Paraneoplastic Pemphigus: Insight into the Autoimmune Pathogenesis, Clinical Features and Therapy

Paraneoplastic pemphigus is a rare autoimmune skin disease that is always associated with a neoplasm. Usually, oral, skin, and mucosal lesions are the earliest manifestations shown by paraneoplastic pemphigus patients. The pathogenesis of paraneoplastic pemphigus is not yet completely understood, although some immunological aspects have been recently clarified. Because of its rarity, several diagnostic criteria have been proposed. Besides, several diagnostic procedures have been used for the diagnosis, including indirect immunofluorescence, direct immunofluorescence, and ELISA. We reviewed the most recent literature, searching on PubMed “paraneoplastic pemphigus”. We included also papers in French, German, and Spanish. We found 613 papers for “paraneoplastic pemphigus”. Among them, 169 were review papers. Because of its varying clinical features, paraneoplastic pemphigus still represents a challenge for clinicians. Furthermore, diagnosis and management of paraneoplastic pemphigus requires close collaboration between physicians, including dermatologist, oncologist, and otorhinolaryngologist.


Introduction
Paraneoplastic pemphigus (PNP) was first reported by Anhalt in 1990 [1]. It is a rare autoimmune skin disease belonging to the group of blistering diseases. PNP is always characterized by an association with neoplasms, including carcinoma of the stomach, lung, and colon [2]. Besides, B-cell lymphomas and hematological malignancies are most frequently reported in association with PNP [3,4]. In 2001
PNP antibodies belong principally to the IgG class. However, IgA class has also been reported in few cases [43][44][45][46]. More specifically, Mentink et al. detected by ELISA IgA autoantibodies only against DSG-3 in a series of four PNP patients [47]. Interestingly, in only one patient the serum showed IgA, but not IgG autoantibodies [43].

Clinical Features
PNP clinical features are extremely polymorphous and PNP lesions can be detected not only on the skin, but also in different mucosae [1][2][3]11]. The presence of different autoantibodies could justify the different clinical features in PNP patients [3].
Usually, a neoplasm is detected before the onset of PNP [1,8,52]. However, PNP is the first clinical manifestation that leads to the detection of an occult tumor in about 30% of cases [3,9].
Because the refractoriness of the oral mucosal and skin lesions to standard immunosuppression therapy, patient should be screened for an underlying occult neoplasm [1,8]. Because of the lack of standardization of a screening protocol for occult neoplasms, a pan-CT scan and flow cytometry on peripheral blood should be mandatory; indeed, PNP is mostly associated to lymphoproliferative or haematological neoplasms [1,8].

Oral Lesions
Oral mucosa is almost always involved [2,21]. Clinically, PNP often presents with oral erosions, which are rarely preceded by vesicles or bulla. The lesions usually involve the vermillion border of the lips (Figures 1 and 2). Ulceration may involve the entire oral mucosal surfaces and may represent the sole manifestation of this condition. A painful stomatitis is also commonly shown caused by massive erosion in the oropharynx [2,21,53,55]. immunoprecipitation-immunoblotting [42]. In addition, the authors showed that epiplakin-negative PNP cases did not develop bronchiolitis obliterans. Indeed, epiplakinis is reported as a target antigen in PNP-related bronchiolitis obliterans [42]. PNP antibodies belong principally to the IgG class. However, IgA class has also been reported in few cases [43][44][45][46]. More specifically, Mentink et al. detected by ELISA IgA autoantibodies only against DSG-3 in a series of four PNP patients [47]. Interestingly, in only one patient the serum showed IgA, but not IgG autoantibodies [43].

Clinical Features
PNP clinical features are extremely polymorphous and PNP lesions can be detected not only on the skin, but also in different mucosae [1][2][3]11]. The presence of different autoantibodies could justify the different clinical features in PNP patients [3].
Usually, a neoplasm is detected before the onset of PNP [1,8,52]. However, PNP is the first clinical manifestation that leads to the detection of an occult tumor in about 30% of cases [3,9].
Because the refractoriness of the oral mucosal and skin lesions to standard immunosuppression therapy, patient should be screened for an underlying occult neoplasm [1,8]. Because of the lack of standardization of a screening protocol for occult neoplasms, a pan-CT scan and flow cytometry on peripheral blood should be mandatory; indeed, PNP is mostly associated to lymphoproliferative or haematological neoplasms [1,8].

Oral Lesions
Oral mucosa is almost always involved [2,21]. Clinically, PNP often presents with oral erosions, which are rarely preceded by vesicles or bulla. The lesions usually involve the vermillion border of the lips (Figures 1 and 2). Ulceration may involve the entire oral mucosal surfaces and may represent the sole manifestation of this condition. A painful stomatitis is also commonly shown caused by massive erosion in the oropharynx [2,21,53,55].   Healy WJ et al. reported the case of a 51-year-old man who was admitted to the emergency department with a 2-3-month history of dyspnea as well as oral ulcerations [55]. The oral ulcerations were biopsied, showing PNP features [55]. In addition, Mahajan et al. described a case of a 58-year-old female with a one-year history of recurrent, multiple flaccid bullae and erosions with oozing, crusting, and severe painful oral ulcerations [56]. Several treatments, including repeated courses of systemic corticosteroids, provided temporary relief, but recurrences and exacerbations caused marked discomfort and disability. The patient had severe erosions/ulceration of the vermilion lips, palate and tongue, and conjunctival congestion (conjunctivitis). The oral lesions preceded the skin lesion appearance by two months.
It should be also noted that extended, painful erosions and crusting on the lips could resemble oral lesions commonly found in erythema multiforme (EM) or Stevens-Johnson's disease [2,21,53].
In children, the stomatitis caused by PNP may be often mistaken for herpetic stomatitis or toxic epidermal necrolysis (TEN), leading to a delay in the diagnosis [57]. In this subset of patients, the most important cause of PNP is Castelman's disease [58].

Secondary Mucosal Lesions
The nasopharynx, ano-genital region, and esophagus could be also affected by large, painful, mucosal lesions [1,3,9,15,59,60]. Yokokura et al. reported the case of a 58-year-old Japanese man with an eight-month history of painful lesions on the lower lip, oral mucosa, and genital region [61]. Physical examination revealed well-demarcated erosion with grayish hyperkeratosis and violet-reddish erythema at the border on the lower lip and glans penis. Multiple erosions were also distributed from the hard palate to the pharynx.
Conjunctival lesions could be also present in PNP patients. Meyers SJ et al. reported two cases of bilateral bulbar conjunctival hyperemia and diffuse papillary tarsal conjunctival reactions [59]. One patient had sloughing of the conjunctival epithelium, and the other ones had tarsal conjunctival scarring and forniceal shortening. Histopathological findings from conjunctivae samples from both patients were consistent with pemphigus vulgaris (PV).

Skin Lesions
Usually, skin lesions appear after the onset of mucosal lesions [3,52]. Most PNP patients show widespread cutaneous involvement, especially on the torso (Figure 3), head ( Figure 4), neck, and proximal extremities [7,63]. PNP onset is highly variable. Patients may present with diffuse erythema, vesiculo-bullous lesions, papules, scaly plaques, exfoliative erythroderma, erosions or Healy WJ et al. reported the case of a 51-year-old man who was admitted to the emergency department with a 2-3-month history of dyspnea as well as oral ulcerations [55]. The oral ulcerations were biopsied, showing PNP features [55]. In addition, Mahajan et al. described a case of a 58-year-old female with a one-year history of recurrent, multiple flaccid bullae and erosions with oozing, crusting, and severe painful oral ulcerations [56]. Several treatments, including repeated courses of systemic corticosteroids, provided temporary relief, but recurrences and exacerbations caused marked discomfort and disability. The patient had severe erosions/ulceration of the vermilion lips, palate and tongue, and conjunctival congestion (conjunctivitis). The oral lesions preceded the skin lesion appearance by two months.
It should be also noted that extended, painful erosions and crusting on the lips could resemble oral lesions commonly found in erythema multiforme (EM) or Stevens-Johnson's disease [2,21,53].
In children, the stomatitis caused by PNP may be often mistaken for herpetic stomatitis or toxic epidermal necrolysis (TEN), leading to a delay in the diagnosis [57]. In this subset of patients, the most important cause of PNP is Castelman's disease [58].

Secondary Mucosal Lesions
The nasopharynx, ano-genital region, and esophagus could be also affected by large, painful, mucosal lesions [1,3,9,15,59,60]. Yokokura et al. reported the case of a 58-year-old Japanese man with an eight-month history of painful lesions on the lower lip, oral mucosa, and genital region [61]. Physical examination revealed well-demarcated erosion with grayish hyperkeratosis and violet-reddish erythema at the border on the lower lip and glans penis. Multiple erosions were also distributed from the hard palate to the pharynx.
Conjunctival lesions could be also present in PNP patients. Meyers SJ et al. reported two cases of bilateral bulbar conjunctival hyperemia and diffuse papillary tarsal conjunctival reactions [59]. One patient had sloughing of the conjunctival epithelium, and the other ones had tarsal conjunctival scarring and forniceal shortening. Histopathological findings from conjunctivae samples from both patients were consistent with pemphigus vulgaris (PV).

Skin Lesions
Usually, skin lesions appear after the onset of mucosal lesions [3,52]. Most PNP patients show widespread cutaneous involvement, especially on the torso (Figure 3), head (Figure 4), neck, and proximal extremities [7,63]. PNP onset is highly variable. Patients may present with diffuse erythema, vesiculo-bullous lesions, papules, scaly plaques, exfoliative erythroderma, erosions or ulcerations. The erythema can be macular, urticarial, targetoid or polymorphous. Patients may initially show erythema, and then develop bullae and erosions [5,10,15,16].   A single patient may present different types of lesions, each of which is able to evolve from one type to another [8,52]. Usually, cutaneous lesions resemble those seen in PV, BP, EM, or graft versus host disease (GVHD) [16]. Furthermore, pustular and psoriasiform lesions have also been reported in the literature [5].
The different clinical features could be due to the predominance of the cell-mediated or humoral-mediated pathogenic mechanism [51]. On the one hand, humoral-mediated cytotoxicity usually causes a prominent PV-like appearance [3,5,52]; on the other hand, cell-mediated cytotoxicity often determines lichenoid lesions [48,50]. Lichenoid lesions are frequently identified in children, predominantly on the torso and limbs [12,13]. Furthermore, a unique case of pemphigus vegetans-like PNP has been described [64,65]. ulcerations. The erythema can be macular, urticarial, targetoid or polymorphous. Patients may initially show erythema, and then develop bullae and erosions [5,10,15,16].  A single patient may present different types of lesions, each of which is able to evolve from one type to another [8,52]. Usually, cutaneous lesions resemble those seen in PV, BP, EM, or graft versus host disease (GVHD) [16]. Furthermore, pustular and psoriasiform lesions have also been reported in the literature [5].
The different clinical features could be due to the predominance of the cell-mediated or humoral-mediated pathogenic mechanism [51]. On the one hand, humoral-mediated cytotoxicity usually causes a prominent PV-like appearance [3,5,52]; on the other hand, cell-mediated cytotoxicity often determines lichenoid lesions [48,50]. Lichenoid lesions are frequently identified in children, predominantly on the torso and limbs [12,13]. Furthermore, a unique case of pemphigus vegetans-like PNP has been described [64,65]. A single patient may present different types of lesions, each of which is able to evolve from one type to another [8,52]. Usually, cutaneous lesions resemble those seen in PV, BP, EM, or graft versus host disease (GVHD) [16]. Furthermore, pustular and psoriasiform lesions have also been reported in the literature [5].
The different clinical features could be due to the predominance of the cell-mediated or humoral-mediated pathogenic mechanism [51]. On the one hand, humoral-mediated cytotoxicity usually causes a prominent PV-like appearance [3,5,52]; on the other hand, cell-mediated cytotoxicity often determines lichenoid lesions [48,50]. Lichenoid lesions are frequently identified in children, predominantly on the torso and limbs [12,13]. Furthermore, a unique case of pemphigus vegetans-like PNP has been described [64,65].
Pediatric patients often show lichenoid PNP lesions rather than bullous skin lesions [57,66]. In a series of 14 PNP pediatric patients, it has been reported that all of them demonstrated antibodies to plakins [12].

Pulmonary Manifestations
PNP affects also the respiratory epithelium in up to 92.8% of cases [11,51], leading to dyspnea, obstructive lung disease, and bronchiolitis obliterans, which is one of the principal causes of death in PNP patients [9,67]. Pulmonary involvement is commoner in children and Chinese patients with Castelman's disease [67]. Recently, Tsuchisaka et al. reported a correlation between epiplakin and bronchiolitis obliterans in Japanese PNP patients [42]. Furthermore, the authors reported that 71% of patients were affected by bronchiolitis obliterans pneumonia that led to a worse prognosis despite treatment of the underlying malignancy [12,42,57].

Pathology
Pathological findings are related to the clinical features, showing different pathological characteristics according to the examined lesion [7,39]. When blisters are present, suprabasal acantholysis with sparse inflammatory infiltrates is usually evident on skin biopsy ( Figure 5) [39], while interface and lichenoid dermatitis are usually detected if inflammatory maculopapular lesions are present [39,50]. In addition, lesions with mixed clinical features might show both acantholysis and lichenoid interface dermatitis [8,53,54]. Dyskeratosis with suprabasal acantholysis is an important finding that leads to the diagnosis of PNP [8]. (Figure 6) However, sometimes the acantholysis is difficult to find, leading to important diagnostic pitfalls with other diseases as erythema multiforme, Stevens-Johnson syndrome, GVHD, and drug reactions. Finally, it is important to highlight that pathological findings could be compatible with a diagnosis of PNP even when the results of direct immunofluorescence (DIF) are negative [8,53]. Indeed, DIF findings are important because the cells involved in PNP include autoantibodies and CD8+ T cells that attack proteins in different layers of the keratin, leading to different DIF features, including intercellular cement substance staining and/or dermo-epidermal junction staining [8,53] (Figure 7). Pediatric patients often show lichenoid PNP lesions rather than bullous skin lesions [57,66]. In a series of 14 PNP pediatric patients, it has been reported that all of them demonstrated antibodies to plakins [12].

Pulmonary Manifestations
PNP affects also the respiratory epithelium in up to 92.8% of cases [11,51], leading to dyspnea, obstructive lung disease, and bronchiolitis obliterans, which is one of the principal causes of death in PNP patients [9,67]. Pulmonary involvement is commoner in children and Chinese patients with Castelman's disease [67]. Recently, Tsuchisaka et al. reported a correlation between epiplakin and bronchiolitis obliterans in Japanese PNP patients [42]. Furthermore, the authors reported that 71% of patients were affected by bronchiolitis obliterans pneumonia that led to a worse prognosis despite treatment of the underlying malignancy [12,42,57].

Pathology
Pathological findings are related to the clinical features, showing different pathological characteristics according to the examined lesion [7,39]. When blisters are present, suprabasal acantholysis with sparse inflammatory infiltrates is usually evident on skin biopsy ( Figure 5) [39], while interface and lichenoid dermatitis are usually detected if inflammatory maculopapular lesions are present [39,50]. In addition, lesions with mixed clinical features might show both acantholysis and lichenoid interface dermatitis [8,53,54]. Dyskeratosis with suprabasal acantholysis is an important finding that leads to the diagnosis of PNP [8]. (Figure 6) However, sometimes the acantholysis is difficult to find, leading to important diagnostic pitfalls with other diseases as erythema multiforme, Stevens-Johnson syndrome, GVHD, and drug reactions. Finally, it is important to highlight that pathological findings could be compatible with a diagnosis of PNP even when the results of direct immunofluorescence (DIF) are negative [8,53]. Indeed, DIF findings are important because the cells involved in PNP include autoantibodies and CD8+ T cells that attack proteins in different layers of the keratin, leading to different DIF features, including intercellular cement substance staining and/or dermo-epidermal junction staining [8,53] (Figure 7).

Immunological Studies
Because PNP represents a pitfall for clinicians, several immunological studies have been developed to better detect different PNP antibodies. In this context, PNP patients show plenty of auto-antibodies, targeting Dsg 3, desmoplakin 1 and 2, envoplakin, periplakin, plectin, BP antigen 230 (BP230), and α-2-macroglobulin-like-1, a 170 kd protease inhibitor found in stratified epithelia and other tissue damaged by PNP, which has been recently identified [16,68].
ELISA is a useful tool to detect anti-Dsg 3 and anti-Dsg 1 auto-antibodies in PNP, although PNP patients usually show only anti-Dsg3 IgG [68]. However, PNP patients lacking anti-Dsg autoantibodies have been also described [68]. In 2009, Probst et al. created a new ELISA based on the recombinant N-terminus of envoplakin, showing a sensitivity of 82% and a specificity of ≥98% [69]. However, it has been reported that ELISA appeared to lack sensitivity in a multi-assay comparison across different assays, because antigen-specific techniques, such as envoplakin ELISA, detect only one of a range of autoantibodies potentially responsible for the disease [70].
Immunofluorescence (IF) is one of the main diagnostic tools for PNP. DIF usually shows IgG and/or C3 deposition in the epidermal intercellular spaces (EIS) alone [71]. The presence of IgG and/or C3 in EIS and at the basement membrane zone (BMZ) has been found in less than 50% of cases [71]. Linear deposits of IgG and/or C3 at the BMZ may also be detected [39]. (Figure 7) This pattern could help to differentiate PNP from other forms of pemphigus that show Ig deposits only between keratinocytes [5]. However, DIF is negative in around 50% of cases [71]. False negatives are often due to necrotic tissue (especially in mucosal specimens) and to lichenoid lesions [50,71]. Different substrates could be used to perform indirect immunofluorescence (IIF), including normal

Immunological Studies
Because PNP represents a pitfall for clinicians, several immunological studies have been developed to better detect different PNP antibodies. In this context, PNP patients show plenty of auto-antibodies, targeting Dsg 3, desmoplakin 1 and 2, envoplakin, periplakin, plectin, BP antigen 230 (BP230), and α-2-macroglobulin-like-1, a 170 kd protease inhibitor found in stratified epithelia and other tissue damaged by PNP, which has been recently identified [16,68].
ELISA is a useful tool to detect anti-Dsg 3 and anti-Dsg 1 auto-antibodies in PNP, although PNP patients usually show only anti-Dsg3 IgG [68]. However, PNP patients lacking anti-Dsg autoantibodies have been also described [68]. In 2009, Probst et al. created a new ELISA based on the recombinant N-terminus of envoplakin, showing a sensitivity of 82% and a specificity of ≥98% [69]. However, it has been reported that ELISA appeared to lack sensitivity in a multi-assay comparison across different assays, because antigen-specific techniques, such as envoplakin ELISA, detect only one of a range of autoantibodies potentially responsible for the disease [70].
Immunofluorescence (IF) is one of the main diagnostic tools for PNP. DIF usually shows IgG and/or C3 deposition in the epidermal intercellular spaces (EIS) alone [71]. The presence of IgG and/or C3 in EIS and at the basement membrane zone (BMZ) has been found in less than 50% of cases [71]. Linear deposits of IgG and/or C3 at the BMZ may also be detected [39]. (Figure 7) This pattern could help to differentiate PNP from other forms of pemphigus that show Ig deposits only between keratinocytes [5]. However, DIF is negative in around 50% of cases [71]. False negatives are often due to necrotic tissue (especially in mucosal specimens) and to lichenoid lesions [50,71]. Different substrates could be used to perform indirect immunofluorescence (IIF), including normal

Immunological Studies
Because PNP represents a pitfall for clinicians, several immunological studies have been developed to better detect different PNP antibodies. In this context, PNP patients show plenty of auto-antibodies, targeting Dsg 3, desmoplakin 1 and 2, envoplakin, periplakin, plectin, BP antigen 230 (BP230), and α-2-macroglobulin-like-1, a 170 kd protease inhibitor found in stratified epithelia and other tissue damaged by PNP, which has been recently identified [16,68].
ELISA is a useful tool to detect anti-Dsg 3 and anti-Dsg 1 auto-antibodies in PNP, although PNP patients usually show only anti-Dsg3 IgG [68]. However, PNP patients lacking anti-Dsg autoantibodies have been also described [68]. In 2009, Probst et al. created a new ELISA based on the recombinant N-terminus of envoplakin, showing a sensitivity of 82% and a specificity of ≥98% [69]. However, it has been reported that ELISA appeared to lack sensitivity in a multi-assay comparison across different assays, because antigen-specific techniques, such as envoplakin ELISA, detect only one of a range of autoantibodies potentially responsible for the disease [70].
Immunofluorescence (IF) is one of the main diagnostic tools for PNP. DIF usually shows IgG and/or C3 deposition in the epidermal intercellular spaces (EIS) alone [71]. The presence of IgG and/or C3 in EIS and at the basement membrane zone (BMZ) has been found in less than 50% of cases [71]. Linear deposits of IgG and/or C3 at the BMZ may also be detected [39]. (Figure 7) This pattern could help to differentiate PNP from other forms of pemphigus that show Ig deposits only between keratinocytes [5]. However, DIF is negative in around 50% of cases [71]. False negatives are often due to necrotic tissue (especially in mucosal specimens) and to lichenoid lesions [50,71]. Different substrates could be used to perform indirect immunofluorescence (IIF), including normal human skin, monkey esophagus, rat bladder, rat myocardium, and rat lung [71]. IIF identifies autoantibodies directed against plakins, among which autoantibodies to envoplakin and periplakin are the most specific [39]. On the one hand, IIF on normal human skin has been reported as positive in up to 50%; on the other hand, IIF on rat bladder urothelium has been found positive in 75% of cases, displaying a better sensitivity [72]. In addition, IIF on rat bladder has a high specificity (83%) [1,72]. Therefore, IIF on rat bladder is now thought as a useful screening test for PNP. However, autoantibodies directed against plakins have been also shown in other dermatoses, including PV, pemphigus foliaceus, and TEN [72][73][74].
Immunoprecipitation (IP) is considered the gold standard for diagnosing PNP [75]. IP can show antibodies against several antigens, including plakins and α-2-macroglobulin-like-1 [40]. In addition, a positive IP has been reported by Camisa et al. as a major criterion for diagnosing PNP [76].

Diagnosis
According to Anahlt et al. [1] the diagnostic criteria include five different points (Table 1). Subsequently, Camisa et al. [76] introduced different criteria, including major and minor ones. According to Camisa et al., three major or two major and two minor criteria are needed to diagnose PNP [76]. More recently, Mimouni et al. [12] revised the original criteria by Anahlt et al. In this new classification DIF was considered as a non-essential criterion for diagnosing PNP, because of its low sensibility [71,72]. IIF on rat bladder urothelium and monkey esophagus are considered useful in screening for detecting PNP [71,72]. A scheme depicting the diagnostic algorithm is shown in Table 2. Table 1. Diagnostic criteria [1,12,75].

Clinical features
Painful erosions involving mucosae with or without a multiform skin eruption producing blisters and erosions, occurring in association with an occult or evident neoplasm

Direct immunofluorescence
Combined presence of IgG and complement (C3) granular-linear deposition within the epidermal intercellular spaces and along the basement-membrane zone

Indirect immunofluorescence
Presence of circulating antibodies that target the intercellular zone of stratified squamous or transitional epithelia

Differential Diagnosis
Differential diagnosis includes PV, BP, and EM. A more complete list is shown in Table 3. Other diseases with mucous skin involvement should always be excluded [77,78]. Clinically, PNP and PV may share many features, but PNP shows blisters developing from inflammatory papules or macules, while PV usually shows blisters on the erythematous background. In addition, PNP could show antibodies anti-A2ML1, anti-envoplakin, and anti-periplakin, highly specific for PNP [71][72][73][74]. Furthermore, DIF shows epithelial cell-surface IgG depositions with concurrent basement membrane zone IgG depositions, considered a hallmark feature of PNP [71,72]. Finally, unlike other autoimmune blistering diseases, PNP antibodies stain rat bladder epithelium. BP also shares some features with PNP, especially when BP230 and BP180 are detected in PNP. However, DIF in PNP could show the presence of epidermal intercellular deposits of IgG and complement C3, which are not found in BP. EM-like lesions resembling TEN could also be present in PNP, but the detection of auto-antibodies leads to PNP diagnosis [1][2][3][71][72][73][74]. It is important to highlight that oral and cutaneous PNP lesions can be variable and resemble many other diseases, both clinically and histologically. Therefore, an otolaryngologic examination is essential for the evaluation of the lesions and of the differential diagnosis with other oral cavity pathologies [2,21,53,55].

Treatment Options
PNP therapy remains challenging because of the rarity of the disease. Although several medical therapies have been suggested in the literature, PNP has been considered as more resistant to medical therapies in comparison to other forms of pemphigus [16,79]. When PNP is suspected, it could be useful to follow the six steps reported by Frew et al. for better management of the patient [80]. This series of steps includes stabilization of vital parameters, evaluation of any underlying malignancy, accurate diagnosis of PNP, removal and medical therapy of the trigger tumor, and treatment of PNP using immunosuppression, immunomodulation, or plasmapheresis. The first step is of capital importance because of the high rate of mortality. Therefore, the stabilization of the patient represents the first measure in the management of PNP patients [80].
High-dose corticosteroids are still considered as first line therapy [81,82]. However, steroids only improve the skin lesion, while mucosal involvement it is not usually affected by steroids [54,83]. Indeed, one of the most important clinical features of PNP is the resistance of the mucosal lesions to most types of therapy [83]. Nevertheless, high-dose prednisolone is still recommended as the first line of treatment [80].
Rituximab, an anti-CD20 monoclonal antibody, has been reported as effective in PNP patients caused by B-cell lymphoma [89,90]. Several rituximab schedules have been reported in the literature, including monotherapy (375 mg/m 2 weekly for four weeks) followed by eight weekly infusions and weekly infusions for four weeks under corticosteroids and other immunosuppressive drugs, such as cyclosporine A [80].
It has been reported that alemtuzumab, a humanized monoclonal antibody that binds CD52, induced long-term remission in a patient with B-cell chronic lymphocytic leukemia [91]. This treatment was used in a patient refractory to many of previous treatment, including corticosteroids, cyclosporine, and IVIG. Alemtuzumab was administered 30 mg intravenously three times a week for 12 weeks, showing improvement of mucosal and cutaneous lesions. Twelve months later, the patient was still in remission on maintenance therapy (500 mg mycophenolate mofetil and 5 mg prednisone) [80,91].
Daclizumab, a humanized monoclonal antibody against the alpha subunit of the IL-2 receptor of T cells, is thought to be a promising therapy for PNP [54].
Early antimicrobial therapy is recommended, because of the risk of sepsis following loss of skin integrity and iatrogenic immunosuppression [16]. Antalgic therapy could be useful in reducing the pain caused by extensive erosions [16].

Prognosis
The prognosis of PNP is generally poor (90% mortality rate) [8,16]. Death is usually due to systemic complications, including sepsis, gastrointestinal bleeding, and bronchiolitis obliterans [8,16]. It has been reported that PNP and underlying malignancy do not have a parallel evolution [3,8,16]. Indeed, PNP lesions generally progress after removing the triggering malignancy or even when the malignancy is under control [11,13,16]. However, it has been shown that outcome is better in PNP patients with concurrent Castleman's disease or benign thymomas after removing the tumor [92].
Finally, the prognosis of PNP depends on an appropriate management, including the effective control of the oral and skin lesions, an adequate treatment of the underlying neoplasm, and prevention of bronchiolitis obliterans. Therefore, it is mandatory to monitor vigilantly the patient and to treat aggressively the disease [8,16].

Conclusions
Because of its various clinical features, PNP represents a challenge for the clinician. In order to detect PNP earlier and to treat patients better, a cooperation between dermatologists, oncologists, otolaryngologists, ophthalmologists, and surgeons is recommended. Although different immunological markers have been found, the pathogenesis of PNP is still unknown. Management of the underlying tumor is of paramount importance. However, several therapies have been attempted to treat this potentially lethal condition.