PNPLA3 polymorphism worsens chemotherapy associated liver injury and affects overall survival in colorectal cancer patients with liver metastasis undergoing hepatic resection

Abstract

Background: Neoadjuvant chemotherapy (NAC) has been controversial for patients with colorecal cancer liver metastasis (CRLM) with resectable disease. Chemotherapy associated liver injury (CALI) may explain the lack of overall survival (OS) benefit in some studies. It remains unclear why CALI severity varies between patients despite receiving the same duration and type of NAC. Single nucleotide polymorphisms (SNPs) have been associated with liver disease and could account for these interindividual differences. Within this study we used the APRI + ALBI score, a non-invasive liver function marker that increases in response to CALI development during NAC, to assess whether preoperative liver function affects OS in CRLM patients undergoing hepatectomy and explore the impact of SNPs on CALI development.

Methods: 551 patients with CRLM undergoing liver surgery after NAC were included. In 149 patients, DNA from histological specimens was genotyped and the presence of SNPs associated with liver disease was assessed.

Findings: Patients with APRI + ALBI scores (≥-2.46), associated with the development of CALI, showed decreased OS after hepatectomy (median OS APRI + ALBI < -2.46 = 46.1 months; median OS APRI + ALBI ≥ -2.46 = 34.3, p = 0.027). The mutated rs738409 variant on the patatin-like phospholipase domain-containing protein 3 (PNPLA3) displayed a close association with chemotherapy-associated steatohepatitis (p = 0.007) as well as intrahepatic fat (p = 0.004). Additionally, all PNPLA3 homozygotes shifted into the high-risk APRI + ALBI group during NAC.

Interpretation: CALI and its effects on liver function not only impact immediate postoperative outcomes but also significantly affect OS in CRLM patients undergoing liver resection. PNPLA3 polymorphism was associated with CALI development. Considering that PNPLA3 polymorphisms are significantly higher in Asian populations, these results could partly explain the heterogeneity in reported effects of NAC in CRLM patients and might improve patient selection.

Funding: None of the authors received funding related to the writing of this manuscript.

Keywords: APRI + ALBI; CASH; Colorectal cancer; Liver metastases; PNPLA3.

Fig. 1

OS in CRLM patients after NAC and after exclusion of 90-day mortality. Patients are grouped according to their APRI + ALBI score. OS, overall survival; NAC, neoadjuvant chemotherapy; CRLM, colorectal liver metastases; APRI + ALBI score = summative combination of the aspartate aminotransferase (AST) to platelet ratio (APRI) and the albumin-bilirubin (ALBI) grade; No. = number; CI, confidence interval; p, p-value.

Fig. 2

(A) Percentage of steatosis, and (B) prevalence of CASH compared among genotypes of their respective single nucleotide polymorphisms. CASH, Chemotherapy associated steatohepatitis.

Fig. 3

The combined APRI + ALBI score was evaluated before (A) and after NAC (B). In addition, the delta between the time points was calculated (C). Furthermore, differences of frequency between PNPLA3 genotypes of patients above the high-risk cut-off of −2.46 were compared among groups before and after NAC (D). Additionally, APRI + ALBI scores post NAC were also compared between PNPLA3 genotypes in subgroups of patients receiving either Oxaliplatin or Irinotecan based NAC (E). APRI + ALBI score = summative combination of the aspartate aminotransferase (AST) to platelet ratio (APRI) and the albumin-bilirubin (ALBI) grade; NAC, neoadjuvant chemotherapy; HR, Hazard ratio; CI Confidence interval.