Statistics from Altmetric.com
In their commentary (Gut 2001;48:149–50), Bruix and Llovet discuss the paper by Bolondi et al (Gut 2001;48:251–9) and emphasise the fact that survival in patients with hepatocellular carcinoma (HCC) is mainly related to tumour stage and degree of liver function impairment at diagnosis. This is most likely true because of the peculiar features of HCC, which almost inevitably arises in the “minefield” of a cirrhotic liver whose residual function is one of the main factors influencing therapeutic options and prognosis.1
Nevertheless, a trend towards increased survival after diagnosis of HCC has recently been observed, although the surveillance programme has not changed over the years (liver ultrasonography and α-fetoprotein determination every six months). As Bruix and Llovet affirm, this increase in survival may be due to advances in diagnosis even in the absence of effective treatment, to the availability of multiple treatment, or both.
However, it must be emphasised that HCC stage (parameter of the tumour) and residual liver function (parameter of the affected patient) are closely related and influence each other, and that both can influence the choice of treatment and prognosis. Therefore, what should improved survival over the years be attributed to since surveillance programmes are only able to detect a minority of “early” HCCs?
Bolondi et al analysed the outcome and cost effectiveness of HCC surveillance programmes. They compared the outcome of a cohort of mixed aetiology cirrhotic patients screened by means of biannual liver ultrasonography and serum α-fetoprotein measurement to the outcome of patients whose HCC had been discovered incidentally. They found that there were no significant differences in eligibility for treatment between patients who had been under surveillance and those who had not (although a higher number of patients in the former group had been transplanted). However, survival at three years was significantly better in the group that had been kept under surveillance. Lastly, both liver function and tumour stage were selected in multivariate analysis as predictors of survival.
We recently performed a similar study in a cohort of hepatitis C virus positive cirrhotic patients. We compared clinical parameters, eligibility for treatment, and survival of patients whose HCC had been discovered during a surveillance programme (biannual liver ultrasonography and α-fetoprotein measurement) with patients whose HCC had been incidentally diagnosed.2 Although age, serum α-fetoprotein levels, and unifocality of the tumour were no different between the two subgroups of patients, we found that more patients in the group under surveillance were eligible for treatment (32/33 v 18/27; p=0.003, Fisher's exact test). Moreover, we found that clinical status at diagnosis was better in the group under surveillance compared with patients with an incidental diagnosis of HCC. Lastly, we observed that longer survival was obtained in treated patients, regardless of diagnosis modality or treatment modality. On the basis of these findings, we attempted to determine whether the longer survival observed in the group under surveillance might be due to better basal conditions, or perhaps they were more likely to benefit from treatment due to their improved clinical status. We thus compared patients treated with the same procedures and analysed the results on the basis of modality of diagnosis. We observed that there was no difference in survival between the groups, and that overall most deaths were liver related (72%) rather than tumour related. Both of these points suggested that the better outcome observed in the group under surveillance was due to the better basal conditions of the patients and not to the procedures themselves. Lastly, multivariate analysis showed that liver function, tumour stage, treatment, and HCC surveillance were independent predictors of better survival.
Thus what emerges from our study as well as from that of Bolondi et al's is that survival of HCC patients is mainly linked to preserved liver function. This probably allows patients to undergo treatment even when this is not classically considered “curative” as even therapeutic options considered “non-curative” have reportedly obtained increasingly positive results in terms of survival.3–5 In an era of multimodal therapeutic approaches to HCCs, these findings further support the results of screening programmes performed almost a decade ago on patients with compensated cirrhosis and whose sole options were liver surgery or percutaneous ethanol injections. No differences were reported regarding survival of patients who developed HCC and those who did not, thus emphasising the importance of residual liver function in relation to survival.6 Therefore, what probably lies beneath these findings is that improved medical therapy of the complications of liver cirrhosis, increased efficacy of HCC treatment, and better management of treatment induced sequelae have led to better care of the patients. This has likely changed both the type of patients who enter HCC surveillance studies and their therapeutic outcomes.