Acquired Perforating Dermatosis as a Paraneoplastic Feature: A Case Report, Literature Review, and Novel Association

Abstract

Kyrle’s disease is an uncommon form of acquired transepidermal elimination dermatosis frequently associated with diabetes mellitus and chronic kidney disease. An association with malignancy has been sporadically reported in the literature. Here, we describe the clinical course of a diabetic patient with end-stage renal disease who developed this disorder as a herald to a regionally advanced renal cell carcinoma. We provide a focused literature review and rationale for the definitive categorization of acquired perforating dermatosis as a potential paraneoplastic manifestation of systemic malignancies. Clinicopathological correlation and prompt communication among clinicians for occult malignancies are always warranted. Furthermore, we describe a novel association of one of the subtypes of acquired perforating dermatosis with such malignancies.

Introduction

Perforating disorders of the skin are a group of diseases with transepidermal elimination of material from the upper dermis. Some authors have classically subdivided the condition into primary and secondary [1, 2]. It is further divided into four types according to the eliminated dermal material: Kyrle’s disease (KD), perforating reactive collagenosis (PRC), elastosis perforans serpiginosa (EPS), and perforating folliculitis (PF) [1–4]. Transdermal elimination of collagen, elastic fibers, and degenerated follicular contents with or without collagen or elastic fibers is seen in PRC, EPS, and PF, respectively. In contrast, transdermal elimination of keratotic material with no collagen or elastic fibers is observed in what is known historically as KD [3]. While the primary form has no clear association, the secondary type, also known as acquired perforating dermatosis (APD) in its four subtypes, is strongly associated with diabetes mellitus (DM) and pruritus of chronic kidney disease (CKD). These acquired perforating eruptions arise in adulthood, are uncommon, and may exhibit an overlap of elimination patterns [5]. The CARE Checklist has been completed by the authors for this case report and is attached as online supplementary material at https://doi.org/10.1159/000530756.

Case Report

We report the case of an African American male in his mid-sixties with an end-stage renal disease (ESRD) history. He had received hemodialysis exclusively for the prior 13 years. He also had a 20-year history of hypertension, DM, anemia of CKD, and secondary hyperparathyroidism of renal origin.

His medical history included renal cell carcinoma treated with a right radical nephrectomy which led to the progression of his baseline CKD stage 3 to stage 5, which later required hemodialysis. He was remarkably compliant with dialysis sessions, with 100% attendance over the past 13 years. His hemodialysis history has been mostly unremarkable except for two angioplasty procedures to his right arteriovenous fistula. He presented with a 2-month history of a darkly pigmented hyperkeratotic patch on the lateral aspect of the proximal right leg. The skin lesion was 4 × 3 cm in size, as shown in Figure 1a, b. Moreover, he also had other smaller lesions on both thighs and the lower back.

Fig. 1.

a The red spot indicates the site of the punch biopsy. The image was taken in February 2017 at the time of initial presentation prior to any treatment. b Modest improvement in the skin lesion after 18 months of keratolytic topical treatment. The image was taken in October 2018.

Lesions were itchy and occasionally painful. There was no history of any local trauma or chronic pruritus before developing these lesions. There was no evidence for Kobner phenomenon either. He tried topical therapy such as antihistamines and corticosteroid creams, with no significant symptomatic relief. Based on the above clinical scenario, a decision to proceed with a skin biopsy was made to arrive at a tissue diagnosis as the major differential diagnostic possibility was calcific uremic arteriolopathy, also known as calciphylaxis, given the locations of his skin lesions, his long ESRD history, and his most recent laboratory tests (Table 1).

Table 1.

Laboratory tests and other studies that were performed 3 months before the presentation

Test	Value	Reference range
Calcium corrected	8.7–10.3	8.4–10.2 mg/dL
Phosphorus	4.7–7.0	2.5–5.0 mg/dL
Albumin	3.8–4.0	3.5–5.7 g/dL
Ca × P product corrected	40.9–73.2	21.0–46.0
PTH intact	754–1,083	11–80 pg/mL
Aluminum	<10	0–9 μg/mL
Vitamin D (25-OH)	15	31–100 Ng/mL
Hemoglobin	10.7–12.0	14.0–18.0 g/dL
URR	70	65–100%
HGB A1c	6.2	0.0–5.6% A1c
Single pool kt/v total	1.31–1.42

Histopathological examination revealed a cup-shaped plug of parakeratosis with neutrophils and melanin pigment (Fig. 2). This lesion was shouldered by mild pseudoepitheliomatous hyperplasia with foci of parakeratosis. The plug’s base was composed of a single layer of degenerating keratinocytes. In addition, nonspecific infrequent perivascular lymphocytes were present in the dermis, along with scattered melanophages. This degenerative process did not involve the follicular structures. The overall clinical and histopathological features were consistent with APD (KD subtype). The lymphocyte infiltrates and inflammatory changes may have been more severe had the biopsy been done before the patient’s repeated steroid use. Histopathologically, the differential diagnosis included PF, reactive perforating collagenosis, and EPS. In histochemical staining for trichrome and elastin, the above variants have been excluded. The high clinical suspicion of calciphylaxis was ruled out given the lack of typical biopsy findings and by performing a Von Kossa calcium stain which was negative for any calcium precipitation in the vessels’ walls.

Fig. 2.

Microscopic appearance of the lesion. a H&E section showing intraepidermal cellular and stromal debris forming a plug and underlying degenerated keratinocytes. Trichrome and elastin histochemical stains shown in b and c, respectively, are negative for collagen and elastic fibers in the degenerated material.

Discussion

Not all darkly pigmented lesions in ESRD patients are calcific uremic arteriolopathy. Careful physical examination aided by tissue sampling is needed to discern different pathologies as the treatment, among other diagnoses, differs, as well as their clinical significance. APD describes a group of disorders with transepidermal elimination of collagen, elastic tissue, or necrotic connective tissue acquired in adulthood. These patients may have systemic diseases such as DM, chronic liver disease, and advanced CKD, including dialysis [6].

It has also been reported in patients with various other disorders, such as hepatic and endocrinological disorders, acquired immunodeficiency syndrome, tuberculosis, pulmonary aspergillosis, neurodermatitis, atopic dermatitis, and scabies [5]. APD disorders can emerge as paraneoplastic diseases, too [7]. There is a growing literature on APD heralding an underlying malignancy, and it has been reported in papillary thyroid carcinoma [8], advanced prostate cancer [9], metastatic breast carcinoma [10], renal cell carcinoma [11], and even in acute myelogenous leukemia [12], which prompted us to screen our patient for occult malignancy. The patients in these reports had no kidney disease, and only two had diabetes. These patients, along with ours, all had an underlying malignancy, and it seems reasonable to hypothesize a rare but existent paraneoplastic origin of perforating dermatosis. Of interest, most of these patients with malignancy had one subtype except for one (Table 2). To our knowledge, this observed novel association has never been reported before. It may potentially be the first evidence that specific APD subtypes may emphasize different clinical significances. For example, many authors reported that the skin lesions’ resolution parallels that of the malignancy and so is their emergence. The PRC subtype was described in most of these cases, and hence, this may necessitate thorough further investigation, especially if the patient has a history of malignancy.

Table 2.

Literature review of all malignancies associated with APD along with subtypes

Malignancy	Number of patients	APD subtype	Authors	Outcome	Temporal relationship
Mixed histiocytic-lymphocytic lymphoma	1	PRC	Henry et al. [13], 1983	No information	Cancer diagnosis predated the dermatosis by 3 years
Hodgkin’s disease	2	PRC	Pedragosa et al. [14], 1987	APD improved with the improvement of the malignancy in 1 case. Persisted with the persistence of the malignancy in the other	Cancer predated the dermatosis by 2 years
Hepatocellular carcinoma	2	PRC	Lee et al. [15], 1996, Kilic et al. [16], 2006	APD improved with the improvement of the malignancy in 1 case (Lee et al. [15]). Persisted with the persistence of the malignancy in the other (Kilic et al. [16])	Dermatosis predated cancer by 3 months [15] Cancer predated the dermatosis by around 2 years [16]
Prostate cancer	1	PRC	Boeck et al. [17], 1997	Topical steroids, topical salicylic acid, and UVA and UVB spectrum were helpful. Skin lesions improved with hormonal control of the malignancy	Cancer predated the dermatosis by around 5 months
Pancreatic cancer (periampullary)	1	PRC	Chae et al. [18], 1998	APD disappeared with the surgical cure of the malignancy	Dermatosis predated the diagnosis of cancer by around 6 weeks
Undifferentiated adenocarcinoma metastasis to liver	1	PRC	Bong et al. [19], 2001	APD persisted with the persistence of the malignancy. A potent topical steroid was only anti-pruritic	Dermatosis predated the diagnosis of cancer by 6 years
Myelodysplastic syndrome/acute myelogenous leukemia	1	PRC	Karpouzis et al. [12], 2004	APD persisted with the persistence of the malignancy	Dermatosis predated the diagnosis of cancer by a few months
Renal cell carcinoma with liver metastasis	1	PF	Kurban et al. [11], 2008	APD persisted with the persistence of the malignancy	Cancer predated the dermatosis by around 8 years

PRC, perforating reactive collagenosis; PF, perforating folliculitis.

Lesions occur predominantly on the trunk and extensor limb surfaces [2, 6, 7]. The incidence in North America ranges between 4.5 and 10% of patients receiving chronic hemodialysis, among whom is a high proportion of black patients [2, 6]. It should be noted that these incidence numbers come from two single-center studies of hemodialysis patients, which can be argued that may not truly reflect the true epidemiology of the disease and simply reflect the low threshold to biopsy all skin lesions. However, given the lack of knowledge of such entities leading to less sampling of such skin lesions, raising awareness through more reporting may lead to identifying more cases and, thus, leading to the true incidence of the disease. Another reason for underreporting is that dialysis patients may perceive an itchy lesion as unimportant in the face of their other health issues and report them less frequently.

Although the exact pathophysiological mechanism for APD in ESRD is unknown, it is thought to result from dermal connective tissue dysplasia and decay [20]. Local trauma and necrosis of the dermal layer may be secondary to injury to the skin induced by frequent scratching in patients with pruritus from CKD [6]. Muller et al. [7] suggested that KD might be a recessively inherited genodermatosis. With our patient, one may postulate that the skin lesions were due to chronic uremic pruritus and not a unique dermatopathological disease such as APD. Three facts in history may suggest that this lesion is an APD lesion rather than a local skin reaction to itching. First, the patient reported that the itching only started after the emergence of the skin lesions, not prior. Second, as mentioned in the case description, local antihistamines were tried to no avail. Third, a single lesion rather than generalized pruritus, often bilateral and not exclusive to a limb, go against uremic or diabetic etiology.

The lesions of APD may resolve spontaneously. Other treatment options have been tried with variable results, including topical retinoids, topical and intradermal steroids, and ultraviolet-B light [6, 8]. In our patient, short-term therapy with topical steroids was unsuccessful.

Rapini et al. [4] suggested “APD” to avoid describing what material is being eliminated throughout the epidermis. However, this nomenclature is more relevant as cases of APD often show perforation of both elastic and collagen fibers. Furthermore, the name encompasses the broad spectrum of perforating disorders in CKD, DM, or both since the different histologic findings probably represent different stages or types of lesions in the same pathologic process [2, 4].

Patient Course

Given the paraneoplastic association that the dermatopathologist highlighted, we evaluated our patient for occult malignancy, guided by standard recommendations for screening based on age and sex as appropriate. In addition, due to his remote history of renal cell carcinoma with unilateral nephrectomy, a computed tomography scan was ordered to evaluate his contralateral kidney. Images have shown a left polycystic kidney disease with large and small malignant masses as well as regional nodal metastasis (Fig. 3).

Fig. 3.

The left kidney is moderately enlarged and deformed due to innumerable small simple cysts, some of which are hyperdense. Additionally, there is a 10 × 7.5 cm heterogeneous, partially calcified, central enhancing mass. There may be at least two to three smaller additional, partially calcified, solid masses in the kidney. An enlarged, para-aortic lymph node adjacent to the left kidney measures 2.2 cm in its greatest dimension (right nephrectomy).

After urological consultation and given our patient’s tumor burden and ESRD status, the decision to proceed with a radical nephrectomy was made. The patient had an uneventful postoperative course. The patient’s nephrectomy revealed renal cell carcinoma, but its final classification proved to be complicated. The tumor demonstrated mixed morphological and immunophenotypical patterns. Some areas showed differentiation of papillary renal cell carcinoma (PRCC), while others showed acquired cystic disease-associated renal cell carcinoma (ACDRCC) characteristics. Therefore, a definitive pathological classification could not be achieved. Genotyping the tumor in an attempt to resolve the different morphology was considered but hurdled by multiple factors. Until this publication, no consensus exists regarding a molecular-aided pathological classification and prognostication system encompassing the full spectrum of RCC, even though many of these questions are being resolved, especially the molecular signature of each tumor [21, 22]. Two possible explanations for the pathological findings can be entertained. These findings can result from intra-tumoral heterogeneity with a preferential selection of two different subclones. Alternatively, two independent tumors could have developed and collided. Regarding our patient, definitive tumor categorization is limited to no clinical implication beyond academic curiosity and information on predicted tumor behavior and outcome.

Nevertheless, two morphologically distinct cancer lineages within the same tumor are well observed in many cancers and renal cell carcinoma [23]. The most common and well recognized is the presence of a sarcomatous component, which correlates with more aggressive behavior and a worse prognosis when exceeding a certain level. Collision tumors are also well documented [24] and most commonly reported to involve a combination of papillary renal cell carcinoma and oncocytomas. In this case, the apparent histological similarities in the two components of cancer argue for clonal selection events in the same cancer. The only true benefit of making the distinction is that our proposed correlation between APD and a particular renal cell carcinoma can be narrowed to a specific pathological entity represented by papillary renal cell carcinoma [25].

The final histopathological examination of the resected kidney demonstrated a mixed pattern containing papillary carcinoma (PRCC) and ACDRCC. As noted by the pathologist, PRCC and ACDRCC share similar morphological and genetic features, making it hard to differentiate between the two tumor patterns [26].

Based on the American Joint Committee on Cancer guidelines (AJCC), the final pathologic staging was pT2a pNx with clinical staging of M0 [27]. Subsequent serial imaging has not shown evidence of locoregional recurrence or distant metastasis, which obviated the need for further chemotherapy. Concerning his dermatosis, he experienced a moderate response to a 6-month topical keratolytic treatment (salicylic acid 5%) with improvement in appearance and symptoms (Fig. 2).

Relevant to our patient, advanced CKD has a well-documented association with acquired cystic kidney disease (ACKD) with malignant transformation [28]. This association was first noted in a CKD patient in 1847 and later documented in an autopsy patient on intermittent maintenance hemodialysis in 1977 [29, 30]. The natural history of ACKD in renal transplant patients who do not undergo native bilateral nephrectomy is less clear [28]. The presence of DM and ESRD in our patient, together with the long vintage of intermittent maintenance hemodialysis, has predisposed him to ACKD with malignant transformation to renal cell carcinoma. His underlying cancer was clinically silent except for KD, which prompted the screening for an underlying malignancy. Here, we hypothesize a paraneoplastic origin of his skin pathology. On the other hand, the commonality of ACKD in HD patients [28] and the increased incidence of KD in ESRD argue for the purely incidental nature of this dermatosis. Nonetheless, paraneoplastic features should be recognized in any high-risk population to pursue further appropriate evaluations, and timely communication between clinicians to raise awareness of any malignancy association is vital.

Statement of Ethics

The publishing of this case report complies with the Wayne State University guidelines regarding IRB approval status and thus is approved for publishing. Ethical approval is not required for this study in accordance with Wayne State University guidelines, given the retrospective nature of the reporting. Written informed consent was obtained from the patient to publish the details of their medical case and any accompanying images.

Conflict of Interest Statement

The author received honoraria from Vifor Pharma within the previous 3 years. The author has no nonfinancial conflicts of interest to declare.

Funding Sources

There were no funding sources for this submission.

Author Contributions

The author confirms sole responsibility for the study conception and design, data collection, analysis and interpretation of results, and manuscript preparation.

Funding Statement

There were no funding sources for this submission.

Data Availability Statement

Further clinical data are not publically available as they are designated health information protected by privacy laws and cannot be shared on data-sharing repository Websites. Further inquiries can be directed to the corresponding author. Supplementary material includes the CARE checklist completed by the author for this case report, a letter to the editor, tables, figures, and responses to reviewers.