December 19, 2019
12 min read

Therapeutic Drug Monitoring: Empirical Evidence Catches up to Common Sense

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

Click Here to Manage Email Alerts

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

Common sense might dictate that, in maintaining health and wellness, being proactive is generally better than being reactive. But the practice of medicine often demands more than just common sense. It demands data and evidence to support clinical decision making. The discussion of therapeutic drug monitoring in gastroenterology hinges on this conundrum. Proactive monitoring appears to be the common-sense approach, but the data supporting its use have not yet closed the case.

At the moment, gastroenterologists largely use reactive therapeutic drug monitoring (TDM). Meaning, when a patient develops symptoms on a therapy, the clinician will assess drug concentrations and antibody levels to help determine what the next best step is (i.e., increasing the dose of the drug or switching therapies).

“Waiting for a patient to fail a therapy — that is, reactive monitoring — is counter intuitive to preventing loss of response and/or anti-drug antibody development,” Marla C. Dubinsky, MD, chief of pediatric gastroenterology and hepatology and co-director of the Susan and Leonard Feinstein IBD Clinical Center at Mount Sinai Hospital, told Healio Gastroenterology and Liver Disease. “Although less costly than proactive as it relates to drug costs and testing costs, the downstream durability of the therapy will be diminished.”

“Overall, we need more data on what a personalized target level should be, at different time of the disease process, before we can justify having proactive TDM standard of practice,” Aline Charabaty, MD, said.
“Overall, we need more data on what a personalized target level should be, at different time of the disease process, before we can justify having proactive TDM standard of practice,” Aline Charabaty, MD, said.

For these reasons, among others, there is a growing movement toward proactive TDM, which essentially calls for initiating the process of monitoring drug concentrations and dosing to an optimal threshold when starting a therapy. The goal is to decrease the risk for primary and secondary loss of response.

But TDM of any kind is not an easy process, and there are several variables to consider, including timing of blood sampling to check serum concentrations; the route, dose, and mechanism of action of the drug being monitored; the accuracy of analyzing concentrations; and the clinical status of the patient, to name a few.

Despite these many considerations, Adam Cheifetz, MD, director of the Center for Inflammatory Bowel Disease at Beth Israel Deaconess Medical Center, remains a strong supporter of proactive TDM. “With biologics in particular, it is critical,” he said, noting that the GI community is increasingly relying on biologic and biosimilar therapies. “If the patient gets too little medication, the drugs are ineffective. Additionally, low drug concentrations are associated with immunogenicity and loss of response.”


But there are skeptics of proactive TDM, and for good reason. There are cost issues to consider, including expensive lab analyses. Further, although drugs like infliximab (Remicade, Janssen) and adalimumab (Humira, AbbVie) have been studied fairly extensively for TDM outcomes, ustekinumab (Stelara, Janssen), vedolizumab (Entyvio, Takeda) and others are only just beginning to undergo rigorous investigation, so it is difficult to generalize. Moreover, the research and development community must perfect assay technology to measure concentrations of all of these drugs. Then there are disease-specific issues to consider, and factors associated with patient selection, because a one-size-fits-all approach is likely not going to work.

It is a multifactorial equation with more questions than answers regarding proactive TDM. But as more data emerge, that tide may be turning.

Landmark Data

Traditionally, the principal argument against the proactive approach was that the available data supporting its use were largely retrospective vs. prospective. That changed with findings from Amit Assa, MD, MHA, and colleagues, who conducted the randomized, controlled PAILOT study. The aim was to assess the primary outcome of corticosteroid-free clinical remission at every time point in a cohort of biologic-naive children with Crohn’s disease who had responded to adalimumab induction. Results showed that 82% of children in the proactive TDM group reached the primary endpoint, compared with just 48% of those in the reactive group (P = .002).

Adam Cheifetz, MD
Adam Cheifetz

At the end of the study, 87% of patients in the proactive TDM group received dose intensification compared with 60% of patients in the reactive TDM group (P = .001), which, according to Cheifetz, indicated that the majority of patients were underdosed with every other week of adalimumab.

Another watershed moment for proactive TDM was the TAXIT trial. The 1-year randomized controlled trial included 263 adults with IBD who had stable responses to maintenance infliximab. Researchers aimed to escalate or reduce doses to reach a target trough concentration of 3 g/mL to 7 g/mL in all patients.

Ninety-one percent of patients (n = 76) with trough concentrations under 3 g/mL achieved the target trough concentration after dose escalation. This resulted in a higher proportion of patients with Crohn’s disease in remission than before dose escalation (88% vs 65%; P = .02).

Cheifetz said the findings are extremely encouraging, but he also understands that other experts may have reservations. “They took patients with Crohn’s or UC who were in remission or stable clinical response and then checked drug concentrations,” he said. “It was only after they put everybody in a therapeutic window that they randomized them to proactive or reactive TDM. This design issue is why the study did not reach its primary endpoint. However, despite this flaw, many secondary endpoints favored continued proactive TDM.”


In short, because patients in the cohort were already in stable condition, it may be difficult to draw the conclusion that proactive TDM is clearly superior to reactive.

That said, given the encouraging results of this trial, it is not hard to imagine that more trials with similar outcomes will emerge in the coming years. Meanwhile, clinicians must try to make sense of retrospective analyses showing that higher drug levels early after induction and during maintenance were associated with clinical, biologic and endoscopic remission.

“A study by Cheifetz showed that proactive TDM in patients in remission after induction with infliximab was associated with less anti-drug antibody formation and infusion reactions, increased drug durability, decreased IBD-related hospitalization and surgeries over time, compared to reactive TDM,” Aline Charabaty, MD, director of the Inflammatory Bowel Disease Center and clinical director of the division of gastroenterology at Johns Hopkins Sibley Memorial Hospital, said in an interview.

Similarly, in the PANTS cohort study, data demonstrated that week 14 infliximab levels greater than 7 µg/mL and adalimumab drug levels greater than 12 µg/mL were associated with clinical remission at weeks 14 and 54.

The proportion of patients who developed anti-drug antibodies was 62.8% for infliximab and 28.5% for adalimumab. For both therapies, suboptimal week 14 drug concentrations predicted immunogenicity, and the development of anti-drug antibodies predicted subsequent low drug concentrations.

“Though most of the data to date are in the maintenance phase, proactive TDM is most important during induction and post-induction,” Cheifetz said. “That is when patients are sickest and have the highest drug clearance. This is when patients are most likely to have sub-therapeutic drug concentrations. The recent trial by Assa and colleagues is an excellent example.”

Charabaty agrees that, “early post-induction is likely the best time to get drug levels, so that therapy can be optimized early and effectively to maximize chances of achieving and maintaining remission.”

Marla C. Dubinsky, MD
Marla C. Dubinsky

For Dubinsky, drug-specific questions also must be answered, and they must be answered for every therapy on the market. “TDM, in theory, can be done for all biologics and agnostic to any previous exposure,” she said. “But getting the dosage right is critical for all biologics. We have the most exposure response data for the anti-TNF drugs. More work needs to be done to understand the exposure response for non-TNF biologics.”

The other question for Charabaty is the frequency with which drug levels should be checked during the maintenance phase after the patient is in clinical and/or endoscopic remission. The research community is zeroing on answers to these questions for infliximab and adalimumab, but if there is another central argument against proactive TDM, it is that few studies have been conducted in other drugs.


Beyond Infliximab, Adalimumab

Waqqas Afif, MD, associate professor of medicine in the division of gastroenterology at McGill University Health Center and GI Site Director at Montreal General Hospital, aimed to rectify the shortage of data for drugs other than infliximab and adalimumab. His group published a review article on the available data on TDM with ustekinumab and vedolizumab.

During maintenance therapy for ustekinumab, drug concentrations of greater than 1µg/mL were associated with improved outcomes, while for vedolizumab there was an association of improved outcomes with concentrations of 5 µg/mL to 11 µg/mL.

“From the pivotal clinical trials of [ustekinumab] and [vedolizumab], we know that during maintenance treatment there is an association between drug concentrations and clinical and endoscopic remission,” Afif said in an interview. “It is important to note that these concentrations are just associations and not an absolute threshold concentration of drug that clinicians should target.”

In the setting of loss of response, dose optimization when concentrations are lower may be of benefit, Afif added. “What is important to note is that the rates of antibody formation with these newer biologic medications are much lower — approximately 10% for vedolizumab and less than 5% for ustekinumab — than we see with anti-TNF medications like infliximab and adalimumab, so TDM testing to detect antibodies for ustekinumab and vedolizumab may be less important,” he said.

In another key data set involving these two drugs, Plevris and colleagues found that a post-induction vedolizumab trough level more than 18 µg/mL was associated with mucosal healing within the first year of therapy. A post-induction ustekinumab trough greater than 3.3 µg/mL was associated with clinical remission.

While Charabaty is encouraged by these findings, she acknowledges their limitations. “Higher drug concentration of vedolizumab or ustekinumab are associated with improved treatment outcomes,” she said. “However, there are no studies comparing treatment outcomes after dose optimization with proactive or reactive TDM with these newer classes of biologics.”

Waqqas Afif, MD
Waqqas Afif

Regarding therapies such as golimumab (Simponi, Janssen) and certolizumab (Cimzia, UCB), Afif said that the data are limited. “Tofacitinib is a small molecule medication and there is no antibody formation,” Afif said. “Drug clearance is stable across patients, so TDM will not be helpful.”

Despite the dearth of data for most drugs in the GI armamentarium, Afif sees a critical sign that researchers are beginning to take the question of TDM seriously. “Interestingly, clinical trials with new biologic medications are being designed to take into account drug concentrations and/or exposure during treatment,” he said. “This should generate a lot of useful data on the clinical utility of TDM.”


Barriers to TDM

Michael J. Rosen, MD
Michael J. Rosen

For all of the movement in the direction of proactive TDM, barriers to its use remain. “TDM is an area where experts interpret the available evidence differently and, therefore, disagree on whether proactive TDM should be broadly applied,” Michael J. Rosen, MD, medical director of the Schubert-Martin Inflammatory Bowel Disease Center at Cincinnati Children’s Hospital Medical Center, told Healio Gastroenterology and Liver Disease. “So, the lack of consensus in the field is an obstacle to uptake.”

Insurance coverage can also be problematic, both for clinical and administrative reasons. “For some reason, proactive monitoring is listed by insurance carriers as ‘experimental’ for most patients despite the cost savings associated with precision dosing,” Dubinsky said.

Although Rosen stated he routinely uses proactive TDM in his own practice, he acknowledged it may add to the administrative workload for many clinicians. “It is simply more work to order, interpret and respond to levels on many patients on a regular basis,” he said.

But there are solutions to this particular set of challenges, according to Dubinsky. “It is a matter of educating third party payers to benefits in outcomes,” she said. “Specifically, they need to understand that proactive monitoring keeps patients on drug longer and requires less health care utilization.”

Cost of TDM laboratory analysis was at the top of Charabaty’s list of obstacles, along with wait time to get results. “This limits the ability to make immediate, quick, decisions regarding drug optimization in our sicker patients,” she said.

This brings the discussion to patient-level questions that remain unanswered. “Most studies looked at the optimal drug level during induction or early maintenance, when the disease is active,” Charabaty said. “Would that be the same drug level we need to keep once the patient is in endoscopic remission for more than a year? Do we need a lower drug level to maintain remission in patients in stable remission?”

At the heart of the issue is that there is no consensus on what exactly makes a therapeutic drug level, according to Charabaty. “We know that our target drug levels are different during induction and early maintenance, for luminal Crohn’s and perianal disease” she said. “It could also vary between early maintenance and long-term maintenance. In addition, target drug level could vary between individuals, and might be different in a patient who is in on his/her first biologic vs. a patient with history of loss of response to several therapies.”


Target drug level might also be different based on disease behavior, severity, and phenotype, and it might be different in patient at higher risk for disease progression and surgery, according to Charabaty. “Overall, we need more data on what a personalized target level should be, at different time of the disease process, before we can justify having proactive TDM standard of practice,” she said.

Reading Assays

If there is another crucial unmet need in the TDM sphere, it is for affordable, reliable technology to measure drug concentrations. Afif was a researcher on a study presented at Digestive Disease Week 2019 that exemplifies this need.

“The thought we had before the study was that drug concentrations across assays were equivalent, but in terms of all the different assays available, we know that antibody detection/reporting is quite variable with the assay being used,” Afif said. The researchers demonstrated that among three assays for measuring ustekinumab concentrations — the Prometheus, Theradiag European ELISA and Dynacare/Progenika products — there was a twofold higher drug concentration reported with Prometheus compared with the other two.

Rosen noted that commercial assays for monitoring adalimumab and infliximab report two results: drug level and anti-drug antibody level. Different assays employ various technologies, meaning that there is not yet standardization or uniformity in the technology clinicians are using to measure drug levels.

“It is important to know whether the anti-drug antibody portion of the assay you use is drug-sensitive or drug-tolerant,” Rosen added. “A drug-sensitive assay can only detect anti-drug antibodies in the absence of drug; that is, they can only report on the presence of absence of antibodies when the drug is undetectable. A drug tolerant assay can detect anti-drug antibodies independent of the presence of drug.”

Some companies employ a drug-tolerant assay but use a reflex system that only runs the anti-drug antibody test if the drug level is low, according to Rosen. “The advantage of a drug-sensitive assay or reflex system is that providers are not at risk of inappropriately switching drugs due to potentially clinically insignificant antibody levels seen when there are detectable drug levels,” he said. “In a sense, this simplifies clinical decision making.”

The advantage of a drug tolerant assay is that it provides more information, Rosen added. “You can be alerted to developing antibodies even the presence of drug,” he said. “In our practice, this may prompt us to monitor drug and antibody levels more frequently for a period, or, if the antibody level is high enough, start an immunomodulator such as methotrexate to reverse the antibodies.”

While drug levels are reported in micrograms per milliliter across assays, anti-drug antibody levels are reported in various absolute or arbitrary units and have various sensitivities for detection across assays, according to Rosen. “It is important to be familiar with how anti-drug antibodies are reported in the assay you use, and what evidence there is for clinically significant antibody levels with that particular assay,” he said.

The take-home message is that some assays are more sensitive than others, and every assay is read differently. The implications in the clinic are clear: the technology does not yet provide clinicians with clear answers as to if, or when, it is necessary to switch drugs. This is just one more uncertainty surrounding proactive TDM.


Looking Ahead

Regardless of these challenges, Dubinsky remains optimistic about proactive TDM. “In my opinion, there are a lot of ‘pros’ and no meaningful ‘cons’ to proactive monitoring,” she said.

For Charabaty, at the current state, it is a matter of patient selection. “I recommend proactive TDM for patients who are likely to clear the drug faster, those with extensive or severe disease burden, with high CRP and low albumin levels,” she said. “Patients who might require higher drug levels to achieve response – such as perianal Crohn’s disease – are also candidates, as are those who have a disease that is difficult to control or who have failed or lost response to more than one biologic.”

The data support many of these recommendations, according to Cheifetz. So now it is a matter of spreading the word. “I do not know what it is going to take to get more widespread use of proactive TDM,” he said. “I will say that I have seen a huge change over the last 5 years. People are talking about it, and when I talk to people about it, it makes sense to them. The bigger thought leaders are coming around. When the next round of studies emerges, hopefully we can put an end to the discussion.” – by Ryan McDonald and Rob Volansky

Disclosures: Afif reports being on the advisory board or speaking for AbbVie, Janssen, Merck, Novartis, Pfizer, and Takeda, and receiving research support from AbbVie, Janssen, Prometheus and Theradiag. Charabaty reports being on an advisory board or speaking for AbbVie, Janssen and Takeda. Cheifetz reports consulting for AbbVie, Janssen, Takeda, Prometheus, Pfizer, and Grifols; research from Inform Diagnostics. Dubinsky reports consulting for AbbVie, Arena, Boehringer Ingelheim, Genentech, Gilead, Janssen, Pfizer, Projections Research, Takeda, and UCB; and being co-founder of MiTest Health and Cornerstones Health. Rosen reports a research collaboration with Prometheus Laboratories Inc.