Inflammation plays a major role in endothelial dysfunction and the development of atherothrombosis. There are several novel targets available to inhibit these inflammatory changes at the vascular level such as tumor necrosis factor-α (TNF-α) blocker (etanercept, rilonacept, infliximab, adalimumab), IL-6 receptor antagonist (tocilizumab), IL-1a receptor antagonist (anakinra) and IL-1β (canakinumab); and steroidogenic acute regulatory protein, so far studied only in rats. Targeting of the IL-1 pathways in disease treatment raises possible safety concerns due to inhibition of innate immunity. This large pooled analysis included more than 1000 patients with type 2 diabetes mellitus (mean duration: ~5 years) treated for up to 1.4 years either with canakinumab, an IL-1β blocker or placebo. It was of particular interest to assess the combined safety data from individual studies to help identify any significant findings, including dose response, from the pooled analysis.
Overall, there were no clinically meaningful differences in AEs and SAEs between canakinumab-treated patients and placebo, with no clear dose-dependent trends in AEs. The observed safety findings and overall tolerability in patients with T2DM are consistent with those in previously reported studies in patients with cryopyrin-associated periodic syndrome, gouty arthritis and systemic juvenile idiopathic arthritis[21–23].
The mode of action of canakinumab is to inhibit innate immunity while binding to IL-1β, therefore, an increased rate of infections was expected in each canakinumab dose group versus placebo. The most commonly reported AEs in this pooled analyses were nasopharyngitis, urinary tract infections and upper respiratory tract infections. The observed increase in rate of infections with canakinumab was, however, not statistically significant. This numerical imbalance is, nevertheless, most likely a true signal due to its biological plausibility. Importantly, the clinical experience with canakinumab differs from that of TNF-α blockers which are associated with an increased risk of tuberculosis reactivation. Studies reveal that tuberculosis risk varies with the use of different anti-TNF-α agents (etanercept, 35 cases per 100,000 patients; infliximab, 144 per 100,000 patients; and adalimumab, 240 per 100,000 patients). Wallis et al. showed that the probability of tuberculosis developing with infliximab treatment is threefold higher than with etanercept. Similar to anakinra, canakinumab has not been associated with the risk of tuberculosis reactivation, indicating that blocking the IL-1 pathway is safer in comparison with blocking TNF-α.
Of the seven patients who experienced serious infections, five were observed on canakinumab and two on placebo. Four of the five canakinumab-treated subjects had a complete recovery and in one case outcome was not reported after the patient was taken off the study drug. No relationship between the events and study drug was reported by the investigators for the patients who reported an outcome. Moreover, there were no significant changes in the WBC counts in these seven patients, with the only value falling below the normal range at month 4 (3.12×109/L; baseline 7.13×109/L) in a patient taking placebo which returned to normal range at month 6 (5.38×109/L). Although none of the patients had WBC values above the upper limit of normal range throughout the study, two patients reported leucocytosis associated with their acute infection. There were no significant changes in hsCRP levels in these patients.
AEs of injection-site reactions, malignancies, neutropenia and changes in renal function were infrequent with no apparent dose-dependent trend. Although IL-1β inhibition increases expression of the pancreatic β-cells in T2DM patients, the present study reported only two hypoglycaemic events with medium-dose canakinumab (versus one on placebo), which were mild-to-moderate in severity and none was suspected by the investigators to be related to the study drug. None of the patients had reported any asymptomatic low blood glucose (<40 mg/dL) event or discontinued their study treatment permanently because of any hypoglycaemic event, thereby suggesting that canakinumab did not result in increased risk of hypoglycaemic events in T2DM patients. Although a lower rate of major adverse cardiovascular events was observed in the canakinumab groups compared with the placebo group, the low number of events prevents drawing conclusions on the significance of this observation.
The decreases in platelet counts and absolute neutrophil counts and the resulting decreased total WBC counts are well-known effects of canakinumab treatment[21–23] due to the role of IL-1β in bone marrow maturation and release of these blood cells. Most laboratory abnormalities observed in this pooled analysis were of CTC grades 1 or 2, and there was no association of any such events either with infections or bleeding/bruising.
The incidence rate of elevations of hepatic enzymes was low and was driven by a small number of patients with a known high prevalence of hepatic steatosis. There were no cases consistent with Hy’s law. Vital signs of the patients treated with canakinumab showed no clinically meaningful difference versus placebo and had no particular dose-related trend. Treatment with canakinumab did not seem to have a clinically meaningful effect on lipid profiles in patients with T2DM or IGT. The increase in TC levels appeared to be due to an increase in HDL-C, because no clear pattern was observed for LDL-C. The clinical significance of the increase in TG levels could not be determined in the absence of increase in LDL-C levels.
Treatment-emergent AEs leading to permanent discontinuation of study treatment were rare and evenly distributed in all treatment groups, demonstrating the good tolerability and safety of canakinumab. No deaths were reported in any of the three studies. The subgroup analyses by age (<65 and ≥65 years) and gender showed no remarkable observations related to the safety findings of canakinumab when compared with overall population. Safety findings from this pooled data set were consistent with the previously reported studies in patients with cryopyrin-associated periodic syndrome, gouty arthritis and systemic juvenile idiopathic arthritis[21–23] with no new safety findings. In addition, a recent study in newly diagnosed type 1 diabetes patients (mostly less than 18 years of age) treated with IL-1Ra (anakinra) and IL-1β (canakinumab) inhibitors revealed no major safety concerns although they failed to prevent deterioration of endogenous pancreatic insulin (C-peptide) secretion in these patients. However, in a preclinical study, the combination of IL-1β blocker with other anti-CD3 monoclonal antibody resulted in significantly greater clinical remission of diabetes suggesting that this combination might be more suitable in new onset type 1 diabetes or in prevention trials in individuals with pre-type 1 diabetes.
By pooling three T2DM studies we obtained a pooled safety database of 1026 patients, which allowed a better characterisation of the safety profile of canakinumab than was possible in each individual trial[24–27]. Nevertheless, this sample size was still too small to conclusively evaluate all potential observed safety issues, which may have included spurious chance findings. It also does not allow a proper evaluation of extremely rare events (such as malignancies, neutropenia or renal function). Another key limitation of this pooled safety analysis is that the duration of treatment and follow-up was on average 6 months and only exceeded 1 year in a small subset of patients. This precludes any definite conclusions regarding the long-term safety of canakinumab in patients with T2DM or IGT. However, CANTOS, an ongoing large outcomes study, is evaluating three doses of canakinumab (50, 150 and 300 mg versus placebo) given subcutaneously on a quarterly basis for up to 5 years in order to assess the effects on cardiovascular risk reduction in post-MI patients and will provide long term safety data on canakinumab.
In conclusion, this pooled safety analysis over a treatment period of 1.4 years indicates that canakinumab across the evaluated doses in T2DM or IGT patients is safe and well tolerated, adding important information to the overall safety profile of the drug.