| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Centre Hépato-Biliaire, Hôpital Paul Brousse, Hôpitaux de Paris Université Paris-Sud, Villejuif, France.
Correspondence: Address correspondence to: René Adam, MD, PhD, Centre Hépato-Biliaire, Hôpital Paul Brousse, Hôpitaux de Paris Université Paris-Sud, Villejuif, France; Fax: 33-01-45-59-38-57; Email: rene.adam{at}pbr.ap-hop-paris.fr
Hepatic resection, the only treatment that offers long-term survival for patients with hepatocellular carcinoma (HCC), has shown significant improvement in results within the past decade.1–5
Although few would contest this well-established fact, the paradox is that no randomized study has clearly demonstrated the benefit of surgery. It is likely that such a study will not be conducted in the future because it will be unethical. On one hand, the disease has overall a dismal prognosis without surgery (<5% survival rate6) compared with 30% to 40% at 5 years after resection.7 On the other hand, the better results of surgery are sometimes argued for and attributed to the positive selection of surgical candidates rather than to the surgical procedure itself. The only way, therefore, to acquire evidence of the benefit of surgery is to compare two similar groups of patients and to demonstrate better outcome with hepatic resection.
One of the main factors of interest in the paper by Liu et al. is the comparison of survival outcomes between patients having surgery and those not having surgery who had tumors of similar sizes and comparable health status.7a By doing this, they show that surgical resection is a significant correlate of survival in the treatment of HCC.
The first point raised by these results, as usual in retrospective studies, is to eliminate any bias that could have favored the surgical groups. For the items evaluated in the study, it appears that patients having surgery were younger and had smaller tumors than patients who did not have surgery. Multivariate analysis, however, is aimed to restore the true influence of surgery.
More critically, a bias could also be related to unstudied items. As the authors themselves mentioned, the factors associated with the decision for surgery were unknown and the data from the registry did not provide information about possible comorbidity. In addition, other major prognostic factors affecting outcome were not evaluated; in particular, whether cirrhosis was present and, if so, its severity and that of vascular invasion. Given that (1) 80% to 90% of HCC developed on cirrhosis, (2) vascular invasion is common,8 and (3) the grade of liver insufficiency has a strong influence on outcome,9–11 it can be hypothesized that patients not having surgery could have been affected by the HCC developing on severe underlying cirrhosis, vascular invasion, or both, factors known to strongly affect survival outcome.
The lack of these data, therefore, is a major limit to this study in reaching its conclusion that surgery improves survival for patients with HCC. The second important point raised by the study is the accessibility of patients to surgical treatment, in relation to race or ethnicity. Interestingly, the authors show that only 30% of black patients received surgery, a proportion significantly lower than that of white, Asian, or Hispanic patients. As a consequence, black patients had a 53% increased risk of dying compared with white patients, a fact that could be interpreted as an indirect demonstration of the role of surgery. Black patients, however, also had a poorer survival rate than white and Asian patients following surgery. As stressed by the authors, race might, therefore, have been a proxy for other important variables (e.g., insurance, socioeconomic factors) not available in the data source.
In summary, although dealing with important problems, the study raised more questions than it brings responses. Although suggestive and logical, the data of the study do not provide a clear evidence-based argument for an improved survival with surgery for patients with HCC.
Should this lack of scientific evidence restrict the role of surgery in clinical practice? Probably not, because of unmentioned arguments suggesting the superiority of surgery over nonsurgical treatments. Because of its key role in the treatment of malignant tumors, surgery suffers from the difficulty in being compared with nonsurgical treatments in a randomized fashion, making its superiority difficult to demonstrate. Such difficulty can also concern nonsurgical treatments. For example, it has taken more than 15 years from the first randomized controlled trial12 to demonstrate that transarterial chemoembolization of HCC prolonged survival in selected patients.13 In the time interval, this has not impeded many groups from recommending such treatment based on the evidence that well-selected patients have clearly benefited from it. Our practice could not always be conducted on evidence-based data and careful use of available information from large cohort experience should guide a reasonable compromise based on common sense. Future studies, however, should be encouraged to bring accurate answers to the questions raised by the study. Ideally, these studies should be prospective to avoid the limiting effect of retrospective ones.14
Another major issue concerns the role of hepatic resection within the panel of other more recent therapies. What should be the respective indications for resection compared with radiofrequency?15 In what circumstances should transplantation be preferred to resection?16–19 Should resection be performed as a bridge to liver transplantation?20 Answers to these questions are as crucial as it is to know if, as logically expected, resection clearly prolongs survival.
Received for publication February 9, 2004. Accepted for publication March 4, 2004.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
