To the Editor:

Severe infections (i.e., infections with a need for antimicrobial therapy or hospitalization) are common side effects in patients on immunosuppression with tumor necrosis factor alpha (TNF-α) inhibitors. For patients with rheumatoid arthritis anti-TNF-α therapy is known to increase the risk of severe infections, with a meta-analysis finding an odds ratio of 2 and a number needed to harm of 59 during a treatment period up to 12 months.1

For patients with Crohn's disease (CD) on anti-TNF-α therapy severe infections are regularly described. For example, a retrospective study with 500 CD patients on infliximab showed that 8,2% of all patients suffered from infections, of which 3% were classified as severe.2

Yet it has not been finally proven that these observations result from a significantly increased risk for severe infections. A meta-analysis in 2008 including 21 studies with 5356 patients did not confirm such an effect.3 A large prospective trial, the TREAT registry,4 enrolled 6290 patients, of whom 3179 received infliximab. The rate of serious infections was significantly higher for patients who had received infliximab (relative risk 2.15, P < 0.01), but as the two groups were different in baseline disease characteristics infliximab could not be identified as an independent risk factor for serious infections. Parameters in that study that were independently associated with such an increased risk were prednisone use, narcotic analgesic use, and moderate-to severe disease activity.

thumbnail image

Figure 1. Acute hydropic cholecystitis due to septic salmonellosis.

Download figure to PowerPoint

Furthermore, it is unknown if there is a difference in the safety of TNF-α inhibitors as monotherapy or in combination with other immunomodulators such as azathioprine. A prospective trial that compared the efficacy of infliximab alone or in combination with azathioprine did not find a difference in the risk for severe infections in the two groups.5

However, it is generally accepted that today's available data from prospective trials are not sufficient to establish a final side effect profile of TNF-α inhibitors.

The considerations above are supported by the pathophysiological knowledge of TNF-α inhibitors. TNF-α plays a key role in host defense of bacterial infections, especially for intracellular infections.


In the following we present the case of a patient with CD on anti-TNF-α therapy. A 35-year-old Caucasian male presented at the emergency room in bad general appearance, with signs of dehydration and strong abdominal pain, described as colicky.

His past medical history included CD, diagnosed at the age of 18. Since then he had suffered from a complicated clinical course, with subtotal colectomy/ileostomy 7 years ago and relapsing perianal fistula. His further medical record included chronic obstructive pulmonary disease (COPD). He smoked cigarettes (about 50 pack-years) and drank alcohol regularly (about 80 g/d).

His medication was azathioprine with a dose of 100 mg/day for several years. He had received the first application of infliximab (250 mg) 20 days before admission to the hospital.

Around 7 days after the infusion the patient started to suffer from strong abdominal pain. There was no relevant change of stool frequency or consistency (via ileostomy), nausea, or vomiting. No other person in his personal surroundings suffered from similar symptoms.

The temperature at admission was 39°C, the pulse 110 beats per minute, and the respiratory rate 20 per minute. The blood pressure was in the normal range. On physical examination the abdominal pain was located in the right upper quadrant, without radiation or aggravation due to palpation. He showed obvious signs of malnutrition (BMI 16 kg/m2) and severe dehydration with acute renal failure and hyperpotassemia. Further notable laboratory findings were increased inflammation parameters (leukocytes 20,000/μL, C-reactive protein [CRP] 280 mg/L, PCT 4 ng/mL). An abdominal ultrasound revealed an acute cholecystitis, with a massively enlarged, hydropic gall bladder (208 mL, normal range <100 mL); thus, an admission diagnosis of a septic cholecystitis was made. No need for an emergency operation was indicated by the surgical consultant (Fig. 1).

The patient was first admitted to the ICU with a need of hemofiltration. After 3 days he was transferred to an internal medicine ward with normalized renal function. The initial empirical antibiotic treatment consisted of piperacilline/sulbactame and ciprofloxacine intravenously. Blood cultures showed a colonization with Salmonella typhimurium. The same agent was found in specimens of stool and urine. In the later course, metronidazole was added to the antibiotic treatment, due to stool specimen positive for Clostridium difficile toxin.

With the antibiotic treatment and supportive care including total parenteral nutrition the patient's general appearance improved quickly. The inflammation parameters normalized and the abdominal pain diminished. Further diagnostic including esophagogastroduodenoscopy, endoscopy via the ileostomy, and magnetic resonance imaging (MRI) enteroclysis showed no acute exacerbation of CD. The histological specimens of the two endoscopic interventions were normal. A proctological examination did not show an exacerbation of the known perianal fistula.


With the drug's increasing use, several case reports published during the last years have shown septic salmonellosis in patients on anti-TNF-α therapy.6, 7 Of note, the patient in this report did not suffer from an essential change of his bowel movement habits (via ileostomy) during a relapse in the prior clinical course of the disease. As the relapses were rather characterized by abdominal pain, comparable to the episode above, an important point of this report is the similarity of the symptoms, due to the acute cholecystitis on the one hand, and those the patient knew from prior relapses on the other.

As attending doctor for a patient on TNF-α inhibitors, one should be aware of therapy side effects simulating aggravations of the underlying disease. Smoking and alcohol consumption always have to be taken into account as immunosuppressive cofactors, as in the patient's case. Patients should be instructed about the general prophylactic behavior (e.g., to heat well putative contaminated food such as eggs, poultry, and fish) to avoid food-borne infections like salmonellosis.


  1. Top of page
  • 1
    Bongartz T, Sutton AJ, Sweeting MJ. Anti-TNF antibody therapy in rheumatoid arthritis and the risk of serious infections and malignancies: systematic review and meta-analysis of rare harmful effects in randomized controlled trials. JAMA. 2006; 295: 22752285.
  • 2
    Colombel JF, Loftus EVJr, Tremaine WJ. The safety profile of infliximab in patients with Crohn's disease: the Mayo clinic experience in 500 patients. Gastroenterology. 2004; 126: 1931.
  • 3
    Peyrin-Biroulet L, Deltenre P, de Suray N. Efficacy and safety of tumor necrosis factor antagonists in Crohn's disease: meta-analysis of placebo-controlled trials. Clin Gastroenterol Hepatol. 2008; 6: 644653.
  • 4
    Lichtenstein GR, Feagan BG, Cohen RD. Serious infections and mortality in association with therapies for Crohn's disease: TREAT registry. Clin Gastroenterol Hepatol. 2006; 4: 621630.
  • 5
    Colombel JF, Sandborn WJ, Reinisch W. Infliximab, azathioprine, or combination therapy for Crohn's disease. N Engl J Med. 2010; 362: 13831395.
  • 6
    Rijkeboer A, Voskuyl A, Van Agtmael M. Fatal Salmonella enteritidis septicaemia in a rheumatoid arthritis patient treated with a TNF-alpha antagonist. Scand J Infect Dis. 2007; 39: 8083.
  • 7
    Rim JY, Tenorio AR. Salmonella septic arthritis in a patient with Crohn's disease on infliximab. Inflamm Bowel Dis. 2010; 16: 545547.

Tobias Kukiolka MD*, Markus F. Neurath MD*, Jonas Mudter MD*, * Department of Internal Medicine 1 University Hospital of Erlangen Erlangen, Germany.