Table 2.
Statistics on NCI CTEP-Sponsored Trials Whose Accrual Was Stopped or Whose Results Were Released Early for Positive Results From 1990 to 2005
| Trial Identifier | Results When Trial Crossed Boundary |
Results When Trial First Published |
Follow-Up Results of Trial |
||||||
|---|---|---|---|---|---|---|---|---|---|
| % Information | Treatment Effect | P | % Information | Treatment Effect | P | % Information | Treatment Effect | P | |
| NCCTG-844652* | 60 | HR = 0.67 | .0064 | Same as when trial stopped | 89 | HR = 0.67 | .0007 | ||
| RTOG-8501† | ∼60 | HR = ∼0.49 | .0045 | 87 | 2-year OS: 38% v 10% | < .001 | 108 | 2-year OS: 36% v 10% | < .001 |
| POG-9006‡ | ∼38 | 2-year CCR: 82% v 70.8% | .0016 | ∼54 | 2-year CCR: 84% v 75% | .006 | 77 | 4-year CCR: 70.6% v 64% | .22 |
| EST-3189§ | 62 | 3-year OS: 84% v 73% | .0050 | Same as when trial stopped | 90 | 4-year OS: 78.1% v 71.4% | .10 | ||
| CLB-9011‖ | 34 | CR rates: 30% v 2% | .00014 | 78 | CR rates: 33% v 8% | < 10−5 | 117 | CR rates: 20% v 4% | < 10−5 |
| SWOG-8892 | 26 | HR = 0.26 | < .0001 | Same as when trial stopped | 50 | HR = 0.31 | < .001 | ||
| ECOG-2491¶ | 50 | 1-year DFS: 92% v 57% | < .0001 | Same as when trial stopped | > 100 | 5-year DFS: 69% v 29% | < .0001 | ||
| CCG-1882 | 68 | 4-yr EFS: 75.4% v 57.2% | .0013 | 80 | 5-year EFS: 75.0% v 55.0%# | < .001 | — | — | — |
| SWOG-8814 | 81 | HR = 0.66 | .002 | Same as when trial stopped | 120 | HR = 0.76 | .002 | ||
| SWOG-8797 | ∼60 | HR = 0.45** | 0.006** | ∼60-63†† | HR = 0.50 | .01 | 63 | HR = .51‡‡ | .007‡‡ |
| CLB-9344§§ | 25 | HR = 0.79 | .013 | Same as when trial stopped | 59 | HR = 0.83 | .0023 | ||
| RTOG-9001 | 58 | 5-year OS: 73% v 58% | .0027 | 59 | 5-year OS: 73% v 58% | .004 | 81 | 5-year OS: 73% v 52% | < .0001 |
| CCG-5942 | 34‖‖ | HR = 0.27 | .0048 | 64 | HR = 0.59 | .057¶¶ | |||
| SWOG-9133 | 46 | 3-year FFS: 93% v 81% | < .001 | Same as when trial stopped | 48 | 3-year FFS: 94% v 81% | < .001 | ||
| RTOG-9413## | 90 | 4-year PFS: 56% v 46% | .014 | Same as when trial stopped | 156 | 4-year PFS: 54% v 54% | NS | ||
| SWOG-S9701 | 15 | HR = 0.43 | .0046 | Same as when trial stopped | 62 | HR = 0.70 | .008 | ||
| NCCTG-N9741*** | 81 | HR = 0.71 | .0009 | Same as when trial stopped | 98 | HR = 0.74 | .0014 | ||
| NCIC-MA17 | 40 | HR = 0.57 | .00008 | Same as when trial stopped | 48 | HR = 0.58 | < .001 | ||
| E-1496 | 49 | HR = 0.42 | .00016 | 59 | HR = 0.5 | .00006 | 83††† | HR = 0.38 | < 10−5 |
| E-E1A00 | 59 | RR: 80% v 53% | .0046 | Same as when trial stopped | 108 | RR: 63% v 41% | .0034 | ||
| CCG-1961‡‡‡ | 84 | 5-year EFS: 78.8% v 69.8% | .0198 | 91 | 5-year EFS: 80% v 71% | .01 | 124 | 5-year EFS: 81.2% v 71.7% | < .001 |
| E-3200§§§ | 90 | HR = 0.74 | .0024 | Same as when trial stopped | 111 | HR = 0.75 | .0011 | ||
| ECOG-2997 | 79 | CR rates: 22.9% v 5.8% | .0008 | 98 | CR rates: 22.4% v 5.8% | .0002 | 107 | CR rates: 23.4% v 4.6% | < 10−5 |
| NSABP-B-31/NCCTG-N9831‖‖‖ | 55 | HR = 0.48 | < .0001 | Same as when trial stopped | 87 | HR = 0.49 | < .0001 | ||
| ECOG-4599 | 72 | HR = 0.78 | .0076 | 100 | HR = 0.79 | .003 | — | — | — |
| ECOG-2100 | 65 | HR = 0.50 | < .001 | 114 | HR = 0.60 | < .001 | — | — | — |
| NCIC-MA21¶¶¶ | 58 | HR = 0.60 | .0006 | Same as when trial stopped | — | — | — | ||
Abbreviations: NCI, National Cancer Institute; CTEP, Cancer Therapy Evaluation Program; NCCTG, North Central Cancer Treatment Group; HR, hazard ratio; RTOG, Radiation Therapy Oncology Group; OS, overall survival; POG, Pediatric Oncology Group; CCR, continuous complete remission; CR, complete response; SWOG, Southwest Oncology Group; ECOG, Eastern Cooperative Oncology Group; DFS, disease-free survival; CCG, Children's Cancer Group; EFS, event-free survival; FFS, failure-free survival; PFS, progression-free survival; NS, not significant; NCIC, National Cancer Institute of Canada; RR, response rate; NSABP, National Surgical Adjuvant Breast and Bowel Project.
NCCTG-844652: We consider only the analysis of the stage C patients here (which was prespecified). The trial was stopped because of differences in the levamisole plus fluorouracil arm compared with the observation arm, which are the results reported here.
RTOG-8501: Information percentage for when the trial crossed the boundary was estimated based on length of accrual/follow-up and observed survival rates. Treatment effect HR for when the trial crossed the boundary was estimated based on P value and estimated number of events.
POG-9006: Information percentage for when trial crossed boundary and when trial was first published was estimated based on accrual/follow-up and observed CCR. In the follow-up period, treatment effect 2-year CCR rates were 85.2% and 80.4%.
EST-3189: OS was one of the two coprimary end points and the one that crossed the interim monitoring boundary. In the follow-up period, the treatment effect– estimated 3-year OS rates from Figure 2 are 83% v 77%. Regarding the P value, the protocol for this trial specified a one-sided 5% level test, which corresponds to a two-sided 10% level test.
CLB-9011: This was initially a three-armed trial, with the experimental fludarabine plus chlorambucil arm later dropped. The results presented here are for the comparison of fludarabine v chlorambucil (control). The P value used for the interim analysis for when trial crossed boundary was P = .0005 based on a comparison of both experimental arms combined compared with the control arm.
ECOG-2491: The results reported here are for the induction therapy comparison (there was also a maintenance treatment randomization in this trial). For the follow-up period, the estimated 1-year DFS rates from Figure 1 of Tallman et al20 are 88% v 57%.
CCG-1882: Regarding treatment effect results when the trial was first published, the estimated 4-year EFS rates from Figure 1 of Nachman et al 21 are 77% v 60%.
SWOG-8797: HR and P value were unadjusted for stratification variables. The values adjusted for the random assignment stratification variables are 0.49 and P = .02.
SWOG-8797: The information fraction is not given in the first publication but must be between when the trial was stopped and the later follow-up.
SWOG-8797: HR and P value were adjusted at follow-up for random assignment stratification factors.
CLB-9344: This trial had a factorial design; comparison of ± paclitaxel was the one that led to the release of the data and is the one reported here. HR for when trial crossed boundary was estimated based on P value and percent information.
CCG-5942: Per protocol, the random assignment was stopped when the boundary was crossed, but the results were not released until there was further follow-up.
CCG-5942: This result would be considered statistically significant because the trial was designed with a one-sided type I error of 0.10, which corresponds to a two-sided type I error of 0.20.
RTOG-9413: This trial had a factorial design; comparison of whole-port v prostate-only radiation was the one that crossed the interim monitoring boundary and is the one reported here. The percent information for when trial crossed boundary was estimated from interim monitoring boundary. The 4-year PFS results in the follow-up period were estimated from curves in Figure 2A of Lawton et al.35
NCCTG-9741: This trial had three arms; fluorouracil/leucovorin + oxaliplatin v fluorouracil/leucovorin + irinotecan is the comparison that crossed the interim monitoring boundary and is the one considered here.
E1496: Information is estimated based on CI width for the HR.
CCG-1961: We are considering the analysis of the intensity of treatment question in the rapid early responder subgroup, which is the one that crossed an interim analysis boundary.
E-3200: Arms being discussed here are the infusional fluorouracil, leucovorin, and oxaliplatin (FOLFOX4) arm (control arm) and FOLFOX4 + bevacizumab arm; the bevacizumab-alone arm closed early for negative results.
NSABP-B-31/NCCTG-N9831: The results reported here are from the combined analysis of the trials NSABP-B-31 and NCCTG-N983, which are considered one trial for the purposes of discussion here.
NCIC-MA21: This was a three-arm trial; results are given for epirubicin-cyclophosphamide/paclitaxel v doxorubicin-cyclophosphamide/paclitaxel as experimental and control treatments, respectively.