Necrotising fasciitis or pyoderma gangrenosum: A fatal dilemma

Abstract

Necrotising fasciitis (NF) is mostly a polymicrobial, severe soft tissue infection that progresses rapidly, penetrating through the subcutaneous tissue to the fascial planes and the muscles. The pyoderma gangrenosum (PG), on the other hand, is a rare, rapidly progressive (except for the post‐surgical PG), autoinflammatory ulcerative skin and soft tissue condition. In this study, we tried to emphasise the importance of diagnosing the NF as well as the PG. Although these two clinical presentations have some standard features, awareness of different symptoms in detail affect the outcome. Any surgical discipline can face NF or PG and, therefore, should be aware of them to decrease the mortality rate. Forty‐five patients with NF and PG who were treated between January 2008 and October 2018 were included in the study and evaluated retrospectively for age, sex, localisation, onset of symptoms and diagnosis, predisposing factors, characteristics of tissue defects, laboratory findings, Laboratory Risk Indicator for Necrotising Fasciitis (LRINEC) scores, isolated microbiological agents, surgical intervention, and mortality rate. Demographic, laboratory, and clinical data were analysed. Among these 45 patients, 14 patients had PG, and 31 patients had NF. The mean age and SD for the NF and PG groups were 50.80 ± 17.67 and 50.78 ± 12.72, respectively. Five patients had rheumatological disorders; four patients had diabetes mellitus (DM) in the PG group. Males had higher risk than females in NF (odds ratio [OR] = 0.077, 95% confidence interval [CI] 0.017‐0.34), and females had higher risk in PG (relative risk [RR] = 5). We compared the LRINEC score of NF patients with PG patients. The mean value of this score was 4.53 for PG patients, and 6.06 for NF patients. Fifteen patients (33.3%) had a radiological evaluation. MRI, CT, and USI were used as imaging modalities. Necrotising fasciitis and PG are two distinct entities that are in general difficult to distinguish. Therefore, differential diagnosis and rapid treatment are crucial for lowering the mortality rate.

1. INTRODUCTION

First introduced by Wilson in 1952 as a more accurate description, necrotising fasciitis (NF) is an uncommon condition with an incidence rate of 4 to 13/1 000 000 annually.1, 2 It is mostly a polymicrobial, severe soft tissue infection that progresses rapidly, penetrating through the subcutaneous tissue to the fascial planes and the muscles. The destruction rate of the fascia could reach up to 2 to 3 cm/h.3 When the vascular supplies of the soft tissues are affected by the infection, skin necrosis and haemorrhagic bullae formation occur. In the literature, the authors reported overall mortality rates differently. The distribution width is approximately 20% to 30%.4, 5 Therefore, early diagnosis and aggressive surgical debridement are crucial for reducing morbidity and mortality. The NF is divided into two categories as primary and secondary, depending on the aetiology.1 Trauma is the most common identifiable aetiology. The affected areas are mostly the lower extremities, the abdomen, the perineum, and the upper extremities.6 Diabetes mellitus (DM) is the most common accompanying disease in NF patients, accounting for 44.5% to 72.3% in various series.7 Clinicians diagnose the NF according to patient status, physical examination, laboratory tests, and radiological assessment. Disproportional severe pain with passive motion is clinically relevant to the NF. Laboratory‐based scoring systems, such as Laboratory Risk Indicator for Necrotising Fasciitis (LRINEC), proposed by Wong et al, can be used as a guide.8 Additionally, there are different scoring systems for evaluation such as the APACHE II scoring system for mortality rate3 and the Fourniers Gangrene Severity Index for determining the surgical debridement.9 Although CT scan is widely used, magnetic resonance imaging (MRI) has the highest sensitivity rate (93%‐100%).1 First step of the treatment is administering broad‐spectrum antibiotics, and taking culture sample. Antibiotics should be revised according to culture results. The infection is mostly polymicrobial, but Methicillin Resistant Staphylococcus Aureus, Clostridium spp., and fungal pathogens must be kept in mind. The primary treatment modality is an aggressive surgical debridement of the affected tissues. As a rule, the surgeon should extend the debridement until the healthy tissue. Usually, multiple surgical interventions with topical wound care are mandatory to control the infection. In addition, vacuum‐assisted closure (VAC) is an ancillary treatment modality. The anatomic location, the extent of the defect, and the surgeons' preference are factors for determining an appropriate reconstructive option.

The pyoderma gangrenosum (PG), on the other hand, is a rare, rapidly progressive (except for the post‐surgical PG), autoinflammatory ulcerative skin and soft tissue condition. It affects approximately three to 10 patients per million.10 The precise aetiology remains unclear, but instability in the immune system, altered interleukin (IL) and tumour necrosis factor (TNF) release, and systemic inflammation under genetic predisposition are thought to be related to the pathogenesis of PG. PG could present in association with most common inflammatory bowel diseases, rheumatologic disorders, and haematological disorders.11, 12 This uncontrollable immune reaction can be seen in any anatomic location, but mostly occur on the lower extremities and the trunk. There are five different types of PG described in the literature; namely, bullous, pustular, ulcerative, vegetative, and atypical bullous types (Figure 1).13 The Delphi Consensus PG Diagnostic Criteria and the PARACELSUS Score are used to avoid misdiagnosis.10 A sterile incisional biopsy from the ulcer border including the subcutaneous tissue is required to rule out an infectious aetiology.11 Surgical debridement is not a recommended treatment modality unless there are skin and soft tissue necroses. Systemic and topical corticosteroids, cyclosporine, TNF‐α inhibitors, IL‐1 receptor antagonists, Janus kinase inhibitors, and PDE4 inhibitors can be administered for regression.10

Figure 1.

Necrotising fasciitis after suicide attempt

In this study, we tried to emphasise the importance of diagnosing the NF as well as the PG. Although these two clinical presentations have some standard features, details affect the outcome. Any surgical discipline can face NF or PG and, therefore, should be aware of them to decrease the mortality rate.

2. MATERIALS AND METHODS

Forty‐five patients with NF and PG who were treated between January 2008 and October 2018 were included in the study and assessed retrospectively for age, sex, localization, onset of symptoms and diagnosis, predisposing factors, characteristics of tissue defects, laboratory findings, LRINEC scores, isolated microbiological agents, surgical intervention, and mortality rate (Table 1).

Table 1.

Laboratory Risk Index for Necrotising Fasciitis (LRINEC) score

Parameter	LRINEC score
CRP (mg/mL)
>150	0
<150	4
WBC (cell/mm3)
<15	0
15‐25	1
25	2
Hb (gr/dL)
>13.5	0
11‐13.5	1
<11	2
Na (mmol/L)
>135	0
<135	2
Cr (mg/dL)
<1.6	0
>1.6	2
Glucose (mg/dL)
<180	0
>180	1

Abbreviations: CRP, C‐reactive protein; WBC, white blood cells.

NF patients took broad‐spectrum antibiotics immediately after referral to the infectious disease unit. Tissue samples were taken from the wound surface for aerobic culture. Surgical debridement was performed until the healthy tissue was located under the necrotic fascia planes. We showed utmost care to protect the functional structures, such as the penis, the scrotum, and the perineum during the initial debridement. Standard wound care was applied to tissue defects. VAC was also used. In most patients, debridement was performed at least two to four times in the first 48 hours.

Contrarily, if there is a rapidly progressive wound that is non‐responsive to surgical intervention, topical wound care, and systemic antibiotics clinical suspicion, the criteria as mentioned above will be necessary for diagnosing the PG. Therefore, we questioned any diseases and conditions associated with PG and took a medical history in detail. Punch or incisional biopsy samples were taken from multiple areas including the subcutaneous tissue.

For statistical analysis, variables were tested for statistical significance with Pearson χ² tests and Fisher exact test, whereas differences in continuous non‐parametric variables were analysed with the Mann‐Whitney U test and independent samples T‐test. Statistical significance was defined by a P value less than .05. Data analysis was carried out using the Statistical Package for Social Sciences version 22 (SPSS Inc., Chicago, Illinois).

All authors hereby confirm that guidelines on patient consent have been met and any details of informed consent obtained are indicated within the text of the submitted manuscript. We assure that informed consent was obtained from each patient and the ethical committee of our hospital approved our clinical study.

3. RESULTS

Forty‐five patients who were diagnosed with NF and PG were assessed retrospectively. Demographic, laboratory, and clinical data were analysed. Among these 45 patients, 14 patients had PG, and 31 patients had NF. Of the 14 patients in the PG group, 10 (71.4%) are female and 4 (28.6%) are male. The mean age was 50.78 ± 12.72. Twenty‐six patients in the NF group (83.8%) are male, and 5 (16.2%) are female. Their mean age was 50.80 ± 17.67. Anatomic distribution of the diseases is shown in Table 2.

Table 2.

Anatomic distribution of necrotising fasciitis and pyoderma gangrenosum patients

Anatomic distribution	Necrotising fasciitis	Pyoderma gangrenosum
Genital region	12	—
Thigh	11	1
Perineal region	10	1
Leg	5	3
Gluteal region	2	1
Foot	2	2
Chest	—	2
Back	—	2
Forearm	1	—
Arm	1	‐
Hand	—	1
Total	44	13

Five patients had rheumatological disorders; four patients had DM in the PG group. Twelve patients had DM in the NF group. Comorbid diseases are shown in Table 3 in detail. In the PG group, 10 patients (71.4%) had negative aerobic wound cultures, overall 22 patients (70.9%) had at least one identified bacteria in their samples. For all patients in the NF group, the LRINEC score was evaluated according to the laboratory results at the time of admission and for the PG group right after the diagnosis. In the NF group, 13 patients (41.9%) had lower than 5 points, 10 patients (32.2%) had over 8 points, and eight patients had 6 to 7 points (25.9%). Only two patients had over 8 points in the PG group.

Table 3.

Comorbid diseases which were accompanying to the necrotising fasciitis and the pyoderma gangrenosum

Comorbid diseases	Necrotising fasciitis	Pyoderma gangrenosum
Diabetes mellitus	12	4
Hypertension	3	3
Rheumatological disorders	2	5
Coronary artery disease	2	2
Congestive heart failure	1	2
Chronic renal failure	1	1
Hypothyroidism	—	2
Total	21	19

We observed the tissue defects from the diagnosis until the treatment process. Mainly, in the NF group there were necrosis, ulceration, hyperaemic zones, crepitation on the affected tissue, and in severe cases there were also purulent drainage and haemorrhagic bullae. Contrarily, in PG group the forefront feature of the wound was the loss of epithelial integrity with inflammatory reactions. Seven patients (50%) had a surgical operation history before PG onset. Pathological findings for the NF group were gangrenous necrosis, fat necrosis, and suppurative inflammation, ulceration, and abscess foci, and for PG group these were liquefaction necrosis, neutrophilic dermatoses, subepidermal lysis, inflammation, and ulceration. In total, 15 patients (33.3%) had a radiological evaluation, such as MRI, computed tomography (CT), and ultrasound imaging (USI). In NF patients, severe oedema, fascial and subcutaneous tissue thickening, and air within the soft tissue were common findings. In PG patients, however, there were no specific findings. While medical therapy, such as antibiotics, local wound therapy, was used in both groups, systemic and local steroids were used only in the PG group. We also used VAC for reducing the exudate, decreasing the bacterial count, and increasing the blood flow and the oxygen concentration. Immediate surgical debridement was performed until clinical regression, and healthy granulation tissue was achieved in the NF group. Mostly, split‐thickness skin graft and fasciocutaneous flaps were used for reconstruction. Three patients passed away because of disease progression. We preferred elective reconstructive surgical operation for the PG group after the regression of the disease.

4. DISCUSSION

NF is a rare, rapidly progressive disease. Progression is usually fatal if the condition is not diagnosed promptly. In the beginning, NF may have similar symptoms with other soft tissue infections, leading the clinician to confusion. Patients with NF have the classical triad of symptoms: disproportionate local pain, swelling, and erythema (Figure 2, Figure 3, and Figure 4), respectively.14 Additionally, associated comorbid diseases, progressive clinical deterioration, laboratory findings, and subcutaneous gas on imaging all help to guide the clinician in early diagnosis and surgical treatment.

Figure 2.

Necrotising fasciitis of genital region

Figure 3.

Necrotising fasciitis after the car accident

Figure 4.

A, Abdominal region after C/S, B, TRAM flap reconstruction C, herbal medicine usage

On the other hand, PG is an auto‐inflammatory, neutrophilic dermatosis without a specific aetiology. Unknown pathophysiology and diverse clinical presentations may lead to misdiagnosis, especially to confusing with other soft tissue infections or NF. Nevertheless, surgical intervention before clinical onset, progressive cellulitis, and ulceration that does not respond to broad‐spectrum antibiotics could be clues for differential diagnosis. The mean age and SD for the NF and PG groups were 50.80 ± 17.67 and 50.78 ± 12.72, respectively. When we performed statistical analysis, we confirmed that there were no any significant differences between age groups (P = .508). As a result, in our study, the age distribution was similar to those reported literature for both groups.1, 13There were 31 male (68.9%) and 14 female (31.1%) patients in total. Based on our results, we concluded that there was an association between genders and diseases (P = .001). Males had a higher risk than females in NF (odds ratio [OR] = 0.077, 95% confidence interval [CI] 0.017‐0.34), and females had a higher risk in PG (relative risk [RR] = 5). It seems that NF has a preference for men and PG inclines for women. Our findings were similar to those reported in the literature.6, 15, 16 These results suggest several assumptions, such as coexisting diseases, for instance, as well as that rheumatologic disorders, are more common in the female population, and males are more prone to trauma than females, and diabetes incidence is higher in the male population.

Anatomic localisation distribution was similar to the literature. Except in one patient who had PG in his hand, the most frequent sites were the lower extremities and the trunk.17Our data promote the previous literature that describes the predominance of these localisations. There are also studies that show different and rare anatomical sites.18

Initial symptoms and clinical features of NF and PG might be similar and the diagnosis could be time‐consuming. The LRINEC proposed by Wong et al⁸ is a laboratory‐based scoring system, which can help to identify NF patients promptly (Table 1). Unfortunately, even if we can use the LRINEC score for NF, it will not applicable for PG. According to this score, patients can be classified into risk categories. We compared the LRINEC score of NF patients with PG patients. The mean value of this score was 4.53 for PG patients, and 6.06 for NF patients. We did not find any significant differences between the two groups (P = .095). As a result, we thought that the LRINEC scoring system is not an appropriate option for differential diagnosis.

Bacterial involvement and identification are essential for antibiotic usage. In our study, 25 (55.5%) of 45 patients had at least one bacterium identified in their tissue culture, and 21 of them were in the NF group. In the present study, we observed that bacterial involvement was significantly higher in the NF group than the PG group (P = .014). Polybacterial infections were the most frequent among the other groups (33%). The latter is expected because NF is, by definition, a form of deep tissue infection whereas PG is a reactive, non‐infectious unlike that reported the literature, in our study cohort, we identified Pseudomonas as a predominant gram‐negative microorganism and Enterobacter as a gram‐positive microorganism.2, 19

Treatment modalities of these two distinct clinical conditions are different from each other. After the initial diagnosis, aggressive surgical debridement up to the healthy tissue borders, and broad‐spectrum antibiotic administration are paramount in NF patients. Besides, surgical debridement could worsen the situation and antibiotics could not enhance the inflammatory reaction in PG patients. Skin biopsies should be performed as soon as possible in the case of clinical suspicion. In NF patients, surgeons take pathological samples in the operation room because of the rapid progress.

Conversely, we observed that the clinical status of PG patients worsens slower than that of NF patients; hence, there is a chance to perform multiple incisional biopsies, after which systemic and local steroid therapy and even immunosuppressive treatment could be an option depending on the results. Several criteria have been proposed15 for avoiding mistreatment as unsuitable or delayed treatment result in devastating outcomes.

Radiological imaging modalities can help to establish the diagnosis of NF and PG. In NF patients, CT can show fascial swelling, oedema, and gas formation, which strongly indicates Clostridium species. An MRI also has high sensitivity and accuracy but is more expensive6 and time‐consuming. USI is a fast and beneficial option in the ICU, providing useful information on Fournier's gangrene patients. Although USI is operator dependent, it can show the extension of oedema, inflammation, and air formation. In PG patients, radiological evaluations show cellulitis, oedema, local fluid accumulation, and while it appears mostly after surgical operations, particularly CT can show any post‐surgical changes. Nevertheless, these findings are very non‐specific and only constitute a clue for a definitive diagnosis.

5. CONCLUSION

NF and PG are two distinct entities that are in general difficult to distinguish. Because of the rapidly progressive nature of these types of diseases, we composed a chart (Table 4) that shows the key diagnostic points to help the clinician for diagnosis.

Table 4.

Differences and similarities between the necrotising fasciitis and pyoderma gangrenosum

Data	Necrotising fasciitis	Pyoderma gangrenosum
Age	40‐60	40‐60
Gender	Male > female	Female > Male
Comorbid diseases	Diabetes mellitus	Inflammatory disorders
Anatomic localization	Genitals, lower extremity	Lower extremity, trunk
Microbial agent	Mostly polymicrobial	Mostly negative tissue culture
Tissue defect type	Necrotic, hyperaemic, ulcerative, purulent drainage	Hyperaemic, ulcerative, necrotic (severe cases)
Radiological evaluation	Oedema, fascial swelling, gas formation	Cellulitis
Pathological findings	Gangrenous necrosis, fat necrosis, ulceration	Liquefaction necrosis, inflammation, ulceration
Diagnosis	Clinical	Incisional biopsy
Treatment modalities	Immediate surgical debridement, broad‐spectrum antibiotic	Steroid, local wound therapy

CONFLICT OF INTEREST

All authors hereby declare that they have no conflicts of interest to disclose.

ETHICS STATEMENT

We confirm that Ethical Committee approval was sought where necessary and is acknowledged within the text of the submitted manuscript.

Demirdover C, Geyik A, Vayvada H. Necrotising fasciitis or pyoderma gangrenosum: A fatal dilemma. Int Wound J. 2019;16:1347–1353. 10.1111/iwj.13196