scholarly journals Progression-free Survival Decreases with Each Subsequent Therapy in Patients Presenting for Phase I Clinical Trials

2012 ◽  
Vol 3 ◽  
pp. 7-13 ◽  
Author(s):  
Christopher H. Bailey ◽  
Gayle Jameson ◽  
Chao Sima ◽  
Sharon Fleck ◽  
Erica White ◽  
...  
2019 ◽  
pp. 1-10 ◽  
Author(s):  
Guillaume Beinse ◽  
Virgile Tellier ◽  
Valentin Charvet ◽  
Eric Deutsch ◽  
Isabelle Borget ◽  
...  

PURPOSE Drug development in oncology currently is facing a conjunction of an increasing number of antineoplastic agents (ANAs) candidate for phase I clinical trials (P1CTs) and an important attrition rate for final approval. We aimed to develop a machine learning algorithm (RESOLVED2) to predict drug development outcome, which could support early go/no-go decisions after P1CTs by better selection of drugs suitable for further development. METHODS PubMed abstracts of P1CTs reporting on ANAs were used together with pharmacologic data from the DrugBank5.0 database to model time to US Food and Drug Administration (FDA) approval (FDA approval-free survival) since the first P1CT publication. The RESOLVED2 model was trained with machine learning methods. Its performance was evaluated on an independent test set with weighted concordance index (IPCW). RESULTS We identified 462 ANAs from PubMed that matched with DrugBank5.0 (P1CT publication dates 1972 to 2017). Among 1,411 variables, 28 were used by RESOLVED2 to model the FDA approval-free survival, with an IPCW of 0.89 on the independent test set. RESOLVED2 outperformed a model that was based on efficacy/toxicity (IPCW, 0.69). In the test set at 6 years of follow-up, 73% (95% CI, 49% to 86%) of drugs predicted to be approved were approved, whereas 92% (95% CI, 87% to 98%) of drugs predicted to be nonapproved were still not approved (log-rank P < .001). A predicted approved drug was 16 times more likely to be approved than a predicted nonapproved drug (hazard ratio, 16.4; 95% CI, 8.40 to 32.2). CONCLUSION As soon as P1CT completion, RESOLVED2 can predict accurately the time to FDA approval. We provide the proof of concept that drug development outcome can be predicted by machine learning strategies.


Blood ◽  
2013 ◽  
Vol 122 (21) ◽  
pp. 1958-1958
Author(s):  
R. Donald Harvey ◽  
Meagan S. Barbee ◽  
Ajay K. Nooka ◽  
Sungjin Kim ◽  
Zhengjia Chen ◽  
...  

Abstract Objectives Categorization of response criteria for multiple myeloma (MM) is based on magnitude of change in serum and urine paraprotein values and normalization of free light chain ratio (rFLC). However, the association between improvements in these surrogate markers and patient outcomes is not validated in the phase I setting. Early measures of response would be beneficial for patients and agents to identify those likely to have prolonged disease-free intervals and to validate agent activity for rapid movement to subsequent development. Methods We identified 31 trials that met enrollment criteria of phase I, relapsed or refractory disease, and non-transplant study population. Clinical and demographic data collected included age, sex, race, ECOG performance status (PS) at entry, myeloma subtype and isotype, prior therapy, cytogenetics at study entry, date of progression, and date of expiration. Patients with t(4;14), del13, del17p, t(14;16), or t(14;20) were considered to have non-standard risk cytogenetics. Evaluation of the relationship between progression free survival (PFS) and change in plasma cell activity by the rFLC and magnitude of response in serum/urine paraprotein per IMWG criteria was performed. Landmark analyses occurred at cycle 2 and 4, 8, and 12 months. Progression free survival (PFS) at 12 months was the primary outcome of interest. Results Among 87 patients; 47 (54%) were female; 56 (64%) white, 29 (33%) black; 27 (31%) non-standard risk cytogenetics; ECOG PS 1 in 73 (84%); and median prior lines 5 (1-11). Eighty were evaluable for paraprotein changes, 71 for rFLC. Normalization of rFLC at 4 months conferred a PFS advantage (11.3 v 2.8 months, p = 0.038) (Table 1). Normalization of rFLC by 12 months was found to predict PFS (6.1 vs. 2.8 months, p = 0.015) and a longer OS (45 vs. 17.4 months, p = 0.002). Magnitude of response in paraprotein was found to predict and correlate linearly with PFS at all time landmarks (r2 = 0.769 to 0.952). Analysis of PFS by IMWG criteria and by quartiles of 50% changes were both linear (p < 0.001) (Figure 1). Conclusion These findings suggest that normalization of the rFLC and magnitude of paraprotein response are viable surrogate disease endpoints in phase I clinical trials of novel agents and combinations. The use of current IMWG criteria in the phase I setting is valuable, but the addition of time of response and alterations of response boundaries should be further evaluated in the setting of improved treatments. Disclosures: Kaufman: Onyx: Consultancy; Celgene: Consultancy, Research Funding; Novartis: Consultancy, Research Funding; Janssen: Consultancy; Millenium: Consultancy; Merck: Research Funding. Lonial:Millennium: Consultancy; Celgene: Consultancy; Novartis: Consultancy; BMS: Consultancy; Sanofi: Consultancy; Onyx: Consultancy.


2013 ◽  
Vol 31 (15_suppl) ◽  
pp. 9509-9509
Author(s):  
Wai Meng David Tai ◽  
Cindy Lim ◽  
Aziah Ahmad ◽  
Whee Sze Ong ◽  
SuPin Choo ◽  
...  

9509 Background: Despite the significant burden of cancer in the older population, their outcomes in the context of phase I studies have been poorly studied. We evaluated the clinical characteristics and outcomes of elderly pts enrolled in phase 1 clinical trials in our centre and evaluate the performance of Royal Marsden Hospital (RMH) prognostic score (albumin, LDH, no of met sites) in this pt population. Methods: 296 consecutive pts who were treated in 20 phase 1 trials from 2005-2012 in our unit were analysed. Clinical characteristics and outcomes between young pts (<65, n=202) and older pts (≥65, n=94) were compared. Results: The median age of the older pts was 69 (65-84), 71% were males. 51% of the pts received chemotherapy based treatment with or w/out biological agents. 61% of the pts had lung cancers and 32% had gastrointestinal cancers. 52% of pts had ≥2 co-morbidities. After median follow up of 7.5 mths (0.36-50.6 mths), the median progression free survival (PFS) and overall survival (OS) were 5.8 and 8.8 mths respectively. Although elderly pts had more co-morbidities and lower albumin levels at baseline, there was no significant difference in survival (8.8 vs 9.9 mths), p=0.68) compared to younger pts. The prognostic factors for OS identified in multivariate analysis were prior lines of chemotherapy (0-2 vs ≥3), baseline sodium levels (≥135 vs <135mmol/L) and platelet levels (≤400 vs >400×10⁹). We developed a risk nomogram based on the factors identified prognostic of OS with concordance(c)-index of 0.65. RMH score (2-3 vs 0-1) predicted for OS with hazard ratio of 2.1, p=0.03 and c- index of 0.63. 26% of elderly pts experienced grade 3/4 toxicities in the first cycle of treatment. Common grade 3/4 toxicities were dermatological (25%), haematological (17%) and gastrointestinal (13%). Both age of pts (p=0.70) and dose levels (p=0.18) did not have any bearing on occurrence of grade 3/4 toxicities. Conclusions: Elderly pts (≥65) enrolled into phase 1 clinical trials had similar survival outcomes and toxicity profiles compared to younger pts. Risk scoring models to aid patient selection need further clarification in this population.


2016 ◽  
Vol 140 (2) ◽  
pp. 480-484 ◽  
Author(s):  
Brian H. Kushner ◽  
Nai-Kong V. Cheung ◽  
Shakeel Modak ◽  
Oren J. Becher ◽  
Ellen M. Basu ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document