Predicting early recurrence after resection of initially unresectable colorectal liver metastases: the role of baseline and pre-surgery clinical, radiological and molecular factors in a real-life multicentre experience
R. Moretto,M.M. Germani,B. Borelli,V. Conca,D. Rossini,P. Boraschi,F. Donati,L. Urbani,S. Lonardi,F. Bergamo,K. Cerma,G. Ramondo,F.E. D’Amico,L. Salvatore,G. Valente,B. Barbaro,F. Giuliante,M. Di Maio,G. Masi,C. Cremolini
DOI: https://doi.org/10.1016/j.esmoop.2024.102991
IF: 6.883
2024-04-18
ESMO Open
Abstract:Early relapse (<6 months) after resection of colorectal liver metastases occurs in 35%-45% of liver-limited mCRC patients. Predictors of early relapse are needed to propose surgery to patients turned resectable after conversion chemotherapy. In this retrospective real-world analysis, no reliable biomarkers were found to build an accurate prognostic model. Novel biomarkers (i.e. ctDNA and radiomic features) are awaited to refine surgical selection after upfront chemotherapy. Advances in surgical techniques and systemic treatments have increased the likelihood of achieving radical surgery and long-term survival in metastatic colorectal cancer (mCRC) patients with initially unresectable colorectal liver metastases (CRLMs). Nonetheless, roughly half of the patients resected after an upfront systemic therapy experience disease relapse within 6 months from surgery, thus leading to the question whether surgery is actually beneficial for these patients. A real-world dataset of mCRC patients with initially unresectable liver-limited disease treated with conversion chemotherapy followed by radical resection of CRLMs at three high-volume Italian institutions was retrospectively assessed with the aim of investigating the association of baseline and pre-surgical clinical, radiological and molecular factors with the risk of relapse within 6 or 12 months from surgery. Overall, 268 patients were included in the analysis and 207 (77%) experienced recurrence. Ninety-six (46%) of them had disease relapse within 6 months after CRLM resection and in spite of several variables associated with early recurrence at univariate analyses, only primary tumour resection at diagnosis [odds ratio (OR) 0.53, 95% confidence interval (CI) 0.32-0.89, P = 0.02] remained significant in the multivariable model. Among patients with resected primary tumours, pN+ stage was associated with higher risk of disease relapse within 6 months (OR 3.02, 95% CI 1.23-7.41, P = 0.02). One hundred and forty-nine patients (72%) had disease relapse within 12 months after CRLMs resection but none of the analysed variables was independently associated with outcome. Clinical, radiological and molecular factors assessed before and after conversion chemotherapy do not reliably predict early recurrence after secondary resection of initially unresectable CRLMs. While novel markers are needed to optimize the cost/efficacy balance of surgical procedures, CRLM resection should be offered as soon as metastases become resectable during first-line chemotherapy to all patients eligible for surgery.
oncology