Bloodstream Infection Raises Suspicion of Colorectal Cancer

Daniel M. Keller, PhD

September 22, 2011

September 22, 2011 (Chicago, Illinois) — A bloodstream infection (BSI), especially with certain species of bacteria, might be an early tip-off to look for colorectal cancer (CRC), Sanchia Warren, MBBS, now at St. Vincent's Hospital in Melbourne, Australia, advised, based on her poster presentation here at the 51st Interscience Conference on Antimicrobial Agents and Chemotherapy.

The research was conducted at the University of Calgary in Alberta, Canada, using a database population of 1.2 million people 18 years and older in the Calgary Health Zone. People diagnosed with a BSI from 2000 to 2007 were identified using an electronic surveillance database. At the end of 2008, investigators used the Alberta Cancer Registry to correlate BSI with CRC appearing within 12 months of the infection.

"We found that of more than 10,000 bloodstream infections [n = 10,121], we had 71 patients who had a [concomitant or] subsequent diagnosis in the 12 months of a colorectal cancer. In looking at the specific organisms, particularly if you have an anaerobic bloodstream infection..., compared to somebody who just has a bloodstream infection, you've got over a 70% increased risk of having a diagnosis in that 12 months," Dr. Warren told Medscape Medical News. "Specifically, with things like Clostridium, there was 115 times [increased risk for CRC], compared with somebody who just had a bloodstream infection." The Bacteroides fragilis group and other anaerobes were also associated with the highest risk for a new CRC diagnosis.

Among people with a BSI, compared with the general population, the researchers found a greater than 14-fold increased risk (standardized incidence ratio, 14.4; 95% confidence interval, 11.3 to 18.3) for a CRC diagnosis concomitantly with or within 1 year of any BSI. Risk factors for a new diagnosis of CRC were age, being male, a higher Charlson comorbidity index score (all P < .001), liver disease, and diabetes with complications (both P = .023). Diabetes without complications was not associated with CRC in the study (P = .588).

Historically, Streptococcus bovis bacteremia has been widely recognized in association with occult colon cancer. In the Calgary Health Zone population, 30 cases of S bovis BSI were observed, but none of the patients developed CRC in the subsequent year. (Most strains of S bovis were reassigned to the species S gallolyticus in 2003.)

Dr. Warren advised that "as an MD or somebody who is looking after these patients, perhaps we should be looking a little bit further, and if we have an anaerobic bloodstream infection, be doing colonoscopy.... This may be their first symptom, and we definitely [should] be having a look at their bowel for that reason."

Although it is not possible to do a colonoscopy on every patient with a BSI, she recommends considering it for older patients, anybody with a BSI caused by anaerobic bacteria, and people without another obvious source of infection.

Although the study involved large numbers of patients, it was retrospective and there were only small numbers of infections with some organisms, Dr. Warren acknowledged.

She thinks that it is more likely that BSI is an early marker of CRC, rather than a cause. David Hooper, MD, chief of the infection control unit and associate chief of the infectious disease division at Massachusetts General Hospital in Boston, and president of the American Society of Microbiology, agrees, pointing out that CRC is an invasive disease. "It can disrupt mucosal structure and in severe cases might, in fact, generate a portal of entry for bowel flora into the bloodstream," he told Medscape Medical News.

Dr. Hooper said S bovis/S gallolyticus, and Clostridium septicum in particular, should raise red flags. "There are certain organisms like S bovis for which there are clinical recommendations. It's enough that you should be looking for [CRC] in an active way, either with colonoscopy or these high-resolution [computed tomography] methods that can sometimes be a substitute for colonoscopy," he advised. "We've had patients who showed up as their first presentation with this bloodstream infection, the bug was identified," and the patients were found that have CRC. "It's quite a striking association."

The study had no commercial funding. Dr. Warren and Dr. Hooper have disclosed no relevant financial relationships.

51st Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC): Abstract K-834. Presented September 18, 2011.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.