Treatment of RA With Etanercept With Reference to DMARDs
Treatment of RA With Etanercept With Reference to DMARDs
Overall, data were available from 8227 patients: 4453 treated with ETN and 3774 treated with conventional DMARDs. Of these, 3529 (79.2%) ETN-treated patients and 2864 (75.9%) conventional DMARD reference patients met the study inclusion criteria (Figure 1). The ETN cohort had longer duration of follow-up (mean 4.8 years; median 5.4 years), with a total follow-up of 16,919 person-years, compared with the conventional DMARD cohort (mean 3.9 years; median 3.9 years) with a total follow-up of 11,095 person-years ( Table 1 ).
As expected, there were notable differences at baseline between the ETN cohort and the conventional DMARD cohort ( Table 1 ). For instance, at baseline the ETN-treated patients were younger (55.3 vs 59.8 years, P < 0.001), with a longer RA duration compared with the conventional DMARD cohort (13.5 vs 9.6 years, P < 0.001). ETN patients had greater disease activity (DAS28 6.6 vs 5.6, P < 0.001) and greater disability (HAQ 2.1 vs 1.6, P < 0.001), but they had less non-RA-related general morbidity as measured by the Charlson index (0.4 vs 0.5, P < 0.001) and other morbidity measures. ETN patients were less likely to smoke (current smokers 21.0% vs 23.8%, P = 0.004).
In the base-case analysis there were 203 deaths in the ETN cohort and 223 in the conventional DMARD cohort ( Table 2 ). There was a lower crude mortality rate in the ETN cohort than in the conventional DMARD cohort (12.0 vs 20.1 events per 1000 person-years, respectively; Table 2 and Figure 2). The difference was significant in the base-case adjusted analysis (aHR = 0.72, 95% CI 0.54, 0.96, P = 0.024; Table 3 ), and remained significant in the sensitivity analysis when restricting follow-up in the ETN cohort to discontinuation plus 90 days (aHR = 0.44, 95% CI 0.31, 0.60, P < 0.001; Table 3 ).
(Enlarge Image)
Figure 2.
Kaplan–Meier survival curves of mortality for ETN and conventional DMARD patients.
There were 538 ETN patients and 375 conventional DMARD patients who experienced at least one serious infection, the most common being pneumonia ( Table 2 ). Although the crude rate was higher in the ETN cohort, this difference was not significant when adjusted for relevant baseline differences (aHR = 1.02, 95% CI 0.83, 1.25, P = 0.86; Table 3 ). When analysis was restricted to ETN exposure plus 90 days follow-up, the corresponding difference between the cohorts remained non-significant (P = 0.49; Table 3 ). When analysed by progressive year of follow-up, the adjusted risk of serious infection was elevated for ETN in the first 2 years of therapy but not in years 3–5: aHR = 1.56 (95% CI 1.16, 2.09) in year 1; aHR = 1.32 (95% CI 1.06, 1.65) in year 2; see Supplementary Table S1, available at Rheumatology Online. Tuberculosis was reported in five ETN-treated patients and in one conventional DMARD patient.
There were 241 first post-index cancers in the ETN cohort and 254 in the conventional DMARD cohort ( Table 2 ). A lower crude rate of cancer was observed in the ETN cohort (14.7 vs 23.9 events per 1000 person-years; Table 2 ), but this was not significant in the analysis following adjustment for age, baseline steroid use, smoking history, previous cancer and BMI (aHR = 0.84, 95% CI 0.68, 1.03, P = 0.08; Table 3 ). In the analysis stratified for patients with and without a history of prior cancer, the aHR for patients with prior cancer was 0.877 (95% CI 0.720, 1.069, P = 0.194); for those without prior history of cancer the aHR was 0.804 (95% CI 0.460, 1.403, P = 0.442). Excluding cases with non-melanoma skin cancer, the aHR was 0.792 (95% CI 0.636, 0.986, P = 0.037). For non-melanoma skin cancer, the aHR was 1.016 (95% CI 0.717, 1.441, P = 0.928).
A total of 47 lymphoproliferative malignancies were reported: 18 in the ETN cohort and 29 in the conventional DMARD cohort ( Table 2 ). This resulted in a lower crude rate of lymphoproliferative malignancy in the ETN cohort (1.1 vs 2.6 per 1000 person-years; Table 2 ) and a significantly lower adjusted rate compared with the conventional DMARD cohort (aHR = 0.51, 95% CI 0.28, 0.95, P = 0.04; Table 3 ).
Overall, other serious adverse events were significantly less frequent in the ETN cohort than in the conventional DMARD cohort (aHR = 0.70, 95% CI 0.56, 0.87, P = 0.001; Table 3 ). However, this apparent difference did not achieve significance at the conventional level in a sensitivity analysis restricted to ETN exposure plus 90 days (aHR = 0.85, 95% CI 0.66, 1.09, P = 0.195; Table 3 ).
There was a much lower overall rate of serious cardiac events among ETN-treated patients (10.0 per 1000 person-years) in comparison with the conventional DMARD cohort (17.9 per 1000 person-years; Table 2 ). When adjusted for age, gender, baseline non-RA drugs, baseline steroid use, number of baseline conventional DMARDs, baseline HAQ and prior angina, the overall rate of serious cardiac events was significantly lower for the ETN cohort in both analytical scenarios (base case: aHR = 0.52, 95% CI 0.37, 0.72, P < 0.001; sensitivity: aHR = 0.56, 95% CI 0.40, 0.79, P = 0.001; Table 3 ).
A sensitivity analysis was conducted comparing ETN patients included in the analysis with those excluded due to a minimum of 90 days' difference between treatment initiation and patient registration. There was no statistically significant difference for any endpoint between the two groups of ETN patients. However, when the ETN patients were combined and compared with the conventional DMARDs, this sensitivity analysis did result in a change in the aHR for lymphoma, which was no longer significantly reduced in ETN vs conventional DMARDs (aHR = 0.649, 95% CI 0.375, 1.124, P = 0.123) ( Table 4 ).
Results
Overall, data were available from 8227 patients: 4453 treated with ETN and 3774 treated with conventional DMARDs. Of these, 3529 (79.2%) ETN-treated patients and 2864 (75.9%) conventional DMARD reference patients met the study inclusion criteria (Figure 1). The ETN cohort had longer duration of follow-up (mean 4.8 years; median 5.4 years), with a total follow-up of 16,919 person-years, compared with the conventional DMARD cohort (mean 3.9 years; median 3.9 years) with a total follow-up of 11,095 person-years ( Table 1 ).
Baseline Characteristics
As expected, there were notable differences at baseline between the ETN cohort and the conventional DMARD cohort ( Table 1 ). For instance, at baseline the ETN-treated patients were younger (55.3 vs 59.8 years, P < 0.001), with a longer RA duration compared with the conventional DMARD cohort (13.5 vs 9.6 years, P < 0.001). ETN patients had greater disease activity (DAS28 6.6 vs 5.6, P < 0.001) and greater disability (HAQ 2.1 vs 1.6, P < 0.001), but they had less non-RA-related general morbidity as measured by the Charlson index (0.4 vs 0.5, P < 0.001) and other morbidity measures. ETN patients were less likely to smoke (current smokers 21.0% vs 23.8%, P = 0.004).
All-Cause Mortality
In the base-case analysis there were 203 deaths in the ETN cohort and 223 in the conventional DMARD cohort ( Table 2 ). There was a lower crude mortality rate in the ETN cohort than in the conventional DMARD cohort (12.0 vs 20.1 events per 1000 person-years, respectively; Table 2 and Figure 2). The difference was significant in the base-case adjusted analysis (aHR = 0.72, 95% CI 0.54, 0.96, P = 0.024; Table 3 ), and remained significant in the sensitivity analysis when restricting follow-up in the ETN cohort to discontinuation plus 90 days (aHR = 0.44, 95% CI 0.31, 0.60, P < 0.001; Table 3 ).
(Enlarge Image)
Figure 2.
Kaplan–Meier survival curves of mortality for ETN and conventional DMARD patients.
Serious Infections
There were 538 ETN patients and 375 conventional DMARD patients who experienced at least one serious infection, the most common being pneumonia ( Table 2 ). Although the crude rate was higher in the ETN cohort, this difference was not significant when adjusted for relevant baseline differences (aHR = 1.02, 95% CI 0.83, 1.25, P = 0.86; Table 3 ). When analysis was restricted to ETN exposure plus 90 days follow-up, the corresponding difference between the cohorts remained non-significant (P = 0.49; Table 3 ). When analysed by progressive year of follow-up, the adjusted risk of serious infection was elevated for ETN in the first 2 years of therapy but not in years 3–5: aHR = 1.56 (95% CI 1.16, 2.09) in year 1; aHR = 1.32 (95% CI 1.06, 1.65) in year 2; see Supplementary Table S1, available at Rheumatology Online. Tuberculosis was reported in five ETN-treated patients and in one conventional DMARD patient.
Cancer
There were 241 first post-index cancers in the ETN cohort and 254 in the conventional DMARD cohort ( Table 2 ). A lower crude rate of cancer was observed in the ETN cohort (14.7 vs 23.9 events per 1000 person-years; Table 2 ), but this was not significant in the analysis following adjustment for age, baseline steroid use, smoking history, previous cancer and BMI (aHR = 0.84, 95% CI 0.68, 1.03, P = 0.08; Table 3 ). In the analysis stratified for patients with and without a history of prior cancer, the aHR for patients with prior cancer was 0.877 (95% CI 0.720, 1.069, P = 0.194); for those without prior history of cancer the aHR was 0.804 (95% CI 0.460, 1.403, P = 0.442). Excluding cases with non-melanoma skin cancer, the aHR was 0.792 (95% CI 0.636, 0.986, P = 0.037). For non-melanoma skin cancer, the aHR was 1.016 (95% CI 0.717, 1.441, P = 0.928).
Lymphoproliferative Malignancy
A total of 47 lymphoproliferative malignancies were reported: 18 in the ETN cohort and 29 in the conventional DMARD cohort ( Table 2 ). This resulted in a lower crude rate of lymphoproliferative malignancy in the ETN cohort (1.1 vs 2.6 per 1000 person-years; Table 2 ) and a significantly lower adjusted rate compared with the conventional DMARD cohort (aHR = 0.51, 95% CI 0.28, 0.95, P = 0.04; Table 3 ).
Other Serious Adverse Events
Overall, other serious adverse events were significantly less frequent in the ETN cohort than in the conventional DMARD cohort (aHR = 0.70, 95% CI 0.56, 0.87, P = 0.001; Table 3 ). However, this apparent difference did not achieve significance at the conventional level in a sensitivity analysis restricted to ETN exposure plus 90 days (aHR = 0.85, 95% CI 0.66, 1.09, P = 0.195; Table 3 ).
There was a much lower overall rate of serious cardiac events among ETN-treated patients (10.0 per 1000 person-years) in comparison with the conventional DMARD cohort (17.9 per 1000 person-years; Table 2 ). When adjusted for age, gender, baseline non-RA drugs, baseline steroid use, number of baseline conventional DMARDs, baseline HAQ and prior angina, the overall rate of serious cardiac events was significantly lower for the ETN cohort in both analytical scenarios (base case: aHR = 0.52, 95% CI 0.37, 0.72, P < 0.001; sensitivity: aHR = 0.56, 95% CI 0.40, 0.79, P = 0.001; Table 3 ).
Sensitivity Analysis for ETN Patients Excluded From the Main Analysis
A sensitivity analysis was conducted comparing ETN patients included in the analysis with those excluded due to a minimum of 90 days' difference between treatment initiation and patient registration. There was no statistically significant difference for any endpoint between the two groups of ETN patients. However, when the ETN patients were combined and compared with the conventional DMARDs, this sensitivity analysis did result in a change in the aHR for lymphoma, which was no longer significantly reduced in ETN vs conventional DMARDs (aHR = 0.649, 95% CI 0.375, 1.124, P = 0.123) ( Table 4 ).
Source...