High inter-patient variability and high spatial heterogeneity are features of colorectal cancer (CRC). This may influence the molecular characterization of tumor tissue, now mandatory for patients with metastatic CRC who are candidates for treatment with an anti-EGFR mAb, as false-negative results can occur, leading to non optimal therapy. Moreover, temporal molecular heterogeneity during treatment is known to influence the response to therapy and prognosis. We present a literature overview of advances made in characterizing molecular heterogeneity in CRC, underlining that the analysis of liquid biopsy could represent an efficient non-invasive tool to overcome the problem. We believe that understanding CRC heterogeneity is fundamental for a more accurate diagnosis, for selecting the best targets to ensure prolonged antitumor response, and for monitoring minimal residual disease and the onset of resistance to therapy, all essential components of successful personalized treatment.

BackgroundAlthough non muscle invasive bladder cancer (NMIBC) generally has a good long-term prognosis, up to 80% of patients will nevertheless experience local recurrence after the primary tumor resection. The search for markers capable of accurately identifying patients at high risk of recurrence is ongoing. We retrospectively evaluated the methylation status of a panel of 24 tumor suppressor genes (TIMP3, APC, CDKN2A, MLH1, ATM, RARB, CDKN2B, HIC1, CHFR, BRCA1, CASP8, CDKN1B, PTEN, BRCA2, CD44, RASSF1, DAPK1, FHIT, VHL, ESR1, TP73, IGSF4, GSTP1 and CDH13) in primary lesions to obtain information about their role in predicting local recurrence in NMIBC.MethodsFormaldehyde-fixed paraffin-embedded (FFPE) samples from 74 patients operated on for bladder cancer were analyzed by methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA): 36 patients had relapsed and 38 were disease-free at the 5-year follow up. Methylation status was considered as a dichotomous variable and genes showing methylation ≥20% were defined as “positive”.ResultsMethylation frequencies were higher in non recurring than recurring tumors. A statistically significant difference was observed for HIC1 (P = 0.03), GSTP1 (P = 0.02) and RASSF1 (P = 0.03). The combination of the three genes showed 78% sensitivity and 66% specificity in identifying recurrent patients, with an overall accuracy of 72%.ConclusionsOur preliminary data suggest a potential role of HIC1, GSTP1 and RASSF1 in predicting local recurrence in NMIBC. Such information could help clinicians to identify patients at high risk of recurrence who require close monitoring during follow up.

No abstract