Z Gastroenterol 2019; 57(01): e43
DOI: 10.1055/s-0038-1677157
2. Clinical Hepatology, Surgery, LTX
Georg Thieme Verlag KG Stuttgart · New York

A simple clinical score to predict survival in patients with hepatocellular carcinoma

J Weiss
1   University Hospital Würzburg, Germany
,
B Datz
1   University Hospital Würzburg, Germany
,
L Kern
1   University Hospital Würzburg, Germany
,
T Kudlich
1   University Hospital Würzburg, Germany
,
M Pinter
2   University Hospital Vienna, Austria
,
M Rau
1   University Hospital Würzburg, Germany
,
O Götze
1   University Hospital Würzburg, Germany
,
I Klein
1   University Hospital Würzburg, Germany
,
T Michael
2   University Hospital Vienna, Austria
,
A Geier
1   University Hospital Würzburg, Germany
› Author Affiliations
Further Information

Publication History

Publication Date:
04 January 2019 (online)

 

Background:

Hepatocellular carcinoma (HCC) is the sixth most common cancer worldwide and the third cause of death due to cancer. To date, the Barcelona Clinic Liver Cancer (BCLC) system is the most commonly used system for staging and prognosis of HCC. We aimed to create a simple score to predict survival in patients with HCC.

Methods:

We retrospectively analysed data of subsequent patients of mostly Caucasian descent treated in the University Hospital Wuerzburg from 2005 to 2014 with proven HCC. We assessed the effect of different clinical and laboratory parameters at baseline on survival and included all statistically significant parameters in a Cox-regression model adjusted for prognostic factors like BCLC, Child-Pugh stage or age at diagnosis. Based on the remaining significant factors and odds ratios for survival, a clinical score was created (GOT (> 50 U/l; 1.5 points), INR (> 1.5; 2 points), Platelet count (> 300,000/µl; 3 points), and presence of Cirrhosis (2 points)) and survival was calculated for the different values. Results were validated in another cohort of HCC patients (University Hospital Vienna). For the Wuerzburg cohort we did a separate survival analysis for platelet count only.

Results:

299 HCC patients (mean age 66.0 years) were included. 83.3% were male, 72.2% had a proven diagnosis of liver cirrhosis. BCLC stages ranged from 0 to D (0: 3.7%, A: 25.8%, B: 40.5%, C: 24.1%, and D: 5.7%). Metastasis and portal vein invasion were observed in 19.7% of the patients. Treatment of HCC generally followed the BCLC algorithm (resection 27.8%, RFA 7.0%, transplant 4.7%, TACE 49.2%, SIRT 7.7% Sorafenib 33.1%). Median survival in the total cohort was 15.0 months (mean 36.3 ± 3.6 months). Patients with a score of 0 had a mean survival of 93.4 ± 15.3 months (score ≤2: 42.9 ± 6.0 months, score ≤4: 25.4 ± 3.1 months score 4: 8.4 ± 2.0 months). Differences between groups were significant. Validation in the Vienna cohort (n = 1050, mean age 62.8 years, 82.5% male, 91.5% cirrhosis) yielded comparable results with statistical significant differences between the different score groups (score 0: mean survival 40.2 ± 6.6 months, score ≤2: 33.9 ± 2.4 months, score ≤4: 28.6 ± 2.3 months, and score 4: 10.3 ± 1.5 months). Separate analysis of the platelet count in the Wuerzburg cohort showed a correlation to survival. Patients with a count > 300*103/mm3 hat the shortest survival with 5.4 months (median) compared to 29.8 months (median) in the group with a count of 150 – 300*103/mm3. Survival for patients with a count of < 50 was 21.4 months, for patients with a count of 50 – 150 15.9 months.

Conclusion:

A simple clinical score containing the parameters GOT, INR, platelet count and cirrhosis state could predict survival in patients with hepatocellular carcinoma. It seems to be driven substantially by the platelet count.