Research letter
Comparison of pyoderma gangrenosum and Martorell hypertensive ischaemic leg ulcer in a Swiss cohort DOI: 10.1111/bjd.15901 Pyoderma gangrenosum (PG) is a rare neutrophilic dermatosis presenting with painful and sterile skin ulcerations.1 Its aetiology remains largely unknown, although an autoinflammatory background seems possible. Several comorbidities, as well as triggering factors such as surgery, trauma or pharmacological therapies, have been associated with the development of PG.2 Different topical and systemic treatments are recommended for PG, most commonly topical steroids or calcineurin inhibitors, as well as systemic steroids, dapsone, infliximab and others. Our group and others have also used canakinumab and ustekinumab.3,4 PG is a diagnosis of exclusion and is often misdiagnosed at initial manifestation. Martorell hypertensive ischaemic leg ulcer (HYTILU) was previously investigated by our group; we showed that 50% of patients with a referral diagnosis of PG were found to have HYTILU.5 HYTILU is caused by ischaemic subcutaneous arteriolosclerosis; all patients show arterial hypertension and up to 58% have diabetes mellitus. Typical clinical presentation of HYTILU is a laterodorsal lower-leg ulceration with central black necrosis and purple inflamed ulcer borders.5,6 Clinically, this appearance can be misleading for PG; owing to the risk of septicaemia, immunosuppression could be fatal in HYTILU.5 In the current study we compared a cohort with PG with a cohort with HYTILU, focusing on clinical data, laboratory findings and comorbidities to develop diagnostic clues for both diseases. A keyword search for ‘pyoderma’ and/or ‘gangrenosum’ was conducted in the files for all patients who were hospitalized between 1 January 2002 and 31 December 2012 at the Department of Dermatology, University Hospital of Zurich (USZ), Cantonal Hospital of Sankt Gallen and at the private practice of W.K. (USZ). In total, 179 patients were identified with an initial suspected differential diagnosis of PG, of whom 38 were diagnosed with PG. We performed a histopathological reassessment of these 38 patients and identified three patients with histopathological signs of HYTILU and one with morphoea, leading to exclusion of these four patients from further analysis. The remaining 34 patients fulfilled the criteria for PG as established by Su et al. and were included in our study.7 These patients with PG were retrospectively compared with a cohort of 32 patients with HYTILU diagnosed at USZ © 2017 British Association of Dermatologists
during the same period. Medical records and laboratory findings were analysed for both cohorts in order to identify features supporting clinical distinction (Table 1). The study was approved by the local ethics committee (KEK-ZH 2014-0432). GraphPad Prismâ 7.0b 2016 (GraphPad, La Jolla, CA, U.S.A.) and Excelâ 14.3.2 2011 (Microsoft, Redmond, WA, U.S.A.) were used for statistical analyses. Compared with PG, HYTILU was less frequent in women and at older age; more common in smokers; showed higher levels of C-reactive protein (CRP) but lower levels of blood leucocytes and neutrophils; showed lesion localization at the lower leg; and showed more cardiovascular comorbidities such as arterial hypertension, diabetes mellitus, peripheral artery occlusive disease and metabolic syndrome, and more microbial superinfection (Table 1). Table 1 Comparison of 32 patients with Martorell hypertensive ischaemic leg ulcer (HYTILU) and 34 patients with pyoderma gangrenosum (PG) HYTILU (n = 32) Female sex Age at manifestation (y) BMI (kg m 2) Smoking Alcohol Laboratory (mean) CRP (mg L 1)b Leucocytes (9 109 cells L 1)c Neutrophils (9 109 cells L 1)d Lesion localization on lower leg Cardiovascular comorbidities Arterial hypertension PAOD Hypertensive heart disease Myocardial infarction Other cardiopathy Cerebrovascular infarction Metabolic syndrome Diabetes mellitus Thrombosis Renal insufficiency Positive microbiological swab
PG (n = 34)
P-valuea
50 73.5 28.3 25 13
62 61.2 24.1 21 12
NS ≤ 0.001 ≤ 0.05 NS NS
31.5 9.9 7.9 100 100 100 62 41 25 12 22 66 53 3 28 91
14.5 10.5 8.4 67 79 29 6 12 6 3 3 3 9 9 21 44
NS NS ≤ 0.05 ≤ 0.001 ≤ 0.01 ≤ 0.001 ≤ 0.001 ≤ 0.01 ≤ 0.05 NS ≤ 0.05 ≤ 0.001 ≤ 0.001 NS NS ≤ 0.001
Data are % unless otherwise indicated. NS, not significant; BMI, body mass index; CRP, C-reactive protein; PAOD, peripheral artery occlusive disease. aUnpaired t-test with Welch correction. b Reference value < 5 mg L 1. cReference value 3.5–9.6 9 109 cells L 1. dReference value 1.4–8.0 9 109 cells L 1.
British Journal of Dermatology (2018)
1
2 Research letter
Our findings of elevated CRP levels and neutrophilia in PG indicates the presence of systemic inflammation, which is increasingly suggested by the field.8 Additionally, the reduced likelihood of pathogenic bacteria in PG lesions could be explained by the exacerbated neutrophil response. Cardiovascular comorbidities like arterial hypertension, peripheral artery occlusive disease, metabolic syndrome and diabetes mellitus appear significantly more often in HYTILU, which has also been reported previously by our group, compared with PG. After age adjustment the prevalence of cardiovascular comorbidities in our small PG cohort seems to correspond to the normal Swiss population; however, this needs to be confirmed in larger studies.5 Furthermore, all patients with HYTILU in our cohort had an ulcer located on the lower leg; as such, a lesion located away from the lower leg could be a further clinical hint favouring the diagnosis of PG (lower leg 67%, trunk 24%, upper limb 5%, head 4%). However, both cohorts were small and larger cohorts are needed to confirm our findings. PG and HYTILU are important differential diagnoses for ulcerative skin disorders and should be considered carefully. In particular, if the biopsy taken in PG and HYTILU is too small, the chance of visualizing ‘PG-like features’ of neutrophilic infiltration and missing the typical features of HYTILU (subcutaneous stenotic arteriolosclerosis in 100% and medial calcification in 71% of cases) is high. Our findings suggest that differences in lesion localization, laboratory parameters and presence of comorbidities could also be utilized in order to differentiate these disorders accurately. We also identified increased inflammatory markers in a considerable proportion of our patients with PG, which is an expected but not formally proven aspect of PG. Larger prospective studies or international registries on these rare ulcerative skin diseases are needed to confirm these findings. 1
Department of Dermatology and Department of Immunology, Z€urich University Hospital, Z€urich, Switzerland 3 Department of Dermatology, Venerology and Allergology, Kantonsspital St Gallen, St Gallen, Switzerland 4 Kempf und Pfaltz Histologische Diagnostik, Z€urich, Switzerland E-mail:
[email protected] 2
References 1 Meier B, Maul JT, French LE. [Pyoderma gangrenosum and Sweet’s syndrome: Cutaneous manifestations of autoinflammatory disorders]. Hautarzt 2016; 67:934–9 (in German). 2 Powell FC, Su WP, Perry HO. Pyoderma gangrenosum: classification and management. J Am Acad Dermatol 1996; 34:395–409. 3 Kolios AG, Maul JT, Meier B et al. Canakinumab in adults with steroid-refractory pyoderma gangrenosum. Br J Dermatol 2015; 173:1216–23. 4 Guenova E, Teske A, Fehrenbacher B et al. Interleukin 23 expression in pyoderma gangrenosum and targeted therapy with ustekinumab. Arch Dermatol 2011; 147:1203–5. 5 Hafner J, Nobbe S, Partsch H et al. Martorell hypertensive ischemic leg ulcer: a model of ischemic subcutaneous arteriolosclerosis. Arch Dermatol 2010; 146:961–8. 6 Weenig RH, Davis MD, Dahl PR, Su WP. Skin ulcers misdiagnosed as pyoderma gangrenosum. N Engl J Med 2002; 347:1412–8. 7 Su WP, Davis MD, Weenig RH et al. Pyoderma gangrenosum: clinicopathologic correlation and proposed diagnostic criteria. Int J Dermatol 2004; 43:790–800. 8 Jockenhofer F, Herberger K, Schaller J et al. Tricenter analysis of cofactors and comorbidity in patients with pyoderma gangrenosum. J Dtsch Dermatol Ges 2016; 14:1023–30. Funding sources: none. Conflicts of interest: none declared.
A . G . A . K O L I O S 1,2 iD J. HAFNER1 C. LUDER1 E. GUENOVA1,3 K. KERL1 W. KEMPF1,4 J. NILSSON2 L.E. FRENCH1 A. COZZIO3
L.E.F. and A.C. contributed equally as final authors of this work.
British Journal of Dermatology (2018)
© 2017 British Association of Dermatologists
