Title: Evaluating Pathologic Complete Response (pCR) as a Surrogate Endpoint for Survival in Early Triple Negative Breast Cancer (TNBC) Treated with Neoadjuvant Chemo-Immunotherapy: A Meta Analysis of Randomized Controlled Studies. Background: Pathologic complete response (pCR) is recognized by regulatory agencies as a surrogate endpoint reasonably likely to predict long term outcomes in neoadjuvant trials for high risk early breast cancer. In triple negative breast cancer (TNBC), the addition of immune checkpoint inhibitors (ICIs) to neoadjuvant chemotherapy significantly increases pCR rates; however, the strength of pCR as a surrogate for event free survival (EFS) or overall survival (OS) in the context of chemo immunotherapy remains uncertain. A formal evaluation of surrogacy is therefore essential to inform clinical development and regulatory decision making. Methods: A systematic literature search identified RCTs evaluating neoadjuvant ICIs in early TNBC. The primary objective was to assess the trial-level association between pCR and EFS and OS. The correlation between hazard ratio (HR) for EFS and OS and the relative risk (RR) for pCR across trials was analyzed using a linear regression model to determine the coefficient of determination (R²). Secondary objectives were to explore heterogeneity of results according to baseline features of interest. Results: Six RCTs, enrolling a total of 4,035 patients, were included. Across studies, the addition of ICIs resulted in a significant increase in pCR, with a pooled relative risk of 1.17 (95% CI, 1.10-1.25; p < 0.0001), and in EFS with a pooled hazard ratio for EFS was 0.73 (95% CI, 0.63–0.85; p < 0.0001). In patients achieving pCR, the HR for EFS was 0.49 (95% CI: 0.22-1.07; p = 0.07). In the arm level weighted analysis, higher pCR rates were associated with higher EFS rates, but the relationship was moderate but non significant (β = 0.35; p = 0.084; R² = 0.27). At the trial level, a modest and non‑significant association was observed between the RR for pCR and the HR for EFS (R² = 0.248, β = -1.01; p = 0.31). The association with OS was even weaker and imprecise (β = -1.98; SE = 2.19; p = 0.46), with limited explanatory power (R² = 0.29; adjusted R² = - 0.065), indicating no consistent relationship between improvements in pCR and OS across studies. Conclusions: The pooled analyses reveal a statistically significant improvement in both pCR and EFS for patients treated with ICI in combination with standard CT. However, across trial‑level analyses, the association between pCR and long‑term outcomes was consistently weak. Overall, the analyses consistently showed that gains in pCR did not reliably anticipate long term benefit. The weak and unstable associations with both EFS and OS suggest that pCR captures only a small portion of the treatment effect, underscoring its limited value as a surrogate endpoint in early TNBC.
Titolo: Valutazione della risposta patologica completa (pCR) come endpoint surrogato per la sopravvivenza nel carcinoma mammario triplo negativo (TNBC) precoce trattato con chemio‑immunoterapia neoadiuvante: una meta‑analisi di studi randomizzati controllati. Background: La risposta patologica completa (pCR) è riconosciuta dalle autorità regolatorie come un endpoint surrogato ragionevolmente predittivo degli esiti a lungo termine negli studi di terapia neoadiuvante condotti sul carcinoma mammario precoce ad alto rischio. Nel carcinoma mammario triplo negativo (TNBC), l’aggiunta degli inibitori dei checkpoint immunitari (ICI) alla chemioterapia neoadiuvante determina un incremento significativo dei tassi di pCR; tuttavia, la robustezza della pCR quale endpoint surrogato della sopravvivenza libera da eventi (EFS) o della sopravvivenza globale (OS) nel contesto della chemio-immunoterapia enoadiuvante rimane non definitivamente chiarita. Pertanto, una valutazione formale della validità surrogata della pCR risulta essenziale per orientare lo sviluppo clinico e il processo decisionale regolatorio. Metodi: È stata condotta una revisione sistematica della letteratura per identificare studi randomizzati controllati (RCT) che valutassero l’impiego neoadiuvante degli ICIs nel TNBC in stadio precoce. L’obiettivo primario era analizzare l’associazione a livello di studio tra pCR ed EFS e OS. La correlazione tra l’hazard ratio (HR) per EFS e OS e il rischio relativo (RR) di pCR tra gli studi è stata valutata mediante modelli di regressione lineare, stimando il coefficiente di determinazione (R²). Gli obiettivi secondari comprendevano l’analisi dell’eterogeneità degli effetti del trattamento in funzione di caratteristiche basali pre-specificate. Risultati: Sono stati inclusi sei RCT, per un totale complessivo di 4.035 pazienti. Complessivamente, l’aggiunta degli ICI alla chemioterapia neoadiuvante è stata associata a un aumento statisticamente significativo dei tassi di pCR, con un RR combinato pari a 1,17 (IC 95%: 1,10–1,25; p < 0,0001), nonché a un miglioramento dell’EFS, con un HR combinato di 0,73 (IC 95%: 0,63–0,85; p < 0,0001). Nei pazienti che hanno ottenuto una pCR, l’HR per EFS è risultato pari a 0,49 (IC 95%: 0,22–1,07; p = 0,07). Nell’analisi ponderata a livello di braccio di trattamento, tassi più elevati di pCR risultavano associati a tassi più elevati di EFS; tuttavia, tale associazione era di entità moderata e non statisticamente significativa (β = 0,35; p = 0,084; R² = 0,27). A livello di studio, è stata osservata un’associazione modesta e non significativa tra il RR di pCR e l’HR per EFS (R² = 0,248; β = −1,01; p = 0,31). L’associazione con l’OS è risultata ulteriormente più debole e imprecisa (β = −1,98; errore standard = 2,19; p = 0,46), con limitata capacità esplicativa (R² = 0,29; R² aggiustato = −0,065), indicando l’assenza di una relazione coerente tra il miglioramento della pCR e l’OS tra gli studi inclusi. Conclusioni: Le analisi aggregate dimostrano un miglioramento statisticamente significativo sia della pCR sia dell’EFS nei pazienti trattati con ICIs in combinazione con la chemioterapia standard. Tuttavia, le analisi a livello di studio evidenziano un’associazione sistematicamente debole tra pCR ed esiti clinici a lungo termine. Nel complesso, i risultati suggeriscono che l’incremento dei tassi di pCR non è in grado di predire in modo affidabile un beneficio a lungo termine. Le associazioni deboli e instabili osservate sia per EFS sia per OS indicano che la pCR cattura solo una quota limitata dell’effetto del trattamento, confermandone il valore limitato come endpoint surrogato nel TNBC in stadio precoce.
Valutazione della risposta patologica completa (pCR) come endpoint surrogato per la sopravvivenza nel carcinoma mammario triplo negativo (TNBC) precoce trattato con chemio‑immunoterapia neoadiuvante: una meta‑analisi di studi randomizzati controllati.
ZURLO, CHRISTIAN
2023/2024
Abstract
Title: Evaluating Pathologic Complete Response (pCR) as a Surrogate Endpoint for Survival in Early Triple Negative Breast Cancer (TNBC) Treated with Neoadjuvant Chemo-Immunotherapy: A Meta Analysis of Randomized Controlled Studies. Background: Pathologic complete response (pCR) is recognized by regulatory agencies as a surrogate endpoint reasonably likely to predict long term outcomes in neoadjuvant trials for high risk early breast cancer. In triple negative breast cancer (TNBC), the addition of immune checkpoint inhibitors (ICIs) to neoadjuvant chemotherapy significantly increases pCR rates; however, the strength of pCR as a surrogate for event free survival (EFS) or overall survival (OS) in the context of chemo immunotherapy remains uncertain. A formal evaluation of surrogacy is therefore essential to inform clinical development and regulatory decision making. Methods: A systematic literature search identified RCTs evaluating neoadjuvant ICIs in early TNBC. The primary objective was to assess the trial-level association between pCR and EFS and OS. The correlation between hazard ratio (HR) for EFS and OS and the relative risk (RR) for pCR across trials was analyzed using a linear regression model to determine the coefficient of determination (R²). Secondary objectives were to explore heterogeneity of results according to baseline features of interest. Results: Six RCTs, enrolling a total of 4,035 patients, were included. Across studies, the addition of ICIs resulted in a significant increase in pCR, with a pooled relative risk of 1.17 (95% CI, 1.10-1.25; p < 0.0001), and in EFS with a pooled hazard ratio for EFS was 0.73 (95% CI, 0.63–0.85; p < 0.0001). In patients achieving pCR, the HR for EFS was 0.49 (95% CI: 0.22-1.07; p = 0.07). In the arm level weighted analysis, higher pCR rates were associated with higher EFS rates, but the relationship was moderate but non significant (β = 0.35; p = 0.084; R² = 0.27). At the trial level, a modest and non‑significant association was observed between the RR for pCR and the HR for EFS (R² = 0.248, β = -1.01; p = 0.31). The association with OS was even weaker and imprecise (β = -1.98; SE = 2.19; p = 0.46), with limited explanatory power (R² = 0.29; adjusted R² = - 0.065), indicating no consistent relationship between improvements in pCR and OS across studies. Conclusions: The pooled analyses reveal a statistically significant improvement in both pCR and EFS for patients treated with ICI in combination with standard CT. However, across trial‑level analyses, the association between pCR and long‑term outcomes was consistently weak. Overall, the analyses consistently showed that gains in pCR did not reliably anticipate long term benefit. The weak and unstable associations with both EFS and OS suggest that pCR captures only a small portion of the treatment effect, underscoring its limited value as a surrogate endpoint in early TNBC.| File | Dimensione | Formato | |
|---|---|---|---|
|
Tesi Christian Zurlo.pdf
Accesso riservato
Dimensione
2.17 MB
Formato
Adobe PDF
|
2.17 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/103291