Introduction of PH (2012) and K (2013) in the US changed the standard-of-care for HER2-positive mBC. RW studies reported decreased K effectiveness in pts with PH pre-tx. We studied whether timing of PH and K availability in clinical practice could introduce selection bias (i.e. pts with K after PH in early yrs of availability may have to have had rapid progression during PH tx to be included in RW analyses).
Using de-identified data from the US-based Flatiron electronic health record-derived database, we selected pts diagnosed with mBC on/after 1 Jan 2011 and initiating K between 1 Feb 2013 and 31 Jul 2018. Primary analysis included all pts receiving K (any tx line). Subgroup analysis focused on pts with 2nd-line K tx after PH tx. Time from index date to next tx/death (TTNT) was used as a proxy for progression-free survival.
Primary cohort (N = 533) demographics were generally consistent across yrs (2013–18); median age was 61 yrs; 89% had prior HER2-targeted tx; 23% had brain metastases (mets); 68% had visceral mets. In pts with (n = 231) vs without (n = 79) prior PH, median age was 61 vs 65 yrs, mean time from mBC diagnosis was 16 vs 18 months, 27% vs 20% had brain mets, 74% vs 58% had visceral mets. Additional findings are shown (Table). Prevalence of prior PH exposure almost doubled over the yrs. In 2nd-line K treated pts with prior PH, time from mBC to K increased over the yrs, paralleled by a numerical increase in median TTNT.Table:
|Primary cohort (N = 533)|
|Prior PH, n (%)||18 (35.3%)||35 (45.5%)||72 (66.7%)||80 (75.5%)||77 (65.8%)||53 (71.6%)|
|TTNT, median months (95% CI)||6.8 (4.9–14.2)||6.5 (4.2–9.2)||7.6 (4.7–9.8)||6.3 (5.1–8.4)||8.4 (6.4–10.3)||Data not yet mature|
|Subgroup analysis (N = 231)|
|Pts with 2nd-line K use after PH, N||13||25||56||51||52||34|
|mBC to K initiation in pts with 2nd-line K use after PH, mean months (SD)||12.7 (6.9)||13.1 (8.5)||13.0 (8.0)||16.0 (12.9)||17.9 (12.5)||18.6 (11.6)|
|TTNT in pts with 2nd-line K use after PH, median months (95% CI)||5.1 (3.0–7.5)||7.2 (4.6–11.7)||6.6 (4.5–8.4)||10.3 (5.6–12.8)||Data not yet mature|
CI, confidence internal; SD, standard deviation.
Prevalence of PH pre-tx increased, yet K effectiveness remained consistent across yrs. With PH launched shortly before K, the initial pts receiving K were possibly selected due to shorter benefit from prior PH. This selection bias may explain reports of lower K effectiveness in pts with prior PH tx. This illustrates the challenges of assessing tx sequence effectiveness shortly after tx approval.
Clinical trial identification
Writing assistance was provided by Meredith Kalish, MD, of Ashfield Healthcare Communications (a UDG Healthcare plc company), and funded by F. Hoffmann-La Roche Ltd.
Legal entity responsible for the study
F. Hoffmann-La Roche Ltd.
F. Hoffmann-La Roche Ltd.
T. Sanglier: Shareholder / Stockholder / Stock options, Full / Part-time employment: Roche. C. Flores: Full / Part-time employment: Genesis Research. E. Flahavan: Shareholder / Stockholder / Stock options: Eli Lilly; Shareholder / Stockholder / Stock options, Full / Part-time employment: Roche. N. Lindegger: Honoraria (self), Leadership role, Shareholder / Stockholder / Stock options, Full / Part-time employment: Roche; Shareholder / Stockholder / Stock options: Novartis; Shareholder / Stockholder / Stock options: Idorsia; Shareholder / Stockholder / Stock options: J&J. F. Montemurro: Speaker Bureau / Expert testimony: Novartis; Speaker Bureau / Expert testimony: AstraZeneca; Speaker Bureau / Expert testimony: Ely Lilly; Speaker Bureau / Expert testimony: Pfizer; Speaker Bureau / Expert testimony, Travel / Accommodation / Expenses: Roche. All other authors have declared no conflicts of interest.