This article is more than 5 years old. Information may no longer be current.

Therapeutic Drug Monitoring in IBD

A paradigm in the evolution of care.

Issue: February 2015

Add topic to email alerts

Receive an email when new articles are posted on

Please provide your email address to receive an email when new articles are posted on .

Subscribe

Added to email alerts

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

Back to Healio

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Back to Healio

Similar to when thiopurine drug monitoring was introduced in the early 2000s, physicians treating patients with inflammatory bowel disease are currently faced with a parallel question regarding how to choose the optimal dosing regimen for patients receiving biologics.

According to experts interviewed by Healio Gastroenterology, however, therapeutic drug monitoring of biologics is far more complex than with thiopurines, largely due to issues stemming from the development of antidrug antibodies that interfere with biologic activity.

Although the utilization of drug monitoring to provide tailored therapies with anti-tumor necrosis factor agents for patients with IBD is on the rise, lack of knowledge among clinicians, insufficient prospective data on outcomes and optimal therapeutic ranges, technological limitations of commercially available assays and cost issues continue to stand in the way of more widespread and optimized implementation. Despite these current barriers and uncertainties, experts said the growing body of evidence supports the rationale for clinicians to improve their understanding of factors that influence the pharmacokinetics of biologics and utilize therapeutic drug monitoring to optimize patient outcomes and reduce the costs of biologic therapies.

Inter-individual Variation in Clearance of Monoclonal Antibodies

Anti-TNF drugs have indeed revolutionized the treatment of IBD, Edward V. Loftus, MD, FACG, AGAF, from the Mayo Clinic in Rochester, Minn., told Healio Gastroenterology. Prebiologic era treatments, such as total parenteral nutrition, bowel rest, IV steroids and antibiotics, were confined to the inpatient setting, he said, and “what the anti-TNFs were able to do was shift a lot of treatments to the outpatient setting.” The major caveat to this is that although many patients initially respond well to anti-TNF therapy, 20% to 30% have primary nonresponse and 10% to 15% have secondary loss of response per year of biologic therapy, he said.

Edward V. Loftus

The incidence of secondary loss of response is variable, with some reports of annual risk at 13% for infliximab (Remicade, Janssen) and 10% to 24% for adalimumab (Humira, AbbVie), the mainstays of biologic treatment for IBD. It also has been estimated that two-thirds of patients develop secondary loss of response within the first year of therapy while the remainder lose response at a significantly less rapid pace.

“We didn’t know why this was happening initially, but then as we began to understand the pharmacokinetics, it became evident that there are huge inter-individual differences in how these drugs are cleared,” Loftus said. “That’s been the big revelation, and that’s why therapeutic drug monitoring is so important in these patients.”

Understanding what causes such marked variability in the rates of biologic drug clearance between patients, as well as knowing what drug concentrations are effective, is essential to providing better patient care in the form of rational evidence-based treatment algorithms, according to William J. Sandborn, MD, from the Inflammatory Bowel Disease Center at University of California, San Diego. “Biologics are highly effective if you sufficiently engage the target, but there are a number of things that can prevent you from effectively engaging the target, one of which is an insufficient drug concentration due to a variety of factors that increase drug clearance.”

William J. Sandborn

According to Sandborn’s 2014 review published in Digestive Diseases, factors that increase clearance of monoclonal antibodies include immunogenicity caused by the development of antidrug antibodies, high TNF concentrations, low serum albumen, high C-reactive protein, high BMI and male gender.

“Together, these factors can result in wide variability in clearance between patients and markedly increased drug clearance in some patients, particularly in the setting of hospitalized patients with severely active ulcerative colitis who are failing intravenous steroids,” Sandborn wrote. Conversely, he added, concomitant immunosuppressive agents decrease clearance. A recent study by Marla C. Dubinsky, MD, MS, chief of pediatric gastroenterology and hepatology, and co-director of the Feinstein Inflammatory Bowel Disease Clinical Center at Icahn School of Medicine, Mount Sinai Hospital, even identified a dose threshold of methotrexate that decreased drug clearance by reducing the production of antidrug antibodies.

“Within those patients who were on concomitant immunomodulators, we looked at both thiopurines as well as methotrexate, and we specifically wanted to know whether it is true that concomitant immunomodulators do indeed lower the risk of antibody formation,” Dubinsky said in an interview at the 2014 Advances in Inflammatory Bowel Diseases meeting. “Within the thiopurine class, we couldn’t find a true dose that was most correlated with reduction in antibody formation. … However, within the methotrexate group, in those patients who were on a dose of … more than 15 mg a week, there was actually zero antibody production.”

Antidrug antibodies are a unique component to therapeutic drug monitoring of biologics, and they “have a profound impact on drug clearance,” Sandborn said.

Antidrug Antibody Interference

According to a review published last year in the Annals of Gastroenterology, antidrug antibodies “interfere with biologic activity by inhibiting the binding of the TNF-alpha inhibitors to both serum and membrane-bound TNF-alpha molecules, and by creating immune complexes that are eliminated by the reticuloendothelial system,” and formation of these antibodies has been demonstrated to correlate with decreased levels of anti-TNFs and reduced clinical response.

“The theory is if you don’t keep measurable drug available to the body it will view the drug as an antigen for which it needs to develop an antibody against,” Dubinsky said. Thus, drug monitoring with biologics uniquely involves not only measuring drug concentration, but also measuring antidrug antibody concentration, as well as understanding the interactions between them, and how this knowledge can be applied in treatment algorithms.

Findings from a study co-written by Loftus and Sandborn published in the American Journal of Gastroenterology in 2010 demonstrated that antidrug antibody and infliximab concentration testing affected treatment decisions in 73% of patients, were a useful adjunct to clinical and endoscopic/radiological assessment, and could optimize treatment algorithms.

Marla C. Dubinsky

In the era before therapeutic drug monitoring, dose escalations or frequency changes were essentially masked, “not driven by any knowledge of the drug concentrations or antibody status,” Dubinsky said. “Since then, we evolved to an [enzyme-linked immunosorbent assay]-based knowledge.” Using enzyme-linked immunosorbent assay (ELISA), the importance of knowing the drug level at the time of trough, as well as antidrug antibody status, became clear, she said. “Preliminary assessments showed strong associations between low drug levels and patients having active disease at the time they were due for infusion, as well as more antibody activity and higher [C-reactive proteins].”

Limitations of Available Assays

Several techniques are commercially available for measuring serum drug levels and antibodies, with options steadily increasing, but the most common is a solid phase double-antigen ELISA, according to a 2014 review in the Annals of Gastroenterology. However, according to David T. Rubin, MD, professor of medicine, section chief of gastroenterology, hepatology and nutrition, at the University of Chicago Medicine, not all assays are created equal.

“There are different assays available in the marketplace, and part of the limitations of therapeutic drug monitoring of biologics is related to the fact that these assays perform differently,” Rubin, a Healio Gastroenterology Peer Perspective Board Member, said. “The assay that Prometheus Labs developed,” — the Prometheus Anser — “is probably the most sensitive assay, has the clearest signal and also has the most evidence behind it because it can detect neutralizing antibodies in the presence of drug.”

The infliximab version of this assay was launched in July 2012, according to a press release, and was “the first commercial test utilizing Prometheus’ proprietary homogenous mobility shift assay (HMSA) platform technology.” An adalimumab version followed in April 2013, according to a subsequent release. According to a validation study published in the Journal of Immunological Methods in 2012, antibodies “as low as 0.036 µg/mL can be measured, even in the presence of 60 µg/mL of infliximab in the serum,” with an HMSA.

“The older ELISA-based assays cannot detect antibodies when drug is present because there is cross-reactivity in the assay,” Rubin said. Unfortunately, he added, although the Prometheus Anser “may be the most sensitive and the assay we know most about, it is also the most expensive.”

David T. Rubin

“As these newer assays have evolved away from the ELISA, there’s been increasing concern about the rising costs,” Dubinsky said. One cost challenge is payers’ unwillingness to cover the cost of testing assays because they consider therapeutic drug monitoring to be experimental. “A major problem is that these newer testing codes are often listed as a miscellaneous code, and because it’s not an ELISA, it often gets red-flagged by the payers, as therapeutic monitoring is not something they would approve,” which she and many experts characterize as short-sighted. “The notion that drug levels — regardless of price — are not even a part of their cost-savings algorithm is unimaginable,” she said.

A widely cited 2013 study by Velayos and colleagues demonstrated that a testing-based strategy for treating patients with IBD assigned infliximab was indeed cost-effective. According to their findings, the drug monitoring compared with an empiric strategy yielded similar treatment efficacy but with reduced costs ($31,870 vs. $37,266 during 1 year). A more recent randomized trial published in Gut confirms these findings by demonstrating that drug monitoring of infliximab in patients with Crohn’s disease who have lost response achieved similar outcomes compared with empiric dose escalation but with reduced costs due to fewer dose escalations

According to Sandborn, who co-wrote the Velayos study, “payers in the United States are increasingly raising the bar for the level of evidence that they require to pay for diagnostic tests, and I think, in part, they are just getting overwhelmed by genomics testing, especially in oncology, and they’re trying to cope with what potentially could be a tidal wave of costs related to diagnostic testing. On the other hand, this is not the usual situation with diagnostic tests. This is a situation where the use of the tests drives down costs, and even if the test is expensive, the alternative is to empirically double the cost of the drug that costs $25,000 a year.”

“The cost study done by Velayos showed that even if a drug test was $5,000, it would still be cheaper than one infusion,” Dubinsky said. “If you get a drug level, you will keep patients on-drug longer and [treatment will be] more durable, which will mean fewer failures and potentially fewer complications of the disease, which will translate to more cost savings down the road for both the payer and the patient. It’s all cyclical; you can’t just take one test out of context and talk about its cost because, unfortunately, the cost to manage an IBD patient is very high when you’re using a biologic therapy.”

Although there are competing assay technologies with variable clinical relevance data that may cause some uncertainty among physicians on which assay is best, an overall lack of competition in the assay market also contributes to cost issues, according to Loftus.

“Currently, in the United States, there are no commercially available assays for certolizumab (Cimzia, UCB) or vedolizumab (Entyvio, Takeda Pharmaceuticals), but … I wouldn’t be surprised if, in the near future, we see a commercially available assay for both,” Loftus said. “But right now, the average clinician doesn’t have access to those assays.” There are other assays in development for infliximab, he added, including one with Mayo Medical Labs. He hopes their availability — as early as March of this year — will lower overall costs.

Dose-Response Relationship

The interconnected relationships between antibody and drug concentration levels, as well as the other previously mentioned clinical factors that contribute to increased drug clearance, should be recognized by clinicians and incorporated into treatment algorithms to optimize treatment for individual patients and reduce costs, according to Sandborn. “When patients experience primary nonresponse or partial response, or respond and then lose response, clinicians should employ therapeutic drug monitoring,” he wrote in a review.

The evidence for how to best utilize drug monitoring in practice, however, is somewhat limited, and varies depending on many factors, including whether a patient has primary nonresponse or secondary loss of response, according to Bruce E. Sands, MD, MS, also from the Icahn School of Medicine at Mount Sinai.

Bruce E. Sands

“As we all learned in medical school, medications of all sorts have a range that is considered to be therapeutic, but for a long time we did not really appreciate that to be the case for biologic therapies,” Sands said. “It’s fair to say now we understand that even biologic therapies probably have a minimum concentration at which we can expect therapeutic efficacy to be best established. There are a number of lines of evidence supporting that, but there are also limitations to what we know with regard to the optimal therapeutic range and also how to insert therapeutic drug monitoring into clinical practice.”

In his presentation at the 2014 Advances in Inflammatory Bowel Disease meeting in Orlando, Fla., Sands discussed a range of data on drug monitoring, including those that indicate that minimum effective concentrations of a drug have been loosely defined, and more broadly, that high levels of a drug are associated with better outcomes, whereas subtherapeutic levels of a drug are associated with worse outcomes because of inadequate suppression of inflammation. Subtherapeutic drug levels are often due to immunogenicity created by antidrug antibodies, but other factors may contribute to rapid drug clearance.

The CLASSIC I and II trials, for instance, which included both induction and maintenance cohorts, demonstrated that median adalimumab levels were significantly higher in patients with clinical remission at week 4 (8.1 vs. 5.05 µg/mL; P < .05). Furthermore, a study published in Clinical Gastroenterology and Hepatology in 2014 showed that higher trough levels of adalimumab are associated with higher rates of mucosal healing. A similar trend has been demonstrated for infliximab concentrations; in the ACT I and II trial cohorts, for example, proportions of patients achieving clinical remission increased with increasing quartiles of infliximab concentrations.  

Accumulating evidence, according to Sands’ presentation, shows that it can be roughly estimated that the minimum effective concentrations for infliximab trough levels are approximately 3 µg/mL and for adalimumab trough levels approximately 5 µg/mL.

Lack of Prospective Data

Despite the wide range of data that are becoming increasingly available, experts agree that more prospective studies actually testing whether therapeutic drug monitoring improves outcomes are needed. “Prospectively, we have little evidence to suggest that therapeutic drug monitoring actually does improve outcomes for patients, although we suspect that may be true,” Sands said. “I’m sure there will be many more studies like TAXIT, which will utilize different algorithms for [therapeutic drug monitoring]. And once we have more of those studies, we’ll have a better idea of how we ought to be incorporating these things into practice on an evidence basis.”

Among the few prospective studies that exist, the TAXIT study is widely cited. This study included an initial optimization phase before the official trial, Loftus said.

“It turned out that by optimizing everyone’s therapy during the optimization phase, they made it more difficult to see any differences in the second phase of the trial,” he said. “It did, however, show that optimizing the dose of drug improves outcomes.”

Sandborn agrees that “the most useful data from the TAXIT trial is not the randomized part, but the observational part that suggests that there’s a considerable number of patients under long-term maintenance therapy who probably should be optimized.”

This trial, according to Dubinsky, provides insight to the question of whether a patient must have repeat drug monitoring. She said proactive optimization as was done in TAXIT lowered the chances of a symptom flare during maintenance therapy.

“If you can optimize them post-induction, the chances of you needing to change their dose again would be very low,” Dubinsky said. “Once you adjust dosing to actually get them to a therapeutic range, you probably don’t need to worry that much about repeat testing because most patients stay well.”

Loftus agrees that one can frame this information in terms of reactive vs. proactive monitoring. “Therapeutic drug monitoring has enabled us to make more informed choices about decisions in patients who are losing response, and now the real appeal (and we only have preliminary data on this) is in the idea of proactive monitoring.” This concept essentially entails checking trough levels proactively to identify patients who are at risk for developing antibodies and proactively escalating the dose to prevent this from happening. Proactive monitoring, Loftus said, may enable physicians to “maintain a much higher percentage of patients on the biologic. Then maybe this ‘primary nonresponse’ problem will go away.”

Dubinsky agrees that proactive monitoring is appealing. “Our goal should be never to let a patient fail, and to be proactive by measuring drug levels after induction to be prepared for drug dosing in maintenance.” She suggested if a physician is aware of low drug levels after the first maintenance infusion, that patient can return to the practice earlier than the recommended 8 weeks.

“If your level was low at the time of first maintenance infusion, it’s going to be low every maintenance infusion,” she added. “The ability to pre-emptively abort someone from getting a very low trough to 0 measurable level is critical to keep a patient on-drug because those are the patients who are getting antibodies.”

Customized Treatment

According to Rubin, the most important thing IBD physicians should understand is that biological therapies are not one-size-fits-all. “The pharmacokinetics and pharmacodynamics of these therapies are variable by patient, by disease and by drug.” Furthermore, primary nonresponse and secondary loss of response, once presumed to occur due to mechanism failure, “we’ve now realized is often a dose or exposure failure instead,” he said. “Recognizing that important point behooves us to understand drug levels and how to optimize or manage them better so that we don’t give up on a therapy that might work if we just change the dosing schedule or do some kind of pharmacokinetic manipulation.”

Sandborn said therapeutic drug monitoring for biologics may seem complicated, but it is not. “To me, this comes down to a simple thing: no drug or low drug, no efficacy. So why guess when you don’t need to?” – by Adam Leitenberger   

References:

For more information: