The FDA approved 26 unique cancer drugs for the treatment of breast cancer across 42 indications between 2000 and 2023. Of these 42 indications for breast cancer, 27 (64%) were small molecules, 8 (19%) antibodies, and 7 (18%) antibody–drug conjugates (Table 1). 6 indications were for early breast cancer and 36 for metastatic breast cancer.
Table 1 Characteristics of drug indications with FDA approval for breast cancer from 2000 to 2023Clinical development timesThe median development time, defined as the time from IND to FDA approval, was 7.8 years (95% CI 6.2–10.8) for drugs treating breast cancer and 6.9 (95% CI 6.0–7.4) for drugs treating other cancers. However, this difference was not significant in the Cox proportional hazard model (0.82, 95% CI 0.56–1.20, p = 0.477) (Fig. 1).
Fig. 1Time from IND to first FDA approval for breast cancer drugs. The graph illustrates the cumulative incidence of first FDA approval for drugs treating breast cancer (blue) and other cancers (red) in their first approved indication. FDA US Food and Drug Administration, HR hazard ratio, IND investigational new drug application
Drug characteristicsAmong the 42 breast cancer indications, 11 (26%) were first-in-indication, 20 (48%) were advance-in-indication, and 11 (26%) were addition-to-indication. Small molecules, antibodies, and ADCs did not significantly differ in their novelty although there was a higher percentage of first-in-indication treatments among ADCs (22% vs. 25% vs. 43%, p = 0.772). The majority of indications acted via a targeted (88%) mechanism of action than immune-regulating (7%) or cytotoxic (5%) mechanisms (Table 1).
Special FDA designations and reviewOut of all the 42 breast cancer indications, 74% received priority review, 44% breakthrough therapy designation, 24% fast track, and 19% received accelerated approval. Accelerated approval was more frequently granted to antibodies and ADCs than small molecules (50% vs. 29% vs. 7%, p = 0.012) (Table e2). Surprisingly, one indication received an orphan designation, granted to tucatinib “in combination with trastuzumab and capecitabine for the treatment of adult patients with advanced unresectable or metastatic HER2-positive breast cancer, including patients with brain metastases, who have received one or more prior anti-HER2-based regimens in the metastatic setting” based on the HER2CLIMB trial [23]. Orphan designations for subsets of common diseases are referred to as “common orphans” and have previously been reported for lung and skin cancer [14]. Importantly, they fall outside the definition and scope of other regulatory agencies’ definition of an orphan disease, e.g. the European Medicines Agency.
Table e3 demonstrates that more special designations were associated with smaller pivotal trials that were less likely to be of randomized controlled phase 3 design. More special FDA designations were associated with a greater efficacy. Benefits for OS (HR: 0.88 vs. 0.81 vs. 0.73 vs. 0.76 vs. 0.66, p = 0.035), PFS (HR: 0.78 vs. 0.61 vs. 0.55 vs. 0.55 vs. 0.54, p = 0.012), and tumor response (RR: 1.17 vs. 1.98 vs. 1.57 vs. 2.25 vs. 1.78, p < 0.001) improved with more special designations. In single-arm trials, the ORR was 12%, 26%, and 60% for indications with one, three and four special designations (p < 0.001). There was no significant association between the number of special designations and drug prices.
Indication characteristicsNew breast cancer treatments were predominantly original indication approvals (60%) for first-line (43%) or second-line (42%) combination treatments (64%). The majority of indications were approved with a biomarker (86%). The most frequently used biomarkers were HER2 and HR (Table 2).
Table 2 Biomarkers supporting the FDA approval of breast cancer drugsPivotal clinical trialsThe FDA approval of breast cancer indications was supported by pivotal clinical trials with a median of 585 patients (IQR: 417–752) with an accrual rate of 34 patients per month (IQR: 17–49). Trials were conducted at a median of 181 sites (IQR: 142–223) across 19 countries (IQR: 17–20). An equal share of trials was double-blinded and open-label. Trials were predominantly phase 3 (88%) concurrent RCTs (95%) assessing PFS (69%) rather than OS (7%) as the primary endpoint (Table 1). The direct comparator was inactive (e.g. placebo or no treatment) for two-third of RCTs, with merely 33% comparing the new drug directly to active comparator (e.g., anti-cancer agent).
ADCs were more frequently supported by single-arm trials (29% vs. 0% vs. 0%, p = 0.024) of open-label blinding (100% vs. 38% vs. 41%, p = 0.012) than antibodies and small molecules (Table e2). RCTs supporting ADCs were more frequently assessing tumor response as the primary endpoint (29% vs. 25% vs. 0%, p = 0.012) and comparing the new drug to an inactive rather than an active comparator (100% vs. 13% vs. 26%, p = 0.003) than antibodies and small molecules. Stratification of the descriptive statistics across biomarkers is presented in Table e4.
Notably, clinical trials evaluating new drugs for early breast cancer were supported by more enrolled patients than those for metastatic breast cancer (median: 1745 vs. 540, p = 0.007). Early breast cancer trials showed a tendency to be conducted at more trial sites (median: 384 vs. 174, p = 0.057) and in more countries (median: 33 vs. 19, p = 0.062) with consequently significantly greater patient enrollment per month (median: 85 vs. 31, p = 0.011). Stratification of the descriptive statistics across treatment setting is presented in Table e5.
Efficacy, clinical benefit, and quality of lifeAmong RCTs with available data, new breast cancer indications significantly reduced the likelihood of death with an HR of 0.78 (95% CI 0.74–0.82) and were associated with a median 2.8 months (IQR: 1.8–5.8) longer OS than control (Fig. 2). For PFS, the average HR was 0.59 (95% CI 0.54–0.64) with a median PFS improvement of 4.4 months (IQR: 2.2–7.1) compared to control. For tumor response, the RR was 1.61 (95% CI 1.46–1.76) with a median duration of response of 2.9 months (IQR: 2.5–3.9) relative to control. In single-arm trials with available data, the average ORR was 31% (95% CI 10–53) with a median duration of response of 8.2 months (IQR: 6.9–11.7). ADCs were associated with a greater OS benefit relative to antibodies and small molecules in terms of HR (0.69 vs. 0.78 vs. 0.81, p = 0.050). There was no difference in PFS (HR: 0.80 vs. 0.78, p = 0.836) and tumor response (RR: 1.83 vs. 1.61, p = 0.600) for drugs treating early and metastatic breast cancer. 60% of treatments had a high-value ESMO-MCBS score with 14% demonstrating improvements in QoL and 41% in OS.
Fig. 2Meta-analysis of overall survival, progression-free survival, and tumor response supporting the FDA approval of breast cancer drugs. For randomized controlled trials, overall survival (graph A), progression-free survival (graph B), and tumor response rates (graph C) are meta-analyzed using random-effects models. For single-arm studies, the overall response rate is meta-analyzed (graph D). A continuity adjustment of 0.5 for control arms with 0 responders was applied for tumor responses. aP-values calculated based on Cochran’s Q test for subgroup differences. bP-values calculated based on Kruskal–Wallis tests. DoR duration of response, FDA US Food and Drug Administration, IQR interquartile range, NR not reported, OS overall survival, PFS progression-free survival
Correlation between OS, PFS, and tumor responseWe conducted a meta-regression of PFS hazard ratios on OS hazard ratios to evaluate the association between the surrogate endpoint PFS and the clinical endpoint OS. A correlation coefficient of 0.34 (95% CI 0.03 to 0.65, p = 0.031) was measured between OS and PFS HRs with an adjusted R2 of 78.4% (Fig. 3). We conducted a WLS regression analysis, weighted by the number of patients enrolled in the pivotal clinical trial, to evaluate the association between median improvements in PFS and OS. A correlation coefficient of 0.20 (95% CI − 37 to 0.78, p = 0.460) was measured between OS and PFS improvements.
Fig. 3Association between overall survival and progression-free survival benefit in clinical trials supporting the FDA approval of breast cancer drugs. In the graph, each indication’s overall survival hazard ratio (y-axis) is mapped against its progression-free survival hazard ratio (x-axis). Within the graphs, the red line presents fitted treatment outcomes of the random-effects meta-regression. Circle sizes are subject to the precision of each treatment outcome, the inverse of their within-study variance. The analysis only includes treatment outcomes from drug indications with available data from randomized controlled trials. In the meta-regression, a correlation coefficient of 0.34 (95% CI 0.03 to 0.65, p = 0.031) was measured with an adjusted R2 of 78.4%. FDA US Food and Drug Administration, HR hazard ratio
Within conducted meta-regressions, there was no significant association between OS HR and tumor response RR (0.01, 95% CI − 0.02 to 0.05, p = 0.435) and PFS HR and tumor response RRs (− 0.03, 95% CI − 0.07 to 0.01, p = 0.100). Within conducted WLS regressions, there was a significant association between OS and duration of response improvements (1.24, 95% CI 0.16 to 2.32, p = 0.032) and PFS and duration of response improvements (0.26, 95% CI 0.13 to 0.40, p = 0.003).
Drug pricesIn 2023, breast cancer drugs cost an average of $16,013 (95% CI 13,097 to 17,617) (Table e6). There was no association between monthly drug prices and the median improvement in OS (ß = 0.83, 95% CI − 3.11 to 4.77, p = 0.661) or PFS (ß = − 0.54, 95% CI − 4.47 to 3.39, p = 0.780) (Fig. 4). Accordingly, there was no significant difference in median prices for drugs that did and did not demonstrate a benefit in OS (USD 16,079 vs. 16,013, p = 0.891), QoL (USD 16,234 vs. 16,013, p = 0.319), or either (USD 16,146 vs. 16,013, p = 0.729). Coherently, there was no significant difference in prices for indications considered high-value vs. low-value based on the ESMO-MCBS (USD 16,013 vs. 16,243, p = 0.299). The median value per life year gained was $62,419 (IQR 25,840–86,062) for OS and $48,053 (IQR 27,265–77,894) for PFS.
Fig. 4Association between monthly prices and improvement in OS for breast cancer drugs. This graph maps the monthly treatment costs of breast cancer indications (y-axis) against the improvement in OS. The improvement in OS was calculated as the different between the experimental and control arm in pivotal randomized-controlled trial. All prices are in 2023 USD. FDA US Food and Drug Administration, OS overall survival
Comments (0)