Introduction

Pyoderma gangrenosum (PG) is a rare neutrophilic dermatosis. Given its clinical overlap with other ulcerative diseases and lack of pathognomonic histopathology or diagnostic markers, PG is frequently misdiagnosed [14]. Patients are often mistreated with surgery including limb amputations, which have been sporadically reported. Thus, in this multicenter case series, we aimed to explore the clinical characteristics of patients with PG undergoing limb amputation.

Methods

Data were retrospectively reviewed from the electronic medical records at 5 US academic hospitals. A search was conducted for patients with a diagnosis of pyoderma gangrenosum who underwent limb amputation between 2010 and 2020. The study was approved or exempted by the institutions’ ethics boards (IRB number 21-006978). Diagnosis of PG was made by dermatologists in academic centers based on their clinical assessment and skin biopsy. We report a descriptive analysis of patients’ clinical characteristics, course, and outcome.

Results

We identified 10 patients with PG who underwent at least one limb amputation. Six were male (60%). The mean age (range) was 65 years (47- 81 years). All patients had ulcerative-type PG on the lower extremities (Fig. 1). Two patients had additional ulcerative lesions on the abdomen. The mean duration of lesions was 30.6 months (1–72 months). Six patients had PG-related comorbidities (ulcerative colitis, myelodysplasia, or arthritis). Other notable comorbidities included diabetes mellitus (DM) (n = 5), coronary artery disease (n = 5), chronic kidney disease (n = 2), and psoriasis (n = 2). Table 1 summarizes the clinical course for this cohort.

Fig. 1
figure 1

A–C Patient number 10 showing progression of ulcerative pyoderma gangrenosum with typical violaceous undermined ulcers on the lower extremities. D Recurrence of ulcers following amputation

Full size image
Table 1 Clinical characteristics of patients with pyoderma gangrenosum who underwent limb amputation in our series
Full size table

The majority of patients (8 out of 10) were correctly diagnosed with PG prior to amputation, whereas two patients were misdiagnosed with necrotizing soft tissue infections (NSTIs) at the time of amputation. When treatment response was incomplete, alternative diagnoses were considered such as venous ulcers, calciphylaxis, or diabetic foot ulcers.

Six patients had positive wound cultures, two patients had evidence of osteomyelitis. Eight patients developed sepsis and shock-like symptoms, and the diagnosis of NSTIs was considered. All patients received intravenous antibiotics without significant improvement.

Driven by the rapid clinical deterioration of shock-like symptoms, limb amputation was necessary as a life-saving procedure. Below-knee amputation was performed in six patients, and above-knee amputation was performed in four patients. Four patients underwent two distinct amputation events due to shock-like symptoms most likely caused by superimposed NSTI of the affected limb. The histopathology of the amputated anatomic areas, which was available for seven patients, showed necrotic tissue, obliterated thrombotic vessels, and atherosclerotic calcification of medium-sized arteries. Healing of PG ulcers occurred in four patients; however, recurrence of PG at the site of amputation was detected in six patients (Fig. 1D). Two patients died within one week of the amputation due to sepsis-related multi-organ failure.

Discussion

In this case series, most patients with PG undergoing limb amputation had an established diagnosis of PG, with non-healing ulcers for longer than 2 years. A recent review showed that among six patients with PG who underwent amputation, half received an alternative diagnosis of NSTI, and only one underwent a lower limb amputation [8]. Another review on necrotizing neutrophilic dermatoses (NND) reported four limb amputations misdiagnosed as NSTI without providing specific clinical information [13]. Three additional case reports of limb amputations have been published since then (Supplementary Table).

The treatment of PG, which is dependent primarily on immunosuppressive drugs, remains challenging with variable response among patients. The estimated average time for the healing of PG ulcers may reach 11 months [3]. Delayed diagnosis and inadequate response to treatment are reflected in the longer disease duration shown in this study. Concomitant PG and infection are not frequently reported [6], but several patients in our cohort demonstrated evidence of superinfection, suggesting that inadequate response to medical treatment should raise suspicion for superinfection. Distinguishing a secondary infection from an exacerbation of PG is challenging. Given that open wounds are portals of entry to microbial agents, patients with PG, who often have prolonged drug immunosuppression, may have increased susceptibility to infections [1]. Colonization with Pseudomonas, Streptococci, and Staphylococci has been documented in over a third of patients [10]. In the setting of clinical overlap of PG and NSTIs, it is recommended to obtain microbiological swabs and cultures and initiate empiric antibiotics along with urgent surgical consultation [7].

Several patients (4/10) in our cohort developed NSTIs as a complication of surgical debridement or first amputation. Given the risk of pathergy, surgical procedures generally have a limited role in the management of PG; postsurgical recurrence of PG has been reported in 15% of cases [15]. In our cohort, 60% of patients had recurrence after amputation, emphasizing that surgery is not curative for PG and should be approached with caution.

Previous reports of NSTIs as a complication of PG with extensive surgical debridement [9] or amputation [8] might suggest that a deep skin biopsy with detection of necrotic subcutis could aid in earlier diagnosis of NSTIs [9]. However, this intervention might be impractical in an acute setting. The distinction between PG and NSTIs is crucial for treatment choice. Unlike PG, which relies chiefly on medical immunosuppressive treatment, NSTIs are treated with urgent surgical debridement, ideally within 6 h of presentation [11]. In the setting of non-diagnostic clinical and histopathologic features, other clues such as negative cultures, presence of underlying autoimmune diseases, and lack of response to antibiotics with improvement on immunosuppressive medications favor a diagnosis of PG over NSTIs [1, 5, 8]. A real-world evaluation of data of 565 patients with suspected PG showed that skin biopsy was obtained from only 9.4% accounting for an elevated rate of misdiagnosis [13]. It is important to note that a diagnosis of a rare presentation of neutrophilic dermatoses such as NND should be considered in these patients [12].

Half of our cohort had DM. Data about the co-occurrence of PG and DM are sparse [4]. In a case series of 103 patients with PG, 28% had DM [4]. Given the older age of patients, and the known elevated incidence of DM above 50 years, no solid association between DM and PG can be confirmed [4]. NSTIs are known complication in patients with DM and uncontrolled DM increases the risk of infections and impairs healing [12]. We also found a high proportion of our cohort with cardiac or chronic renal disease or both. In a Danish study, cardiac and renal diseases were frequently associated with PG [2]. Cardiac diseases were reported in 45.5% of patients with PG [10], with a possible link of metabolic syndrome and atherosclerosis contributing to systemic inflammation [2, 10].

The multifactorial nature of the ulcers in patients with DM, cardiac and renal disease adds to the complexity of management and might have contributed to inadequate response to PG and infection treatments. Clinicians should be aware of patients with PG who may be at risk of developing NSTIs with subsequent limb amputation. An algorithm for management of PG in elderly patients (above 65 years) is proposed in Fig. 2.

Fig. 2
figure 2

Schematic approach to patients with pyoderma gangrenosum who may be at risk of limb amputation. ABI ankle brachial index; DM diabetes mellitus; CHF chronic heart failure; CKD chronic kidney disease; ICU intensive care unit; SIRS systemic inflammatory response syndrome, defined as occurrence of at least 2 of the following: fever > 38 °C or hypothermia < 36 °C, tachycardia > 90 beats/min, tachypnea > 20 breaths/min, leukocytosis > 12 × 109/l or leukopenia < 4 × 109/l

Full size image

Limitations

The retrospective nature of the study limited collection of exact durations between diagnosis of PG and timing of amputation. Wound culture results were not available for all patients. Diagnosis of PG was made by dermatologists in academic centers based on their clinical assessment and expertise. Additionally, conclusions are limited by the small sample size.

Conclusion

This case series highlights practice gaps in the management of PG, in that amputation may result not only from misdiagnosis but also from superimposed severe infections. Prompt and tailored treatments in more vulnerable patients seem to be essential to prevent devastating complications such as limb amputations. Patients with coexisting lower extremity PG, DM, and/or chronic cardiac or renal disease should be managed with particular care toward preventing infection/NSTIs.