ystemic therapy and 29 with 2 or more prior systemic therapies) and 61 patients in the placebo arm (including 25 with 1 prior systemic therapy and 36 with 2 or more prior systemic therapies) who had received prior systemic therapy, the hazard ratio for PFS was 0.456 (95% CI 0.264, 0.787), p=0.0036.
A sensitivity analysis of PFS was conducted where progression was based upon investigator-reported tumour measurements and where all subjects censored for reasons other than study termination were treated as PFS events. This analysis provided a conservative estimate of the treatment effect of sunitinib and supported the primary analysis, demonstrating a hazard ratio of 0.507 (95% CI 0.350, 0.733) and p=0.000193. The pivotal study in pancreatic NET was terminated prematurely at the recommendation of an independent Drug Monitoring Committee, and the primary endpoint was based upon investigator assessment, both of which may have affected the estimates of the treatment effect.
In order to rule out bias in the investigator-based assessment of PFS, a blinded independent central review of scans was performed and supported the investigator assessment, as shown in Table 6.
Table 6 - pNET Efficacy Results from the Phase 3 Study
Efficacy Parameter
SUTENT
(n=86)
Placebo
(n=85)
HR
(95% CI)
P-value
Progression-Free Survival [median, months (95% CI)] by Investigator Assessment
11.4
(7.4, 19.8)
5.5
(3.6, 7.4)
0.418
(0.263, 0.662)
0.0001a
Progression-Free Survival [median, months (95% CI)] by derived tumour response assessment based upon application of RECIST to investigator tumour assessments
12.6
(7.4, 16.9)
5.4
(3.5, 6.0)
0.401
(0.252, 0.640)
0.000066a
Progression-Free Survival [median, months (95% CI)] by blinded independent central review of tumour assessments
12.6
(11.1, 20.6)
5.8
(3.8, 7.2)
0.315
(0.181, 0.546)
0.000015 a
Overall Survival
[median, months (95% CI)]
20.6
(20.6, NR)
NR
(15.5, NR)
0.409
(0.187, 0.894)
0.0204a
Objective Response Rate
[%, (95% CI)]
9.3
(3.2, 15.4)
0
NA
0.0066b
CI=Confidence interval, HR=Hazard ratio, NA=Not applicable, NR=Not reached
a2-sided unstratified log-rank test
bFisher's Exact test
Figure 1 - Kaplan-Meier Curve of PFS in the pNET Phase 3 Study
OS data were not mature at the time of the analysis. There were 9 deaths in the sunitinib arm and 21 deaths in the placebo arm. A statistically significant difference in ORR favouring sunitinib over placebo was observed.
Upon disease progression, patients were unblinded and placebo patients could have been offered access to open-label SUTENT in a separate extension study. As a result of the early study closure, remaining patients were unblinded and offered access to open-label SUTENT in an extension study. A total of 59 patients from the placebo arm received SUTENT in an extension study.
Results from the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQC-30) showed that the overall global health-related quality of life and the five functioning domains (physical, role, cognitive, emotion