- Email: [email protected]
No treatment, resection and ethanol injection in hepatocellular carcinoma: a retrospective analysis of survival in 391 patients with cirrhosis
Journal of Hepntology 1995; 22: 522-526 Printed in Denmark . All rights reserved
Copyrtght 8 Journal of Heporology 1995
Journal of Hepatology ISSN 0168-8278
No treatment, resection and ethanol injection in hepatocellular carcinoma: a retrospective analysis of survival in 391 patients with cirrhosis Tito Livraghi’,
Luigi Bolondi2, Luigi Buscarini 3, Mario Cottone4, Alighieri Mazziotti5, Guido Torzilli7 and the Italian Cooperative HCC Study Group*
Albert0
Morabito6,
‘Serv. Radiologia, Hospital of Vimercate (Milano). 21st Clinica Medica, University of Bologna, 3Div. Medicina I, Hospital of Piacenza, 4Clinica Medica R. University of Palermo, 5Clinica Chirurgica, University of Bologna, 61st di Biometria, University of Milano, and ‘Div. Chirurgia Generale. Hospital of L.odi, Italy
This retrospective study was Background/Aims: undertaken to obtain information relevant to the therapeutic strategy in single hepatocellular carcinoma associated with Child’s A and B cirrhosis. Methods: From a total of 1108 consecutive patients with hepatocellular carcinoma, 391 patients with single, small (55 cm) hepatocellular carcinoma (260 in Child A class and 131 in Child B class) were observed: 120 were treated by surgical resection, 155 by percutaneous ethanol injection and 116 were untreated. The end point of the study was 3-year survival. The log rank test was used to compare survival among the different groups. Results: In the Child A group the cumulative 3-year survival was 79% for surgery, 71% for percutaneous ethanol injection and 26% for no treatment (p
comparable to the surgical group, i.e. potentially operable, survival was 80% for percutaneous ethanol injection and 30% for no treatment. In the Child B group the 3-year survival was 40% for surgery, 41% for percutaneous ethanol injection and 13% for no treatment QKO.01 for surgery vs no treatment andp
of hepatocellular carcinoma (HCC) associated with cirrhosis is based on the results of open studies. Surgery, percutaneous ethanol injection
(PEI) and liver transplantation (OLT) seem to play a role in prolonging survival of patients with this neoplasia (l-8). No controlled randomized studies have
T
E TREATMENT
Key words.- Liver tumors; Natural history; Percutaneous ethanol injection; Surgery. 0 Journal of Hepatology.
Received 30 December 1993
Correspondence: Tito Livraghi, M.D., Serv. di radiologia, Ospedale Civile, 20059 Vimercate, Milan, Italy. 8The following are the participating centers and members of the Italian Cooperative HCC Study Group: Clinica Medica I, University of Bologna: L. Barbara; Div. Chirurgia Gen. II, “Mauriziano” Hospital of Torino: H. Bouzari and L. Capussotti; Div. Chirurgia II, Hospital of Varese: A. Calvi; Div. Gastroenterologia, Hospital of San Giovanni Rotondo (FG): E. Caturelli; 1st di Medicina Interna, University of Milano: M. Colombo, M. Tommasini; Div. Oncologia Chirurgica A, 1st Naz. Tumori, Milano: R. Doci and L. Gennari; 1st Chir. Sperimentale, University of Milano: L. Fassati; Div. Medicina I, Hospital of Piacenza: E Fornari; 1st Radiologia, 1st Naz. Tumori, Milano: E Frigerio and E Garbagnati; Clinica Chirurgica, University of Bologna: G. Gozzetti; Div. Medicina, Hospital of Clusone (BG): S. Lazzaroni; Clinica Chirurgica, University of Verona: M. Marchiori and N. Nicoli; Div. Medicina II, Hospital of Padova: G. Marin; 1st Chirurgia Generale, University of Milano: M. Montorsi and G. Pezzuoli; Clinica Medica R, University of Palermo: F! Parisi; Clinica Medica, Catholic University of Roma: GL. Rapaccini; Serv. Radiologia, “Bellaria” Hospital of Bologna: S. Ricci and RA. Roversi; Div. Chirurgia II, “Niguarda” Hospital of Milano: E Romani; 1st Radiologia, University of Milano: l? Rossi; Div. Medicina Generale, “Fatebenefratelli” Hospital of Brescia: A. Salmi; 1st Radiologia, University of Verona: G. Taddei; Serv. Radiologia, Hospital of Vimercate (MI): C. Vettori.
522
Retrospective analysis of survival in HCC
been carried out on these procedures for different reasons (difficulties in recruiting large series of patients, difficulties in randomization, difficulties in obtaining the informed consent of the patients for such a study). To be evaluable, uncontrolled studies which compare the different procedures should be carried out in homogeneous patients, in large numbers of patients, and should have a control population in which the main variables associated with prognosis are comparable to the treated group. In many studies the stage of cirrhosis, and the size and the number of lesions have been shown to be the main, easily evaluable prognostic variables in the outcome of HCC (9-12). We therefore conducted a multicentric review of a consecutive large series of patients with HCC, focusing on the small (55 cm) single HCC, associated with either Child A or Child B cirrhosis and looking at the 3-year survival in the different subgroups of treatment and in the untreated patients.
Material and Methods Twenty-three centers took part in the study From January 1984 to January 1992, 1108 patients with HCC associated with cirrhosis were consecutively observed. The diagnosis of HCC was confirmed in all cases by biopsy or by alphafetoprotein assay >200 rig/ml, plus the presence of focal lesions by ultrasound and/or CT examination. The cirrhosis was posthepatitic in 455 patients (41.0%) 22% of whom were HBsAg+ and 78% HCV+, alcoholic in 292 (26.3%) and cryptogenetic or not investigated in 362 (32.7%); the majority of the patients with cryptogenetic cirrhosis were evaluated when no anti-HCV test was available. Eight hundred and fifty-six patients were men and 252 were women (ratio 3.4: 1); their mean age was 63.1 years (62.7 for men and 64.8 for women). Six hundred and sixty-four out of 1108 had a single lesion, whereas 444 had multiple or advanced and infiltrating lesions. The analysis was limited to the patients with single lesions and with Child A and B cirrhosis, i.e. the potentially treatable population. Fifty-six patients with Child C cirrhosis were thus excluded. Among patients with single lesions 484 lesions were 15 cm, whereas 180 were >5 cm. The mean follow up was 34.2 months. The distribution of different treatments is shown in Table 1. In patients with lesions 55 cm and Child A cirrhosis the indication was either PEI or surgery in 123 patients (82 operated on and 41 treated by PEI), according to the availability in the center of an expert in one of the two procedures. In 64 patients PEI was carried out because of contraindication to surgery (associated disease) or advanced age. One hundred and sixteen patients with Child A or B cirrhosis and a lesion 55 cm
TABLE 1 Distribution of different modalities of treatments according to the cirrhosis Child’s class in patients with single HCC HCC
Child
NT
Surg
PEI
TACE
OLT
Total
<5cm
A B C
73 43 15
82 38 2
105 50 13
30 16 1
4 8 4
294 155 35
>5cm
A B C
33 25 17
28 17 1
6 5 1
20 23 0
0 2 2
87 72 21
HCC=hepatocellular carcinoma; NT=not treatment; PEI=percutaneous ethanol injection; TACE=transarterial chemoembolization; OLT=orthotopic liver transplantation.
were untreated for the following reasons: patient refusal of any kind of treatment, or the decision of the attending physician, even if the patient was treatable (47 patients), or contraindication to surgery but unavailability of PEI in the center (69 patients). In patients with lesions 55 cm and Child B cirrhosis, surgery was indicated in 38 patients and PEI in 50 patients. The choice of treatment was related to the presence of an expert in the procedures. The control group (43 patients) consisted mostly of patients who refused any intervention because of the advanced state of their disease. The mean age by group was 65.8 for no treatment (NT), 61.2 for surgery and 63.7 for PEI. OLT (12 patients) or transcutaneous intrarterial chemoembolization (46 patients) were rare indications; therefore, because of the small number of patients submitted to these procedures, they were not included in the survival comparison. In patients with single lesions >5 cm the indication was surgical resection (45 patients) or transarterial chemoembolization (TACE) (43 patients), whereas PEI (11 patients) and OLT (2 patients) were rare indications. The control population consisted of 58 untreated patients. The number of patients, if subgrouped on the basis of Child class, was very small and therefore not evaluable. The final analysis of survival was thus limited to patients with HCC 55 cm in Child A and B cirrhosis in untreated patients and in patients treated by surgery or PEI (391 patients). The end-point of the study was the 3-year survival study. Statistical
analysis
Survival in the control group was calculated from the time the tumor was first diagnosed. In treated cases survival was calculated from the time of the first ethanol injection or from the time of tumor resection.
523
T. Livraghi et al.
The actuarial survival of the patients was calculated according to the method of Kaplan-Meier (13). The actuarial method was used because 43.6% of the patients were still living at the end of the study. The standard errors of the 3-year survival rate give a measure of the confidence interval of the estimates obtained. The log-rank test was used to test the survival differences (14). The p-values for the probabilities of rejection of the null hypothesis have been cited, despite the fact that many test repeats lower the power of the test of the hypothesis. Actuarial survival of treated and untreated patients is given as the median.
Three-year survival for untreated inoperable patients was 82, 51, and 17%. Table 3 gives the p-values for the comparison of survival among the treatment subgroups. There was a significant difference between surgery and no treatment @
Results
Discussion
Table 2 gives the survival rate and median survival of the NT group, surgical, and PEI groups of patients with HCCs 55 cm and Child A or B cirrhosis. In the PEI group of patients with Child A cirrhosis the 3-year survival was 97, 91 and 71%; in the surgical group the 3-year survival was 87, 82 and 79%, with a perioperative mortality of 2.5%, whereas in the NT group the 3year survival was 86,51 and 26%. In the NT group the 5-year survival rate was also available: 5% for the group as a whole, 11% for Child A and 0% for Child B, respectively. Fig. la, b, show the survival and confidence interval for patients with HCC 55 cm Child A in the surgical group and for those comparable, i.e. potentially operable, in the NT and PEI group. Threeyear survival for these patients was 97, 87, and 80% in the PEI group and 89, 5 1, and 30% in the NT group.
Among 1108 consecutive patients with HCC associated with cirrhosis, 701 were treated by different procedures and only in 391 was it possible to perform a survival analysis with comparison between subgroups. This result indicates how difficult it is to perform therapeutic studies large enough to be analyzable in HCC associated with cirrhosis. This difficulty is due to the variables which are associated with a different outcome (912). The main variables, however, which seem to influence prognosis are the status of cirrhosis and the size and number of the HCC. Therefore large studies on the different treatment procedures on patients stratified for these factors may allow conclusions to be drawn on the role of different procedures. This paper shows that in a large consecutive retrospective series surgery or PEI determined an approximately 75% 3-year survival of patients with small HCC and Child A cirrhosis and 40% in patients with small HCC and Child B cirrhosis. These survivals are clearly superior to a population of untreated patients of the same Child class. Although the untreated patients were a heterogeneous group, most of them had refused surgery and were therefore quite comparable. In fact, if analysis was limited to the potentially treatable patients, the difference in survival was better in the PEI and surgical groups than in the untreated patients. Studies on HCC treatment are generally uncontrolled because of difficulties in randomization to placebo. Most of the studies on HCC treatment are open, without a control population. To our knowledge, only two studies have been published using a control population (15,16). In both studies the control populations were matched for Child class and dimension, but the number of controls was very small. Our series of untreated patients in the Child A class is the largest reported in the literature. Many patients were potentially treatable and refused surgery; only a
TABLE 2 3-year survival of patients with single small HCC (~5) according to the different modalities of treatment and on the basis of Child class Child
Treatment
Survival
Median
1Y
2Y
3Y
A
NT Surgery PEI
86 87 97
51 82 91
26 79 71
26 67 52
B
NT Surgery PEI
65 86 86
22 60 66
13 40 41
17 26 30
Abbreviations as in Table 1.
TABLE 3 P-value of the subgroups considered in Table 2
Surgery vs NT PEI vs NT Abbreviations as in Table 1.
524
Child A
Child B
co.01
Retrospective analysis of survival in HCC
Fig. la. Kaplan-Meier survival curves and 95% confidence intervals at 1, 2 and 3 years between group of patients treated with percutaneous ethanol injection ( P EI) and patients untreated (NT). b: KaplanMeier survival curves and 95% confidence intervals at 1, 2 and 3 years between group of patients treated with surgery and untreated.
78 S U R V I V A L
80
% 25
0 I
I
,I
0
1
2
3
I
I
2
3
a
,
I
YEARS
75 S U R V I V A L
50
% 28 1'" SURGERY (82 pie) O I
b
0
small proportion of patients were ineligible for surgery because of associated disease which contraindicated the resection, and these patients were comparable to patients submitted to ethanol injection where the indication, in some cases, was ineligibility for surgery. The 3-year survival of patients treated by surgical resection in this series is in agreement with some of the literature results. These results vary, according to author, from 41% to 79% for an aggregate of more
~K NT (47 pte)
I
1 YEARS
than 700 patients with cirrhosis and single HCC <5 cm (16-24). The wide range is due chiefly to variability in the selection criteria, in the accuracy of preoperative staging and in the intraoperative mortality, which varies from 1.9 to 19%. The 3-year survival rate for PEItreated patients in this series is much closer to the literature results where the range, on an aggregate of 350 patients, varies from 63 to 68% in single HCC <5 cm with no mortality (5-7). 525
T. Livraghi et al.
It is important to underline that patients submitted to surgery or to PEI had the same survival in the two Child classes. In many cases PEI was performed in patients who were not considered suitable for surgery (age, associated disease), and therefore had a theoretically shorter survival; despite this drawback, this procedure gave the same survival as surgery These results seem to indicate that PEI is an alternative to surgical resection, as has been shown also by Castells et al. (25) and Kotoh et al. (26). In our opinion, the comparability of PEI and resection in their 3-year survival reflects a balancing between the greater radicality rate of surgery and the lesser damage of PEI. The planning of a controlled study to determine the factors that may differentiate the choice of treatment is therefore justified.
Acknowledgements The authors are grateful to BRACCO SPA. (Milano) for their support during this study
References 1. Ohsinishi K, Tanabe Y, Ryu M, Isono K, Yamamoto Y, Usui S, Hiyama Y, et al. Prognosis of hepatocellular carcinoma smaller than 5 cm in relation to treatment: study of 100 patients. Hepatology 1987; 7: 1285-90. 2. Franc0 D, Capussotti L, Smadja C, Bouzari H, Grance D, Delle Piane M. Resection of hepatocellular carcinoma. Results in 72 European patients with cirrhosis. Gastroenterology 1990; 98: 733-8. 3. Macintosh EL, Minuk GY. Hepatic resection in patients with cirrhosis and hepatocellular carcinoma. Surg Gynecol Obstet 1992; 174: 245-54. 4. Bismuth H, Chiche L, Adam R, Castaing D, Diamond T, Dennison A. Liver resection versus transplantation for hepatocellular carcinoma in cirrhotic patients. Ann Surg 1993; 218: 145-51. 5. Shiina S, Tagawa K, Unuma T, Fujino H, Uta Y, Niwa Y, Hata Y, et al. Percutaneous ethanol injection therapy of hepatocellular carcinoma: analysis of 77 patients. AJR 1990; 155: 1221-6. 6. Ebara M, Ohto M, Sogiura N, Kita K, Yoshikawa M, Okuda K, Kondo F, et al. Percutaneous ethanol injection for the treatment of small hepatocellular carcinoma. Gastroenterol Hepatol 1990; 5: 61626. 7. Livraghi T, Bolondi L, Lazzaroni S, Marin G, Morabito A, Rapaccini GL, Salmi A, et al. Percutaneous ethanol injection in the treatment of hepatocellular carcinoma: a study on 207 patients. Cancer 1992; 68: 926-9. 8. Vilana R, Bruix J, Bru C, Ayuso C, Sole M, Rodes J. Tumor size determines the efficacy of percutaneous ethanol injection for treatment of small hepatocellular carcinoma. Hepatology 1992; 16: 353-7. 9. Ricca RS, Arienti V, Nanni 0, Ugenti E Tassinari M, Camporesi C, Boriani K, et al. Prognostic factors and survival analysis in 135 Italian patients with hepatocellular carcinoma. J Hepatol 1992; 16: 166672.
526
10. Calvet X, Bruix J, Gines P Bru C, Sole M, Vilana R, Rodes J. Prognostic factors of hepatocellular carcinoma in the west: a multivariate analysis in 206 patients. Hepatology 1990; 12: 753-60. 11. Zarski JP, Lutz JM, Froissart B, Rachail M. Proposition for a survival prognostic classification in patients with primary liver cancer. J Hepatol 1989; 9: 249-57. 12 Cottone M, Virdone R, Fusco G, Orlando A, Turri M, Caltagirone M. Asymptomatic hepatocellular carcinoma in Child’s A cirrhosis. A comparison of natural history and treatment. Gastroenterology 1989; 96: 1566-71. 13. Kaplan EI, Meier I? Non-parametric estimation from incomplete observation. J Am Stat Assoc 1958; 53: 457-81. 14. Mantel N. Evaluation of survival data and two new rank order statistics arising in its considerations. Cancer Chemother Rep 1966; 50: 163-70. 15. Bruix J, Cirera I, Calvet X, Fuster J, Bru C, Ayuso C, Vilana R, et al. Surgical resection and survival in Western patients with hepatocellular carcinoma. J Hepatol 1992; 15: 350-5. 16. Tang ZJ, Yu YQ, Zhou XD, Ma ZC, Yang R, Lu JZ, Lin ZY, et al. Surgery of small hepato-cellular carcinoma. Cancer 1989; 64: 53641. 17. Kimani Y, Shigeki T, Miyazaki I. Hepatic resection for hepato-cellular carcinoma associated with liver cirrhosis. World J Surg 1986; 10: 294301. 18. Nagao T, Goto S, Kawamo N, Inoue S, Mizuta T, Morioka Y, Omori Y. Hepatic resection for hepatocellular carcinoma. Ann Surg 1987; 205: 3340. 19. Hsu HC, Wu TT, Sheu JC, Cee CS, Chen DS. Tumor invasiveness and prognosis in resected hepatocellular carcinoma. Cancer 1988; 61: 2095-99. 20. Takayasu K, Muramatsu Y, Moriyama N, Hasegawa H, Makuuchi M, Okazaki N, Hirohashi S, et al. Clinical and radiologic assessments of the results of hepatectomy for small hepatocellular carcinoma and therapeutic arterial embolization for postoperative recurrence. Cancer 1989; 64: 1848-52. 21. Makuuchi M, Takayama T, Kosuge T. The value of ultrasonography for hepatic surgery. Hepato-gastroenterol 1991; 38: 6470. 22. Paquet K, Kaussouris P Mercado MA, Kalk JF, Mutang D, Rambach W. Limited hepatic resection for selected cirrhotic patients with hepatocellular or cholangiocellular carcinoma: a prospective study. Br J Surg 1991; 78: 45962. 23. Gozzetti G, Belli L, Capussotti L, Di Carlo V, Gennari L, Maffei Faccioli A, et al. Liver resection for hepatocellular carcinoma in cirrhotic patients. Ital J Gastroenterol 1992; 24: 105-10. 24. Nagasue N, Kohno H, Chang Yu C, Hiroyuki T, Yamanoi A, Uchida M, Kineato T, et al. Liver resection for hepatocellular carcinoma. Results of 229 consecutive patients during 11 years. Ann Surg 1993; 217: 375-84. 25. Castells A, Bruix J, Bru C, Fuster J, Vilana R, Navasa M, Ayuso C. Treatment of small hepatocellular carcinoma in cirrhotic patients: a cohort study comparing surgical resection and ethanol injection. Hepatology 1993; 18: 11216. 26. Kotoh K, Sakai H, Sakamoto S, Nakayama S, Satoh M, Motoromi I, Nawata H. The effect of percutaneous ethanol injection on small solitary hepatocellular carcinoma is comparable to that of hepatectomy. Am J Gastroenterol 1994: 89: 1948.
Copyrtght 8 Journal of Heporology 1995
Journal of Hepatology ISSN 0168-8278
No treatment, resection and ethanol injection in hepatocellular carcinoma: a retrospective analysis of survival in 391 patients with cirrhosis Tito Livraghi’,
Luigi Bolondi2, Luigi Buscarini 3, Mario Cottone4, Alighieri Mazziotti5, Guido Torzilli7 and the Italian Cooperative HCC Study Group*
Albert0
Morabito6,
‘Serv. Radiologia, Hospital of Vimercate (Milano). 21st Clinica Medica, University of Bologna, 3Div. Medicina I, Hospital of Piacenza, 4Clinica Medica R. University of Palermo, 5Clinica Chirurgica, University of Bologna, 61st di Biometria, University of Milano, and ‘Div. Chirurgia Generale. Hospital of L.odi, Italy
This retrospective study was Background/Aims: undertaken to obtain information relevant to the therapeutic strategy in single hepatocellular carcinoma associated with Child’s A and B cirrhosis. Methods: From a total of 1108 consecutive patients with hepatocellular carcinoma, 391 patients with single, small (55 cm) hepatocellular carcinoma (260 in Child A class and 131 in Child B class) were observed: 120 were treated by surgical resection, 155 by percutaneous ethanol injection and 116 were untreated. The end point of the study was 3-year survival. The log rank test was used to compare survival among the different groups. Results: In the Child A group the cumulative 3-year survival was 79% for surgery, 71% for percutaneous ethanol injection and 26% for no treatment (p
comparable to the surgical group, i.e. potentially operable, survival was 80% for percutaneous ethanol injection and 30% for no treatment. In the Child B group the 3-year survival was 40% for surgery, 41% for percutaneous ethanol injection and 13% for no treatment QKO.01 for surgery vs no treatment andp
of hepatocellular carcinoma (HCC) associated with cirrhosis is based on the results of open studies. Surgery, percutaneous ethanol injection
(PEI) and liver transplantation (OLT) seem to play a role in prolonging survival of patients with this neoplasia (l-8). No controlled randomized studies have
T
E TREATMENT
Key words.- Liver tumors; Natural history; Percutaneous ethanol injection; Surgery. 0 Journal of Hepatology.
Received 30 December 1993
Correspondence: Tito Livraghi, M.D., Serv. di radiologia, Ospedale Civile, 20059 Vimercate, Milan, Italy. 8The following are the participating centers and members of the Italian Cooperative HCC Study Group: Clinica Medica I, University of Bologna: L. Barbara; Div. Chirurgia Gen. II, “Mauriziano” Hospital of Torino: H. Bouzari and L. Capussotti; Div. Chirurgia II, Hospital of Varese: A. Calvi; Div. Gastroenterologia, Hospital of San Giovanni Rotondo (FG): E. Caturelli; 1st di Medicina Interna, University of Milano: M. Colombo, M. Tommasini; Div. Oncologia Chirurgica A, 1st Naz. Tumori, Milano: R. Doci and L. Gennari; 1st Chir. Sperimentale, University of Milano: L. Fassati; Div. Medicina I, Hospital of Piacenza: E Fornari; 1st Radiologia, 1st Naz. Tumori, Milano: E Frigerio and E Garbagnati; Clinica Chirurgica, University of Bologna: G. Gozzetti; Div. Medicina, Hospital of Clusone (BG): S. Lazzaroni; Clinica Chirurgica, University of Verona: M. Marchiori and N. Nicoli; Div. Medicina II, Hospital of Padova: G. Marin; 1st Chirurgia Generale, University of Milano: M. Montorsi and G. Pezzuoli; Clinica Medica R, University of Palermo: F! Parisi; Clinica Medica, Catholic University of Roma: GL. Rapaccini; Serv. Radiologia, “Bellaria” Hospital of Bologna: S. Ricci and RA. Roversi; Div. Chirurgia II, “Niguarda” Hospital of Milano: E Romani; 1st Radiologia, University of Milano: l? Rossi; Div. Medicina Generale, “Fatebenefratelli” Hospital of Brescia: A. Salmi; 1st Radiologia, University of Verona: G. Taddei; Serv. Radiologia, Hospital of Vimercate (MI): C. Vettori.
522
Retrospective analysis of survival in HCC
been carried out on these procedures for different reasons (difficulties in recruiting large series of patients, difficulties in randomization, difficulties in obtaining the informed consent of the patients for such a study). To be evaluable, uncontrolled studies which compare the different procedures should be carried out in homogeneous patients, in large numbers of patients, and should have a control population in which the main variables associated with prognosis are comparable to the treated group. In many studies the stage of cirrhosis, and the size and the number of lesions have been shown to be the main, easily evaluable prognostic variables in the outcome of HCC (9-12). We therefore conducted a multicentric review of a consecutive large series of patients with HCC, focusing on the small (55 cm) single HCC, associated with either Child A or Child B cirrhosis and looking at the 3-year survival in the different subgroups of treatment and in the untreated patients.
Material and Methods Twenty-three centers took part in the study From January 1984 to January 1992, 1108 patients with HCC associated with cirrhosis were consecutively observed. The diagnosis of HCC was confirmed in all cases by biopsy or by alphafetoprotein assay >200 rig/ml, plus the presence of focal lesions by ultrasound and/or CT examination. The cirrhosis was posthepatitic in 455 patients (41.0%) 22% of whom were HBsAg+ and 78% HCV+, alcoholic in 292 (26.3%) and cryptogenetic or not investigated in 362 (32.7%); the majority of the patients with cryptogenetic cirrhosis were evaluated when no anti-HCV test was available. Eight hundred and fifty-six patients were men and 252 were women (ratio 3.4: 1); their mean age was 63.1 years (62.7 for men and 64.8 for women). Six hundred and sixty-four out of 1108 had a single lesion, whereas 444 had multiple or advanced and infiltrating lesions. The analysis was limited to the patients with single lesions and with Child A and B cirrhosis, i.e. the potentially treatable population. Fifty-six patients with Child C cirrhosis were thus excluded. Among patients with single lesions 484 lesions were 15 cm, whereas 180 were >5 cm. The mean follow up was 34.2 months. The distribution of different treatments is shown in Table 1. In patients with lesions 55 cm and Child A cirrhosis the indication was either PEI or surgery in 123 patients (82 operated on and 41 treated by PEI), according to the availability in the center of an expert in one of the two procedures. In 64 patients PEI was carried out because of contraindication to surgery (associated disease) or advanced age. One hundred and sixteen patients with Child A or B cirrhosis and a lesion 55 cm
TABLE 1 Distribution of different modalities of treatments according to the cirrhosis Child’s class in patients with single HCC HCC
Child
NT
Surg
PEI
TACE
OLT
Total
<5cm
A B C
73 43 15
82 38 2
105 50 13
30 16 1
4 8 4
294 155 35
>5cm
A B C
33 25 17
28 17 1
6 5 1
20 23 0
0 2 2
87 72 21
HCC=hepatocellular carcinoma; NT=not treatment; PEI=percutaneous ethanol injection; TACE=transarterial chemoembolization; OLT=orthotopic liver transplantation.
were untreated for the following reasons: patient refusal of any kind of treatment, or the decision of the attending physician, even if the patient was treatable (47 patients), or contraindication to surgery but unavailability of PEI in the center (69 patients). In patients with lesions 55 cm and Child B cirrhosis, surgery was indicated in 38 patients and PEI in 50 patients. The choice of treatment was related to the presence of an expert in the procedures. The control group (43 patients) consisted mostly of patients who refused any intervention because of the advanced state of their disease. The mean age by group was 65.8 for no treatment (NT), 61.2 for surgery and 63.7 for PEI. OLT (12 patients) or transcutaneous intrarterial chemoembolization (46 patients) were rare indications; therefore, because of the small number of patients submitted to these procedures, they were not included in the survival comparison. In patients with single lesions >5 cm the indication was surgical resection (45 patients) or transarterial chemoembolization (TACE) (43 patients), whereas PEI (11 patients) and OLT (2 patients) were rare indications. The control population consisted of 58 untreated patients. The number of patients, if subgrouped on the basis of Child class, was very small and therefore not evaluable. The final analysis of survival was thus limited to patients with HCC 55 cm in Child A and B cirrhosis in untreated patients and in patients treated by surgery or PEI (391 patients). The end-point of the study was the 3-year survival study. Statistical
analysis
Survival in the control group was calculated from the time the tumor was first diagnosed. In treated cases survival was calculated from the time of the first ethanol injection or from the time of tumor resection.
523
T. Livraghi et al.
The actuarial survival of the patients was calculated according to the method of Kaplan-Meier (13). The actuarial method was used because 43.6% of the patients were still living at the end of the study. The standard errors of the 3-year survival rate give a measure of the confidence interval of the estimates obtained. The log-rank test was used to test the survival differences (14). The p-values for the probabilities of rejection of the null hypothesis have been cited, despite the fact that many test repeats lower the power of the test of the hypothesis. Actuarial survival of treated and untreated patients is given as the median.
Three-year survival for untreated inoperable patients was 82, 51, and 17%. Table 3 gives the p-values for the comparison of survival among the treatment subgroups. There was a significant difference between surgery and no treatment @
Results
Discussion
Table 2 gives the survival rate and median survival of the NT group, surgical, and PEI groups of patients with HCCs 55 cm and Child A or B cirrhosis. In the PEI group of patients with Child A cirrhosis the 3-year survival was 97, 91 and 71%; in the surgical group the 3-year survival was 87, 82 and 79%, with a perioperative mortality of 2.5%, whereas in the NT group the 3year survival was 86,51 and 26%. In the NT group the 5-year survival rate was also available: 5% for the group as a whole, 11% for Child A and 0% for Child B, respectively. Fig. la, b, show the survival and confidence interval for patients with HCC 55 cm Child A in the surgical group and for those comparable, i.e. potentially operable, in the NT and PEI group. Threeyear survival for these patients was 97, 87, and 80% in the PEI group and 89, 5 1, and 30% in the NT group.
Among 1108 consecutive patients with HCC associated with cirrhosis, 701 were treated by different procedures and only in 391 was it possible to perform a survival analysis with comparison between subgroups. This result indicates how difficult it is to perform therapeutic studies large enough to be analyzable in HCC associated with cirrhosis. This difficulty is due to the variables which are associated with a different outcome (912). The main variables, however, which seem to influence prognosis are the status of cirrhosis and the size and number of the HCC. Therefore large studies on the different treatment procedures on patients stratified for these factors may allow conclusions to be drawn on the role of different procedures. This paper shows that in a large consecutive retrospective series surgery or PEI determined an approximately 75% 3-year survival of patients with small HCC and Child A cirrhosis and 40% in patients with small HCC and Child B cirrhosis. These survivals are clearly superior to a population of untreated patients of the same Child class. Although the untreated patients were a heterogeneous group, most of them had refused surgery and were therefore quite comparable. In fact, if analysis was limited to the potentially treatable patients, the difference in survival was better in the PEI and surgical groups than in the untreated patients. Studies on HCC treatment are generally uncontrolled because of difficulties in randomization to placebo. Most of the studies on HCC treatment are open, without a control population. To our knowledge, only two studies have been published using a control population (15,16). In both studies the control populations were matched for Child class and dimension, but the number of controls was very small. Our series of untreated patients in the Child A class is the largest reported in the literature. Many patients were potentially treatable and refused surgery; only a
TABLE 2 3-year survival of patients with single small HCC (~5) according to the different modalities of treatment and on the basis of Child class Child
Treatment
Survival
Median
1Y
2Y
3Y
A
NT Surgery PEI
86 87 97
51 82 91
26 79 71
26 67 52
B
NT Surgery PEI
65 86 86
22 60 66
13 40 41
17 26 30
Abbreviations as in Table 1.
TABLE 3 P-value of the subgroups considered in Table 2
Surgery vs NT PEI vs NT Abbreviations as in Table 1.
524
Child A
Child B
co.01
Retrospective analysis of survival in HCC
Fig. la. Kaplan-Meier survival curves and 95% confidence intervals at 1, 2 and 3 years between group of patients treated with percutaneous ethanol injection ( P EI) and patients untreated (NT). b: KaplanMeier survival curves and 95% confidence intervals at 1, 2 and 3 years between group of patients treated with surgery and untreated.
78 S U R V I V A L
80
% 25
0 I
I
,I
0
1
2
3
I
I
2
3
a
,
I
YEARS
75 S U R V I V A L
50
% 28 1'" SURGERY (82 pie) O I
b
0
small proportion of patients were ineligible for surgery because of associated disease which contraindicated the resection, and these patients were comparable to patients submitted to ethanol injection where the indication, in some cases, was ineligibility for surgery. The 3-year survival of patients treated by surgical resection in this series is in agreement with some of the literature results. These results vary, according to author, from 41% to 79% for an aggregate of more
~K NT (47 pte)
I
1 YEARS
than 700 patients with cirrhosis and single HCC <5 cm (16-24). The wide range is due chiefly to variability in the selection criteria, in the accuracy of preoperative staging and in the intraoperative mortality, which varies from 1.9 to 19%. The 3-year survival rate for PEItreated patients in this series is much closer to the literature results where the range, on an aggregate of 350 patients, varies from 63 to 68% in single HCC <5 cm with no mortality (5-7). 525
T. Livraghi et al.
It is important to underline that patients submitted to surgery or to PEI had the same survival in the two Child classes. In many cases PEI was performed in patients who were not considered suitable for surgery (age, associated disease), and therefore had a theoretically shorter survival; despite this drawback, this procedure gave the same survival as surgery These results seem to indicate that PEI is an alternative to surgical resection, as has been shown also by Castells et al. (25) and Kotoh et al. (26). In our opinion, the comparability of PEI and resection in their 3-year survival reflects a balancing between the greater radicality rate of surgery and the lesser damage of PEI. The planning of a controlled study to determine the factors that may differentiate the choice of treatment is therefore justified.
Acknowledgements The authors are grateful to BRACCO SPA. (Milano) for their support during this study
References 1. Ohsinishi K, Tanabe Y, Ryu M, Isono K, Yamamoto Y, Usui S, Hiyama Y, et al. Prognosis of hepatocellular carcinoma smaller than 5 cm in relation to treatment: study of 100 patients. Hepatology 1987; 7: 1285-90. 2. Franc0 D, Capussotti L, Smadja C, Bouzari H, Grance D, Delle Piane M. Resection of hepatocellular carcinoma. Results in 72 European patients with cirrhosis. Gastroenterology 1990; 98: 733-8. 3. Macintosh EL, Minuk GY. Hepatic resection in patients with cirrhosis and hepatocellular carcinoma. Surg Gynecol Obstet 1992; 174: 245-54. 4. Bismuth H, Chiche L, Adam R, Castaing D, Diamond T, Dennison A. Liver resection versus transplantation for hepatocellular carcinoma in cirrhotic patients. Ann Surg 1993; 218: 145-51. 5. Shiina S, Tagawa K, Unuma T, Fujino H, Uta Y, Niwa Y, Hata Y, et al. Percutaneous ethanol injection therapy of hepatocellular carcinoma: analysis of 77 patients. AJR 1990; 155: 1221-6. 6. Ebara M, Ohto M, Sogiura N, Kita K, Yoshikawa M, Okuda K, Kondo F, et al. Percutaneous ethanol injection for the treatment of small hepatocellular carcinoma. Gastroenterol Hepatol 1990; 5: 61626. 7. Livraghi T, Bolondi L, Lazzaroni S, Marin G, Morabito A, Rapaccini GL, Salmi A, et al. Percutaneous ethanol injection in the treatment of hepatocellular carcinoma: a study on 207 patients. Cancer 1992; 68: 926-9. 8. Vilana R, Bruix J, Bru C, Ayuso C, Sole M, Rodes J. Tumor size determines the efficacy of percutaneous ethanol injection for treatment of small hepatocellular carcinoma. Hepatology 1992; 16: 353-7. 9. Ricca RS, Arienti V, Nanni 0, Ugenti E Tassinari M, Camporesi C, Boriani K, et al. Prognostic factors and survival analysis in 135 Italian patients with hepatocellular carcinoma. J Hepatol 1992; 16: 166672.
526
10. Calvet X, Bruix J, Gines P Bru C, Sole M, Vilana R, Rodes J. Prognostic factors of hepatocellular carcinoma in the west: a multivariate analysis in 206 patients. Hepatology 1990; 12: 753-60. 11. Zarski JP, Lutz JM, Froissart B, Rachail M. Proposition for a survival prognostic classification in patients with primary liver cancer. J Hepatol 1989; 9: 249-57. 12 Cottone M, Virdone R, Fusco G, Orlando A, Turri M, Caltagirone M. Asymptomatic hepatocellular carcinoma in Child’s A cirrhosis. A comparison of natural history and treatment. Gastroenterology 1989; 96: 1566-71. 13. Kaplan EI, Meier I? Non-parametric estimation from incomplete observation. J Am Stat Assoc 1958; 53: 457-81. 14. Mantel N. Evaluation of survival data and two new rank order statistics arising in its considerations. Cancer Chemother Rep 1966; 50: 163-70. 15. Bruix J, Cirera I, Calvet X, Fuster J, Bru C, Ayuso C, Vilana R, et al. Surgical resection and survival in Western patients with hepatocellular carcinoma. J Hepatol 1992; 15: 350-5. 16. Tang ZJ, Yu YQ, Zhou XD, Ma ZC, Yang R, Lu JZ, Lin ZY, et al. Surgery of small hepato-cellular carcinoma. Cancer 1989; 64: 53641. 17. Kimani Y, Shigeki T, Miyazaki I. Hepatic resection for hepato-cellular carcinoma associated with liver cirrhosis. World J Surg 1986; 10: 294301. 18. Nagao T, Goto S, Kawamo N, Inoue S, Mizuta T, Morioka Y, Omori Y. Hepatic resection for hepatocellular carcinoma. Ann Surg 1987; 205: 3340. 19. Hsu HC, Wu TT, Sheu JC, Cee CS, Chen DS. Tumor invasiveness and prognosis in resected hepatocellular carcinoma. Cancer 1988; 61: 2095-99. 20. Takayasu K, Muramatsu Y, Moriyama N, Hasegawa H, Makuuchi M, Okazaki N, Hirohashi S, et al. Clinical and radiologic assessments of the results of hepatectomy for small hepatocellular carcinoma and therapeutic arterial embolization for postoperative recurrence. Cancer 1989; 64: 1848-52. 21. Makuuchi M, Takayama T, Kosuge T. The value of ultrasonography for hepatic surgery. Hepato-gastroenterol 1991; 38: 6470. 22. Paquet K, Kaussouris P Mercado MA, Kalk JF, Mutang D, Rambach W. Limited hepatic resection for selected cirrhotic patients with hepatocellular or cholangiocellular carcinoma: a prospective study. Br J Surg 1991; 78: 45962. 23. Gozzetti G, Belli L, Capussotti L, Di Carlo V, Gennari L, Maffei Faccioli A, et al. Liver resection for hepatocellular carcinoma in cirrhotic patients. Ital J Gastroenterol 1992; 24: 105-10. 24. Nagasue N, Kohno H, Chang Yu C, Hiroyuki T, Yamanoi A, Uchida M, Kineato T, et al. Liver resection for hepatocellular carcinoma. Results of 229 consecutive patients during 11 years. Ann Surg 1993; 217: 375-84. 25. Castells A, Bruix J, Bru C, Fuster J, Vilana R, Navasa M, Ayuso C. Treatment of small hepatocellular carcinoma in cirrhotic patients: a cohort study comparing surgical resection and ethanol injection. Hepatology 1993; 18: 11216. 26. Kotoh K, Sakai H, Sakamoto S, Nakayama S, Satoh M, Motoromi I, Nawata H. The effect of percutaneous ethanol injection on small solitary hepatocellular carcinoma is comparable to that of hepatectomy. Am J Gastroenterol 1994: 89: 1948.