In August 2022, an 80-year-old man presented to the emergency department, directly after having returned by plane from a Berber mountain village, his birth region in Morocco. He presented with a 2-week history of watery diarrhea 6 times daily, vomiting, fever and chills. Four days prior to the complaints, he consumed fresh water from a local well. Initially, the symptoms improved after a few days, but the fever persisted with nocturnal temperature spikes, followed by worsening of diarrhea and vomiting. Due to the deterioration of symptoms and fear for dehydration, the patient decided to seek medical attention in his place of residence, Amsterdam, The Netherlands.
His medical history consisted of a HIV-1 infection with suppressed viral load (recent CD4+ T cell count: 826 (norm 404-1612) with an undetectable viral load), and gastroesophageal reflux disease for which he used a proton pump inhibitor.
On presentation, the patient was confused and physical examination revealed a temperature of 39.0 °C with chills, a pulse rate of 120 beats per minute, and a blood pressure of 110/70 mm Hg. Apart from a capillary refill time of > 2 s, mildly dry mucous membranes and profound perspiration, no abnormalities were observed. The laboratory tests (Table 1) showed signs of inflammation with an elevated C-reactive protein (218), leukocytosis (18,4) and elevated liver enzymes.
Table 1 Laboratory results on admissionThe urine analysis and a chest X-ray revealed no abnormalities.
The preliminary diagnosis was sepsis of unknown origin, with a gastro-intestinal focus deemed probable, given the history of diarrhea and vomiting. Specifically, the differential diagnosis included cholangitis and cholecystitis (due to the elevate liver enzymes), other common foci of infections, such as lung and urinary tract, were deemed less probable. After samples for microbiological diagnostics (blood cultures, urine, stool) were collected, treatment with intravenous (IV) fluids, antibiotics (ceftriaxone 2 g iv qd) and antidiabetics was started.
Twenty-four hours after admission, blood cultures were positive for gram negative rods in 3 out of 4 vials. Following standard laboratory procedures, colonies grew on solid culture media (COS, PVX, CLED; bioMerieux, Marcy l’Etoile, France), which were initially identified as Vibrio albensis (score 1.85), and subsequently identified as Vibrio cholerae (score 2.1) by Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) on a MALDI-ToF Biotyper (Bruker Daltonics, Bremen, Germany) (See Fig. 1). The strain was further preliminary typed as V.cholerae non-O1 by agglutination with V. cholerae O1 polyvalent anti-serum (ThermoFisher Scientific, USA).
Fig. 1Gram stain of V.cholerae from positive blood culture (i); colonies of V.cholerae on blood agar plate, showing hemolytic colonies (ii), on Cysteine-Lactose-Electolyte-Deficient agar plate (iii), and on PolyViteX agar (iv)
The strain was sent to the Dutch National Institute for Public Health and the Environment (RIVM) where the identification of V. cholerae was confirmed by MALDI-ToF, analyzing our institute’s sequencing data in their sequencing pipeline, and subsequent serotyping with V. cholerae O1 and O139 antiserum agglutination test, revealed a non-O1/non-O139 V. cholerae strain (NOVC). Doxycycline 100 mg IV twice a day was added to the treatment regimen as soon as the identification of the blood culture strain was made, awaiting antimicrobial susceptibility results. Using a 0.5 McFarland suspension to inoculate Mueller Hinton E agar plates (bioMerieux, Marcy l’Etoile, France) with the V. cholerae strain, a standardized set of antimicrobial agents for Vibrio species was tested for potential drug resistance, by using paper discs (Becton Dickinson and Company, Sparks, MD, USA). MHE-plates were incubated under ambient conditions at 35–37 °C for 18 h before reading. The V. cholerae strain was susceptible for all drugs tested, using clinical breakpoints by the European Committee on Susceptibility Testing (EUCAST Breakpoint table version 12.0, 2022), i.e. ciprofloxacin (33 mm; S ≥ 23 mm), cotrimoxazol (25 mm; S ≥ 18), cefotaxim (34 mm; S ≥ 21 mm), ceftazidim (25 mm; S ≥ 22 mm), tetracycline (26 mm; S ≥ 20 mm), and erythromycin (14 mm; S ≥ 12 mm); azithromycin susceptibility was inferred from erythromycin.
Fecal molecular diagnostic screening for bacterial causes of gastrointestinal infections (Salmonella-, Shigella-, Yersinia-, Campylobacter species), showed positive results for Campylobacter spp. After the identification of V. cholerae in de blood culture, the fecal sample was also analyzed for V. cholerae by culture using Thiosulfate Citrate Bile Salts Sucrose (TCBS) selective medium and molecular diagnostics, of which only the PCR was positive. Parasitic causes of infection, such as Entamoeba histolytica and Strongyloides species were not detected in feces and serum. Urine culture was negative.
Prompted by the elevated liver enzymes, an ultrasound sonography of the abdomen was performed which showed multiple abscesses spread throughout the liver (see Fig. 2A).
Fig. 2Ultrasound shows multiple liver abscesses (A), and residual abnormalities at the place of previous liver abscessed (B)
Although the patient refused a diagnostic biopsy of the liver abscesses, our diagnosis was a NOVC bacteremia with multiple liver abscesses.
On day 5, the patient’s clinical situation had improved substantially, whereupon ceftriaxone and doxycycline IV were switched to oral ciprofloxacin 500 mg twice daily. On day 8 he was discharged in good clinical condition.
Six weeks after discharge, ciprofloxacin was stopped during the follow-up visit to the outpatient clinic, after ultrasound sonography of the abdomen showed only residual abnormalities at the site of previous liver abscesses (see Fig. 2B). Laboratory results showed normalization of inflammation parameters and of liver enzymes and function. See timeline for summarized clinical course (Fig. 3).
Fig. 3Timeline of clinical course
Whole Genome Sequencing of the isolate revealed a high similarity to V.cholerae using KmerFinder [10]. The genome was subsequently uploaded to PathogenWatch (https://pathogen.watch) and analyzed using the VibrioWatch database (https://vibriowatch.readthedocs.io/en/latest/index.html). This analysis showed that the strain belonged to V. cholerae ST438, a rare sequence type phylogenetically distinct from the 7th pandemic cluster (see Fig. 4).
Fig. 4No antimicrobial resistance gene was identified. The genome contained several known pathogenicity-related genes but not the genes encoding for the cholera toxin (CT) or the colonization factor toxin-coregulated pilus (TCP), two of the major virulence factors of highly pathogenic serotypes.
Comments (0)