Gastrointestinal ultrasound (GIUS) has the potential to predict steroid-responsiveness and need for early salvage treatment in patients with flares of ulcerative colitis.
A Melbourne pilot study of 10 consecutive patients incorporated ultrasounds by two independent gastroenterologists on days 1 and 3 of IV steroid treatment, and on day 7 if the patient remained in hospital.
The study found a significant difference at baseline in the mean bowel wall thickness (BMT) between four steroid responders and six non-responders who later required infliximab salvage therapy (4.6 v 6.2 mm). Three non-responders eventually required colectomy.
As well, patients with a BMT > 6 mm at any colonic segment required infliximab therapy whereas patients with BMT <6mm at all colonic segments were more likely to be steroid-responsive.
“Finally, if only the most affected colonic segment was analysed, the median BWT again remained significantly lower in the steroid-responsive group (4.7 mm [4.6-6.4 mm]) compared with those requiring salvage therapy (7.4 mm [6.2-8.9mm]) (p = 0.019; Mann-Whitney U test).”
Similar findings were observed on day 3.
“On the day 3 GIUS, the difference in median BWT between the steroid-responsive group (4.0 mm [3.5-4.5 mm]) and the group requiring infliximab (6.3 mm [5.5-6.9 mm]) was more apparent (p = 0.009).”
The study, published in Ultrasound in Medicine and Biology, said there was no difference in BWT between those who responded to salvage therapy and those who failed salvage therapy and underwent colectomy.
Patients who clinically improved had a BWT of 2.6 mm at two weeks.
Other GIUS measures such as the number of colonic segments with severe hyperaemia, wall stratification and extraintestinal features of inflammation did not predict steroid responders from non-responders.
“The key finding was that the simple measurement of BWT of affected colonic segments at admission provided a clear guide to subsequent failure of steroids,” the study said.
“Furthermore, the predictive power of BWT was retained at day 3, implying that the scan could be performed at any time during the first 3 d without loss of discriminating power.”
“While the median BWT of the involved colonic segments was the best method to discriminate steroid responsiveness, a BWT cut-off value of 6 mm and analysing the single most affected colonic segment also yielded statistically significant and clinically relevant differences between steroid responders and non-responders.”
The researchers, including senior investigator Professor Peter Gibson from Alfred Health and Monash University, said the GIUS findings were in direct contrast to endoscopic scores or appearances, which provided no discriminatory value.
“GIUS value adds to clinical management in additional ways. It non-invasively provides accurate information on disease distribution, which may be useful in risk-stratifying individuals given that extensive disease suggests a more aggressive disease course than those without.”
They said endoscopic assessment of the rectum and distal colon still remained essential in patient management.
A large, multi-centre investigation of the potential of GIUS was clearly warranted.