A woman in her thirties with facial swelling and necrotic lesions

A woman in her thirties with known atopic dermatitis was admitted via the out-of-hours primary care service to the department of medicine at her local hospital with suspected facial cellulitis. She was not taking any regular medication. She had a cat at home. She had been on holiday in southern Europe one month before the onset of symptoms. In the few days prior to admission, she had developed increasing pain, swelling, redness and warmth in the left side of her face. The day before admission, she had started to take oral phenoxymethylpenicillin 1 g four times daily after being assessed by an ear-nose-throat specialist.

On admission to the department of medicine, she was in a slightly diminished general condition. Her blood pressure was 137/77 mmHg, pulse 92 bpm, respiratory rate 19 breaths per minute, oxygen saturation 97 % and temperature 36.3 °C. She had clear swelling and diffuse redness on the left side of her face, particularly around the eye and on the cheek. Several small crusted lesions had also appeared in this area. Blood test results found CRP 70 mg/L (reference range <5) and leukocytes 5.5 × 10⁹/L (4.1–9.8 × 10⁹/L). Her condition was considered to be cellulitis, with eczema as a possible portal of entry. After blood cultures had been ordered, intravenous antibiotic treatment consisting of cloxacillin 2 g four times daily was initiated.

Blood cultures were negative. Due to a lack of improvement and to improve tissue penetration, intravenous treatment with clindamycin 600 mg four times daily was added on day 5. A swab sample of wound secretion was sent for culture and sensitivity testing on day 6, which showed growth of Staphylococcus aureus. Her leukocyte count had increased to 14.9 × 10⁹/L. An automated differential count was not possible, and there was an alert about potentially abnormal lymphocytes. Despite this, CRP fell from 70 mg/L to 38 mg/L.

After a few more days of clinical deterioration, oral ciprofloxacin 750 mg twice daily was also added to cover any gram-negative bacteria. A clearly enlarged gland was discovered at the corner of her left jaw, and when taken together with lymphocytosis of 10.0 × 10⁹/L (1.2–3.1 × 10⁹/L), this led to cutaneous manifestation of lymphoma or lymphoproliferative disease being considered. Blood smear revealed lymphocytosis with atypical cells, but findings of immunophenotyping of the blood were normal. The diagnosis of mpox (formerly known as monkeypox) was briefly discussed due to the ongoing outbreak in non-endemic countries in Europe and other parts of the world. However, since the patient was not in a known risk group, a bacterial infection was considered more likely.

A dermatologist was contacted on day 9 due to the uncertain diagnosis and clinical deterioration. The review took place the same day, and the dermatologist recommended transfer to the dermatology department at a regional hospital. At the time of transfer, the patient had several large lesions with necrosis and central tissue loss located on the left cheek and temple. There were similar, but smaller, lesions on the right eyebrow and the back of the right hand. There was considerable swelling of the left side of the face, and her left eye was almost sealed shut. The lymph nodes under the corner of her left jaw were clearly enlarged (Figure 1). In the previous 24 hours, she had also developed generalised widespread maculopapular exanthem, as is seen with a drug reaction.

Cloxacillin, clindamycin and ciprofloxacin were all discontinued because it was not possible to determine which medicinal product was causing the rashes. Blood test results found leukocytes 16.1 × 10⁹/L (4.1–9.8 × 10⁹/L), of which lymphocytes were 9.1 × 10⁹ (1.2–3.1 × 10⁹/L), neutrophils 5.6 × 10⁹ (1.8–6.9 × 10⁹), CRP 29 mg/L (<5 mg/L), alanine aminotransferase (ALT) 608 U/L (10–45), gamma-glutamyl transferase (GGT) 186 U/L (10–45) and bilirubin 6 µmol/L (> 20).

For diagnostic elucidation, a swab sample of wound secretion was taken on virus transport medium for polymerase-chain reaction (PCR) testing in the Department of Microbiology. A biopsy from the wound edge was placed in a sterile container with formalin and sent to the Department of Pathology at Haukeland University Hospital for haematoxylin-eosin (HE) staining. A biopsy placed in sodium chloride was sent to the National Reference Centre for Molecular Biological Parasite Diagnosis at Oslo University Hospital for leishmaniasis PCR testing.

PCR analysis for orthopoxviruses was positive, while specific PCR analysis for mpox was negative. Diagnostic testing for further differentiation within the orthopox group is not available in Norway. Since we wanted to confirm the diagnosis for academic reasons, another swab sample of wound secretion was sent to the State Serum Institute in Denmark for further investigation. The sample was analysed with PCR and sequencing there, and cowpox was confirmed. Histological findings were also consistent with cowpox (Figure 2). Leishmaniasis DNA was not detected in the skin biopsy examined at the reference laboratory at Oslo University Hospital.

Oral prednisolone 40 mg once daily was administered, with a tapering plan, primarily for the patient's generalised drug reaction, but also to reduce the localised facial inflammation. The exanthematous drug eruption gradually subsided over the first few days, and liver function tests returned to normal. Several attempts were made to mechanically remove the necrotic lesions on the face. Various wound dressings and antiseptic agents were used. Eventually, ultrasound of the left facial region was performed due to increasingly swollen and taut skin around the lesions on the left cheek. This examination revealed considerable inflammation and a large lymph node, which was considered to be reactive.

On day 30, the issue was discussed with a plastic surgeon, who recommended that conservative management be continued. In the following days, increasing undermining of the skin at the wound peripheries was discovered, as well as increasing deep necrosis at the wounds beyond the left lateral canthus. Repeat ultrasound revealed an incipient abscess. On day 37, another plastic surgery assessment was requested with regard to wound revision and assistance with planning further treatment. The patient was then transferred to the plastic surgery department. Surgical treatment was found to be indicated with opening of the skin and removal of fat necrosis and dead tissue.

The skin defects were reconstructed with local skin flaps. Repeat samples sent for culture and sensitivity testing revealed growth of Staphylococcus aureus, and treatment with intravenous cloxacillin 2 g four times daily was initiated despite an earlier drug reaction to an unclear agent. The patient was in a good general condition and afebrile, but she had elevated CRP at 71 mg/L, while her leukocyte count was 9.1 × 10⁹/L. Clinical examinations found increasing signs of localised infection, pus under the skin flaps, followed by ischaemia and ensuing necrosis of the flaps. Revision of the necrosis and wound care was performed on several occasions. During infection control and examination of clean wound surfaces, it was decided to perform reconstructive surgery with a full-thickness skin graft from a donor site on the left upper arm. One of the reasons for selecting the donor site was the patient's history of atopic eczema (Figure 3). Intravenous cloxacillin 2 g four times daily was administered for a total of nine days, before switching to oral dicloxacillin 500 mg three times daily for a total of ten days. The patient did not develop any skin reaction to antibiotics this time.

The patient was discharged home after a total of 55 days in hospital. She received follow-up with regular check-ups at the plastic surgery outpatient clinic. The graft took well, and two small residual wounds healed by secondary intention with a conservative wound care regimen.