Table 1 Characteristics of oncology drug studies reporting rank preserving structural failure time analysis
All studies (N=65) | Fundamental, no SOC (N=12) | Fundamental with SOC (N=34) | Sequential (N=19) | p-value1 | |
|---|---|---|---|---|---|
Tumor type, n (%) | <0.001 | ||||
Breast | 7 (10.8) | 0 | 7 (20.6) | 0 | |
Colorectal | 3 (4.6) | 0 | 0 | 3 (15.8) | |
GIST | 4 (6.2) | 4 (33.3) | 0 | 0 | |
CLL/SLL | 4 (6.2) | 0 | 4 (11.8) | 0 | |
Melanoma | 5 (7.7) | 1 (8.3) | 4 (11.8) | 0 | |
Myelofibrosis | 2 (3.1) | 2 (16.7) | 0 | 0 | |
Multiple myeloma | 5 (7.7) | 0 | 3 (8.8) | 2 (10.5) | |
Neuroendocrine | 3 (4.6) | 3 (25.0) | 0 | 0 | |
Non-small cell lung cancer | 13 (20.0) | 0 | 6 (17.6) | 7 (36.8) | |
Ovarian | 2 (3.1) | 0 | 0 | 2 (10.5) | |
Prostate | 5 (7.7) | 0 | 3 (8.8) | 2 (10.5) | |
Renal cell carcinoma | 4 (6.2) | 1 (8.3) | 2 (6.1) | 1 (5.3) | |
Thyroid | 3 (4.6) | 1 (8.3) | 2 (5.9) | 0 | |
Other | 5 (7.7) | 0 | 3 (8.8) | 2 (10.5) | |
Randomization ratio, n (%) | 0.18 | ||||
1:1 | 37 (56.9) | 5 (41.7) | 17 (50.0) | 15 (78.9) | |
2:1 | 27 (41.5) | 7 (58.3) | 16 (47.1) | 4 (21.1) | |
3:1 | 1 (1.5) | 0 | 1 (2.9) | 0 | |
Open label, n (%) | 29 (44.6) | 2 (16.7) | 17 (50.0) | 10 (52.6) | 0.08 |
FDA approved for indication (yes), n (%) | 54 (83.1) | 11 (91.7) | 26 (76.5) | 17 (98.5) | 0.33 |
Medical writers, n (%) | 0.33 | ||||
Yes | 44 (67.7) | 7 (58.3) | 26 (76.5) | 11 (57.9) | |
No | 3 (4.6) | 0 | 1 (2.9) | 2 (10.5) | |
Not indicated | 18 (27.7) | 5 (41.7) | 7 (20.6) | 6 (31.6) | |
Author as employee of company (yes), n (%) | 58 (89.2) | 10 (82.3) | 31 (91.2) | 17 (89.5) | 0.26 |
Age of participants, median (Q1, Q3) | 61 (58 to 64) | 62 (58 to 64) | 62 (55 to 64) | 60 (58 to 62) | 0.82 |
Total number of participants, median (Q1, Q3) | 361 (233 to 512) | 311 (214 to 373) | 392 (259 to 519) | 359 (242 to 690) | 0.20 |
Difference in adjusted and unadjusted overall survival hazard ratio, median (Q1, Q3) | -0.1 (-0.3 to -0.6) | -0.3 (-0.3 to -0.02) | -0.1 (-0.2 to -0.3) | -0.09 (-0.2 to -0.4) | 0.02 |
% Trials with numerically higher unadjusted overall survival hazard ratio, n (%) | 60 (92.3) | 12 (100) | 31 (91.2) | 17 (89.5) | 0.39 |
% Crossover, median (Q1, Q3) | 55.7 (36.5 to 71.6) | 75.3 (59.3 to 84.2) | 53.0 (38.0 to 67.4) | 46.6 (33.2 to 67.4) | 0.02 |
Median overall survival for intervention arm, median (Q1, Q3) | 23.9 (17.8 to 36.8) | 18.4 (17.6 to 33.6) | 24.7 (16.2 to 42.6) | 24.0 (22.9 to 32.8) | 0.74 |
Median overall survival for control arm (uncorrected), median (Q1, Q3) | 20.5 (14.5 to 36.3) | 17.4 (16.2 to 36.3) | 19.4 (12.0 to 22.3) | 30.3 (21.5 to 36.7) | 0.11 |
Median overall survival for control arm (corrected), median (Q1, Q3) | 15.9 (9.8 to 27.4) | 10.9 (9.6 to 17.6) | 12.3 (7.8 to 18.3) | 26.6 (17.1 to 35.1) | 0.04 |
Change in significance of overall survival hazard ratio (corrected vs. uncorrected), yes n (%) | 41 (64.1) | 3 (25.0) | 22 (66.7) | 16 (84.2) | 0.003 |
RPSFT analysis stipulated in protocol | 0.36 | ||||
Yes | 22 (33.8) | 2 (16.7) | 13 (38.2) | 7 (36.8) | |
No | 29 (44.6) | 5 (41.7) | 16 (47.1) | 8 (42.1) | |
No protocol available | 14 (21.5) | 5 (41.7) | 5 (14.7) | 4 (21.1) | |
Year of study publication, n (%) | 0.07 | ||||
2008 | 1 (1.5) | 1 (8.3) | 0 | 0 | |
2011 | 1 (1.5) | 0 | 0 | 1 (5.3) | |
2012 | 3 (4.6) | 2 (16.7) | 1 (2.9) | 0 | |
2013 | 2 (3.1) | 0 | 2 (5.9) | 0 | |
2015 | 4 (6.2) | 1 (8.3) | 3 (8.8) | 0 | |
2016 | 11 (16.9) | 4 (33.3) | 4 (11.8) | 3 (15.8) | |
2017 | 6 (9.2) | 2 (16.7) | 0 | 4 (21.1) | |
2018 | 5 (7.7) | 0 | 4 (11.8) | 1 (5.3) | |
2019 | 5 (7.7) | 0 | 3 (8.8) | 2 (10.5) | |
2020 | 7 (10.8) | 0 | 5 (14.7) | 2 (10.5) | |
2021 | 11 (16.9) | 2 (16.7) | 7 (20.6) | 2 (10.5) | |
2022 | 9 (13.8) | 0 | 5 (14.7) | 4 (21.1.) |