Hans Herfarth, MD, PhD on Biomarkers for the Optimization of IBD: Finding Precision

The advent of tests for the protein coding gene nudix hydrolase 15 (NUDT15) and the thiopurine S-methyltransferase (TPMT) phenotype has advanced the precision that can be achieved in the management of inflammatory bowel disease (IBD), Hans Herfarth, MD, PhD, said at the virtual regional meeting of Advances in IBD (AIBD) on September 25.

Dr Herfarth is the codirector of the Multidisciplinary Center for IBD Research and Treatment at the University of North Carolina at Chapel Hill.

He related the case of a male patient aged 22 years who presented with diarrhea and recurrent anal fissures. Colonoscopy and computed tomography (CT) enterography revealed severe Crohn’s disease (CD). The patient began treatment with adalimumab and 50 mg of 6-mercaptopurine (6-MP). The patient’s TPMT was normal.

However, Dr Herfarth said, 10 days later this patient was hospitalized with fever and malaise.

Testing showed a white blood cell count, which had been totally normal before, of 0.5 and a total lymphocyte count 0.2. The patient’s C-reactive protein (CRP) 170 mg/dl. “Further tests showed heterogeneity for NUDT15,” he said.

Both TPMT and the NUDT15 polymorphism are involved in the metabolism of thiopurines, Dr Herfarth explained, as thiopurine is processed through enzymatic steps to 6-thioguanine nucleotides (6-TGN), the active metabolite that treats IBD.

“If TPMT activity—one of the major enzymes in this process — is low or the NUDT15 polymorphism is present, the 6-TGN levels significantly increase and can cause side effects such as leukopenia,” Dr Herfarth said.

While NUDT15 is more prevalent among patients of Asian ancestry, he said, “it is important actually to determine test for NUDT15 among all patients.” An analysis of TPMT and NUDT15 genome-wide and exome-wide association studies among patients of European ancestry treated with thiopurines showed an odds ratio for NUDT15 and thiopurine-induced myelosuppression (TIM) was 38.2.  The presence of NUDT15 causes rapid onset of myelosuppression; TPMT and NUDT15 together result in even faster onset.

Is it worth it to screen every patient for NUDT15? Dr Herfarth asked. Noting that the estimated number of patients needed to test to prevent 1 patient from developing TIM, he said, “Given that we have a huge population in the United States of patients with Asian ancestry, I think it is worth testing every patient for NUDT15 and TPMT before we start them on azathioprine or 6-MP.”

Dr Herfarth reported on a second case of a woman aged 23 years with left-sided ulcerative colitis (UC) who had been diagnosed 2 years earlier. For 12 months the patient had been treated with adalimumab every 2 weeks and had achieved initial remission. However, the patient presented with new onset of bloody diarrhea, 4 stools per day, and abdominal pain beginning 4 weeks earlier.

Testing found the trough level of adalimumab was 6 mg/ml and a high anti-adalimumab antibody titer of 261 ng/ml. The patient was found to be positive for the HLA-DQA1*05 gene. Her treatment was changed to weekly adalimumab with the addition of long-term azathioprine and remission was re-established, Dr Herfarth said.

“The PANTS study showed importance of HLA-DQA1*05 gene association with immunogenicity to anti-TNF therapy,” he explained. The PANTS—Personalized Anti-TNF Therapy in Crohn  Disease—study was a 3-year observational prospective study among 1240 bio-naïve patients designed to investigate loss of response, primary nonresponse, and adverse drug reactions to adalimumab and infliximab.

The hazard ratio for immunogenicity to adalimumab and infliximab among patients with HLA-DQA1*05 was found to be 1.9, Dr Herfarth noted. Approximately 40% of patients in the study had HLA-DQA1*05.

“What does this mean? Almost 100% with this genotype developed immunogenicity on infliximab monotherapy,” he explained. “The presence of 1 or 2 copies of HLA- DQA1*05 creates a high probability of developing antidrug antibodies. Combination therapy is clearly superior, with a higher persistence of therapy and less loss of response.”

Among patients treated with adalimumab, he pointed out, 50% of those with HLA-DQA1*05 on monotherapy developed antidrug antibodies. “With combination therapy you can essentially compensate for the presence” of this genotype.

“This is a feasible biomarker to determine risks of antibody formation before start of anti-TNF therapy,” Dr Herfarth said. “The test informs about the need to use combination therapy with thiopurines or methotrexate, and importantly, the risk of antibody formation if you want to stop thiopurines.” However, he said, “I don’t stop combination therapy with my patients if they have this gene.”

There are open questions to be addressed on this topic, including, “What is the mechanism? Does this genotype have a similar impact with certolizumab or golimumab, or for other biologics like ustekinumab and vedolizumab? No anti-TNF first-line therapy for HLA-DQA1*05 patients? This has to be evaluated in the future.”

His third case involved a male patient aged 45 years with colonic Crohn disease, stable on adalimumab and thiopurines for 2 years. A colonoscopy 6 months prior showed mucosal healing. “Now the decision is to monitor disease with fecal calprotectin every 6 months,” Dr Herfarth related. He explained that calprotectin is a cytosolic protein complex, mainly expressed in neutrophils, and has a good correlation with endoscopic and histologic inflammation. Normal levels are 10 to 50 mcg/g stool. A level of >250 mcg/g is associated with inflammation.

For this patient, “the first calprotectin comes back at 370 mcg. The patient reports he has been doing fine other than having some diarrhea after eating pizza, around the time he sent in the test.” A repeat fecal calprotectin level 4 weeks later is 45 mcg/g stool.

“You have to be aware of the confounders” when monitoring with fecal calprotectin, Dr Herfarth stressed. “A single test is not diagnostic; there is a high variability of results.” Among the confounders, according to the International Organization for IBD’s STRIDE 2 guidance, are gastrointestinal infections and bleeds; microscopic colitis; obesity, bowel preparation for colonoscopy, rheumatic conditions, and the use of some drugs, including nonsteroidal anti-inflammatory drugs and proton pump inhibitors (PPIs).

“You can get to a gray zone just with PPIs,” Dr Herfarth said. The patient in this case was taking PPIs for heartburn and had a body mass index of 42. “You have to see that in context.”

With high levels of fecal calprotectin but no symptoms, he cautioned, “don’t automatically continue to colonoscopy and treatment changes; of course if there are symptoms, and a red zone reading, then you need evaluation. But don’t trust a single test.” He also noted that fecal calprotectin is not optimal for monitoring predominant small bowel disease and that the patient’s insurance coverage should be checked to determine coverage.

Oncostatin M “is a promising cytokine indicating refractory disease in UC, and perhaps recurrence of CD after surgery,” Dr Herfarth stated. “It is correlated with histologic  inflammation.” This marker has been elevated in refractory and postoperative CD and also in nonresponders to vedolizumab, he said. “Serum OSM in one study was found to be a better marker of predicting endoscopic recurrence of CD after resection than fecal calprotectin—more sensitive and specific. This is a very interesting marker to watch in the future.”

But for now, Dr Herfarth stated, testing for TPMT and NUDT15 “should be the new standard before starting a thiopurine.” In addition, he said, “Determining the presence of HLADQA1*05 should be standard of care before starting anti-TNF monotherapy or before considering stopping combination therapy, but insurance coverage may be tricky.”

--Rebecca Mashaw

Herfarth, H. Biomarkers for optimization of IBD: finding precision. Presented at: Advances in Inflammatory Bowel Diseases regional meeting. September 25, 2021. Virtual.