Everolimus for the treatment of advanced, non-functional neuroendocrine tumours of the lung or gastrointestinal tract (RADIANT-4): a randomised, placebo-controlled, phase 3 study

Abstract

Background: Effective systemic therapies for patients with advanced, progressive neuroendocrine tumours of the lung or gastrointestinal tract are scarce. We aimed to assess the efficacy and safety of everolimus compared with placebo in this patient population.

Methods: In the randomised, double-blind, placebo-controlled, phase 3 RADIANT-4 trial, adult patients (aged ≥18 years) with advanced, progressive, well-differentiated, non-functional neuroendocrine tumours of lung or gastrointestinal origin were enrolled from 97 centres in 25 countries worldwide. Eligible patients were randomly assigned in a 2:1 ratio by an interactive voice response system to receive everolimus 10 mg per day orally or identical placebo, both with supportive care. Patients were stratified by tumour origin, performance status, and previous somatostatin analogue treatment. Patients, investigators, and the study sponsor were masked to treatment assignment. The primary endpoint was progression-free survival assessed by central radiology review, analysed by intention to treat. Overall survival was a key secondary endpoint. This trial is registered with ClinicalTrials.gov, number NCT01524783.

Findings: Between April 3, 2012, and Aug 23, 2013, a total of 302 patients were enrolled, of whom 205 were allocated to everolimus 10 mg per day and 97 to placebo. Median progression-free survival was 11·0 months (95% CI 9·2-13·3) in the everolimus group and 3·9 months (3·6-7·4) in the placebo group. Everolimus was associated with a 52% reduction in the estimated risk of progression or death (hazard ratio [HR] 0·48 [95% CI 0·35-0·67], p<0·00001). Although not statistically significant, the results of the first pre-planned interim overall survival analysis indicated that everolimus might be associated with a reduction in the risk of death (HR 0·64 [95% CI 0·40-1·05], one-sided p=0·037, whereas the boundary for statistical significance was 0·0002). Grade 3 or 4 drug-related adverse events were infrequent and included stomatitis (in 18 [9%] of 202 patients in the everolimus group vs 0 of 98 in the placebo group), diarrhoea (15 [7%] vs 2 [2%]), infections (14 [7%] vs 0), anaemia (8 [4%] vs 1 [1%]), fatigue (7 [3%] vs 1 [1%]), and hyperglycaemia (7 [3%] vs 0).

Interpretation: Treatment with everolimus was associated with significant improvement in progression-free survival in patients with progressive lung or gastrointestinal neuroendocrine tumours. The safety findings were consistent with the known side-effect profile of everolimus. Everolimus is the first targeted agent to show robust anti-tumour activity with acceptable tolerability across a broad range of neuroendocrine tumours, including those arising from the pancreas, lung, and gastrointestinal tract.

Funding: Novartis Pharmaceuticals Corporation.

Figure 1. Trial profile

^*The full analysis set contained all 302 randomized patients (205 patients in the everolimus arm and 97 in the placebo arm). Two patients randomized to everolimus were not treated and one patient randomized to everolimus received only placebo treatment, therefore the safety set contains 202 patients in the everolimus arm and 98 patients in the placebo arm. ^†At the time of data cut-off (November 28, 2014).

Figure 2. Progression-free and overall survival

Kaplan–Meier curves are shown for progression-free survival as assessed by central radiology review (Panel A) and by local investigators (Panel B). Forest plots show the effect of study treatment on progression-free survival in predefined patient subgroups based on stratification factors (Panel C), major demographic and prognostic subgroups (Panel D) and a retrospective analysis in subgroups by primary tumour origin (Panel E). Subgroup results are reported based on central review. In the retrospective post hoc analysis, stomach, colon, rectum, appendix, caecum, ileum, duodenum, and jejunum are grouped under GI. The HRs in all subgroups are obtained from unstratified Cox model. Kaplan–Meier curves are also shown for overall survival (Panel F). ^*Based on prognostic level, patients were divided into two strata. Stratum A (better prognosis) consisted of patients with tumour sites originating from appendix, cecum, jejunum, ileum, duodenum, and NET of unknown primary and stratum B (worse prognosis) comprised patients with primary tumours of lung, stomach, rectum, and colon (except the cecum). ^†Included Black. ^‡Defined as no prior chemotherapy or no SSA therapy continuously for ≥12 weeks any time before study. ^**One patient with thymus as primary tumour origin was not included. ^¶p-value boundary for significance at this interim analysis was 0.0002. CI=confidence interval. CgA=chromogranin A. GI=gastrointestinal. NA=not available. NET=neuroendocrine tumours. NSE=neuron-specific enolase. SSA=somatostatin analogue. ULN=upper limit of normal. WHO=World Health Organization.

Figure 2. Progression-free and overall survival

Kaplan–Meier curves are shown for progression-free survival as assessed by central radiology review (Panel A) and by local investigators (Panel B). Forest plots show the effect of study treatment on progression-free survival in predefined patient subgroups based on stratification factors (Panel C), major demographic and prognostic subgroups (Panel D) and a retrospective analysis in subgroups by primary tumour origin (Panel E). Subgroup results are reported based on central review. In the retrospective post hoc analysis, stomach, colon, rectum, appendix, caecum, ileum, duodenum, and jejunum are grouped under GI. The HRs in all subgroups are obtained from unstratified Cox model. Kaplan–Meier curves are also shown for overall survival (Panel F). ^*Based on prognostic level, patients were divided into two strata. Stratum A (better prognosis) consisted of patients with tumour sites originating from appendix, cecum, jejunum, ileum, duodenum, and NET of unknown primary and stratum B (worse prognosis) comprised patients with primary tumours of lung, stomach, rectum, and colon (except the cecum). ^†Included Black. ^‡Defined as no prior chemotherapy or no SSA therapy continuously for ≥12 weeks any time before study. ^**One patient with thymus as primary tumour origin was not included. ^¶p-value boundary for significance at this interim analysis was 0.0002. CI=confidence interval. CgA=chromogranin A. GI=gastrointestinal. NA=not available. NET=neuroendocrine tumours. NSE=neuron-specific enolase. SSA=somatostatin analogue. ULN=upper limit of normal. WHO=World Health Organization.

Figure 2. Progression-free and overall survival

Kaplan–Meier curves are shown for progression-free survival as assessed by central radiology review (Panel A) and by local investigators (Panel B). Forest plots show the effect of study treatment on progression-free survival in predefined patient subgroups based on stratification factors (Panel C), major demographic and prognostic subgroups (Panel D) and a retrospective analysis in subgroups by primary tumour origin (Panel E). Subgroup results are reported based on central review. In the retrospective post hoc analysis, stomach, colon, rectum, appendix, caecum, ileum, duodenum, and jejunum are grouped under GI. The HRs in all subgroups are obtained from unstratified Cox model. Kaplan–Meier curves are also shown for overall survival (Panel F). ^*Based on prognostic level, patients were divided into two strata. Stratum A (better prognosis) consisted of patients with tumour sites originating from appendix, cecum, jejunum, ileum, duodenum, and NET of unknown primary and stratum B (worse prognosis) comprised patients with primary tumours of lung, stomach, rectum, and colon (except the cecum). ^†Included Black. ^‡Defined as no prior chemotherapy or no SSA therapy continuously for ≥12 weeks any time before study. ^**One patient with thymus as primary tumour origin was not included. ^¶p-value boundary for significance at this interim analysis was 0.0002. CI=confidence interval. CgA=chromogranin A. GI=gastrointestinal. NA=not available. NET=neuroendocrine tumours. NSE=neuron-specific enolase. SSA=somatostatin analogue. ULN=upper limit of normal. WHO=World Health Organization.

Figure 2. Progression-free and overall survival

Kaplan–Meier curves are shown for progression-free survival as assessed by central radiology review (Panel A) and by local investigators (Panel B). Forest plots show the effect of study treatment on progression-free survival in predefined patient subgroups based on stratification factors (Panel C), major demographic and prognostic subgroups (Panel D) and a retrospective analysis in subgroups by primary tumour origin (Panel E). Subgroup results are reported based on central review. In the retrospective post hoc analysis, stomach, colon, rectum, appendix, caecum, ileum, duodenum, and jejunum are grouped under GI. The HRs in all subgroups are obtained from unstratified Cox model. Kaplan–Meier curves are also shown for overall survival (Panel F). ^*Based on prognostic level, patients were divided into two strata. Stratum A (better prognosis) consisted of patients with tumour sites originating from appendix, cecum, jejunum, ileum, duodenum, and NET of unknown primary and stratum B (worse prognosis) comprised patients with primary tumours of lung, stomach, rectum, and colon (except the cecum). ^†Included Black. ^‡Defined as no prior chemotherapy or no SSA therapy continuously for ≥12 weeks any time before study. ^**One patient with thymus as primary tumour origin was not included. ^¶p-value boundary for significance at this interim analysis was 0.0002. CI=confidence interval. CgA=chromogranin A. GI=gastrointestinal. NA=not available. NET=neuroendocrine tumours. NSE=neuron-specific enolase. SSA=somatostatin analogue. ULN=upper limit of normal. WHO=World Health Organization.

Figure 2. Progression-free and overall survival

Kaplan–Meier curves are shown for progression-free survival as assessed by central radiology review (Panel A) and by local investigators (Panel B). Forest plots show the effect of study treatment on progression-free survival in predefined patient subgroups based on stratification factors (Panel C), major demographic and prognostic subgroups (Panel D) and a retrospective analysis in subgroups by primary tumour origin (Panel E). Subgroup results are reported based on central review. In the retrospective post hoc analysis, stomach, colon, rectum, appendix, caecum, ileum, duodenum, and jejunum are grouped under GI. The HRs in all subgroups are obtained from unstratified Cox model. Kaplan–Meier curves are also shown for overall survival (Panel F). ^*Based on prognostic level, patients were divided into two strata. Stratum A (better prognosis) consisted of patients with tumour sites originating from appendix, cecum, jejunum, ileum, duodenum, and NET of unknown primary and stratum B (worse prognosis) comprised patients with primary tumours of lung, stomach, rectum, and colon (except the cecum). ^†Included Black. ^‡Defined as no prior chemotherapy or no SSA therapy continuously for ≥12 weeks any time before study. ^**One patient with thymus as primary tumour origin was not included. ^¶p-value boundary for significance at this interim analysis was 0.0002. CI=confidence interval. CgA=chromogranin A. GI=gastrointestinal. NA=not available. NET=neuroendocrine tumours. NSE=neuron-specific enolase. SSA=somatostatin analogue. ULN=upper limit of normal. WHO=World Health Organization.

Figure 2. Progression-free and overall survival

Kaplan–Meier curves are shown for progression-free survival as assessed by central radiology review (Panel A) and by local investigators (Panel B). Forest plots show the effect of study treatment on progression-free survival in predefined patient subgroups based on stratification factors (Panel C), major demographic and prognostic subgroups (Panel D) and a retrospective analysis in subgroups by primary tumour origin (Panel E). Subgroup results are reported based on central review. In the retrospective post hoc analysis, stomach, colon, rectum, appendix, caecum, ileum, duodenum, and jejunum are grouped under GI. The HRs in all subgroups are obtained from unstratified Cox model. Kaplan–Meier curves are also shown for overall survival (Panel F). ^*Based on prognostic level, patients were divided into two strata. Stratum A (better prognosis) consisted of patients with tumour sites originating from appendix, cecum, jejunum, ileum, duodenum, and NET of unknown primary and stratum B (worse prognosis) comprised patients with primary tumours of lung, stomach, rectum, and colon (except the cecum). ^†Included Black. ^‡Defined as no prior chemotherapy or no SSA therapy continuously for ≥12 weeks any time before study. ^**One patient with thymus as primary tumour origin was not included. ^¶p-value boundary for significance at this interim analysis was 0.0002. CI=confidence interval. CgA=chromogranin A. GI=gastrointestinal. NA=not available. NET=neuroendocrine tumours. NSE=neuron-specific enolase. SSA=somatostatin analogue. ULN=upper limit of normal. WHO=World Health Organization.

Figure 3. Percentage change from baseline in size of target lesion, central review (full analysis set)

The plot shows the best percentage change from baseline in the size of the target lesion (ie, the best response in each patient) in the everolimus arm (left) and placebo arm (right). Fourteen patients (7.6%) in the everolimus arm and 13 patients (15.3%) in the placebo arm showed a change in the available target lesion that contradicted the overall response of progressive disease (marked as * in the graph). Patients for whom the best percentage change in target lesion was not available (21 patients receiving everolimus and 12 patients receiving placebo) or was available but contradicted by overall lesion response of unknown (none), were excluded from the analysis. Percentage above used N as denominator.