Pharmacology and Metabolism

6-mercaptopurine (6MP) and its prodrug, (AZA), are purine analogs, which are metabolized via three pathways. The majority of 6MP is catabolized by xanthine oxidase. Xanthine oxidase hydroxylates 6MP preferentially at the 8- position to form 8-oxo-6MP and then the 2- position to form 6-thiouric acid ( Fig. 1 ). The remaining substrate is anabolized through two competing pathways: (1) TPMT converts 6MP to 6-methyl-mercaptopurine ribonucleotides (6-MMPR), which are inactive metabolites. (2) In a pathway involving several enzymatic steps, 6MP is converted to 6-thioguanine nucleotides (6-TGN; the sum of 6-thio- guanine monophosphate, -guanosine diphosphate and -guanosine triphosphate), which are believed to be the active metabolites.

Pathways for thiopurine metabolism. The majority of 6MP is catabolized by xanthine oxidase. The remaining substrate is anabolized to 6-MMPR and 6-TGN. DPK, diphosphate kinase; GMPS, guanine monophosphate synthetase; HGPRT, hypoxanthine guanine phosphoribosyl transferase; IMPDH, inosine monophosphate dehydrogenase; ITPase, inosine triphosphate pyrophosphatase; MPK, monophosphate kinase; XO, xanthine oxidase.

TPMT

The activity of TPMT is critically important in determining the balance between the production of 6-TGN and that of 6-MMPR. In 1980, Weinshilboum and Sladek demonstrated that TPMT activity is determined by a genetic polymorphism. Alleles conferring normal (TPMT ^N ) or low/absent TPMT activity (TPMT ^L ) are inherited in codominant fashion. Approximately 90% of whites are homozygous for wild type alleles (TPMT ^N /TPMT ^N ) and therefore exhibit normal or high enzymatic activity. The 10% of individuals who are heterozygotes (TPMT ^N /TPMT ^L ) have intermediate enzyme activity. Finally, approximately 0.3% are homozygous for low activity alleles (TPMT ^L /TPMT ^L ) and have low to absent TPMT activity. TPMT ^L /TPMT ^L individuals generate very high 6-TGN concentrations and invariably develop profound leukopenia when treated with standard doses of AZA (2.5–3.0 mg/kg/day) or 6MP (1.0–1.5 mg/kg/day). Although such individuals have rarely been treated with miniscule thiopurine doses, treatment is best avoided. Conversely, individuals with normal or high TPMT activity achieve lower 6-TGN concentrations and lower rates of treatment response. Individuals with TPMT heterozygosity/lower TPMT activity display the optimal AZA/6MP pharmacokinetics, achieving higher 6-TGN concentrations and response rates. Nonetheless, other studies have not found a relation between TPMT activity and treatment outcome, It must be kept in mind that although TPMT genotype generally correlates with enzymatic activity, some patients with two normal alleles have intermediate TPMT activity, whereas some heterozygotes actually have normal TPMT activity. The TPMT enzymatic assay is therefore more informative than TPMT genotyping.

Whereas decreased TPMT activity is associated with early leukopenia, myelosuppression commonly occurs independently of a decreased TPMT enzymatic activity. One study found that only 25% of leukopenic events were related to low-intermediate activity, whereas the remaining 75% were sporadic events unrelated to the TPMT activity.

Standard dosing of thiopurines is based on weight (AZA 2.5–3.0 mg/kg/day and 6MP 1.0–1.5 mg/kg/day). TPMT measurements may help to guide thiopurine dosing, in other words with standard dosage reserved for normal TPMT metabolizers but lower dosages used in intermediate metabolizers. A recent study suggested the utility of this approach. Consecutive inflammatory bowel disease (IBD) patients starting AZA or 6MP were followed up for 9 months. The thiopurine dose was individualized using 6-TGN concentrations (range, 235–450 pmol/8 × 10 ⁸ red blood cells [RBCs]) and clinical status. The mean initial dose (as AZA equivalents) was similar (approximately 1 mg/kg/d) for the two TPMT genotypes, but after 9 months the dose was 50% lower in the heterozygous group (0.9 vs 1.8 mg/kg/d, P <.0001). Despite lower doses, heterozygotes had two-fold higher median 6-TGN concentrations at the end of the follow-up period (505 vs 273 pmol/8 × 10 ⁸ RBCs, P = .02). This difference was three-fold when the concentration was adjusted for dose (578 vs 183 pmol/8 × 10 ⁸ per mg/kg/d, P = .0007). The results were similar if TPMT phenotype was used instead of genotype. The investigators concluded that initial AZA doses to attain therapeutic 6-TGN concentrations (>235 pmol/8 × 10 ⁸ RBCs) in patients with IBD might be 1 and 3 mg/kg/d in intermediate and normal metabolizers, respectively.

Although some current guidelines do not endorse pretreatment TPMT testing, avoiding thiopurine therapy in patients with absent or very low TPMT activity should prevent life-threatening toxicities and associated costs of hospitalizations ( Table 1 ). Indeed, two modeling studies found that TPMT testing is highly cost-effective or even cost-saving compared with a policy of no testing. The Food and Drug Administration recommends consideration of TPMT genotype or phenotype testing before initiating thiopurine therapy but also states that TPMT testing cannot substitute for blood count monitoring . In regards to other enzymes in the thiopurine pathway, there are no confirmed pharmacogenetic variations predicting clinically relevant differences in effectiveness and toxicity.

6-TGN

Studies addressing the correlation of 6-TGN concentrations with thiopurine effectiveness have yielded conflicting results. A metaanalysis found that patients with 6-TGN concentrations higher than 230 to 260 pmol/8 × 10 ⁸ were more likely to be in remission (62%) than those with concentrations below the threshold value (36%) (pooled odds ratio, 3.3; 95% confidence interval [CI], 1.7–6.3; P <.001).

The potential value of 6-TGN (and 6-MMPR) measurements was bolstered by a study that assessed the effects of dose escalation on 6-TGN concentrations in 51 IBD patients for whom therapy failed. Metabolite measurements were repeated at a median interval of 13 weeks after dose increase. All but 4 patients had “subtherapeutic” 6-TGN concentrations (<250 pmol/8 × 10 ⁸ RBC) at baseline. The median starting 6MP dose was 0.87 mg/kg/day, with only 22 patients receiving more than the recommended daily dose of 1.0 mg/kg/day. Only 14 of 51 patients (27%) achieved remission (responders) with dose escalation. The change in dose (mg/kg/day) did not differ between responders and nonresponders (median [range]: 0.4 [−0.1 to +0.9] vs 0.56 [−0.15 to +1.2]; P = .1). Remission in responders coincided with significant increases in 6-TGN levels (183 [39–298] to 306 [168–853]; P = .03). There was a small increase in 6-TGN levels among nonresponders (136 [50–378] to 155 [90–707]; P = .046). 6-TGN levels were higher at baseline among the responders than the nonresponders ( P = .006). The absolute change in 6-TGN levels from baseline to follow-up was significantly higher among the responders versus nonresponders (median: 122 [−2 to +627] vs 26 [−177 to +503]; P = .0003). Dose escalation had strikingly different effects on 6-MMPR levels in responders versus nonresponders. Median 6-MMPR levels rose insignificantly among responders from 498.5 (0– 2000) to 2345 (264–11,591). Among nonresponders, median 6-MMPR levels increased from 2201 (0–14,412) at baseline to 9346 (0–45,095). The median changes of 6-MMPR were 1908 (+264 to +9591) versus 7986 (−201 to +35,134) ( P = .0057). Thus, both at baseline and upon dose escalation, responders were characterized by 6-TGN production and nonresponders by 6-MMPR production. These differences were also reflected on the values of the metabolite ratio (6-MMPR÷6-TGN). Responders had lower ratios both at baseline (2.5 vs 18; P <.0007) and at follow-up (9.1 vs 66; P <.0001). Of note, dose escalation produced a similar 3.7-fold increase in the metabolite ratio of both responders and nonresponders. Surprisingly, the baseline TPMT activities did not differ between preferential 6-TGN producers (responders) and preferential 6-MMPR producers (nonresponders). The investigators concluded that metabolite testing identifies IBD patients who are resistant to 6MP/AZA therapy and are biochemically characterized by suboptimal 6-TGN and preferential 6-MMPR production upon dose escalation These findings will require replication in prospective studies. In the future, it will also be important to assess the predictor value of the metabolite ratio by measuring it at shorter time intervals from any dosing changes.

Two clinical trials have compared individualized with standard, weight-based dosing. An unblinded, randomized trial compared standard AZA dosing (2.5 mg/kg/d; n = 33) to 6-TGN–adjusted dosing (target 6-TGN 250-400 mol/8 × 10 ⁸ RBC; n = 25) in patients with active CD (Crohn disease activity index 150–450). The primary end point of steroid-free remission at week 16 was achieved by 43.8% (14 of 32) patients in the standard group versus 44% (11 of 25) in the adjusted group (intent-to-treat analysis). Median 6-TGN concentrations in the adjusted arm were below the target range throughout the trial, and in fact were lower than those in the standard arm. The unblinded design, the selection of an initial dose without account for the baseline TPMT activity, and the nonattainment of target 6-TGN concentrations in the adjusted arm are significant limitations of the trial.

A multicenter double-blind, randomized controlled trial compared the efficacy and safety of weight-based AZA dosing (WD) to individualized AZA dosing (ID) in inducing and maintaining remission in adults and children with steroid-treated CD. In the WD arm, the AZA dose was 2.5 mg/kg/d. In the ID group, the initial AZA dose was 1.0 mg/kg/d (if intermediate TPMT) or 2.5 mg/kg/d (if normal TPMT). Starting at week 5, the dose was adjusted to target 6-TGN concentrations of 250 to 400 pmol/8 × 10 ⁸ RBC, or up to a maximal dose of 4 mg/kg/d. The primary outcome was clinical remission (CR) at 16 weeks. The trial was stopped prematurely because of insufficient enrollment. In the intention-to-treat analysis, CR rates at week 16 were 40% (10 of 25) for the ID and 16% (4 of 25) WD groups ( P = .11). In the per-protocol (PP) analysis (noncompliers excluded), week 16 CR rates were 60% (9 of 15) for the ID and 25% (3 of 12) WD groups respectively ( P = .12). Median 6-TGN concentrations at week 16 among normal TPMT metabolizers in the ID arm (198 pmol/8 × 10 ⁸ RBCs) were below the target of 250 pmol/8 × 10 ⁸ RBCs, and not significantly higher than the corresponding concentrations of the normal TPMT metabolizers of the WD arm (150 pmol/8 × 10 ⁸ RBCs). It must be kept in mind that these (albeit nonsignificant) differences in 6-TGN concentrations reflected differences in the final doses at week 16 (median doses of 3.4 vs 2.3 mg/kg/day in the ID and WD arms respectively, P <.0001). As may have been predicted, intermediate TPMT metabolizers in the ID arm achieved higher 6-TGN concentrations than the normal TPMT metabolizers in the same arm (despite a much lower AZA dose: 1.5 vs 3.4 mg/kg/day): median 6-TGN concentrations 212 (88–413) versus 198 (81–246) pmol/8 × 10 ⁸ RBCs. At week 16, median (range) 6-TGN concentrations in PP remitters (n = 12) and nonremitters (n = 15) were 216 (127–413) and 154 (81–972) pmol/8 × 10 ⁸ RBCs respectively ( P = .07). There were no differences in the frequencies of adverse events, including myelosuppression and hepatotoxicity. In summary, despite a trend in favor of ID over WD, there was no statistically significant difference in efficacy. In the PP analysis, remitters had higher 6-TGN concentrations than nonremitters, but this missed statistical significance ( P = .07). The trial was limited by low statistical power and inability to achieve the target 6-TGN concentrations in the ID arm.

6-MMPR

Elevated 6-MMPR concentrations have been associated with hepatotoxicity, prompting questions regarding the role of routine 6-MMPR monitoring. Dubinsky and colleagues reported increased risk of hepatotoxicity with 6-MMPR concentrations greater than 5700 pmol/8 × 10 ⁸ RBC, and Nygaard and colleagues found that increasing 6-MMPR levels correlated with rises in alanine aminotransferase. However, elevated 6-MMPR concentrations are expected in many patients taking 6MP or AZA, and most of these patients will have normal liver biochemistries. Goldenberg and colleagues found that 12% of patients on thiopurines had 6-MMPR levels greater than 5700, but none had liver chemistry abnormalities. Because measurement of 6-MMPR for purposes of hepatotoxicity monitoring has a low predictive value, patients are simply followed for symptoms and abnormal liver chemistries. Presently, the value of 6-MMPR metabolite measurements is limited to identifying noncompliers. As noted, 6-MMPR (and 6-TGN) testing may have a role in identifying patients with a 6-MMPR dominant metabolism who are likely to have failure of thiopurine dose escalation and may instead benefit from addition of allopurinol ( Table 2 ; also see Table 1 ).

Allopurinol Therapy

A pioneering study of the effects of allopurinol on metabolite concentrations and thiopurine effectiveness further supports the concept of a therapeutic window for 6-TGN concentrations. Sparrow and colleagues treated 20 thiopurine nonresponders who were 6-MMPR preferential metabolizers with combination allopurinol 100 mg orally daily and AZA or 6MP at 25% to 50% of the original dose. 6-TGN concentrations increased from 191 ± 17 to 400 ± 37 pmol/8 × 10 ⁸ RBCs ( P <.001), whereas 6-MMPR decreased from 10,605 ± 1278 to 2001 ± 437 pmol/8 × 10 ⁸ RBCs ( P <.001). In CD patients, the mean partial Harvey Bradshaw Index decreased from 4.9 ± 1.0 to 1.5 ± 0.3 ( P = .001). In patients with ulcerative colitis, the mean Mayo Score decreased from 4.1 ± 0.7 to 2.9 ± 0.7 ( P = .13). The addition of allopurinol also enabled a reduction in mean daily prednisone dosage from 17.6 ± 3.9 to 1.8 ± 0.7 mg ( P <.001) and led to normalization of transaminases. Allopurinol thus (a) shifted thiopurine metabolism toward production of 6-TGN in nonresponding, 6-MMPR preferential producers, (b) improved clinical responses, and (c) improved liver biochemistries. Subsequent, open-label studies have replicated these findings. Randomized, placebo-controlled trials are needed to establish the effectiveness and safety of allopurinol in 6-MMPR preferential metabolizers. Allopurinol inhibits xanthine oxidase, but this activity would not be expected to produce simultaneous increases in 6-TGN concentrations and decreases in 6-MMPR concentrations. TPMT inhibition would intuitively seem the most likely mechanism. Investigators have recently postulated that, in the setting of xanthine oxidase inhibition by allopurinol, aldehyde oxidase hydroxylates 6MP at the 2 position to form 6-thio-xanthine, which is a potent inhibitor of TPMT ( Fig. 2 ).

Hypothesis for increased 6-TGN and decreased 6MMPR on thiopurine-allopurinol cotherapy. In the setting of xanthine oxidase inhibition by allopurinol, aldehyde oxidase hydroxylates 6MP at the 2 position to form 6-thio-xanthine, which is a potent inhibitor of TPMT. The formation of 6-Me-MP from 6MP is suppressed, and the synthesis of 6-TGN is enhanced. AO, aldehyde oxidase; XO, xanthine oxidase.

( From Blaker PA, Monica Arenas A, Fairbanks L, et al. A Biochemical mechanism for the role of allopurinol in TPMT inhibition. Gastroenterology 2011;140:S–769.)

TPMT Activity and Metabolite Concentrations in Clinical Practice

Testing of baseline TPMT enzyme activity is critical in avoiding thiopurines in patients with absent TPMT activity. TPMT testing may have a role in selecting the initial thiopurine dose, in other words 1.0 mg/kg/d in intermediate TPMT metabolizers versus 3.0 mg/kg/d normal TPMT metabolizers. 6-TGN measurements are helpful in characterizing thiopurine failures. Patients for whom therapy fails fall into four groups: (1) noncompliers (absent-low 6-TGN and 6-MMPR concentrations); (2) underdosed patients (low but detectable 6-TGN and 6-MMPR concentrations) who will benefit from higher dosing; (3) patients with “therapeutic” 6-TGN concentrations, who should be given alternate therapies; and (4) patients with preferential 6-MMPR producers, for whom thiopurine dose escalation will likely fail and who should receive alternate therapies. Based on early data, another option in preferential 6-MMPR producers is combination thiopurine-allopurinol therapy (see Tables 1 and 2 ). A recent study demonstrated the clinical usefulness of metabolite monitoring in patients for whom thiopurines fail. Based on weight-based criteria, 50% of 63 patients were underdosed. However, metabolite patterns showed that only 29% of patients were actually underdosed. Nine percent of patients were noncompliant, 53% had either appropriate (40%) or elevated (13%) 6-TGN concentrations, and 9% were preferential 6-MMPR producers. The clinical outcome improved in 40 of 46 (87%) of patients in whom the course of action taken followed the metabolite-directed algorithm, whereas 3 of 17 patients (18%) improved where discordant actions were taken ( P = .0001; Fisher exact test). Fifteen patients (24%) avoided inappropriate escalation of therapy.

Biology and Pharmacokinetics of the Anti-TNF Monoclonal Antibodies

The introduction of the anti-TNF mAb’s (infliximab [IFX], adalimumab [ADA] and certolizumab pegol [CZP]) revolutionized the therapy of CD. Successive trials proved the efficacy of these agents in the induction and maintenance of remission of luminal and perianal CD. IFX and ADA consist of two TNF-binding domains linked to human immunoglobin G [IgG]1 Fc. ADA is humanized, whereas IFX is chimeric, containing a human Fc region and a murine variable region. CZP consists of a pegylated Fab fragment of a humanized IgG1 mAb. Lacking the Fc region, CZP cannot mediate antibody-dependent cell-mediated cytotoxicity and complement-dependent cytotoxicity. These mAb’s interfere with the TNFα system through several mechanisms, including binding and neutralization of soluble as well as membrane-bound TNF (all three agents), apoptosis of T cells and monocytes (IFX and ADA), antibody-dependent cell-mediated cytotoxicity (IFX and ADA), complement-dependent cytotoxicity (IFX and ADA), reverse signaling via membrane TNF (IFX and ADA), and blockade of the CD40/CD40L pathway (IFX). Suppression of mucosal inflammation is ultimately mediated by neutralization of soluble and membrane-bound TNF, death of activated immune cells by cytotoxicity and apoptosis, and downstream effects on inflammatory cytokines, adhesion molecules, and regulatory subpopulations of T cells and macrophages.

A comprehensive discussion of the pharmacokinetics of the anti-TNF mAb’s is beyond the scope of this review. In the context of anti-TNF mAb therapy for IBD, pharmacokinetic studies in this patient population, focused on IFX, have provided the following insights: (1) IFX pharmacokinetics conform to a two-compartment model with first-order distribution and elimination constants. (2) IFX clearance is increased in the presence of antibodies to infliximab (ATI). ATIs form complexes with IFX, which are likely cleared by the reticuloendothelial system. Interestingly, one group reported that patients with higher serum albumin concentrations (SAC) maintained higher IFX concentrations, lower clearance, and longer half-life than patients with lower SAC. These investigators postulated that alterations in the neonatal Fc receptor, which protects both albumin and IgG from intracellular degradation, may explain the observed SAC-IFX relationship. (3) IFX clearance is decreased with coadministration of immunomodulators.

There are no confirmed pharmacogenetic variations associated with response (or nonresponse) to the anti-TNF mAb’s. (For a review, see Vermeire and colleagues. ) Ultimately, and despite a tendency to “lump” IFX, ADA, and CZP together, differences in the mechanisms of action and pharmacokinetics may translate into differences in short- and long-term response, as well as safety.

Immunogenicity of the Anti-TNF Monoclonal Antibodies

All three anti-TNF mAb’s are immunogenic, irrespective of chimeric versus humanized structure. Concerning IFX, the development of ATIs is mitigated by a policy of scheduled maintenance therapy and by the concomitant administration of an immunomodulator, both among patients receiving IFX episodically, as well as among patients receiving scheduled maintenance therapy. Similar to IFX, the frequency of anti-CZP antibodies is reduced by administering CZP regularly and by co-treating with immunomodulators. In PRECISE 2, the rates of antibody formation to CZP were lower in patients who received continuous therapy with CZP (18% among subjects who received induction CZP and maintenance placebo, vs 8% among subjects who received induction and maintenance CZP and maintenance placebo). Anti-CZP antibodies were also less frequent among subjects who received concomitant immunosuppressive agents: 24% versus 8% among subjects who only received induction CZP, and 12% versus 2% among subjects who received induction and maintenance CZP. In PRECISE 3, anti-CZP antibodies were observed in 8.0% of patients in the continuous group versus 17.7% in the drug-interruption group. In PRECISE 1, the frequencies of CZP antibodies did not vary according to concomitant immunomodulatory therapy. In an open label study of ADA, concomitant immunomodulator therapy at baseline did not did not influence the development of antibodies against ADA (AAA). Because circulating IFX and ADA interfere with the ATI and AAA assays, these assays are indeterminate when the anti-TNF mAb is detected.

Optimizing Response to the Anti-TNF Monoclonal Antibodies

For the currently approved anti-TNF mAb’s, response rates to induction therapy approximate 60% at 2 to 6 weeks, with higher response rates seen in patients with shorter disease duration. Despite the high rates of short response, 30% to 50% of the initial responders lose response over time while on scheduled maintenance therapy and require dose intensification. Loss of response can occur through several mechanisms, including altered pharmacokinetics (for example, via increased clearance by antibodies to the anti-TNF mAb or a greater burden of inflammation), changes in the dominant mechanism of inflammation, or the development of other processes that decrease response, such as strictures, irritable bowel syndrome, or small bowel bacterial overgrowth.

Maximizing initial response and mitigating loss of response to anti-TNF mAb’s are therefore critically important. Clinical studies have revealed a number of approaches that improve outcomes. Higher rates of initial remission and response, and lower rates of loss of response are possible via a policy of high dose induction therapy, followed by scheduled (rather than episodic) maintenance therapy.

Importantly, IFX trough concentrations correlate with IFX effectiveness in CD. Lower IFX trough concentrations are associated with a shorter duration of response while on episodic therapy and with loss of response while on maintenance therapy. Higher anti-TNF mAb trough concentrations are observed with concomitant immunomodulator therapy among patients treated episodically, as well as in patients on scheduled maintenance therapy in the recent SONIC trial. IFX trough concentrations are lower in patients with detectable ATIs. Importantly, these data parallel those in rheumatoid arthritis (RA). In RA as in CD, low IFX concentrations correlate with treatment failure and ATIs are associated with lower IFX concentrations. It must be emphasized that the rates of clinical remission in CD are significantly lower in patients with undetectable trough IFX, whether or not ATIs are present . Therefore, trough IFX concentration is a primary determinant of efficacy/effectiveness independent of ATI formation. There are several corollaries to this observation: (1) in addition to ATIs, any processes that lead to suboptimal IFX concentrations, such as individual pharmacokinetic variability, will lead to lower IFX effectiveness ; (2) minimizing the development of ATIs should enhance efficacy; (3) theoretically, initial IFX response can be optimized and loss of IFX response can be restored by targeting trough concentrations above the threshold value. Only one commercial assay is available for measurement of IFX concentrations in the United States (Prometheus, San Diego, CA). Patients with IFX at or above 12 mcg/mL at 4 weeks after infusion, or a detectable IFX (>1.4 mcg/mL) at trough are considered to have therapeutic concentrations.

In ADA-treated patients with rheumatoid arthritis, low ADA concentrations correlate with treatment failure and AAA. Few data are available on the clinical relevance of ADA concentrations in CD. In an observational study, ADA primary nonresponders and patients who lost response had lower trough serum concentration compared with those who maintained response. The presence of AAA was associated with lower ADA concentrations. In a retrospective study, AAA was associated with lack of response. ADA concentrations were not measured in that study. An analysis of CLASSIC I found a dose-exposure-response relationship, but the overlap in serum ADA concentrations precluded the delineation of a predictive trough concentration. Serum ADA concentrations did not correlate with clinical remission in CLASSIC II. In CLASSIC I ⁵⁹ and in one observational study, concomitant AZA or 6MP therapy was not associated with any differences in ADA trough concentrations. Antibodies to CZP are associated with lower drug concentrations, but there are no data correlating CZP concentrations with efficacy.

Whether concomitant immunomodulator therapy enhances anti-TNF efficacy was controversial until recently. One trial and two observational studies found higher response rates in patients treated with combination immunomodulator-IFX therapy. However, post-hoc analyses of all but one of the pivotal IFX, ADA, and CZP trials did not demonstrate superior efficacy in subjects who also received an immunomodulator. Observational studies also found no benefit to concomitant immunomodulator therapy compared with IFX alone or ADA alone.

The SONIC trial has settled the question of combination AZA-IFX therapy in one group of patients, namely those who have active CD and are naïve to both AZA and IFX. The trial randomized 508 immunomodulator- and biologic-naïve patients with active CD (40% on prednisone or budesonide) to three arms: AZA (2.5 mg/kg/d) monotherapy, IFX monotherapy, or combination therapy for 26 weeks. The primary end point of clinical remission at week 26 was observed in 30.0%, 44.4%, and 56.8% of the AZA, IFX, and combination groups, respectively (AZA vs IFX, P = .006; IFX vs combination, P = .02; and AZA vs combination, P <.001). Mucosal healing rates at week 26 were 16.5%, 30.1%, and 43.9% (AZA vs IFX, P = .02; IFX vs combination, P = .06; and AZA vs combination, P <.001). Patients were followed to week 30 and given the option of continuing in the extended phase of the trial through week 50, which 280 patients did. Patients who did not enter the study extension were assumed to have had treatment failure at week 50. In this analysis of the entire 508-patient study population, steroid-free remission rates at 50 weeks were 46.2% in the combination group, 34.9% in the IFX group, and 24.1% in the AZA group (combination vs AZA, P <.001; IFX vs AZA, P = .03; combination vs IFX, P = .04). For the subgroup of 280 patients in the extension phase, 72.2% of those in the combination group, 60.8% of those in the IFX group, and 54.7% of those in the AZA group maintained steroid-free remission at week 50 (combination vs AZA, P = .010; IFX vs AZA, P = .32; combination vs IFX therapy, P = .07). The fraction of patients with infusion reactions and the fraction with antibodies to IFX at week 30 were significantly lower in the combination arm compared with IFX monotherapy (infusion reactions: 5.0% versus 16.6%; P <.001 and ATIs 0.9% vs 14.6%). Conversely, median trough IFX concentrations were significantly higher in the combination arm compared with the IFX arm both at week 30 (3.5 vs 1.6 μg/mL; P <.001) and at week 46 (3.8 vs 1.0 μg/mL; P <.001). Finally, infections and serious infections did not differ among the groups.

It must be noted that the SONIC trial likely underestimated the efficacy of AZA therapy for several reasons. Intermediate TPMT metabolizers, who represent 10% of the population and probably achieve higher response rates compared with normal TPMT metabolizers, were excluded. Moreover, more than 60% of subjects in the AZA arm were not receiving steroids; in other words, were not on any inductive therapy. Finally, evaluating mucosal healing at 26 weeks is probably too soon for a slow-acting drug, like AZA. Unfortunately, the SONIC investigators did not perform a multivariate analysis to identify the relative contribution of AZA cotherapy and higher IFX concentrations toward treatment success. This is an important question because the thiopurines probably enhance IFX efficacy by suppressing inflammation directly, as well as by augmenting IFX concentrations. Combination therapy was also found superior to IFX monotherapy in a recent trial in ulcerative colitis.

In summary, the following conclusions can be drawn regarding means of optimizing anti-TNF therapy: (1) With all three anti-TNF mAb’s, higher rates of short- and long-term response are possible via high dose induction therapy, followed by scheduled maintenance therapy. (2) Trough concentrations of IFX correlate with effectiveness. (3) AZA-IFX combination therapy is superior to IFX monotherapy in patients with active CD naïve to immunomodulators and biologics. (4) AZA-IFX combination therapy leads to higher IFX concentrations than IFX monotherapy in patients with active CD naïve to immunomodulators and biologics. It is not known whether concomitant immunomodulator therapy enhances the efficacy of ADA and CZP.

Managing Loss of Response to Anti-TNF mAb’s

As reviewed, low IFX trough concentrations are associated with loss of response. There are conflicting data on the value of ADA concentrations, and there are no data on CZP concentrations. In addition, there are no commercially available assays for the measurement of ADA and CZP. The following discussion thus focuses on managing patients with loss of response to IFX.

Loss of response to IFX should never be equated with bowel inflammation that has become refractory to IFX therapy. After intercurrent infection, particularly with Clostridium difficile , is considered, disease activity should be assessed with standard tests, such as colonoscopy, imaging studies, and serum and fecal inflammatory markers. If there is no active disease, or if the symptoms are out of proportion to the objectively assessed disease activity, then the clinician must consider other processes, such as a flare of irritable bowel syndrome, bile acid diarrhea, or the development of stricture(s) or small bowel bacterial overgrowth. Intuitively, individuals with active disease and low IFX concentrations, or undetectable circulating IFX and no detectable ATIs, would benefit from dose escalation. Conversely, individuals with IFX within the therapeutic window (IFX ≥12 mcg/mL at 4 weeks after infusion, or detectable trough IFX ) would benefit from non-IFX-based therapy. Although switching to ADA or CZP may be effective, non-anti-TNF therapy may be preferable in these patients who may have developed inflammation resistant to TNF suppression. The management of individuals with undetectable trough IFX and detectable ATIs may involve increasing the IFX dose to circumvent the effects of the ATIs, or switching to another therapy.

The clinical utility of IFX and ATI concentrations was addressed in a retrospective study of 155 patients. The management of individuals with detectable ATIs (increasing the IFX dose vs switching to another therapy) was also assessed. The main indications for testing were loss of response to IFX (49%), partial response after initiation of IFX (22%), and possible autoimmune/delayed hypersensitivity reaction (10%). ATIs were identified in 35 patients (23%) and therapeutic IFX concentrations in 51 (33%). Of 177 tests assessed, the results impacted treatment decisions in 73%. In ATI-positive patients, change to another anti-TNF mAb was associated with a complete or partial response in 92%, whereas dose escalation produced a response in only 17%. In patients with subtherapeutic IFX concentrations, dose escalation was associated with complete or partial clinical response in 86%, whereas changing to another anti-TNF mAb yielded a response in 33% of patients. The investigators concluded that (1) increasing the IFX dose in ATI-positive patients is ineffective, and (2) in patients with subtherapeutic IFX concentrations, dose escalation is a good alternative to changing to another anti-TNF mAb. The investigators did not address the predictive value of the actual ATI titer or the time-point of ATI detection. A high ATI titer at 4 weeks precludes continued IFX therapy, whereas a low ATI titer at 8 weeks may not obviate continued IFX therapy. In contrast, Baert and colleagues found that it was the titer , not the mere presence of ATIs that mattered. The median duration of response among patients with ATI concentrations less than 8.0 μg/mL was 71 days (95% CI, 57–88), as compared with 35 days (95% CI, 28–42) among those with ATI concentrations at or above 8.0 μg/mL ( P <.001). Duration of response was the same in patients with undetectable ATI (<1.7 μg/mL) and patients with ATI concentrations between 1.7 and 7.9 μg/mL. Further data are needed on patients with detectable ATI at trough. Table 3 summarizes the interpretation of IFX and ATI concentrations and suggested course of action in patients with active CD and loss of response to IFX.

Table 3

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

The Role of Capsule Endoscopy in Evaluating Inflammatory Bowel Disease Health Maintenance in the Inflammatory Bowel Disease Patient Disability in Inflammatory Bowel Disease Chromoendoscopy in Inflammatory Bowel Disease Evaluating Pouch Problems Imaging for Luminal Disease and Complications: CT Enterography, MR Enterography, Small-Bowel Follow-Through, and Ultrasound

Stay updated, free articles. Join our Telegram channel

Join