A 54-year-old Asian man with IgA nephropathy underwent living-donor kidney transplantation 14 years previously.

His medical condition had been almost stable for 14 years.

He developed multifocal salmon-pink skin discoloration, and swelling and spontaneous pain in the left knee and leg.

The symptoms had gradually expanded across the right forearm and outside of the right thigh.

He was admitted to our hospital for the evaluation of fever (39 °C) and multifocal cellulitis (Fig.1).

Two months before admission, he had developed chronic diarrhea.

His serum creatinine level was stabilized at 1.7 mg/dL with maintenance immunosuppressive therapy comprising tacrolimus (3 mg/day), mycophenolate mofetil (1500 mg/day), and prednisone (4 mg every other day).

The tacrolimus trough concentration was 6.3 ng/mL.

The patient had been a dog breeder for 12 years.

On admission, his white blood cell count was 12,400/μL and his C-reactive protein level was 3.9 mg/dL.

The patient was initially treated with ampicillin/sulbactam (9 g/day intravenously).

Two days after the initiation of this therapy, he was afebrile.

Four days later, an automated blood culture (Bactec FX®; Becton–Dickinson and Company, Sparks, MD, USA) showed positivity for gram-negative spiral bacilli.

A colony obtained from the patient’s blood culture was analyzed by MALDI-TOF MS (Biotyper ver. 3.0®; Bruker Corporation, Germany).

The identification score was 2.064 (>2.0), indicating accurate identification of H. cinaedi (Fig.2).

Additional evaluation of the patient’s blood specifically for H.

cinaedi by means of gyrase subunit B (gyrB)-targeted PCR assays (using the forward primer AGGGATTCCACAAAGTGAGC and the reverse primer TCTTGTCCTGTGCGTTCATC to amplify the region of the gyrB gene) yielded positive results.

We performed gyrB-targeted PCR using a single colony isolated from a blood culture.

We used distilled water (DNA- and DNase-free water) as a negative control to prevent contamination of gyrB-targeted PCR (Fig.3).

In addition, we performed 16S rRNA gene sequencing of this strain, and the results were > 99% consistent with the existing sequence (Fig.4).

Given these results, we diagnosed the patient with H.

cinaedi bacteremia with cellulitis.

We examined the genomic heat shock protein (HSP) 60 sequence of the blood culture isolate, which resulted in identification of cluster B H. cinaedi.

A blood culture obtained at 11 days was positive for H.

cinaedi, but a culture obtained at 15 days was negative.

The patient’s cellulitis gradually resolved.

The patient continued antibiotic treatment for a total of 6 weeks (ampicillin-sulbactam in a drip for 2 weeks and oral levofloxacin for 4 weeks), and he had no recurrence 6 months after this therapy.

His stool culture was negative, although it was taken after treatment.

We did not apply enteric bacteria elimination in this case.