In this report, data from the RAPID 2 trial confirm that CZP provides broad relief of the multiple burdens imposed by RA. CZP was associated with rapid, clinically meaningful improvements in multiple PROs, observed as early as week 1, and maintained until study end at 6 months. These results mirror those observed in the RAPID 1 trial,6
and data reported for other TNF inhibitors in randomised controlled trials,1 18 27–31
although the rapid onset of action in all PROs observed with CZP (by week 1) has not been previously reported with the other agents.1
To further demonstrate the benefits of CZP on PROs, improvements ≥MCID in NNT and predictive analyses were used. Both methods can provide patients and physicians with meaningful information on treatment effect, such as how likely an individual patient is to benefit from treatment with CZP and whether the level of response within the first 12 weeks can determine longer term clinical outcomes.
To the authors' knowledge, this is the first study evaluating the benefit of an anti-TNF in RA by calculating the NNT using a variety of PROs. Previous work has investigated the NNT for other TNF inhibitors based on clinical responses such as ACR 20/50/70.32–39
One study of TNF inhibitors (etanercept, infliximab or adalimumab), analysing NNT according to HAQ-DI, reported an NNT of 1.94.37
In the present analysis, the NNT to achieve clinically meaningful improvements in up to five of six PROs was similarly low; only two to three additional patients needed to be treated to have at least one patient report simultaneous relief in five PROs, and the NNT for improvements in all six PROs was five to six patients. Further analyses indicated that it was more difficult for patients to achieve improvements ≥MCID in SF-36 PCS and more particularly MCS than in pain, PtGA, fatigue and physical function. This is not surprising as SF-36 measures additional aspects of the burden of RA and reporting improvements in complex concepts included in the mental component score implies a broader and holistic relief.
Further support for the use of all PRO measures including individual SF-36 domains to fully assess treatment benefit is provided by the correlation analyses. There were high correlations between pain, disease activity, fatigue and physical function; however, these improvements were less well correlated with those reported for HRQoL. As expected, correlations between mTSS and the PROs or other clinical parameters were low. Physical function scores (HAQ-DI) showed strong correlations with SF-36 PF and PCS scores as expected, but also BP domain and only moderate correlations with fatigue (FAS) scores and low correlations with SF-36 domains assessing MH. High correlations were also observed between PF and BP, and between VT and SF and MH. Interestingly, strong correlations were observed between certain ‘physical’ and ‘mental’ domains (RP and RE; BP and SF). Evaluating treatment-associated changes across all PROs is thus a better method for assessing change in multidimensional function. Careful evaluation of all SF-36 domain scores offers more information than the summary scores alone. For example, correlations between FAS and SF-36 MCS scores were lower than those between FAS and SF-36 VT domain scores. As VT is one of the domains scored positively when calculating MCS scores, these results indicate that the other three ‘mental’ domains (SF, and particularly MH and RE) provide the greatest contribution to changes in SF-36 MCS scores. FAS had high/marked correlations with pain VAS and PtGA followed by VT, which could indicate that FAS and SF-36 VT scores provide complementary assessments for measuring fatigue/tiredness; FAS may be more appropriate to evaluate physical aspects of fatigue, whereas SF-36 VT more fully assesses mental aspects such as pep, motivation and energy. Therefore, although SF-36 PCS and MCS scores provide a good summary metric of HRQoL, they do not fully reflect the overall magnitude of changes within individual domains.17 19
Physicians should consider data from all SF-36 domains as well as PCS and MCS scores in their decision-making process. To facilitate analysis and interpretation of treatment-related effects across the individual domains, ‘spydergrams’ were used to depict SF-36 results, which allowed a more global comparison of disease-related decrements in HRQoL versus normative data.19
A limitation of the NNT analysis is that patients with active disease in the placebo plus MTX group were used as the comparator and, therefore, the analysis only relates to this comparison. Nevertheless, the low NNT confirms the efficacy of CZP regarding reported improvements in PROs, indicating that relatively few patients need to be treated with CZP to achieve relief from the burdens of RA.
More patients with responses by week 6 had LDA at week 24 than those with responses at week 12. Similarly, those patients achieving pain MCID by week 12 were more likely to have better outcomes than those patients who did not achieve pain MCID at week 12, including LDA, remission and acceptable HAQ-DI scores. Taken together these results indicate that early improvement in PROs predict both clinical and PROs at 6 months. They complement previous analyses from the RAPID 1 trial demonstrating that patients who reported more rapid improvements in disease activity had higher chances of achieving better clinical outcomes at 1 year.40
Overall, these results suggest that patients treated with CZP who achieve early responses according to either PROs or LDA have a greater chance of avoiding long-term disability. Because of the short 6-month study duration of RAPID 2, longer term data (1–2 years) are needed to confirm the use of time to responses by PROs as a predictor of better clinical outcomes.
In conclusion, CZP 200 or 400 mg plus MTX administered every 2 weeks provides broad relief from the multiple burdens of RA. Significant and clinically meaningful improvements, low NNTs and increased likelihood of achieving longer term outcomes with an earlier PRO response indicate that CZP is an effective treatment option, offering substantial benefits to patients.