FROM HAZARD RATIO TO CLINICAL IMPACT: NUMBER NEEDED TO TREAT IN PLATINUM-RESISTANT OVARIAN CANCER
Author(s)
Anju Parthan, PhD1, Gaurav Suri, MSc2, Rob Musci, MSc3, Sonja Sorensen, MPH4.
1Senior Director, Health Economics & Outcomes, Corcept Therapeutics, Redwood City, CA, USA, 2Corcept Therapeutics, Redwood City, CA, USA, 3Evidera, San Francisco, CA, USA, 4Evidera, Bethesda, MD, USA.
1Senior Director, Health Economics & Outcomes, Corcept Therapeutics, Redwood City, CA, USA, 2Corcept Therapeutics, Redwood City, CA, USA, 3Evidera, San Francisco, CA, USA, 4Evidera, Bethesda, MD, USA.
OBJECTIVES: To contextualize survival outcomes in platinum-resistant ovarian cancer (PROC), we estimated number needed to treat (NNT) using phase 3 trial data of approved therapies.
METHODS: NNT was calculated as the inverse of the absolute risk difference in 12- and 18-month overall survival (OS) probabilities. Estimates were derived from Kaplan-Meier (KM) curves for relacorilant + nab-paclitaxel (ROSELLA; comparator: nab-paclitaxel), mirvetuximab (MIRASOL; FRɑ-positive population; comparator: investigator’s-choice chemotherapy, including a paclitaxel subgroup), and pembrolizumab + paclitaxel ± bevacizumab (KEYNOTE-B96; CPS ≥ 1 population; comparator: paclitaxel ± bevacizumab), using populations corresponding to each therapy’s approved indication. When OS probabilities were unavailable, KM curves were digitized to estimate survival probabilities. NNTs were calculated within trials and are presented descriptively. Indirect comparisons were not performed.
RESULTS: At 12 months, NNTs were ~10 across therapies, meaning 10 patients need treatment for one additional survivor at 12 months. Differences widened at 18 months: NNTs were 5 for relacorilant + nab-paclitaxel, 9 for mirvetuximab, and 8 for pembrolizumab + paclitaxel ± bevacizumab. These findings suggest greater absolute survival benefit at later time points versus respective control arms. Within the MIRASOL paclitaxel subgroup, mirvetuximab NNTs were 21 and 17 at 12 and 18 months, respectively, versus paclitaxel alone. Results are descriptive and should not be interpreted as comparative treatment effects across trials.
CONCLUSIONS: NNT provides a clinically intuitive metric for contextualizing treatment benefit in PROC. Although NNTs were similar at 12 months, this descriptive analysis estimated lower NNT values for relacorilant + nab-paclitaxel at 18 months, suggesting greater long-term survival impact relative to the corresponding control arm. Lower NNT values indicate fewer patients would need treatment to achieve one additional survivor relative to the control arm. While economic outcomes were not assessed, lower NNT values could potentially translate into a lower cost per additional survivor, which should be explored in future economic analyses.
METHODS: NNT was calculated as the inverse of the absolute risk difference in 12- and 18-month overall survival (OS) probabilities. Estimates were derived from Kaplan-Meier (KM) curves for relacorilant + nab-paclitaxel (ROSELLA; comparator: nab-paclitaxel), mirvetuximab (MIRASOL; FRɑ-positive population; comparator: investigator’s-choice chemotherapy, including a paclitaxel subgroup), and pembrolizumab + paclitaxel ± bevacizumab (KEYNOTE-B96; CPS ≥ 1 population; comparator: paclitaxel ± bevacizumab), using populations corresponding to each therapy’s approved indication. When OS probabilities were unavailable, KM curves were digitized to estimate survival probabilities. NNTs were calculated within trials and are presented descriptively. Indirect comparisons were not performed.
RESULTS: At 12 months, NNTs were ~10 across therapies, meaning 10 patients need treatment for one additional survivor at 12 months. Differences widened at 18 months: NNTs were 5 for relacorilant + nab-paclitaxel, 9 for mirvetuximab, and 8 for pembrolizumab + paclitaxel ± bevacizumab. These findings suggest greater absolute survival benefit at later time points versus respective control arms. Within the MIRASOL paclitaxel subgroup, mirvetuximab NNTs were 21 and 17 at 12 and 18 months, respectively, versus paclitaxel alone. Results are descriptive and should not be interpreted as comparative treatment effects across trials.
CONCLUSIONS: NNT provides a clinically intuitive metric for contextualizing treatment benefit in PROC. Although NNTs were similar at 12 months, this descriptive analysis estimated lower NNT values for relacorilant + nab-paclitaxel at 18 months, suggesting greater long-term survival impact relative to the corresponding control arm. Lower NNT values indicate fewer patients would need treatment to achieve one additional survivor relative to the control arm. While economic outcomes were not assessed, lower NNT values could potentially translate into a lower cost per additional survivor, which should be explored in future economic analyses.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
CO62
Topic
Clinical Outcomes
Topic Subcategory
Clinical Outcomes Assessment, Comparative Effectiveness or Efficacy, Performance-based Outcomes
Disease
No Additional Disease & Conditions/Specialized Treatment Areas, Oncology