CT angiography accurately localizes lower gastrointestinal hemorrhage

Last Updated: 2015-05-29

By Will Boggs MD

NEW YORK (Reuters Health) - CT angiography (CTA) is far better than nuclear bleeding scans for localizing lower gastrointestinal hemorrhage (LGIH), according to a before-and-after study.

"CTA should supplant bleeding nuclear scans almost completely," Dr. Jose L. Pascual from University of Pennsylvania Perelman School of Medicine in Philadelphia told Reuters Health by email. "It's faster, more specific, and likely more sensitive. And only where CTA is found to be positive should patients be considered for formal angiography."

Nuclear bleeding scans (NBS) and CTA have both been used for diagnosing LGIH. But which is the superior pre-arteriogram test remains to be seen, Dr. Pascual and his colleagues write in JAMA Surgery, online May 20.

The team investigated the diagnostic value of the Hospital of the University of Pennsylvania's adoption of an algorithm that incorporated the use of CTA into their multidisciplinary protocol for management of patients with LGIH.

In their protocol, patients presenting with LGIH first undergo gastric lavage to rule out upper GI hemorrhage. If clinically stable, they are referred for colonoscopy; if unstable, they undergo CTA. Further angiographic or surgical intervention depends on the results of diagnostic testing.

After implementation of the protocol, CTA use increased 15-fold, and NBS use decreased by almost half, the researchers report.

While the sensitivity and specificity of NBS and CTA as pre-visceral arteriogram diagnostic tests were similar, NBS usually localized the site of hemorrhage only to a particular abdominal quadrant, while CTA localized the hemorrhage to a specific segment of the colon or small bowel.

Similarly, although the proportion of hemorrhage identification was similar in patients who underwent visceral arteriogram (VA) with or without prior CTA, identification of the hemorrhage site more than doubled when CTA preceded VA.

CTA resulted in administration of more overall contrast than did NBS, but without greater increases in renal dysfunction and with a tendency to reduce overall imaging studies used.

At $407 per study, CTA was slightly more expensive than NBS ($382).

"RBC (nuclear) scans are rarely helpful but are time consuming," Dr. Pascual said. "Higher-generation CTs should allow for rapid screening study using CTA for lower GI bleeds -- from there consider surgery, colonoscopy, or interventional angiography."

Dr. Pascual added, "This needs a more structured randomized study to be definitive for recommendations."

Dr. Amy L. Lightner and Dr. Marcia M. Russell from David Geffen School of Medicine, University of California, Los Angeles, California agree in an invited commentary in the journal.

They conclude: "With faster time to administer and interpret CTA, patient care can be expedited. However, as a medical community, more work is needed with respect to clinical outcomes, including contrast-induced nephropathy and costs, before routinely using CTA prior to VA."

Dr. Pasteur Rasuli from the University of Ottawa in Ontario, Canada, recently reviewed the factors that influence the yield of mesenteric angiography in patients with LGIH.

"In our institution we are currently using CTA to screen for patients who are bleeding from anywhere in the GI tract upper and lower. We will not put the patient through angio unless CT is positive," he told Reuters Health by email.

"But we have had a few cases that the angio was negative despite positive CT, and we attribute that to the fact that the bleeding has stopped in the interval," explained Dr. Rasuli, who was not involved in the new study. "So that means that given the intermittent nature of GI bleed we lost the window of high yield by placing a CTA ahead of angio. But my colleagues and I accept this as being a fair deal, i.e., more patients will benefit from having a CTA than lose due to delay caused by CTA."

"I think the referring physician should consider CTA in any case of brisk or fast bleeding patient as the modality of choice for diagnosis and localization of bleeding source and consider angio as a tool for treatment via embolization and not diagnosis," Dr. Rasuli said. "RBC scan and colonoscopy should be reserved for cases of very slow bleed in which repeated CTA is negative despite clinical evidence of GI bleed."

SOURCE: http://bit.ly/1Rub2aH and http://bit.ly/1Ay4lQI

JAMA Surg 2015.