Serious adverse events

Over the course of the trial, at least one SAE – defined as an event that is life-threatening, causes hospitalisation (initial or prolonged), disability or permanent damage, a congenital anomaly or birth defect, or death – was reported for 171 participants and, of these, 30 participants died. The frequency of the primary safety end point, an ATE or HF or death from a vascular cause, did not differ significantly between the drugs (OR 1.69, 95% CI 0.80 to 3.57; p = 0.16) and between the treatment regimens drugs (OR 0.56; 95% CI 0.27 to 1.19; p = 0.13) (Table 16 and Figure 28). Of the 30 deaths, 15 participants were the ranibizumab group (4.8%) and 15 in the bevacizumab group (5.1%). By treatment regimen, there were more deaths in the discontinuous treatment group [20 (6.6%) vs. 10 (3.2%), OR 0.47, 95% CI 0.22 to 1.03; p = 0.05] (see Figure 28). Causes of death, determined from the participant death certificates are summarised in Table 17.

TABLE 16

Vascular events, HF and death

FIGURE 28

Safety outcomes at 2 years (a) by drug and (b) by treatment regimen. ORs of < 1 reflect fewer SAEs during the 2 years of follow-up in the ranibizumab or continuous groups. Circles indicate ORs and bars indicate 95% CIs. Vascular events (more...)

TABLE 17

Causes of death by drug and regimen combination

Serious adverse events grouped by MedDRA system organ class are shown in Table 18. The most commonly reported type of SAE was cardiac disorders. Tests of the frequency of SAEs for different organ systems (when there were more than 10 participant-specific events) showed a statistically significant difference in SAEs coded as general disorders and administration site conditions (which includes all deaths) by treatment regimen (continuous 3.2%; discontinuous 7.0%; p = 0.03) but not for any other organ system by drug or treatment regimen (Figure 29). SAEs coded as gastrointestinal disorders were more frequent with bevacizumab but the difference was less marked than at 1 year.²⁶ Serious ocular AEs in the study eye were extremely rare (14 participants; see Table 18), as were events in the non-study eye (five participants; Table 19).

TABLE 18

Serious adverse events by system organ class

FIGURE 29

Safety outcomes classified by MedDRA system organ class at 2 years (a) by drug and (b) by treatment regimen. ORs of < 1 reflect fewer SAEs during the 2 years of follow-up in the ranibizumab or continuous groups. Circles indicate ORs and (more...)

TABLE 19

Ocular SAEs in the non-study eye

The percentages of patients having any systemic SAE were very similar by drug (ranibizumab 26%; bevacizumab 27%; OR 0.96; 95% CI 0.66 to 1.39; p = 0.82) and by treatment regimen (continuous 24%; discontinuous 29%; OR 0.77; 95% CI 0.53 to 1.11; p = 0.16). In total, 39 non-ocular SAEs were classified as possibly, probably or definitely related to treatment (Table 20).

TABLE 20

Serious adverse events classified as possibly, probably or definitely related to treatment

Non-serious adverse events

Over the course of the trial, non-serious AEs were reported for 570 (93.4%) of the 610 participants. Almost 83% of participants had one or more non-ocular event and 69% had at least one ocular AE. Frequencies were similar by drug and treatment regimen (Table 21). Details of the events, by MedDRA system organ class, are given in Table 21. Ocular AEs in the non-study eye are given in Table 22.

TABLE 21

Adverse events by system organ class

TABLE 22

Ocular AEs in the non-study eye

Meta-analyses

Figure 30 shows meta-analyses of safety outcomes from the CATT, GEFAL, IVAN, LUCAS (presented at the 2013 meeting of the American Association for Ophthalmology), MANTA and Subramanian et al. trials comparing ranibizumab and bevacizumab^22,42–44 (available information for the BRAMD trial reported only by MedDRA System Organ Class and did not distinguish deaths from other SAEs). The CATT²² and IVAN trials followed participants for 2 years, and the other four trials reported data to 1 year after recruitment. Pooled estimates of safety outcomes showed no differences by drug for deaths (heterogeneity χ² = 0.76; df = 5; p = 0.98; I² = 0%) or ATEs (heterogeneity χ² = 3.21; df = 3; p = 0.36; I² = 7%) but a significantly increased risk of any systemic SAE for bevacizumab (OR 0.77, 95% CI 0.64 to 0.92; p = 0.004; heterogeneity χ² = 1.85; df = 3; p = 0.61; I² = 0%) (see Figure 30). The comparison by treatment regimen (IVAN 2-year data and CATT 1-year data only²²) showed consistent increases in mortality (OR 0.49, 95% CI 0.27 to 0.86; p = 0.014) and the risk of any systemic SAE (OR 0.81, 95% CI 0.65 to 1.01; p = 0.063) with discontinuous treatment (see Figure 30).

FIGURE 30

Meta-analysis of safety outcomes (a) by drug and (b) by treatment regimen. Ratios of < 1 reflect better outcomes in the ranibizumab or continuous groups. 95% CIs are given in parentheses and illustrated by horizontal bars. Source: © (more...)

A post hoc meta-analysis of gastrointestinal SAEs, including the studies listed above, with the addition of BRAMD for the drug comparison, was performed. This meta-analysis was prompted by a report from the CATT²² that such events occurred more often when participants were treated with bevacizumab than when treated with ranibizumab; the IVAN interim analysis was also consistent with this observation.²⁶ The pooled estimate for the drug comparison showed a significantly increased risk of gastrointestinal SAEs in the bevacizumab group (OR 0.53, 95% CI 0.33 to 0.85; p = 0.009; heterogeneity χ² = 5.13, df = 5; p = 0.40; I² = 3%) but no difference by treatment regimen (OR 0.88, 95% CI 0.44 to 1.78; p = 0.73) (Figure 31).

FIGURE 31

Meta-analysis of gastrointestinal SAEs (a) by drug and (b) by treatment regimen. Ratios of < 1 reflect better outcomes in the ranibizumab or continuous groups. 95% CIs are given in parentheses and illustrated by horizontal bars.

Summary

At least one SAE was reported for 171 participants, and 30 participants died. The frequency of the primary safety end point did not differ by drug or treatment regimen. There were 15 deaths in each of the ranibizumab and bevacizumab groups, but twice as many deaths with discontinuous as continuous treatment regimens (20 deaths compared with 10). We examined the reasons for death by both drug and regimen and we did not observe any clustering of cause for mortality by drug or by regimen. There was also a statistically significant difference in the frequency of SAEs coded as general disorders and administration site conditions (which includes all deaths), with about twice as many of these SAEs observed to occur in the discontinuous treatment group. SAEs coded as ‘gastrointestinal’ were more frequent with bevacizumab in the IVAN trial. Serious ocular AEs were extremely rare. The number of participants having any systemic SAE was similar across the groups. Over 90% of participants reported at least one non-SAE.

Meta-analyses of safety outcomes showed a significantly increased risk of any systemic SAEs and gastrointestinal events for bevacizumab, and consistent increases in mortality and the risk of any systemic SAE with discontinuous treatment across trials.