also in patients with systemic vasculitis, such as Wegener granulomatosis or Takayasu arteritis (11,12). Four of our patients with RP had deep splenic or mesenteric abscesses associated with various dermatologic manifestations, that is, pseudofolliculitis, cutaneous abscesses, aphthosis, erythema nodosum-like lesions; they preceded the onset of chondritis, which provided the diagnosis of RP. In these patients, no inflammatory bowel disease was detected by endoscopy, unlike the majority of patients with aseptic visceral abscesses reported in our group (1). Palpable purpuric and urticarial lesions corresponding to leukocytoclastic vasculitis or necrotizing venulitis may be drug-induced or encountered in many diseases, such as infections, connective tissue diseases, systemic vasculitis, inflammatory bowel diseases, and hematologic and other malignant disorders. Only 1 patient with IgA myeloma had IgA skin vessel deposits excluding in others the diagnosis of IgA-mediated vasculitis. No correlation was found between skin vasculitis and the presence of antineutrophil cytoplasm antibodies previously detected in 24% of RP patients (26). The importance of vasculitis in the pathogenesis of RP has been emphasized by several authors, especially with regard to its dermatologic manifestations (17,23,24). Our data confirm that a vasculitic process is indeed frequently present, but other aspects also may be discovered by histologic examination. Among them, thrombosis of skin vessels was observed in 4 patients without associated disorders and in 3 patients with myelodysplasia. The mechanism of thrombosis remains unclear. These patients did not have antiphospholipid antibodies (lupus anticoagulant or anticardiolipin antibodies), monoclonal cryoglobulin, or cryofibrinogen. Extensive investigations to detect other thrombophilic abnormalities were not performed in the oldest cases. Antiphospholipid antibodies were mostly detected when RP overlapped with another disorder, such as systemic lupus erythematosus in women and myelodysplasia in older men (27). The lack of specificity of the diverse dermatologic manifestations frequently encountered in patients with RP explains why these manifestations were not included in the diagnostic criteria proposed by Michet et al in 1986 (25). Nonetheless, their development coincided with chondritis in half the cases. Furthermore, they may be observed without any associated disorder and therefore are probably directly related to RP. Other dermatologic manifestations, such as psoriasis (7,9,16), Reiter syndrome (31), lichen planus, alopecia areata (20), or vitiligo, probably result from associated immune disturbances. In conclusion, dermatologic manifestations of RP are frequent, especially in association with myelodysplasia. They are nonspecific, similar to those observed in Behçet and inflammatory bowel diseases. Their presence in the elderly warrants repeated blood cell counts to detect a smouldering myelodysplasia. Summary Dermatologic manifestations of relapsing polychondritis (RP) have been relatively poorly studied compared to other manifestations. In this study we describe dermatologic manifestations in a large series of patients with RP and the corresponding pathologic findings. In this retrospective, single-center review of 200 patients diagnosed with RP according to Michet’s criteria, we analyzed separately those suffering from associated diseases with potential dermatologic involvement or chronic dermatitis. Skin or mucosal biopsies taken from 59 patients were examined without knowledge of the clinical data. Among the 200 patients with RP, 73 had chronic dermatitis or associated diseases with potential dermatologic involvement, especially hematologic disorders (n = 24) and connective tissue diseases (n = 22). Among the other 127 patients, 45 (35.4%) had dermatologic manifestations: aphthosis (n = 21; oral in 14 and complex in 7), nodules on the limbs (n = 19), purpura (n = 13), papules (n = 10), sterile pustules (n = 9), superficial phlebitis (n = 8), livedo reticularis (n = 7), ulcerations on the limbs (n = 6), and distal necrosis (n = 4). Dermatologic manifestations were the presenting feature of RP in 15 cases (12%), and appeared concomitantly (n = 23) or not (n = 22) with attacks of chondritis. Histologic findings included vasculitis (n = 19, leukocytoclastic in 17 and lymphocytic in 2), neutrophil infiltrates (n = 6), thrombosis of skin vessels (n = 4), septal panniculitis (n = 3), and minor changes (n = 2). Patients with and without dermatologic manifestations did not differ with regard to male/female ratio; age at RP onset; frequency of auricular, nasal, or tracheobronchial chondritis; or frequency of rheumatologic, ocular, audiovestibular, renal, arterial, or venous involvement. The frequency of dermatologic manifestations (91% versus 35.4%; p < 0.0001), sex ratio (18 male/4 female versus 44 male/83 female, p < 0.0001), and age at first chondritis (63.3 ± 14 yr versus 41.4 ± 17 yr; p < 0.0002) were significantly higher in the 22 patients with myelodysplastic syndrome than in the 127 patients without any associated disease. In conclusion, although dermatologic manifestations occur frequently in patients with RP, especially in association with myelodysplasia, they are nonspecific and sometimes resemble those observed in Behçet disease or inflammatory bowel diseases. Their presence in the elderly warrants repeated blood cell counts to detect a smouldering myelodysplasia.