Avoid Abatacept in RA patients with a history of cancer? Save

The management of rheumatoid arthritis patients with a history of cancer continues to be an area of concern for rheumatologists. General recommendations are to treat patients with a history of solid organ cancer as no different to any other RA patient. The evidence is especially strong for this when the cancer was over five years ago. However, despite this, concern and uncertainty remain for physician and patient alike. 

More recently the much-discussed ORAL-Surveillance study threw another spanner into the works demonstrating evidence of a greater risk of cancer with the JAK inhibitor tofacitinib than with TNF inhibitors. It is TNF inhibitors themselves that often cause the biggest concern, mainly due to the name – you can’t inhibit the “tumour” necrosis factor, that couldn’t be good for cancer!

It is in this setting that two posters at Saturday’s poster session both reassure and raise a new cause for concern.

The first of these was a study from the BIOBADASER Spanish biologic registry presented by Molina-Collada et al, Abstract#0267. 352 patients of 9,129 in BIOBADASER had a history of malignancy. Rates of incident cancer varied from 0 (for IL-17i) to 51.7 (for abatacept) per 1000 patient years. The rate ratio for mortality compared to TNFi was 0.3 (0.0-2.0) for JAKi, 1.1 (0.4-2.5) for rituximab, 0.7 (0.2-2.4) for tocilizumab, 0.8 (0.2-3.3) for IL-17i, and 1.8 (0.8-4.3) for abatacept. Overall there was no significant increase risk in new cancer in those with a history of malignancy for other agents compared to TNFi (which we have good evidence for safety). JAKi looked good in this analysis. While there was no significant difference however, the abatacept point estimate of 1.8 (and associated confidence intervals) concerns me a bit. Targeting CTLA-4, abatacept is essentially a reverse ipilimumab – a potent anti-cancer agent. So, there is a hypothetical pathogenic basis to go with this potential clinical concern.

In the same poster session, Kunishita et al presented the results of a Japanese study, Abstract#0255. This was a single centre retrospective chart review including 276 RA patients treated with abatacept. 61 had prior malignancy and 215 did not. Abatacept was equally efficacious in both groups. The 5-year malignancy rates were 16.5% in those with prior malignancy and 5.8% in those without. The p-value was 0.18, so nominally not significant, and the authors concluded there was no difference. However, the raw numbers have a 10% absolute risk difference and the curves shown in the poster are diverging further over time, both causes for concern.

Overall, while not definitive, for me these two clinical studies raise concerns over incident cancer risk in RA patients with a prior history of cancer treated with abatacept. Given the theoretical supportive rationale for this, and the availability of other equally efficacious pharmacologic agents, I think this argues for caution in the use of abatacept in patients with a prior history of malignancy unless all other agents are exhausted.