Pancreatic Panniculitis Secondary to Pancreatic Cyst With Lower Extremity Wounds as a Primary Presenting Feature: A Case Report and Literature Review

Pancreatic panniculitis is a rare skin finding that occurs in up to 3% of patients with pancreatic pathologies, most commonly due to pancreatitis.^1,2 It may also develop in individuals secondary to cholelithiasis, acinar carcinoma of the pancreas, or pancreatic pseudocysts. Patients present with painful, erythematous nodules that can progress to an ulcer with drainage of an oily liquid.^1,3 The pathogenesis of pancreatic panniculitis is currently unknown but is thought to involve the release of amylase and lipase into the bloodstream. These digestive enzymes then enter adipose tissue and cause necrotic destruction. Cutaneous lesions may manifest before, during, or after clinically detectable pancreatic disease.⁴ Nodules resembling chronic wounds may be the chief presenting symptom and should prompt suspicion and further work-up to accelerate the diagnosis of the underlying cause.^5,6 Here, the authors report a case of pancreatic panniculitis caused by a pancreatic cyst.

A 68-year-old White man presented to the emergency department with bilateral leg swelling up to the knees and painful, burning wounds on his lower extremities of 2 months’ duration. The patient was previously prescribed antibiotics and steroids for suspected pyoderma gangrenosum, which had not led to any improvement. He also described intermittent, bloating epigastric pain that started after his lesions first appeared.

Physical examination revealed large, tender, erythematous nodules scattered bilaterally over the ankles and anterior legs as well as ulceration on the left forefoot (Figure 1). Arterial and vascular ultrasounds appeared normal. Results of laboratory studies revealed elevated serum lipase (>4400 U/L; reference range, 17–40 U/L), amylase (2495 U/L; reference range, 28–100 U/L), erythrocyte sedimentation rate (48 mm/hour; reference range, 0–20 mm/hour), and CRP (1.6 mg/dL; reference range, 0–0.744 mg/dL). CT of the abdomen and pelvis revealed portal vein thrombosis and intrahepatic bile duct dilation (Figure 2A); no evidence of pancreatitis was noted.

The patient was admitted with a diagnosis of acute pancreatitis as well as COVID pneumonia, although the patient had no respiratory symptoms. Multiple 3- to 6-mm punch biopsies of the lower extremity wounds were obtained, which revealed a neutrophilic and eosinophilic infiltrate within the small vessels, venules, and arterioles of the deep dermis as well as extensive extravasation of erythrocytes. The patient was prescribed high-dose dexamethasone, which led to mild improvement in his lesions. His condition stabilized, and he was discharged home.

Three months after the initial emergency department visit, the patient presented to the hospital again with worsening rash over the lower extremities. He reported increased nodules, pain, and ulceration. Results of laboratory studies showed serum lipase of 561 U/L and CRP of 21.6 mg/dL. Contrast-enhanced CT of the abdomen showed a heterogenous pancreatic mass resembling a cyst with apparent tumor thrombus extending into the portal vein without occlusion. Extensive retroperitoneal heterogenous soft tissue and fluid and adenopathy in the pelvis were described as worrisome for metastatic disease (Figure 2B). Antigen testing showed elevated levels of CA 19-9 (119.3 U/mL) and carcinoembryonic antigen (7.0 ng/mL). However, cytology revealed necrosis with no malignant cells. Excisional biopsy was then performed, which demonstrated marked saponification of the subcutaneous fat associated with ghost adipocytes, neutrophilic infiltrate, hemorrhage, basophilic calcifications, and hematoidin pigment (Figure 3). No vasculitis was found.

A diagnosis of pancreatic panniculitis was made. Because the patient’s wounds had developed increased purulence and a muddy brown drainage, an excisional debridement was performed. He was seen at the outpatient wound clinic for further management of his panniculitis, which consisted of additional sharp debridements, manuka honey wound gel and calcium alginate daily dressing changes, and soap and water cleansing.

One week later, the patient was readmitted to the hospital for worsening abdominal pain, anorexia, and chills. Results of laboratory studies showed a leukocyte count of 21 300 cells/µL (reference range, 4500-11 000 cells/µL) and serum lipase concentration of 296 U/L (reference range, 17-40 U/L). Due to the symptom progression and unresolving necrotizing pancreatitis, the patient was taken to the operating room for a pancreatic debridement and Whipple procedure. Postoperative complications included abdominal wound dehiscence treated with dressing changes using Manuka honey dressings then switched to negative pressure wound therapy. While old lesions were still present, the patient did not develop any new lesions.

On follow-up 1 month after debridement, the wounds had improved dramatically, with most healing except for a distal lesion on the lateral aspect of the right leg. However, with ongoing gauze pack and diluted sodium hypochlorite solution, the lesion continued to slowly improve (Figure 4).

Panniculitis refers to a broad category of disorders involving inflammation of the subcutaneous fat.³ The authors highlight this case of pancreatic panniculitis caused by a pancreatic cyst because of the varied histopathologic findings on skin biopsies and delayed discovery of the underlying etiology.

Pancreatic panniculitis can be a challenging diagnosis to make for several reasons: (1) the type and timing of the skin biopsy influences diagnostic accuracy; (2) the histopathology is often similar to other diseases; and (3) the histopathology alone does not identify the root cause, so skin biopsy must be combined with the full clinical picture and other extracutaneous findings.

First, it is important that the diagnosis be made through deep skin biopsy via excision.³ Punch biopsy is a first-line diagnostic tool for skin disorders and provides a full-thickness (4 mm deep), cone-shaped tissue specimen with the widest diameter along the skin surface and the narrowest dimension at the base. Even through this method, however, the biopsy may not contain a sufficient sample of subcutaneous fat, which is the layer most affected in pancreatic panniculitis.⁷ Inconclusive biopsy is most likely to occur with smaller punch biopsies. This patient received multiple tissue punch biopsies ranging from 3 mm to 6 mm in size, but this was not enough to make the diagnosis. Only after excisional biopsy did the diagnosis become apparent. To achieve the highest diagnostic yield, the biopsy specimen must contain an adequate amount of subcutaneous fat. This is best accomplished by a large punch biopsy 6 mm to 8 mm in diameter, but even more preferably by elliptical excision with scalpel.⁷

Aside from overall biopsy depth and quality, the timing of specimen collection influences diagnostic accuracy. On early histology, pancreatic panniculitis is a lobular and nonvasculitic pathology that presents with neutrophilic infiltrate, septal involvement, and liquefactive fat necrosis. As the disease progresses, histology reveals a granulomatous pattern.^8,9 Chronic pancreatic panniculitis demonstrates less fat necrosis and fewer ghost cells with infiltrate consisting of lymphocytes, macrophages, and Langerhans giant cells.¹⁰ Eosinophilia, if identified, is associated with underlying pancreatic carcinoma.¹⁰

At 2 months after symptom onset, this patient’s skin biopsy demonstrated neutrophilic and eosinophilic infiltrate within the deep dermis, but there was no evidence of adipocyte necrosis. It was not until 5 months from symptom onset that the patient’s skin biopsy produced hallmark findings of ghost adipocytes, which indicate saponification of the subcutaneous fat. The discovery of this histopathologic feature was delayed, but this was more likely due to poor specimen collection rather than a true change in the disease course. It should be noted that a lack of ghost cells and fat necrosis does not exclude the diagnosis of pancreatic panniculitis because chronic pancreatitis may present this way. Repeat biopsy may be required throughout disease progression (for example, during flares or acute episodes) before the most definitive changes characterizing pancreatic panniculitis are captured on histopathologic evaluation.

Importantly, hallmark histological features alone are insufficient in determining the underlying cause of pancreatic panniculitis. This is because pancreatic panniculitis shares histopathological similarities with several other diseases.⁸ For instance, one study found that histologic examination of subcutaneous mucormycosis contained structures that resembled ghost adipocytes.¹¹ Erythema nodosum, calciphylaxis, malignancy, leukocytoclastic vasculitis, infection, and trauma can also mimic pancreatic panniculitis.¹² To narrow down the differential diagnosis, panniculitis can be further classified histologically based on the involvement of the lobules or septa and the presence or absence of vasculitis.

Systemic symptoms and extra-cutaneous findings are even more critical for determining the underlying cause because histopathologic pancreatic panniculitis can present in the absence of pancreatic disease.¹² For example, a case series of 3 patients reported by Milani-Nejad et al demonstrated that ghost cells, fat necrosis, calcification, and basophilic degeneration were present on histopathology, yet none of the patients had underlying pancreatic disease. Enzyme levels were unremarkable, and pancreatic imaging was normal.¹³ In the current case report, the initial working diagnosis was pyoderma gangrenosum versus vasculitis. The first CT scan of the abdomen was inconsistent with pancreatitis despite significantly elevated lipase levels. The patient’s worsening abdominal pain while under care at the wound clinic prompted a second CT and excisional biopsy. In the end, it was the excisional biopsy in combination with the patient history, second CT scan showing a pancreatic mass, and laboratory findings of persistently elevated inflammatory markers and pancreatic enzyme levels that helped to clarify the underlying cause.

The existing literature on pancreatic panniculitis discusses common etiologies, symptoms, patient demographics, and diagnostic timeline. A study of 36 pancreatitis and 17 pancreatic carcinoma cases found that skin lesions were the presenting complaint in 22% and 65% of the patients in these groups, respectively.⁶ In another review, almost 50% of the 131 pancreatic panniculitis cases analyzed involved cutaneous lesions that were noted before the diagnosis of the underlying disease.⁵ On average, it took approximately 2.7 months from the appearance of the first lesion to diagnose the patient with pancreatic panniculitis, and the mean patient age was 54.8 years.⁵

The patient’s presentation in the current case report is consistent with these literature findings. Lower extremity skin nodules preceded any abdominal symptoms by several weeks and were the patient’s initial chief complaint. His skin lesions did in fact indicate underlying pancreatic disease. Delayed excisional biopsy led to a diagnosis 5 months after the first lesion appearance. While pancreatic acinar cell carcinoma is the most common neoplastic origin of pancreatic panniculitis, pancreatic panniculitis itself is not pathognomonic for malignancy.⁵ This patient had elevated CA 19-9, but cytology showed no malignant cells. Instead, pancreatitis secondary to pancreatic cyst necrosis was the etiology of the panniculitis.

While the exact pathophysiology of pancreatic panniculitis is unclear, a widely accepted theory is that pancreatic enzymes can reach the skin and cause subsequent subcutaneous fat necrosis. Out of 112 pancreatic panniculitis case reports containing information on serum enzyme levels, more than 98% indicated elevated levels of lipase and/or amylase.⁵ The median serum lipase concentration was 3942.5 U/L, slightly less than the 4400 U/L measured in this patient.⁵ Interestingly, however, in vitro experiments seem to suggest a pathologic mechanism involving more than just enzymatic destruction by amylase or lipase alone. When human adipose tissue was incubated with serum containing pancreatic enzymes from 2 patients—one healthy and one unhealthy; both were exposed and none resulted in histopathological fat necrosis, even at the highest concentrations.^13,14 In addition to elevated pancreatic enzyme levels, other factors, including previous vascular damage or cytokine release, may also contribute to the development of pancreatic panniculitis.¹³

Treatment of pancreatic panniculitis involves proper management of the underlying disorder and supportive care with such agents as non-steroidal anti-inflammatory drugs and corticosteroids.⁹ Wounds in patients with a treatable pancreatic pathology may resolve with treatment.¹² However, pancreatic panniculitis caused by pancreatic malignancy tends to be associated with a poor prognosis.⁵ This patient was initially treated with high-dose dexamethasone, which led to mild improvement. He then underwent multiple sharp wound debridements followed by manuka honey wound gel and calcium alginate dressing changes. Ultimately, the wound was cleansed with diluted sodium hypochlorite solution and packed with gauze, a strategy that was continued throughout his last hospitalization until discharge. Lesion improvement was noted after he underwent surgery to treat his underlying pancreatic disease.

Although not trialed in this patient, the somatostatin analogue octreotide has been reported to provide symptomatic relief in some neoplastic-related cases of pancreatic panniculitis.⁵ Overall, therapeutic algorithms for pancreatic panniculitis are hard to establish because there are few reported cases in the literature. The majority of the published data are from single case reports, and malignancy-associated cases often result in early mortality.

This paper represents the authors’ personal experience with this case; the current report is limited in its nature of a single case. Larger clinical series are needed for more comprehensive review and understanding of this condition.

Erythematous, ulcerating skin nodules over the lower extremities can be an early manifestation and presenting concern in patients with underlying pancreatic disease. Cutaneous findings consistent with pancreatic panniculitis are also a possible “facultative” paraneoplastic syndrome of pancreatic and pancreatic-type cancers.⁵ Therefore, persistent skin nodules should not be interpreted in isolation but rather further investigated to identify the etiology. If panniculitis is suspected, early excisional biopsy should be obtained so that there is enough deep tissue sample for analysis. Hallmark histopathologic features of ghost adipocytes with surrounding granular basophilic calcifications may warrant pancreatic enzyme level measurements, abdominal imaging, and/or procedures. When combined with other aspects of the clinical presentation, particularly gastrointestinal and constitutional symptoms, biopsy findings may help guide the diagnosis toward pancreatic panniculitis caused by an underlying pancreatic condition.