Heterogeneity of pathological tumour response in patients with multiple colorectal liver metastases undergoing resection after preoperative chemotherapy.

BACKGROUND

Pathological tumour response to preoperative chemotherapy is a key prognostic factor in patients with colorectal liver metastases (CRLM). Multiple CRLM may present different responses, but the prognostic significance of such heterogeneity remains uncertain. The present study aims to elucidate this issue.

PATIENTS AND

METHODS

We conducted a retrospective analysis of patients undergoing liver resection for multiple CRLM after neoadjuvant chemotherapy. We analysed up to five metastases per patient. Pathological response was assessed for each lesion using tumour regression grade (TRG) and the percentage of viable tumour cells (PVTCs). Overall survival (OS) and recurrence behaviour were compared between homogeneous and heterogeneous responders; in heterogeneous cases, prognosis was stratified according to the largest, best-responding, and worst-responding lesions.

RESULTS

Overall, 180 patients (705 nodules) were analysed. A heterogeneous response was observed in 45% of patients according to TRG and in 53% according to PVTC. Taken as a whole, heterogeneous TRG had 3-year OS comparable to homogeneous TRG4-5 patients (43.6% versus 43.4%) and heterogeneous PVTC had 3-year OS similar to homogeneous major responders (47.6% versus 47.4%). The recurrence analysis showed similar results. Heterogeneous TRG patients with at least one TRG1-2 metastasis had significantly better survival (56.4% versus TRG4-5 43.4%, P = 0.033), overlapping with the OS of homogeneous TRG1-2. Likewise, heterogeneous PVTC with at least one metastasis showing a complete response had significantly better OS, similar to homogeneous complete responders.

CONCLUSIONS

Heterogeneous pathological response is common in multiple CRLM. For a personalised treatment, lesion-specific assessment is essential since the presence of a single responder nodule is associated with a more favourable outcome.