Adalimumab Trough Levels and Response to Biological Treatment in Patients With Inflammatory Bowel Disease: A Useful Cutoff in Clinical Practice

Adalimumab Trough Levels and Response to Biological Treatment in Patients With Inflammatory Bowel Disease: A Useful Cutoff in Clinical Practice

tory body that oversees herbal dietary supplements, the Center for Food Safety and Nutrition, they have a diff erent statutory defi nition, as Dr Levy points out. Speci fi cally, a medical food as defi ned by the Orphan Drug Act is "a food which is formulated to be consumed or administered enterally under the supervision of a physician and which is intended for the specifi c dietary management of a disease or condition for which distinctive nutritional requirements, based on recognized scientifi c principles, are established by medical evaluation".
Th e authors re-emphasize, as stated in the Guidelines, that medical foods are not subject to the same rigorous premarket review and approval as drugs. Given the derivation of fl avocoxid from herbal (botanical) products, one must acknowledge that the potential exists for medical foods to be subject to variability in constitution as well as to contamination and adulteration. Paradoxically, although stipulated to be administered under the supervision of a physician, they do not require a prescription.
In response to Dr Levy's second issue in which he stated that fl axocoxid was used as the primary example of drug-induced liver injury (DILI) associated with herbal products, we wish to clarify that it was not proposed in this way. Rather, it was presented as an example of a medical food that has been linked to hepatotoxicity by some authors of the Guidelines (with cooperation and assistance from Dr Levy himself).
Finally, Dr Levy takes the issue with fl avocoxid being singled out as an injurious substance when, in comparison, nonsteroidal anti-infl ammatory drugs (NSAIDs) are more likely to lead to signifi cant morbidity. We acknowledge that liver injury from fl avocoxid and herbal dietary supplements is likely to be rare; however, the purpose of the Guidelines was not to compare the risks of injury from the agents discussed with those from commonly used over-the-counter medications, such as NSAIDs, or to weigh in on the overall risk to benefi t of various medications. Th e purpose of the Guidelines was to suggest preferred approaches to the diagnosis and management of DILI.

CONFLICT OF INTEREST
Th e authors declare no confl ict of interest. To the Editor: We read with great interest the article by Roblin et al. ( 1 ) about the development of an algorithm incorporating pharmacokinetics of adalimumab in the management of patients with infl ammatory bowel disease (IBD). Th e fi ndings of their study emphasize the importance of therapeutic drug monitoring during biological therapy in IBD patients so as to improve the workup of loss of response to treatment. Th ey confi rm previous fi ndings by showing that the absence of mucosal healing was associated with adalimumab trough levels below 4.9 μ g/ml, and elaborate more by proposing to optimize adalimumab treatment schedule or to switch to another antitumor necrosis factor-α drug when antidrug antibodies are absent or present, respectively, and to switch to another therapeutic class when high adalimumab trough levels are found in patients who show loss of response during treatment ( 1,2 ).

Adalimumab Trough Levels and Response to Biological Treatment in Patients With Infl ammatory Bowel Disease: A Useful Cutoff in Clinical Practice
We feel that these results may greatly improve the management of patients with IBD under adalimumab treatment, as they propose a pragmatic and evidence-based algorithm that can be used in clinical practice in order to optimize the available resources. However, before an algorithm can be generalized to clinical practice, its results need to be reproduced in independent series. Th erefore, we sought of interest to evaluate how assessment of adalimumab trough levels were associated with loss of response in our series of patients with Crohn's disease on adalimumab treatment.
We assessed adalimumab trough levels during the follow-up of 23 patients with Crohn's disease who were infl iximab-naive and who responded to adalimumab induction therapy. In these patients, median Harvey-Bradshaw Index before starting adalimumab was 10 (range 5-17) and significantly decreased to 4 (range 3-8; P <0.0001) following adalimumab induction dose at 160/80 mg. Median follow-up was 48 weeks, and during this period 13 patients (56.5%) showed loss of response to adalimumab. In these patients, adalimumab trough levels were assessed using an ELISA Kit (Matriks Biotek, Ankara, Turkey). We found that median adalimumab trough levels in patients who showed loss of response were 4.8 μ g/ml (range 2.4-7.2 μ g/ml), and 7.5 μ g/ml (range 6.6-8.4 μ g/ml) in patients who maintained remission ( P =0.01). Noteworthy, these results are quite similar to those obtained by Roblin et al. in their preliminary study, and further confi rm the relevance of biologic drugs' pharmacokinetic assessment so as to improve the management of patients aff ected by IBD, and in order to tailor adalimumab treatment to individual patient's profi le ( 2,3 ). In conclusion, our preliminary results support the usefulness of therapeutic drug monitoring as an on-treatment tool able to optimize therapy, and confi rm that an adalimumab trough level of 4.9 μ g/ml is a valid cutoff to assess loss of response to treatment. Th is study adds to the one we used in our meta-analysis, which also clearly shows the interest of monitoring adalimumab trough levels (TRA) for predicting remission ( 3 ). In this meta-analysis, patients with IBD with TRA over a predefi ned cutoff were more likely to be in clinical remission with an odds ratio (OR) of 2.6 (95% confi dence interval (CI): 1.79-3.77, P <0.0001). Th e association was stronger if the analysis was limited to the adult population ( N =3, OR: 7.05, 95% CI: 3.58-13.9, P <0.0001). In this meta-analysis, three other studies, in addition to our describes a cutoff for TRA between 4.5 and 5.5 μ g/ml ( 4-7 ).

Response to Bodini
Th e results obtained in this letter are highly consistent with those already published. Of these fi ve studies, the four of them based on an enzyme-linked immunosorbent assy (ELISA) technique allowed to determine TRA cutoff values that are quite similar (4)(5)(6).
In a comparison study of assays for infliximab (IFX) and antibodies to infl iximab measurement by ELISA, a very good linear correlation was obtained ( 8 ). Th ese data on IFX and those now described for adalimumab seem to demonstrate that the situation should be the same for adalimumab.
Regarding the homogeneous mobility shift assay assay, it is still diffi cult to have cutoff values close or similar to those reported in ELISA. However, several recent studies describe very similar cutoff values than those obtained by ELISA ( 7,9 ).
In conclusion, the values of cutoff s for TRA associated with clinical remission seem fairly consistent across each study and this is regardless of the biological assay used. To improve the quality of our decision-making algorithm, it is now important to report the data on the induction of antiadalimumab antibodies including their kinetics and stability over time (permanent vs transient). Indeed, the immunogenicity data are particularly important in the case of IFX. Th e discussion about the assay used for the determination of antiadalimumab antibodies, however, remain open and the availability of international controls should allow a better standardization of data ( 10 ).