Therapeutic DMARD Monitoring Has Potential but Also Hurdles

Therapeutic drug monitoring (TDM) — the practice of using laboratory testing to measure blood levels of drugs — has garnered growing interest among rheumatologists in managing patients on disease-modifying antirheumatic drugs (DMARDs), but that hasn’t exactly translated to widespread practice.

While TDM has made some inroads with patients taking monoclonal antibodies, specifically infliximab, its uptake has encountered a number of headwinds, not the least of which is a lack of evidence and clinical guidelines, uneven access and standards of assays, and even an uncertainty about how to interpret laboratory results.

“In some fields, such as neurology, TDM is accepted for antiepileptics,” Michelle Petri, MD, MPH, director of the Johns Hopkins Lupus Center, Baltimore, told Medscape Medical News. “In rheumatology, though, TDM is underutilized and not adequately championed by the American College of Rheumatology.”

She noted that TDM is most acutely needed for management of systemic lupus erythematosus, where nonadherence is a major problem. “Whole blood hydroxychloroquine monitoring has proven beneficial for identifying nonadherence, but also to pinpoint patients who are on too much, a risk factor for retinopathy,” Petri said.

“The state of therapeutic drug monitoring in general has been interesting when you think about its use in autoimmune disease because it’s very much used in gastroenterology and it’s been much less used in rheumatology,” Zachary Wallace, MD, codirector of the Rheumatology & Allergy Clinical Epidemiology Research Center at Massachusetts General Hospital in Boston, told Medscape Medical News. “Some of that may have to do with the interpretation of the availability of evidence, but I think it’s something clinicians will come across more and more often in their practice and wondering what its role might be,” he added.

The movement to precision medicine also portends to grow interest in TDM in rheumatology, said Stephen Balevic, MD, PhD, a rheumatologist and pharmacologist at Duke University and director of pharmacometrics at the Duke Clinical Research Institute, Durham, North Carolina.

“It’s a very exciting time for rheumatologists to begin thinking outside box on what it means to study precision medicine, and I think pharmacology is one of the most overlooked aspects of precision medicine in our community,” he told Medscape Medical News.

That may be because older DMARDs, namely hydroxychloroquine and methotrexate, came to market when regulatory requirements were different than they are today, Balevic said. “Many of the older conventional DMARDs were discovered incidentally and never really had the traditional pharmacokinetic-pharmacodynamic trials to determine optimal dosing, or perhaps that was extrapolated from other populations,” he said.

So, the “one-size-fits-all” approach does not work for prescribing older or even some of the newer DMARDs for rheumatologic disorders, Balevic said.

Reactive vs Proactive TDM

Among the few trials that examined TDM in rheumatology patients are the NOR-DRUM A and B trials in Norway. Marthe Brun, MD, PhD, a rheumatologist at the Center for Treatment of Rheumatic and Musculoskeletal Diseases at Diakonhjemmet Hospital in Oslo, Norway, and a coauthor of the NOR-DRUM trials, told Medscape Medical News that the trials found an overall benefit to TDM during infliximab maintenance therapy. The trials included not only patients with inflammatory arthritis (rheumatoid arthritis, psoriatic arthritis, and spondyloarthritis) but also patients with inflammatory bowel disease and psoriasis, Brun said.

Brun explained that two types of TDM exist: Reactive and proactive. “Reactive TDM is when you use it to find the reason for a patient having a flare or disease worsening,” she told Medscape Medical News. “Proactive TDM would be regular testing to keep a patient within a therapeutic range to avoid flare because of low drug concentrations.”

Gastroenterologists are more inclined than rheumatologists and dermatologists to use reactive TDM, she said. “There have been no recommendations regarding proactive TDM because of the lack of data.”

In Europe, Wallace noted that European Alliance of Associations for Rheumatology (EULAR) recommendations consider the use of TDM in specific clinical scenarios, such as when treatment fails or to evaluate immunogenicity of a reaction, but they are limited. The American College of Rheumatology (ACR) does not have any recommendations for the use of TDM.

Based on the NOR-DRUM trials, rheumatologists in Norway have published their own guidelines for TDM for infliximab in rheumatologic disease, but they are in Norwegian and have not yet been taken up by EULAR, Brun noted. Publication of those recommendations in English is pending, she said.

“But for other subcutaneously administered TNF inhibitors, there’s a lack of data,” Brun added.

The State of the Evidence

NOR-DRUM A did not support the use of proactive TDM in the 30-week induction period as a way to improve disease remission in patients with chronic immune-mediated inflammatory disease. NOR-DRUM B, which evaluated TDM over a year, found the approach was more likely to lead to sustained disease control for that period.

Brun’s group recently published an analysis of the trials. “We did not find an overall effect during the initial phase of the treatment, the first 30 weeks,” she told Medscape Medical News.

“Then we looked at subgroups, and we found that the patients that developed antidrug antibodies [ADAs] had an effect, and ADA are associated with poorer outcomes as well as infusion reactions for patients treated with infliximab.

“So, it’s probably a benefit to be able to detect these ADA early before the patient experiences a disease flare or infusion reaction,” Brun added. “It facilitates for the clinician to take action to, for example, increase the dosing or switch therapy.”

However, the quality of the data supporting TDM in rheumatology is limited, Balevic said. “There’s very good observational data, but we have very few clinical trials that actually leverage TDM,” he said.

NOR-DRUM is the exception, he said. “Ideally, we need more of these dose-optimization trials to help guide clinical practice,” he said. But it stands alone.

Wallace noted several take-home messages from the NOR-DRUM trials, namely that using TDM to prevent ADA may be more effective during the maintenance phase of treatment than the induction phase. However, he said, the evidence is still emerging.

“It’s reasonable to say that we’re at an early stage of the evidence,” he said. “If you look at the large trials that have been done in rheumatology, they’ve combined patients with many different types of conditions, and a lot of our recommendations in rheumatology are disease-specific — in rheumatoid arthritis, in vasculitis. There’s a lack of data in specific diseases to guide or examine what the role of TDM might be.”

In the meantime, no fewer than four clinical trials evaluating TDM with tumor necrosis factor (TNF) inhibitors in rheumatologic diseases are ongoing or have completed but not yet released results, according to Wallace. Three Adalimumab Drug Optimization in Rheumatoid Arthritis trials are underway: The first is evaluating drug tapering vs disease activity score; the second is testing low or usual drug concentration; and the third is studying switches to etanercept or a non-TNF inhibitor drug (abatacept, rituximab, tocilizumab, or sarilumab) in patients failing treatment. Another trial called Tocilizumab Drug Levels to Optimized Treatment in RA is randomizing patients with high drug levels to dose maintenance or dose reduction. All four trials are sponsored by the Reade Rheumatology Research Institute, Amsterdam, the Netherlands.

Until clearer answers emerge from clinical trials, a number of barriers to and questions about the potential for TDM in rheumatology persist.

Barriers to Wider Use of TDM

“The biggest barrier with TDM is simply just a lack of what to do with the data,” Balevic said. “The clinician needs clear-cut guidance on what to do with the drug level. So, in other words, what is the target concentration for the drug? And if that target is not the goal, how should that dose be adjusted?”

The optimal drug levels, particularly for the older conventional synthetic DMARDs, simply have not been validated by clinical trials, he said.

“Different studies may report different target drug levels, and this could be due to different underlying population, or a different matrix — a measure of whole blood vs plasma — or even the timing of the sample,” he said. Balevic led a pharmacokinetic study earlier this year that proposed an algorithm for determining the number of missed hydroxychloroquine doses.

“This really goes back to the clinician needing to draw on a lot of pharmacology training to interpret the literature,” Balevic added.

That gets to the need for more education among rheumatologists, as Brun pointed out. “The physician needs to be educated about therapeutic ranges, when to assess concentrations of drug antibodies, and how to react to the results,” Brun said.

Which ADAs to identify is also problematic. “For antidrug antibodies, it’s especially challenging because there are so many assay formats in use, and it’s a bit complicated to analyze these antidrug antibodies,” Brun said. “There’s no consensus on what calibrators to use, and there’s no standardization of how to report the results, so you can’t really compare results from different assays. You need to know what your laboratory is using and how to interpret results from that particular assay, so that’s a challenge.”

Variability in drug tolerance also exists across assays, Wallace noted. “One of the challenges that have come up in the discussion of therapeutic drug monitoring is understanding what the target level is,” he said. “Defining what the target level might be for a specific condition is not something that’s well understood.”

Breaking down the science, he noted that an ADA can bind to a monoclonal antibody, forming an immune complex that avoids detection. Drug-sensitive assays may detect high concentrations of ADAs but miss low or moderate concentrations. Drug-tolerant assays may be more likely to detect low concentrations at ADAs, but the clinical significance is unclear.

Cost and Patient Trust as Barriers

“The costs vary a lot from assay to assay,” Brun said. “Some commercial assays can be really expensive.” In Norway, a dedicated lab with its own in-house assays helps to keep costs down, she said.

But that’s not the case in the United States, where insurance coverage can be a question mark, Shivani Garg, MD, a rheumatologist at the University of Wisconsin (UW)-Madison and director of the UW-Madison Health Lupus and Lupus Nephritis Clinics, told Medscape Medical News. “A lot of insurances are covering therapeutic drug monitoring, but for the high-deductible plans, there should be a way to offer these important tests to patients at a lower cost or figure out a way for coverage for those patients so that they can show that there are benefits of therapeutic drug monitoring without being sent a really big bill,” she said.

Patient trust could be another potential barrier, Garg said. “A lot of times there is not shared decision-making involved in why this test is being done, how those tests will help us as clinicians, and [patients’ understanding of] the use of the medicine,” Garg said.

“If the shared decision-making to build trust is not there, a lot of times patients worry that they’re being under surveillance or they’re being watched, so that might add to the lack of trust in the core issues that are critical threats to patients with chronic diseases because this is a lifelong partnership,” she said.

Convenience is another issue. “Particularly with mycophenolate levels, a lot of studies have used area under the curve, so getting an area under the curve level over a period of 12 hours would require several samples,” Garg said.

Testing protocols are also uncertain, Garg added. “A few data points…are missing, like how we use the data over time,” she said. “If you do it for a given patient over several years, how often should you do it? How often do the levels fluctuate? How are the data used to inform dosing changes or monitoring changes?

“When those pieces are put together, then we are more likely to build up an intervention that clinicians can use in clinical practice, so they know how to order it and how frequently do it — every 6 months, 3 months, or every month. And then, over a period of time, how to adjust the dosing. That’s the big question.”

Who May Benefit Most From TDM?

In the NOR-DRUM trials, patients at risk of developing ADA early on, before a disease flare or infusion reaction, seemed to benefit most from TDM. But who are those patients?

“We looked at risk factors for developing antidrug antibodies, and we found that that patients with high disease activity when starting treatment, smokers, and patients with rheumatoid arthritis had a higher risk than other patients, as did patients who are not using concomitant immunosuppressive therapy,” Brun said.

“During treatment, we also found that low serum drug levels and drug holidays above 11 weeks were also risk factors,” she added.

The NOR-DRUM researchers also evaluated genetic risk factors and found that patients with the HLAD22 gene variant were also at increased risk of developing ADA.

While NOR-DRUM evaluated only infliximab, some of its lessons may be applied to other DMARDs, Brun said. “We think that for other subcutaneously administered TNF inhibitors, you would probably see the same effect of proactive TDM, but we currently do not have data on that,” she said. A study similar to the NOR-DRUM design will evaluate this in Norway, Brun added.

She explained why the findings with infliximab may extend to adalimumab, which may be the second most immunogenic TNF inhibitor after infliximab. “The administration is different; it’s administered more often than infliximab; that would also make the results more uncertain to generalize to the other treatments, but I would guess there are also benefits of using TDM in other treatments.”

Potential Risks for TDM

Wallace has noted that TDM, with the current state of evidence, carries a number of potential risks. “The potential risks might be that you unnecessarily discontinue a medication because you detected an antibody, or the level seems low and you’re not able to get it higher, but the patient is otherwise doing fine,” he said. “You might end up increasing doses of the medicine that would put the patient at potentially increased risk of infection, as well as obviously more costs.”

That would also lead to more utilization of resources and costs, he said. “Some of those reasons are why there has been hesitation with therapeutic drug monitoring,” Wallace added.

A number of questions also surround the use of biosimilars and ADA levels, Wallace said. While a review of clinical trials found no meaningful differences in terms of immunogenicity between biosimilars and reference products, it did note discrepancies in how the agents were evaluated.

What DMARDs Are Most Suitable for TDM?

Petri said TDM would be useful for monitoring patients on mycophenolate mofetil. “A trough level can at least tell us if a patient is taking it,” she said. “Tacrolimus, used for lupus nephritis, has well-accepted peak and trough trends due to widespread use in transplant.”

Drugs with a wide variability in pharmacokinetics may also be suitable for TDM, Balevic said. That would include hydroxychloroquine, azathioprine, mycophenolate, or even cyclophosphamide. Drugs that have a narrow therapeutic index, such as tacrolimus, cyclosporine, or again, cyclophosphamide, might also be amenable to TDM, he said.

Why Do TDM?

“The two main reasons why somebody would go on to detect drug levels: The first may be to assess medication adherence, and this applies virtually to any drug that rheumatologists use; the second reason is to optimize dozing, either for efficacy purposes or to prevent toxicity,” Balevic said.

“When it comes to optimizing dosing, you should really think about TDM as one tool in our toolbelt,” he said.

Dose is “just a surrogate,” he said. “When we prescribe a drug, what truly matters is the amount of active unbound drug at the site of action. That’s what’s responsible for a drug’s pharmacologic effect.”

However, the same dose, or even the same weight-based dose, does not necessarily mean similar patients will achieve the same amount of exposure to the drug, but TDM can help determine that, he said.

What’s Next

Studies into the use of TDM in rheumatology are ongoing. Brun said her group is currently conducting a cost-effective analysis from the NOR-DRUM trials.

“There’s going to be more studies coming out in the next few years, looking at what impact the use of therapeutic drug monitoring might have on outcomes,” Wallace said.

“As we accumulate more and more evidence, we might see organizations like ACR and EULAR start to weigh in more on whether or not therapeutic drug monitoring can or should be used.”

Petri, Brun, and Garg had no relevant disclosures.

Wallace disclosed financial relationships with Amgen, Alexion, BioCryst, Boehringer Ingelheim, Bristol Myers Squibb, Medpace, Novartis, Sanofi, Viela Bio, Visterra, Xencor, and Zenas.

Balevic disclosed relationships with the National Institutes of Health, the Childhood Arthritis and Rheumatology Research Alliance, and UCB.

Richard Mark Kirkner is a medical journalist based in the Philadelphia area.