Interleukin 4 and interferon-gamma expression of the dermal infiltrate in patients with erythroderma and mycosis fungoides. An immuno-histochemical study

BACKGROUND: Erythroderma, or generalized erythema of the skin, may result from different causes. At present it is unclear whether the underlying patho-mechanisms that lead to erythroderma are identical or different depending on the original disease. The aim of this study was to investigate the dermal cytokine profile in different types of erythroderma and mycosis fungoides. METHODS: Snap-frozen skin biopsy specimens from 33 patients with erythroderma were studied. Thirteen had idiopathic erythroderma, 7 erythrodermic atopic dermatitis, 5 Sézary syndrome and 8 had erythroderma from miscellaneous causes. We also studied 6 patients with mycosis fungoides (5 plaques and 1 tumor) and 5 healthy non-atopic volunteers. The biopsies were immunohistochemically stained for interleukin 4 (IL-4) and interferon gamma (IFN-gamma). All positive cells for IL-4 and IFN-gamma in the dermis were counted and the number of positive cells was calculated per mm2. IL-4/IFN-gamma ratio was calculated for each biopsy. RESULTS: The patients with idiopathic erythroderma, atopic dermatitis and miscellaneous erythroderma, all showed more IFN-gamma-positive cells than IL-4-positive cells in the dermis. The median IL-4/ IFN-gamma ratio for these three groups was 0.6, 0.9 and 0.45, respectively. These differences were not statistically significant. All patients with Sézary syndrome however, showed more IL-4-positive cells than IFN-gamma-positive cells. The median IL-4/IFN-gamma ratio was 1.8, which is significantly higher than in the other groups p<0.05). In mycosis fungoides roughly the same number of cells expressed IL-4 and IFN-gamma. The median IL-4/IFN-gamma ratio was 1.0, which is significantly lower than in Sézary syndrome (p<0.05). CONCLUSIONS: The dermal infiltrate in patients with Sezary syndrome mainly shows a T-helper 2 (Th2) cytokine profile, this in contrast to T-helper 1 (Th1) cytokine profile in benign reactive erythroderma. This indicates that although a relative uniform clinical picture of erythroderma is obvious, a different patho-mechanisms may be underlying.     

Histopathologic studies in Sézary syndrome and erythrodermic mycosis fungoides: a comparison with benign forms of erythroderma

Histologic sections from eleven patients with Sézary syndrome were reviewed and compared with those of four patients with erythrodermic mycosis fungoides and twenty-four patients with a benign form of erythroderma, including fifteen patients with chronic dermatitis, four with a generalized drug eruption, and five with an erythrodermic psoriasis. The most important discriminating histologic feature in patients with Sézary syndrome was the presence of a monotonous bandlike or perivascular infiltrate in the papillary dermis, mainly composed of large cerebriform-mononuclear cells, as seen in seven of eleven Sézary syndrome patients. Pautrier's microabscesses were observed in seven of eleven Sézary syndrome patients, two of four patients with erythrodermic mycosis fungoides, but not in any of the patients with a benign form of erythroderma; their presence was therefore considered a reliable criterion in differentiating erythrodermic cutaneous T cell lymphoma from benign forms of erythroderma. However, features of chronic dermatitis were often found superimposed on those of Sézary syndrome and were even predominating in four of eleven Sézary syndrome patients. Moreover, four patients with a benign form of erythroderma showed a histologic picture suggestive of cutaneous T cell lymphoma. Therefore, in dubious cases repeated skin biopsies, additional investigations of lymph nodes and peripheral blood, and careful follow-up are mandatory for the achievement of a correct diagnosis.     

182例红皮病病因与临床特征分析

目的 探讨红皮病的病因与临床特征的相关性。方法 回顾性分析182例红皮病患者的临床资料。结果 男女之比为2.8 ∶ 1;发病年龄（58.6 ± 14.6）岁。继发于原有皮肤病135例（74.2%）,药物过敏14例（7.7%）,恶性肿瘤8例（4.4%）,原因不明25例（13.7%）。红皮病诱发原因中系统用药诱发比例最高,共52例（28.6%）,其中最常见的药物为糖皮质激素。随访76例,58例痊愈后未再发红皮病,再发患者14例,死亡5例,2例不明原因红皮病患者经过多次组织病理检查确诊为蕈样肉芽肿。结论 继发于原有皮肤病是红皮病的最常见原因。原因不明红皮病患者易复发,有伴随恶性肿瘤可能,应密切随访。     

PUVA treatment of erythrodermic and plaque-type mycosis fungoides. Ten-year follow-up study

Since our preliminary report of psoralen plus long-wave ultraviolet A (PUVA) therapy in ten patients with erythroderma-type or plaque-type mycosis fungoides (MF), we have treated 38 patients with biopsy-proved MF. Approximately one third, mostly patients with erythroderma, received PUVA as primary therapy; the remainder had recurrent disease following electron beam irradiation or topical mechlorethamine (Mustargen) hydrochloride. Follow-up data are presented in 29 patients who completed an initial course of PUVA given two to three times weekly. A complete clinical response was observed in ten patients with plaque-type MF and seven with erythroderma without Sézary syndrome. The PUVA therapy was palliative for patients with advanced disease, in combination with other therapies. The mean observation period was approximately five years. Despite maintenance PUVA, most patients relapsed between ten and twenty months and were treated with another intensive course. Long-term maintenance therapy with PUVA was necessary to control the disease.     

Sézary syndrome without erythroderma: A case report and review of published work

Sézary syndrome (SS) is defined by erythroderma and circulating atypical T cells, with or without lymphadenopathy. Recently, Thompson et al. identified a distinct population of SS patients with an atypical presentation: a high blood tumor burden of Sézary cells fulfilling criteria for SS but without fulfilling the criteria for erythroderma at the diagnosis. Here, we report a case of a 49‐year‐old Japanese man with SS who did not present with erythroderma initially, but exhibited erythematous itchy papules symmetrically located on the legs and arms. We also reviewed reported cases of SS without initial erythroderma. The skin manifestations at diagnosis varied from patches to tumors often seen in mycosis fungoides, and other rarer findings such as excoriation, palmoplantar keratoderma and alopecia. Pruritus was reported in most patients (86%), unlike early mycosis fungoides, and could be the main clue to the diagnosis of SS. Notably, three patients were reported to have presented with papular lesions, similar to our case. Little is known about why skin lesions in SS without erythroderma vary and why these cases did not exhibit erythroderma initially. Attenuated stimulation by colonized Staphylococcus aureus, impairment in recruitment of malignant T cells and suppression of inflammatory response induced by malignant T cells with regulatory phenotype may be associated with skin manifestations. Further studies are necessary to elucidate the etiology of this entity.     

Heteroduplex analysis of T-cell receptor γ gene rearrangement as an adjuvant diagnostic tool in skin biopsies for erythroderma

BACKGROUND: Erythroderma, defined as red skin covering most of the body surface often accompanied or followed by exfoliation, is the clinical manifestation of at least six different underlying etiologies with allergic or irritant contact dermatitis, atopic/asteotic dermatitis, pityriasis rubra pilaris (PRP), psoriasis, and seborrheic dermatitis accounting for the majority of cases. Approximately 10% of cases are due to adverse drug reactions with roughly another 10% due to cutaneous T-cell lymphoma (CTCL), predominantly mycosis fungoides, or leukemia. It is clear from multiple studies that the clinical diagnosis of the underlying entity is often difficult, as these diseases can present in a very similar fashion. A skin biopsy is usually employed in this setting as a diagnostic tool. However, the histopathologic diagnosis of the underlying cause is complicated by the subtlety of the distinguishing histologic features. In this situation, an ancillary technique demonstrating the presence of a monoclonal T-cell proliferation could help to rule in or out CTCL in cases that clinically and histopathologically do not allow a definitive diagnosis. METHODS: We retrospectively studied 25 biopsies from sixteen patients who presented to the Stanford Dermatology Clinic with erythroderma. We examined the specimens morphologically and analyzed the gamma chain of the T-cell receptor (TCR- gamma) by polymerase chain reaction (PCR) followed by heteroduplex analysis for clonality. We then correlated the results of our PCR and heteroduplex analyses with the patients' clinical outcomes. RESULTS: Four biopsies, from three patients, contained clonal TCR-gamma rearrangements; the four biopsies, all of which were equivocal histologically, correlated to diagnoses of mycosis fungoides (MF) or Sézary syndrome (SS). Twenty-one biopsies contained polyclonal T-cell populations. Eighteen of these biopsies represent patients with inflammatory dermatoses. Three of these biopsies, all of which were taken from a single patient, correlate to a diagnosis of MF. CONCLUSION: TCR-gamma PCR heteroduplex analysis seems to represent an important adjuvant diagnostic tool that, used in conjunction with histopathology and clinical history, could help to clarify the underlying etiology of erythroderma.     

Sézary Cell Count in Exfoliative Dermatitis

Sézary cells were present in a number of less than 1000/mm3 in 52 of 78 cases of various benign diffuse dermatoses and mycosis fungoides. The count was higher than 1000/mm3 in two cases of Sézary syndrome in which the histopathology was that of mycosis fungoides and in six cases of exfoliative erythroderma of undetermined etiology with the histopathology of nonspecific chronic dermatitis. One of these six cases developed palmoplantar keratoderma, but abnormal cells were not present in the tissue. The diagnosis of Sézary syndrome in these cases of exfoliative erythroderma could not be established, even though the Sézary cell count was more than 1000/mm3, because the abnormal cells were absent in the skin biopsy specimen. It is concluded that the number of Sézary cells present is not significant for the diagnosis of Sézary syndrome. Our cases with the diagnosis of Sézary syndrome are an erythrodermic variant of mycosis fungoides rather than a progressive conversion from exfoliative erythroderma.     

Cutaneous histopathology of Sézary syndrome: a study of 41 cases with a proven circulating T-cell clone

Sézary syndrome is an uncommon variant of cutaneous T-cell lymphoma (CTCL) characterized by erythroderma, pruritus, adenopathy, and circulating atypical T-lymphocytes with cerebriform nuclei. The definition of Sézary syndrome can be further refined by including only patients with a circulating peripheral blood population of clonal T-cells. We have evaluated 79 skin biopsies from such a group of 41 erythrodermic patients with circulating Sézary cells and a clonal population of T-cells detected by T-cell receptor-figene rearrangement on Southern analysis of peripheral blood mononuclear cells. Histopathologic features consistent with chronic dermatitis were observed in 26/79 (33%) skin biopsy specimens, emphasizing that a non-specific histologic appearance is common. Evidence of CTCL was lacking in 11/41 patients on biopsy of their erythrodermic skin. The survival of these patients was not significantly different from 30/41 patients in whom skin biopsies revealed changes diagnostic of CTCL, such as a dermal lymphocytic band with atypical lymphocytes (18/79, 23%) or a mycosis fungoides-like infiltrate (30/79, 38%). This study confirms that non-specific cutaneous hlstopathologic findings are common in Sézary syndrome, even when a circulating T-cell clone is present. This stresses the need for peripheral blood genetic analysis and for multiple or repeat skin biopsies in erythrodermic patients when there is high clinical suspicion of CTCL.     

Benign and malignant forms of erythroderma: cutaneous immunophenotypic characteristics

In order to determine if immunohistologic features are useful in distinguishing benign from malignant types of erythroderma, we studied the immunophenotype of lesional T cells in 20 patients (8 mycosis fungoides/Sézary syndrome, 12 benign) and found them to be generally similar. In all cases, the majority of T cells were Leu-1+, Leu-4+, and Leu-5+, as is typical of mature T cells. Although in most cases a majority of Leu-3+ (helper/phenotype) T cells were present, in 2 there was a majority of the Leu-2+ (cytotoxic/suppressor) subset and in 12 others, a significant minority (20%-40%) of these cells. Low percentages of Leu-2+ cells (less than or equal to 10%), resulting in high Leu-3+/Leu-2+ ratios, did not distinguish benign from malignant erythroderma. Leu-8 antigen deficiency was common in both mycosis fungoides/Sézary syndrome and benign cases (62% vs 75%, respectively). In contrast, Leu-9 antigen deficiency was present in only one patient in each group. The lack of combined Leu-8/9 antigen deficiency in our patients may be due to a heavy inflammatory T cell component, obscuring the antigen deficiencies seen in most nonerythrodermic mycosis fungoides cases. We conclude that immunophenotypic studies with the use of the current antibody panel show many similarities between benign and malignant forms of erythroderma, as well as some minor differences that may prove diagnostically useful if corroborated by future studies.     

Lymphomatoid papulosis. A follow-up study of 30 patients

Thirty patients, 13 female and 17 male, have been followed from 3 months to 22 years (mean, 81 months; median, 63 months) and special studies have been performed on a proportion of these in order to try to predict malignant evolution. Age at onset was from 20 to 70 years (mean, 43 years; median, 42 years). Duration of disease was from 1 to 30 years (mean, 119 months; median 161 months). Seven patients also had parapsoriasis en plaque or plaque-stage mycosis fungoides at the time of diagnosis and one patient had erythroderma. None of the 22 uncomplicated lymphomatoid papulosis patients developed malignant cutaneous lymphoma during the period of observation, while the remaining 8 patients who had concurrent parapsoriasis en plaque, mycosis fungoides, or erythroderma did not deteriorate further. Single-cell deoxyribonucleic acid (DNA) measurements on the dermal infiltrate were done in 13 patients and were abnormal in 7 patients. Two of these had greatly abnormal DNA histograms and at the same time an abnormal clinical presentation with multiple nodules and tumors. The remaining five patients had DNA histograms that indicated a potential for malignancy. Monoclonal antibody studies were performed on skin biopsy specimens of 10 patients. The dermal infiltrate was dominated by T-helper lymphocytes and some Hodgkin and Reed-Sternberg cells could be detected by the antibodies Ki-1, Ki-24, and Ki-27. Human T-lymphotropic virus type I (HTLV-I) antibodies were found in 3 of 18 patients examined. Treatment with psoralens plus ultraviolet A (PUVA) was effective (partial or complete remission) in six patients but they relapsed at cessation of therapy.(ABSTRACT TRUNCATED AT 250 WORDS)     

Expression of VCAM-1, ICAM-1, E-selectin, and P-selectin on endothelium in situ in patients with erythroderma, mycosis fungoides and atopic dermatitis

BACKGROUND: Erythroderma may result from different causes. At present it is unclear whether the patho-mechanisms that lead to these different types of erythroderma are identical or different. Adhesion molecules and their ligands play a major role in endothelial-leukocyte interactions, which affect the binding, transmigration and infiltration of lymphocytes and mononuclear cells during inflammation, injury, or immunological stimulation. The aim of this study was to investigate the adhesion molecule expression on endothelial cells in erythroderma in situ. METHODS: Snap-frozen skin biopsy specimens from 23 patients with erythroderma were studied. Eight had idiopathic erythroderma, 5 erythrodermic atopic dermatitis, 4 Sézary syndrome and 6 had erythroderma from miscellaneous causes. As a control we studied skin specimens from 10 patients with mycosis fungoides, 5 patients with atopic dermatitis and 5 healthy non-atopic volunteers. To determine adhesion molecule expression on endothelial cells in situ, sections were immuno-histochemically double stained with biotinylated Ulex Europaeus agglutinin 1 as a pan-endothelial cell marker, and for the adhesion molecules VCAM-1, ICAM-1, E-, and P-selectin. All double- and single-stained blood vessels in the dermis were counted. RESULTS: Mean endothelial expression in erythroderma was as follows: VCAM-1 51.4%, ICAM-1 70.1%, E-selectin 43.5%, and P-selectin 52.6%. There was no statistical difference between different groups of erythroderma. Mean expression of all adhesion molecules tested, was in Sézary syndrome higher than in mycosis fungoides albeit not significant. In erythrodermic atopic dermatitis only VCAM-1 expression was significantly higher than in lesional skin of atopic dermatitis. No differences were observed in expression of the other three adhesion molecules. CONCLUSIONS: There is no difference regarding adhesion molecule expression on endothelial cells between different types of erythroderma.     

Erythroderma: a clinico-etiologic study of 90 cases

Background The difficulty with erythroderma lies in finding the underlying cause. Almost all the published original clinical series of erythroderma originate from western countries. Our aim was to evaluate various causes of exfoliative dermatitis in our community and compare the findings with previous studies.
Methods Ninety patients with erythroderma of either sex and any age, were studied at the Dermatology Department, Mayo Hospital, Lahore, Pakistan, between April 1992 and May 1995. A thorough clinical examination and relevant laboratory investigations, including skin biopsies, were performed.
Results The mean age of onset was 41.6 years with a male : female ratio of 2.8 : 1. The onset was acute in 69% of patients. The salient features included nail changes (80%), mucosal involvement (36.6%), alopecia (30%), islands of normal skin (14.4%), and the "deck chair sign" (5.5%). The most frequent cause of erythroderma was pre-existing dermatoses (74.4%), including psoriasis (37.8%), dermatitis (12.2%), ichthyoses (7.8%), and pemphigus foliaceus (5.6%). Drugs and malignancy each induced erythroderma in 5.5% of cases. No cause could be found in 14.6% of cases.
Conclusions Etiologically, pre-existing dermatoses showed the highest incidence and drugs the lowest compared with previous studies. Dermatitis was less common. There was a greater variety of causes of erythroderma in our series. Hair and nails were more frequently involved. Mucosal involvement, not mentioned in other studies, was present in 36.6% of our cases. The "deck chair sign" and islands of normal skin were seen in dermatoses not reported previously.     

Aneuploidy in cutaneous T-cell lymphoma

A new method is described which makes it possible using skin specimens to perform flow cytometric analysis of DNA content. DNA content analysis was performed on 28 skin specimens and 9 blood samples from 18 patients with mycosis fungoides and Sézary syndrome. The reproducibility was fair, with almost identical results in 6 cases (mycosis fungoides and Sézary syndrome) where two samples (skin specimens or blood samples) were taken 6 hours to 10 days apart. Hyperdiploidy was found in 7 of 11 skin specimens from patients with mycosis fungoides stage I with negative histology. In 13 skin specimens and 3 blood samples from patients with mycosis fungoides stages II and IV, abnormalities including hyperdiploidy, near-tetraploidy and near-hexaploidy were found in 8 of 13 skin specimens and in 2 of 3 blood samples. Four patients with Sézary syndrome were studied: 2 patients in remission showed normal DNA histograms (2 skin specimens, 3 blood samples) and 2 patients with active disease showed aneuploidy in the 2 skin specimens examined and in 1 of 3 blood samples. These studies demonstrate: 1) the importance of flow cytometry as a diagnostic tool for use on skin specimens in the early stages of mycosis fungoides where routine histology is non-diagnostic; 2) the diagnostic and prognostic aid of flow cytometry during the course of mycosis fungoides and Sézary syndrome in addition to the probability of measuring the effect of treatment.     

Life-threatening erythroderma: diagnosing and treating the "red man"

Exfoliative erythroderma, or diffuse erythema and scaling of the skin, may be the morphologic presentation of a variety of cutaneous and systemic diseases. Establishing the diagnosis of the underlying disease is often difficult and, not uncommonly, erythroderma is classified as idiopathic. Several cases are presented to demonstrate the diversity of presentation of this disease. Laboratory findings are typically unhelpful in establishing the etiology of erythroderma. Clinical data combined with multiple skin biopsies over time are necessary. Systemic complications of erythroderma include infection, fluid and electrolyte imbalances, thermoregulatory disturbance, high output cardiac failure, and acute respiratory distress syndrome. The initial approach to the management of erythroderma of any etiology includes attention to nutrition, fluid and electrolyte replacement, and the institution of gentle local skin care measures. Oatmeal baths and wet dressings to weeping or crusted sites should be followed by application of bland emollients and low-potency topical corticosteroids. Systemic dermatologic therapy may be required to maintain improvement achieved with local measures or to control erythroderma refractory to local measures. The prognosis of erythroderma is dependent on the underlying etiology.     

An Aggressive Spindle Cell Squamous Carcinoma Arising in a Patient with Long-standing Erythroderma

Spindle cell squamous carcinoma (SCSC) of the left hand arising in a patient with longstanding erythroderma is reported. Histopathologically, spindle shaped atypical cells were observed neighboring the cells of well differentiated squamous cell carcinoma. These two types of tumor cells, spindle cells and well differentiated cells, were present side by side and merged into each other. The erythroderma had been present for over 20 years, and both clinical and histopathological findings suggested cutaneous T cell lymphoma, but were not diagnostic for mycosis fungoides, Sézary syndrome, or adult T cell lymphoma. Flow cytometry of peripheral blood cells showed a low CD4/CD8 ratio which suggested impaired T cell function. Multiple metastases of SCSC occurred in a short period and the patient died ten months after his first visit to us. The aggressive course of this case was unusual, and may be due to immunological abnormalities associated with the long standing erythroderma with impaired T cell function.     

Lymphomatoid papulosis: a follow-up study

A follow-up study has been performed on 16 patients with lymphomatoid papulosis diagnosed at the Finsen Institute during the years 1970-81. In none of the patients did malignant lymphoma develop during the observation period (7 months to 22 years). During this period the nature of the lesions and the tendency to recurrence were unchanged in 11 patients, spontaneous remission took place in 4, and 1 patient went into complete remission after PUVA treatment (8-methoxsalen followed by UVA). The histological material (32 punch biopsies) could be divided into two major groups diagnosed as either typical (16 biopsies) or consistent with lymphomatoid papulosis (16 biopsies). Based on our present knowledge, we suggest the following classification of lymphomatoid papulosis: 1) "classical" lymphomatoid papulosis, 2) lymphomatoid papulosis associated with parapsoriasis en plaque or mycosis fungoides and 3) primary cutaneous T-cell lymphoma.     

Actinic reticuloid with Sézary cells

A 65-year-old man presented with erythroderma, Sezary cells were present in the peripheral blood on electron microscopy and a skin biopsy revealed mycosis fungoides-like histology. A provisional diagnosis of Sézary syndrome was made. He proved to be exquisitely sensitive to light with markedly abnormal reactions to UV-B, UV-A and visible radiation. By carefully avoiding exposure to light, he has remained well for 2 years and the Sézary cells have not been subsequently detected. The clinical features, histology, photo-test results and course were all in keeping with a diagnosis of actinic reticuloid and this would appear to be a further benign disorder in which Sézary cells may be found.     

A modified staging classification for cutaneous T-cell lymphoma

BACKGROUND: Despite refinements in the diagnosis of cutaneous T-cell lymphoma (CTCL), since 1979 there have been no changes to the staging of CTCL used to classify mycosis fungoides and Sézary syndrome. OBJECTIVE: We reviewed the current staging of CTCL and examined the usefulness of a new staging scheme for mycosis fungoides and Sézary syndrome. METHODS: We determined overall survival of 450 patients with mycosis fungoides and Sézary syndrome using the current and modified staging classifications. RESULTS: There were no significant differences between survival of patients with stage IB (patches/plaques involving greater than 10% body surface area) and IIA (peripheral adenopathy) disease and of patients with stage IIB (tumor) and III (erythroderma) disease. There was a significant difference in survival between patients with extensive patch versus extensive plaque stage disease. Modification of the current classification by splitting T2 into patch versus plaque stage disease and incorporating tumors and erythroderma into stage III proved superior to the current scheme in predicting overall survival. CONCLUSION: Modification of the current staging classification for CTCL yields subgroups useful in the prognostic assessment of CTCL.     

Mycosis fungoides: a retrospective study

The clinical course of ninety patients with the histopathologic diagnosis of mycosis fungoides is reviewed. An attempt is made to correlate sex, age, of onset, lesion type, and form of therapy with outcome of the disease. The data indicate that mycosis fungoides affects predominantly middle-aged males. Patients developing tumors or erythroderma in middle age or later tend to have a shorter survival. More aggressive therapy is associated with shortened survival. Most deaths were due to unknown or unrelated processes and therapeutic complications. The data support the theory that mycosis fungoides can be a relatively nonaggressive cutaneous lymphoma. In an attempt to treat aggressively, we may be exposing patients to increased mortality.     

Neonatal and infantile erythrodermas: a retrospective study of 51 patients

OBJECTIVE: To determine the frequency of the various underlying causes of erythroderma in newborns or infants, as well as which clinical or laboratory findings were relevant for the etiological diagnosis. PATIENTS: Fifty-one patients who presented with exfoliative erythroderma during their first year of life were included in this retrospective study. SETTING: Department of Pediatric Dermatology at a university hospital. RESULTS: On average, the etiological diagnosis was established 11 months after the onset of erythroderma. The underlying causes observed included immunodeficiency (30%), simple or complex ichthyosis (24%), Netherton syndrome (18%), and eczematous or papulosquamous dermatitis (20%). Five patients (10%) had erythroderma of unknown origin. The following parameters were of value in determining the underlying cause of erythroderma: congenital onset, skin induration and the presence of large scaling plaques, alopecia with or without hair dysplasia, evolution, response to topical corticosteroid therapy, presence of infections, and failure to thrive. Histological analysis confirmed the diagnosis in only 19 (45%) of 42 cases. However, it proved of great value for the detection of significant lymphocyte infiltration or keratinocyte necrosis indicating a diagnosis of Omenn syndrome or immunodeficiency. The prognosis was poor in this series: the mortality rate was 16%, and severe dermatosis persisted in 29 (67%) of the survivors. CONCLUSIONS: The etiological diagnosis of neonatal erythroderma is difficult to make; some clinical features may be helpful, but no one feature is characteristic of a cause. An immunodeficiency must be suspected in cases of severe erythroderma with skin induration, severe alopecia, failure to thrive, infectious complications, or evocative histological findings. The prognosis is poor, with a high rate of mortality in immunodeficiency disorders and severe chronic disease in Netherton syndrome and psoriasis.