Valutazione della velocità di progressione di malattia come biomarker e sviluppo di nomogramma clinico per predire l'outcome alla seconda di linea di trattamento nei pazienti affetti da tumore del colon-retto metastatico trattati con bevacizumab beyond progression o aflibercept

Backgroud In metastatic colorectal cancer (mCRC) there is the unmeet clinical need to predict outcome to 2nd line therapy. There aren't yet standardized prognostic or predictive parameters of outcome to the second line treatment that can predict at progression which are patients who can most benefit from systemic treatment next to the first line in mCRC patients treated with bevacizumab beyond progression or aflibercept. The aim of our study was to test if the velocity rate of progressive disease (VRPD) may be a new dynamic marker, can predict outcome in terms of 2nd line progression free survival (PFS) in mCRC patients (pts). Patients and Methods We enrolled 167 pts treated with bevacizumab beyond progression or aflibercept in 2nd line. For each patient, VRPD was calculated as the ratio between the sum of the longest diameters (SLD) for target lesions evaluated by CT scan, according to RECIST criteria v 1.1, at the time of progression and the sum of the longest diameters (SLD) for target lesions evaluated by CT scan before progression multiplied for difference in months. For VRPD cut-off point estimation a mixture model of two Gaussian distributions was used. The optimal cutoff is determined as the value where the probability density functions of the mixing distribution coincide. Results We enrolled 167 mCRC pts treated with bevacizumab beyond progression or aflibercept in second line. In second line 63% of pts were treated with bevacizumab plus doublet, 12% with bevacizumab in association with triplet and 25% of patients received aflibercept Folfiri. A cut-off value of 1.19 was unbiasedly generated using a mixture model of two Gaussian distributions. This cut-off discriminated pts with low (VRPD <1.19; 91% of pts) vs high (VRPD >1.19; 9% of pts) VRPD. At univariate analysis, high VRPD pts had a non-significant longer PFS than low VRPD pts (HR: 0,59 [95%CI: 0.34-1.03], p = 0.06). We collected data on 22 variables, which were chosen based on the previous literature. All the 22 available variables, including VRPD (considered either as a continuous or as a dichotomized variable) were included in the multivariate analysis to determine which factors could better predict PFS in second line. Six variables were selected and used to build a Nomogram. These six variables were: gender (0 = F; 1 = M); primary tumour location (0 = right; 1 = left), baseline Performance Status (PS), presence of liver metastases (0 = no; 1 = yes), first line PFS in weeks, sum of the longest diameters at time of progression (SLD0). Conclusion VRPD is a parameter never tested in metastatic colorectal cancer. Sex, primary tumor location, baseline PS, presence of liver metastases, first line PFS and sum of the longest diameters at time of progression, resulted as prognostic factors for determine second line PFS and were included in our nomogram. Validation of our nomogram using an external cohort is ongoing