Dose de-escalation strategies and role of therapeutic drug monitoring of biologics in RA

Evolving new paradigms Treatment for RA has been evolving over the past decade from traditional step-up regimens to more aggressive strategies. Pivotal to these changes are the early start of treatment (hit early), the use of combination therapy including steroids and rapid escalation to biologics (hit hard) and most importantly, frequently assessing disease activity and acting accordingly. Strategies incorporating these concepts lead to the achievement of low disease activity or remission in most patients, which prevents joint damage and improves function and quality of life [1]. An important disadvantage of the hit-hard approach compared with the traditional step-up approach, however, is that the former method does not allow for individual titration of the minimal effective treatment. Indeed, the traditional step-up approach largely prevents overtreatment, but the high(er) disease activity in the beginning of the disease has to be accepted. To prevent overtreatment when using high-dose or multidrug strategies, treatment must be tapered down when low disease activity is reached up to the point that either the disease activity increases again or medication can be stopped. This way, the minimal effective dose is found and overtreatment is prevented. We, therefore, argue that the switch from traditional step-up to hit-hard strategies inevitably requires proactive down-titration of treatment in a later phase because information about the minimal effective dose for the individual patient is lacking. The paradigm ‘hit hard, hit early’ should therefore be amended with ‘how low can you go’. We will now focus on treatment with biologics, as optimal dosing of these agents is especially important due to the risk of dose-dependent adverse effects and moreover risk of low cost-effectiveness due to high cost [2]. Active dose reduction strategies in RA patients who are doing well on a biologic are not widely adopted by rheumatologists in clinical practice. This is exemplified by the high median of anti-TNF dosages in biologic registries [3], which had several causes. First, an important prerequisite for this approach, frequent visits and monitoring of disease activity, is often not met. Second, rheumatologists, and presumably patients also, are generally reluctant to change treatment, let alone change ‘a winning team’, a patient who is doing well on a particular drug [4]. Let us examine this in more detail. When a patient starts using a biologic and reaches a low disease activity, most commonly, the rheumatologist does not change the medication because the conclusion is that the patient must be responding. However, the relation between the (dose of) medication and improvement in disease activity is more complicated than a simple cause–effect relation. Two alternative scenarios being true for this patient: (i) the patient would have responded to a lower dose (the dose is a supratherapeutic dose) or (ii) the patient would also have improved even without medication (the drug is non-therapeutic). This well-established phenomenon called spontaneous improvement or regression to the mean is often not recognized [5]. The reason is probably because of a form of irrational thinking or cognitive bias and is known as a missing negative fallacy (you do not know what would have happened if medication had not been given—the negative case) and the post hoc ergo propter hoc fallacy (something that happens after something else must be caused by it) [6]. It must be noted that cognitive bias is inherent to human nature and is not specific by any means to rheumatologists. If anything and if this thought might provide some comfort, cognitive bias seems to be reported or at least studied more in surgical specialties than in internal specialties [7]. Regression to the mean is also not rare. The improvement that is seen in 20–40% of the placebo-treated patients in randomized controlled trials concerning biologics in RA, although often called placebo response, is not only caused by placebo effect sensu strictu (expectation bias) but also due to a large extent by regression to the mean [8, 9]. Our own data from a cohort of RA patients who are being treated with infliximab showed that in 37% of patients with low disease activity, low or no infliximab is present in serum with 25% of the patients having human anti-chimeric antibodies (A.vdM, unpublished data 2010). Regarding alternative scenario number one: 30% of these patients had high to very high infliximab trough levels (>5 mg/ml). Thus, both supratherapeutical and non-therapeutical dosing is prevalent but remains largely unrecognized. The solution to this problem of overtreatment, which will inevitably increase when more intensive multidrug strategies find their way into clinical practice, is systematic measurement of disease activity and lowering the dose of the drug until either a flare in disease activity occurs or medication can be stopped completely. This has been shown to be a feasible approach and to result in a sizeable reduction in dosage [10, 11] without persistent increase in disease activity in nearly all patients. This kind of biologic step-down protocol has, however, not been studied extensively in rheumatic diseases or indeed other chronic diseases. The reason could be as mentioned before: the approach is contraintuitive, both for physicians and patients, as the treatment is changed in patients who are doing well. Another important factor, however, is the lack of incentive for the pharmaceutical industry to drive this kind of research because of the imbalance between the very real risk of loss of revenue and the possible gain in cost-effectiveness and safety profile. The latter is underscored by van Vollenhoven [12] who estimated that 1 billion dollars have been wasted worldwide because of unwarranted use of infliximab. One drawback of the widespread use of dose de-escalation strategies that remains is the temporary flare in disease activity that by definition will occur in patients treated with (supra)therapeutical dosages, although of course not in patients who are doing well even without the drug. This flare in disease activity not only affects the quality of life of the patient but is also associated with more radiological damage [13]. A promising solution is the use of therapeutic drug monitoring (TDM). Prerequisites for TDM are that a drug assay is available; a large variation in both serum levels and the outcome that is to be predicted is present; a clear association exists between drug levels and clinical effect; and there is a small range of therapeutic dosages (limited either by the toxic effect or costs) [14]. The use of TDM by measurement of serum (trough) levels of biologics fulfils all these criteria [15]. So far, however, TDM has not been widely used in clinical practice of RA. This is largely because the test characteristics (cut-off threshold, sensitivity and specificity) of biologic serum levels to predict successful de-escalation have not been established. Therefore, for a successful dose reduction, it is not clear in which patient at what moment serum levels can be measured and what the positive and negative predictive values of these levels are. Future research should aim at providing answers to these important questions. In conclusion, we argue that overtreatment is prevalent in RA and will increase when more intensive multidrug strategies and rapid escalation to biologics find their way into clinical practice. To prevent toxicity and, especially in the case of biologics, high costs and consequently unacceptable cost-effectiveness, we strongly recommend proactive down-titration of treatment in all RA patients who are given biologics and in whom the treatment goal has been reached. TDM seems a promising tool to improve identification of patients who are being overtreated. Disclosure statement: The authors have declared no conflicts of interest.