Degele, Meghan

PURPOSEThere are numerous barriers to enrollment in oncology biomarker-driven studies.METHODSThe ELAINE 2 study (ClinicalTrials.gov identifier: NCT04432454 ) is an open-label phase 2 study of lasofoxifene combined with abemaciclib in patients with advanced or metastatic estrogen receptor–positive/human epidermal growth factor receptor 2–negative breast cancer with an ESR1 mutation. ELAINE 2 opened clinical sites by using a Traditional approach, which activated a site before patient identification, and the Tempus TIME Trial network, which opened a site only after identifying an eligible patient. This manuscript presents the operational metrics comparing the Traditional and TIME Trial site data.RESULTSThe study enrolled patients over 34 weeks and 16 sites (six Traditional and 10 TIME Trial) participated. Duration for full clinical trial agreement execution for Traditional sites and TIME Trial sites averaged 200.5 (range, 142-257) and 7.6 days (range, 2-14), respectively. Institutional review board approval time for Traditional sites and TIME Trial sites was 27.5 (range, 12-71) and 3.0 days (range, 1-12), respectively. Duration from study activation to first consent was 33.3 (range, 18-58) and 8.8 days (range, 1-35) for Traditional and TIME Trial sites, respectively. The first patient on study was at a TIME Trial site 115 days before a Traditional site and the first seven patients enrolled were at TIME Trial sites. Traditional sites consented 23 and enrolled 16 patients, while TIME Trial sites consented 16 and enrolled 13. The trial enrolled 29 patients in 8.5 months with the anticipated enrollment duration being 12-18 months.CONCLUSIONThe TIME Trial network opened earlier and enrolled the first study patients. These results demonstrate that the Just-in-TIME model, along with a Traditional model, can improve enrollment in biomarker-driven studies.

1504 Background: Trial recruitment that requires specific actionable mutations based on next-generation sequencing (NGS) is challenging. Barriers can include competing studies, physician study awareness, site proximity, mutation incidence, among other concerns. Methods: This study (NCT04432454) opened clinical sites using two methods during the COVID-19 pandemic. The “Traditional” approach included site selection, IRB and contract approval, and trial activation prior to a patient being identified for enrollment. The second approach used the Tempus “TIME” Trials network that would only open a site after identifying a patient with a mutation of interest and eligible for the trial. Results: The first patient enrolled was on 10/12/20 and the last patient was on 6/24/21. A total of 16 sites (6 Traditional and 10 TIME) participated. All Traditional sites, and none of the TIME sites, were affiliated with major academic institutions. Duration for full CTA execution for Traditional sites averaged 200.5 days (range 142 to 257) and for TIME sites averaged 7.6 days (range 2 to 14). IRB approval time average for Traditional sites was 27.5 days (range 12 to 71) and TIME sites was 3.0 days (range 1 to 12 days). Days from site selection to activation letter for Traditional sites was on average 250.0 days (range 187 to 281) and for TIME sites was 131.6 days (range 22 to 248). Time from study activation to first consent was 33.3 days (range 18 to 58) for Traditional sites and 8.8 days (range 1 to 35) for TIME sites. The first patient on-study was at a TIME site 115 days prior to a Traditional site and the first 7 patients enrolled were at TIME sites. Traditional sites consented 23 and enrolled 16 patients while the TIME sites consented 16 and enrolled 13. The trial enrolled all 29 patients in 8 months with the anticipated enrollment duration being 12 to 18 months. Conclusions: Although the Traditional and TIME programs had different operational models, they both contributed a significant number of patients and reduced the projected enrollment timeline. TIME sites enrolled the initial patients. These results demonstrate that the “Just-in-Time model,” in conjunction with a Traditional model, can reduce projected overall time to enrollment in biomarker-driven studies. [Table: see text]