Rachel Miest was a student at Mayo Medical School, College of Medicine, Mayo Clinic, Rochester, MN at the time the manuscript was initially written.
Nneka I. Comfere, md Department of Dermatology, Mayo Clinic, 200 First St SW, Rochester, MN 55905, USA E-mail: email@example.com
Background Polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes syndrome is a rare multisystem paraneoplastic condition associated with plasma cell dyscrasia.
Methods From our institution’s dysproteinemia database, 107 patients met criteria for polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes syndrome between January 1, 2000, and October 1, 2009. Medical records were reviewed for documented syndrome features at diagnosis. We assessed prevalence of skin findings and associations between dermatologic and other characteristic disease findings.
Results Of the 107 patients, 96 (90%) had a recognized cutaneous manifestation. Hyperpigmentation and hemangioma were most common (47%), followed by hypertrichosis (38%). Vascular skin changes – acrocyanosis (34%), Raynaud phenomenon (20%), hyperemia/erythema (20%), flushing (16%), or rubor (11%) – occurred in 62%; white nails, sclerodermoid changes, and clubbing occurred in 30%, 26%, and 6%, respectively. Mean number of skin findings per patient was 2.9 (median, 3.0; range, 0–7). Presence of cutaneous manifestation was associated with abnormal pulmonary function tests (P < 0.001); immunoglobulin G gammopathy was associated with hyperpigmentation and hypertrichosis. No other significant associations were seen.
Conclusions The high prevalence of skin findings (90%) shows the value of dermatologic evaluation in diagnosis of polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes syndrome. Our data indicate new associations between skin findings and other disease characteristics.
Polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes (POEMS) syndrome is a rare multisystem paraneoplastic condition associated with plasma cell dyscrasia.1,2 Other names for the syndrome include osteosclerotic myeloma, Crow–Fukase syndrome, and Takatsuki syndrome.3,4 All persons with POEMS syndrome have polyneuropathy and a monoclonal plasma cell proliferative disorder. In addition, a diagnosis of POEMS syndrome requires one other major criterion and one minor criterion. Major criteria of POEMS syndrome include osteosclerotic bone lesions, Castleman disease, and elevated vascular endothelial growth factor (VEGF) levels. Minor criteria include endocrinopathy, extravascular volume overload, papilledema, organomegaly (e.g., lymphadenopathy), thrombocytosis, polycythemia, and skin changes.2 Pulmonary hypertension and restrictive lung disease have also been associated with POEMS syndrome.2,5,6 Previously identified dermatologic findings linked with the syndrome include hyperpigmentation, hemangiomas, flushing, white nails, clubbing, dependent rubor, skin thickening, acrocyanosis, and hypertrichosis.2,7,8
The pathogenesis of POEMS syndrome is incompletely understood. However, increased levels of proinflammatory and proangiogenic cytokines, including interleukin (IL)-1β, tumor necrosis factor α, IL-6, and VEGF, have been implicated. VEGF, a multifunctional cytokine that induces angiogenesis and increases microvascular permeability, has emerged as an important component in the syndrome’s pathogenesis. Bone marrow stromal cells, plasma cells, and platelets secrete VEGF. Serum VEGF levels are often elevated in affected patients, correlate with disease activity, and decrease with successful therapy.9–14
Although POEMS syndrome is well described in the medical literature, 1–4,15,16 most information regarding its skin findings is from isolated case reports or small case series.17–20 To take advantage of our institution’s experience with a large series of patients, we sought to determine the prevalence and characteristics of cutaneous manifestations on diagnosis of POEMS syndrome and to identify correlations between these skin findings and other clinical manifestations of the disease.
Patients and methods
We identified 107 patients within the Mayo Clinic dysproteinemia database who met appropriate criteria for POEMS syndrome. Given the variability in reporting its cutaneous manifestations and lack of serum VEGF and IL-6 testing before the year 2000, we elected to conduct our investigation on patients seen between January 1, 2000, and October 1, 2009. The majority of patients (76%) were seen by a sole hematologist with expertise in the evaluation and management of patients with POEMS syndrome (A.D.).
Using a standardized data extraction form, we obtained information from the medical records of these patients regarding cutaneous manifestations and other manifestations of POEMS syndrome at their presentation to Mayo Clinic. Vascular skin changes were assessed individually and as a group, given that they occur along a continuum with varying degrees of severity but have common etiologic features. VEGF and IL-6 levels were recorded when available. To qualify as a baseline measurement, VEGF and IL-6 levels had to be obtained before therapy with lenalidomide, thalidomide, bortezomib, cyclophosphamide, melphalan, irradiation, or transplantation. Patients receiving any of these therapies before VEGF and IL-6 measurements were excluded from subsequent analysis. VEGF and IL-6 levels obtained after therapy with mycophenolate mofetil, cyclosporine, intravenous immunoglobulin, plasma exchange, and rituximab were included in the analysis. In addition, given the large number of patients who had received corticosteroids for a suspected inflammatory condition before diagnosis, VEGF and IL-6 levels obtained after therapy with corticosteroids were included. Potential associations between dermatologic findings and other characteristic features of the disease were assessed. Associations between features were evaluated with χ2 or Fisher exact test. All tests were two-sided, and P values <0.05 were considered statistically significant.
The Mayo Clinic Institutional Review Board approved the study in accordance with Minnesota state law that regulates research from medical records.
Of patients with POEMS syndrome, 68 were male and 39 were female (mean age at initial diagnosis, 50 years [median, 49 years; interquartile range, 42–58 years; range, 19–82 years]). In total, 96 patients (90%) had a recognized cutaneous manifestation (Table 1, Fig. 1). Hyperpigmentation and hemangioma were the most common cutaneous manifestations, followed in frequency by hypertrichosis. Vascular skin changes presented as acrocyanosis, Raynaud phenomenon, hyperemia/erythema, flushing, or rubor, with one or more of these vascular skin changes noted in 62% of patients. White nails, sclerodermoid changes, and clubbing occurred in 30%, 26%, and 6%, respectively. The mean number of skin findings among the 107 patients was 2.9 (median, 3; range, 0–7) (Fig. 1). A total of 77 different combinations of presenting skin changes, without identifiable groupings, were reported.
Table 1. Prevalence of cutaneous manifestations in POEMS syndrome
Patients, no. (%) (n = 107)
POEMS, polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes.
Prevalence of cutaneous manifestations
Hyperpigmentation was the most common cutaneous manifestation, reported in 50 patients (47%). Twenty-four patients (48%) had hyperpigmentation of the extremities; other areas of hyperpigmentation included generalized (nine patients; 18%), the torso (eight patients; 16%), areola complex (six patients; 12%), and head or neck (four patients; 8%). Eight patients (16%) had hyperpigmentation in an unspecified area.
Along with hyperpigmentation, hemangioma was the most common cutaneous manifestation in our series of patients. Fifty patients (47%) had at least one hemangioma, and 46 of these patients had two or more hemangiomas. The hemangiomas were biopsied in 10 patients, with the biopsies of four patients showing histological evidence of glomeruloid hemangioma. Other biopsy-based diagnoses included angioma, hemangioma, capillary hemangioma, benign fibrous histiocytoma, one descriptive diagnosis, and one unavailable diagnosis. Clinical terms applied to identified vascular proliferations (Fig. 2) included cherry hemangioma, angioma, and hemangioma.
Hypertrichosis was recorded as an increase in amount or coarseness of hair (Fig. 3). In the cohort, 41 patients (38%) had reported hypertrichosis. Among these patients, 28 (68%) had hypertrichosis of the extremities. Other areas of hypertrichosis included the trunk (10 patients; 24%), head or neck (five patients; 12%), and generalized locations (one patient; 2%). In addition, 7 of the 41 patients (17%) had hypertrichosis in an unspecified area.
Thirty-six patients (34%) presented with acrocyanosis. Extremities were often described as having a blue or “dusky” appearance (Fig. 3).
In 32 patients (30%), white nails were described as diffuse, opaque whitish discoloration involving most of the fingernail plate (Figs. 4 and 5).
Sclerodermoid changes occurred in 28 patients (26%). These changes were described as skin thickening. In 14 (50%) of these patients, the location was not specified. Areas specified included extremities (11 patients; 39%), trunk (two patients; 7%), and generalized (one patient; 4%).
For 21 patients (20%), Raynaud phenomenon was noted or symptoms were described as classic for Raynaud phenomenon.
Hyperemia/erythema was described in 21 patients (20%), with 14 (67%) patients having affected hands or feet. Other affected areas included the face and trunk.
Described in 17 patients (16%), flushing typically involved predominantly the head and neck regions.
Twelve patients (11%) had rubor. Of these patients, six (50%) had reported dependent rubor of the feet, five (42%) had unspecified rubor, and one (8%) had rubor involving the face.
Clubbing of the fingernails was reported in six patients (6%).
Vascular endothelial growth factor levels
The mean baseline serum VEGF level documented for 40 patients meeting inclusion criteria was 702 pg/ml (median, 417 pg/ml; range, 48–4591 pg/ml); three patients had a level in the reference range (31–86 pg/ml); and 37 patients had an abnormally high level (>86 pg/ml). Given our strict inclusion criteria for baseline VEGF and the few patients with baseline VEGF levels in the reference range, associations with this feature were limited because of low statistical power. Interestingly, when looking at all available VEGF levels for patients in our series (n = 61) irrespective of inclusion criteria, we found that 34 of 36 patients with vascular skin changes (i.e., acrocyanosis, Raynaud phenomenon, hyperemia/erythema, flushing, or rubor) had abnormally high VEGF levels compared with 19 of 25 patients without vascular skin changes (94% vs. 76%; P = 0.054).
The mean baseline IL-6 level documented for 27 patients was 11.80 pg/ml (median, 4.12 pg/ml; range, 0.92–140); 14 patients had a value in the reference range (0.31–5 pg/ml), and 13 patients had an abnormally high value (> 5 pg/ml). No statistically significant associations were seen between baseline IL-6 levels and cutaneous manifestations of POEMS syndrome.
Other disease characteristics
Given that POEMS syndrome affects multiple systems, we sought to determine whether any cutaneous manifestations were associated with other findings characteristic of the disease (Table 2). Several statistically significant associations were seen (Table 3). Notably, abnormal pulmonary function test (PFT) results were associated with the presence of any cutaneous manifestation (P < 0.001), white nails (P < 0.001), hypertrichosis (P = 0.002), and vascular skin changes (P = 0.001), including acrocyanosis (P = 0.04) and hyperemia/erythema (P = 0.045). In addition, patients with an immunoglobulin G gammopathy were more likely to have hyperpigmentation (P = 0.02) or hypertrichosis (P = 0.03). A dermatologist examination as part of the diagnostic evaluation and presence of hemangiomas were also significantly associated (P = 0.007). No significant associations were seen between cutaneous manifestations of POEMS and presence of peripheral neuropathy or endocrinopathy.
Table 2. Characteristic findings in POEMS syndrome
Patients, no. (%) (n = 107)
IgA, immunoglobulin A; IgG, immunoglobulin G; IgM, immunoglobulin M; PFT, pulmonary function test; POEMS, polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes.
aOf the patients without polyneuropathy on presentation, two patients with biopsy-proven Castleman disease received a diagnosis of “atypical” POEMS syndrome. Polyneuropathy is considered a major criterion; however, given the clinical presentation of these patients we included them in our data analysis.
Sclerotic bone lesions
Table 3. Cutaneous manifestations of POEMS syndrome and associated findings
Presence of manifestations, no. of patients (%)
AA, both cutaneous manifestation and associated finding absent; AP, cutaneous manifestation absent and associated finding present; IgG, immunoglobulin G; PA, cutaneous manifestation present and associated finding absent; PFT, pulmonary function test; POEMS, polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes; PP, both cutaneous manifestation and associated finding present.
aDefined as the presence of at least one of the following symptoms or signs: acrocyanosis, Raynaud phenomenon, hyperemia/erythema, flushing, or rubor.
Presence of any cutaneous manifestation
Abnormal PFT results
Abnormal PFT results
Vascular skin changesa
Abnormal PFT results
Abnormal PFT results
Abnormal PFT results
Abnormal PFT results
Response to treatment
Some patients had a cutaneous response to treatment, with improvement of hemangiomas (Fig. 6), white nails, sclerodermoid changes, hyperpigmentation, hypertrichosis, and vascular skin changes.
Skin findings in POEMS syndrome were common, with a prevalence of 90% in our series of 107 patients. These manifestations included hyperpigmentation, hypertrichosis, hemangioma, and vascular skin changes presenting as acrocyanosis, Raynaud phenomenon, hyperemia/erythema, or rubor. Other manifestations included sclerodermoid changes, white nails, and clubbing. We found multiple associations between skin findings and other characteristic findings of POEMS syndrome. Although skin changes and abnormal PFTs were both very prevalent in the patient population, there appeared to be a particularly strong association between abnormal PFTs and vascular skin changes, hypertrichosis, and white nails. We also identified an association between immunoglobulin G gammopathy and hyperpigmentation or hypertrichosis.
Causes of cutaneous manifestations
The mechanisms underlying POEMS syndrome pathogenesis and their roles in the development of cutaneous manifestations have yet to be elucidated. Serum levels of VEGF and proinflammatory cytokines such as IL-6 are known to be elevated in patients with POEMS syndrome and have been shown to serve as biomarkers correlating with disease activity.9,10,12–14,21–25 In addition, the VEGF level is obtained to distinguish the peripheral neuropathy in POEMS syndrome from chronic inflammatory demyelinating polyneuropathy, peripheral neuropathy associated with monoclonal gammopathy of undetermined significance, and amyloidosis.24,26,27 Previous reports have suggested a causative relationship between VEGF levels and the skin changes of POEMS syndrome.2,9,14,28 For example, in a recent case series of 21 patients with POEMS syndrome, Barete et al.20 demonstrated increased serum VEGF levels in patients with hypertrichosis (P = 0.04) or more than 10 hemangiomas but not in patients with <10 (P = 0.06) or no hemangiomas (P = 0.08), lending support to the theory that VEGF may have a role in these skin manifestations. However, although patients with any cutaneous manifestation had higher VEGF levels than patients without skin involvement, the difference was not significant (P = 0.45), thus complicating the relationship.
VEGF is a logical candidate for contributing to the skin changes of POEMS syndrome. VEGF has known angiogenic properties and has been shown to be expressed with its receptor VEGFR-1 in glomeruloid hemangiomas, which is suggestive of a potential role in the development of the distinct lesions seen in POEMS syndrome.29–31 It has also been suggested that an elevated VEGF level may lead to hypertrichosis, either by acting directly on dermal papilla cells or by stimulating the local vasculature.32 The expression of VEGF has been found to be significantly increased in melasma lesions, which may be suggestive of a mechanistic role in the development of hyperpigmentation.33 Patients with systemic sclerosis have been reported to have significantly higher serum VEGF levels than healthy control subjects, suggesting a possible relationship between VEGF and the sclerodermoid changes seen in patients with POEMS syndrome.34 Finally, VEGF, along with platelet-derived growth factor, may also induce the stromal and vascular changes present in digital clubbing.35
A major variable complicating an association between VEGF and skin changes in POEMS syndrome is the effect of previous therapy on VEGF levels, making it difficult to determine a true VEGF baseline at diagnosis. It is well documented that VEGF levels are sensitive to many therapies, including corticosteroids and antineoplastic therapies. Because of these inclusion requirements, only 40 patients met the criteria for baseline VEGF levels that could be included in our study, which limited the power of our statistical analysis. First, this illustrates the difficulty in obtaining true baseline VEGF levels at diagnosis. Second, when inclusion criteria were removed and all patients were analyzed, we found a nonsignificant trend (P = 0.054) between vascular skin changes and abnormally high levels of VEGF similar to previous studies.20 However, the majority of these patients had received therapies known to affect VEGF levels at outside institutions before their first visit to Mayo Clinic. Previous studies frequently define baseline VEGF levels as those drawn at the time of diagnosis. However, therapy received before diagnosis is often not well defined. Future studies are needed as serum VEGF becomes a standard measurement in patients with POEMS syndrome to better correlate baseline VEGF levels with clinical findings before therapy is initiated. VEGF has complex actions on multiple systems, making it an important candidate, likely in combination with other factors, potentially underlying many of the hallmark signs and symptoms of POEMS syndrome.6,36
Of interest, we identified a strong association between the cutaneous manifestations of POEMS syndrome and abnormal PFTs (P < 0.001). The most common pattern of PFT abnormalities identified in POEMS syndrome is a restrictive pattern associated with diminished diffusion capacity of the lung for carbon monoxide.6 Given the nature of the observed association between cutaneous manifestations of POEMS syndrome and abnormal PFT results (P < 0.001), it is possible to speculate that the presence of specific cutaneous manifestations should raise concern for potential respiratory involvement and prompt additional diagnostic studies, including PFTs and imaging studies. Although speculation about the exact nature of these associations is difficult, they may serve as important observations that drive further investigation and clarification of the underlying disease mechanisms manifested in patients with POEMS syndrome.
As with VEGF levels, increased statistical power with larger patient numbers in future studies may well yield a significant association with these manifestations and provide further insight into the pathogenesis of POEMS syndrome.
Finally, our observation of a cutaneous response to treatment, with improvement of hemangiomas (Fig. 6), white nails, sclerodermoid changes, hyperpigmentation, hypertrichosis, and vascular skin changes, provides further evidence that common etiologic factors may underlie many of the characteristic findings of POEMS syndrome.
In this retrospective review of medical records, we encountered limitations common to this type of study – notably, limits to data reporting and interpretation. However, most patients (76%) were seen by a single hematologist with expertise in POEMS syndrome (A.D.). In addition, although many of the cutaneous manifestations of POEMS syndrome may be seen in the general population, studies have demonstrated that an abrupt or eruptive onset of skin manifestations is typical in patients with POEMS syndrome.2,15 Resolution or stabilization of cutaneous lesions or changes occurs after appropriate POEMS syndrome therapy,16,20 increasing the likelihood that cutaneous manifestations specific for POEMS syndrome were documented in our study. Nevertheless, identification of a statistically significant association between the presence of hemangiomas and patients seen by a dermatologist as part of their diagnostic work-up (P = 0.007) lends evidence to the importance of multispecialty evaluation, including dermatologic care, for patients with POEMS syndrome. As physicians become more aware of the constellation of cutaneous findings associated with POEMS syndrome, more frequent and detailed dermatologic evaluations will help to more precisely define the skin symptoms and pathologic characteristics of this disease.
This retrospective review, to our knowledge, is one of the largest published series of POEMS syndrome cases looking specifically at the cutaneous manifestations of the disease. We identified novel and potentially important associations between skin findings and other aspects of the disease (i.e., type of immunoglobulin heavy chain contributing to the monoclonal gammopathy and abnormal PFT results). As the importance of skin findings in POEMS syndrome and value of dermatologic evaluation in the diagnosis of these patients become more widely appreciated, future studies will be empowered to better define pertinent associations between dermatologic manifestation and other findings of the disease, ultimately leading physicians closer to a mechanistic understanding of this complex, multisystem disease.