Invasive shigellosis in MSM

Presentation

A 43-year-old man of Colombian origin and a resident of the UK presented to casualty with a three-day history of constipation associated with intermittent severe abdominal pain, nausea and rectal bleeding. He had fever and rigors. He was admitted and on clinical examination he looked unwell, febrile (temperature 39°C), vital parameters were stable and abdominal examination revealed a slightly distended abdomen with generalised tenderness and guarding. There was no hepatosplenomegaly and bowel sounds were present. He was investigated for possible bowel perforation and sepsis. He was started on broad spectrum antibiotics and blood cultures were taken. He had no ill contacts and had travelled to Spain about a month prior to admission and had been well following his return. Of note in his medical history, he was a man who had sex with men (MSM), was HIV-positive, well controlled on antiretroviral drugs with CD4 406 (21%) and viral load <40 copies/ml; he had acquired hepatitis C infection within the past year for which he was not yet being treated. He had unprotected anal intercourse regularly with his male partner. He denied having other sexual partners or using recreational drugs for at least a year prior to this hospital admission. He was transferred to the HIV tertiary centre for further medical management.

Investigations

Complete blood count showed an elevated white cell count of 13.8 × 10⁹/L, platelet count 444 × 10⁹/L and C-reactive protein (CRP) 25 mg/L. Renal function, liver function and coagulation times were all within normal limits. Computed tomogram of the abdomen with contrast showed free fluid within the pelvis and evidence of colitis with no bowel perforation.

His sexual health screen, rectal and throat swabs for Chlamydia trachomatis and Neisseria gonorrhoeae and syphilis serology, were negative. Treponema pallidum polymerase chain reaction (PCR) was not done.

Shigella flexneri serotype 2 a, sensitive to ceftriaxone, azithromycin, ciprofloxacin was isolated from two sets of blood cultures as well as stool cultures.

Treatment

He was treated with intravenous ceftriaxone for four days and responded symptomatically, became afebrile and diarrhoea settled after 48 h. CRP (<5) and white cell count improved (6.0 × 10⁹/L). He was subsequently changed to oral azithromycin to complete 10 days of antimicrobial therapy.

Discussion

S. flexneri is a known cause of infectious diarrhoea the MSM population, with outbreaks reported across the UK, rest of Europe, USA and Australia.^1–6 While a vast majority of cases have gastroenteritis, reports of bacteraemia are rare. A large study from South Africa reported S. flexneri 2 a as the most common serotype, causing invasive shigellosis including bacteraemia. There was a high incidence of multidrug resistance (41%), and HIV confection was associated with higher mortality. ⁷ HIV-infected MSM are more likely to have an increased period of shedding ⁴ and therefore increase onward transmission.

Between 1 January 2004 and 31 August 2014, 16,562 Shigella species isolates were referred from England and Wales to the Gastrointestinal Bacteria Reference Unit (GBRU), Public Health England, Colindale. S. flexneri accounted for 6193 (37%) of the isolates (2406 had documented history of travel) and 1884 (30%) belonged to serogroup 2 a. While the vast majority of isolates were cultures from faeces, 50 (0.8%) S. flexneri bacteraemias were noted. These were predominantly seen in males (33) and 15 were S. flexneri serotype 2 a, of which six were related to foreign travel. Shigella bacteraemia is associated with significant morbidity (acute renal failure, overwhelming Gram-negative sepsis and, rarely, disseminated intravascular coagulation [DIC]). ⁸ In a study of systemic shigellosis in South Africa, HIV-infected patients were substantially more likely to die than HIV-uninfected individuals (29 of 78 versus five of 40 fatal cases, respectively). ⁹ It is important that appropriate microbiological specimens are taken in patients presenting to genitourinary clinics or casualty with appropriate risk factors, prior to administration of antibiotics, as the accurate diagnosis has a direct impact on both medical management of the patient and public health actions. Culture from a stool sample gives a better yield than culture from a rectal swab. ¹⁰ Empirical antibiotic is warranted in ill patients with diarrhoea requiring hospitalisation, immunocompromised patients, food handlers and health care workers. Empirical antibiotic of choice in uncomplicated bacillary dysentery is oral ciprofloxacin for three days.^11,12 Alternative oral agents used in uncomplicated shigellosis include azithromycin and cefixime. ¹² Intravenous ceftriaxone is preferred in invasive infections like bacteraemias and severely unwell patients. Treatment should be tailored according to antimicrobial susceptibility results as increasing azithromycin resistance is well documented in S. flexneri 3 a and maybe emerging in serotype 2 a. ¹³ The duration of therapy is usually three to five days; intravenous therapy should be administered to patients with immunosuppression or presence of HIV co-infection.^14–16

Learning points

The ongoing outbreak of S. flexneri in MSM is proving to be difficult to control. Though invasive shigellosis is rarely seen in the UK, it is important to highlight the continuing risk of complications associated with this pathogen in the MSM population.