Albumin-bilirubin versus Child-Pugh Grade as a Predictor of Survival after Individualised Hypofractionated Radiotherapy for Hepatocellular Carcinoma

Hong Kong J Radiol 2020 Sep;23(3):185-90

 

 

¹ Department of Clinical Oncology, Tuen Mun Hospital, Tuen Mun, Hong Kong

² Department of Clinical Oncology, The University of Hong Kong, Pokfulam, Hong Kong

³ Department of Clinical Oncology, The University of Hong Kong–Shenzhen Hospital, Shenzhen, China

⁴ Department of Clinical Oncology, Queen Elizabeth Hospital, Jordan, Hong Kong

 

  Full Article in PDF

Most patients with hepatocellular carcinoma (HCC) have compromised liver function due to pre-existing liver diseases. Treatment decisions depend very much on hepatic function as estimated by Child-Pugh (CP) grade.[1] However, its calculation involves subjective variables. Albumin-bilirubin (ALBI) grade is an objective test.[2] ALBI grade has recently been shown to be better than CP grade as a predictor of overall survival (OS) in patients with HCC undergoing surgical resection,[3] [4] transarterial chemoembolisation (TACE),[5] yttrium-90 radioembolisation,[5] sorafenib,[6] [7] or stereotactic ablative radiation therapy (SBAR).[8]

 

In our centre, we offer patients with HCC individualised hypofractionated radiotherapy (IHRT) with stereotactic body radiotherapy. IHRT is administered to patients with HCC who are not eligible for, or have failed, standard locoregional therapy. In contrast to SBAR, which delivers a high dose to a relatively small tumour in three to five fractions, IHRT delivers doses to HCC that are individualised to each patient according to the normal tissue constraints, rather than relative to size, in five to 10 fractions. This permits more patients with HCC to benefit from high-precision radiotherapy.

 

In a previous study, we investigated the predictive value of platelet-albumin-bilirubin, ALBI, and CP scores for patients with advanced HCC receiving radiotherapy.[9] However, to the best of our knowledge there are no studies that have investigated the prognostic value of ALBI grade in routine daily practice for patients with HCC treated with IHRT. The purpose of the present study was to compare the ability of CP versus ALBI grade to predict OS in patients with HCC with CP grade A5 or A6 treated with IHRT.

 

Data from consecutive patients who were treated between 2006 and 2015 with our institutional IHRT protocol were retrospectively extracted from medical records in Tuen Mun Hospital, Hong Kong. Those patients who fulfilled the inclusion criteria for this study were included for analysis. The diagnosis of HCC was established either by biopsy or by the American Association for the Study of Liver Diseases criteria. The eligibility criteria for this study were: unsuitable for, failed or refused standard local treatments (including resection, liver transplantation, local ablation therapies); a minimum of 700 mL of uninvolved liver; an Eastern Cooperative Oncology Group performance score ≤2; CP grade A5 or A6; and adequate liver function and complete blood counts. Patients with small-volume distant metastases were included. Patients were excluded if they had infiltrative disease or more than five tumour nodules. There was no limit on tumour size.

 

Different techniques of stereotactic body radiotherapy have been used in our institution over the years. Patients were immobilised with a customised device (Vac-Lok™; MEDTEC, Orange City [IA], United States). Computed tomography (CT) with multiphasic intravenous contrast was used to define gross tumour volume (GTV). Breath-hold CT or four-dimensional CT (average phase or respiratory phase sorted) was used to determine internal target volume and/or planning target volume. Motion management was done with maximum intensity projection, respiratory gating, active breathing control, or abdominal compression. IHRT was delivered using dynamic conformal arc, intensity-modulated radiotherapy, or volumetric arc therapy. Image guidance was done with an infrared body positioning system (ExacTrac™; BrainLab AG, Feldkirchen, Germany) or cone beam CT.

 

Our IHRT protocol divides patients into ‘favourable’ and ‘unfavourable’ groups depending on dose and fractionation. Individuals who meet all the following criteria are classified into the favourable group: tumour size ≤10 cm, Eastern Cooperative Oncology Group score 0-1, and liver volume minus GTV ≥700 mL. For 47 patients classified into the favourable group, 5-9 Gy for six fractions was prescribed; for the other 127 patients classified into the unfavourable group, 4 Gy for five to 10 fractions was prescribed. The dose was individualised by normal tissue constraints, as described previously.[9] The normal liver was allowed to receive a biological effective dose with α/β = 3 of 30 Gy₃ <40% and mean dose <28 Gy₃. Minor dose constraint violation was allowed in patients who were not carriers of hepatitis B or C and were without evidence of cirrhosis.

 

Baseline liver function data were obtained within 1 week prior to IHRT. The ALBI score was calculated using the formula: ALBI score = (log 10 bilirubin [μmol/L] × 0.66) + (-0.085 × albumin [g/L]). The ALBI scores are stratified into: ALBI grade A1 (≤-2.60); ALBI grade A2 (>-2.60 to ≤-1.39); and ALBI grade A3 (>-1.39).

 

We assessed the patients weekly during IHRT, every 6 weeks after treatment, every 3 months for the first 2 years, and every 4 months thereafter. Physical examinations and blood tests were done at every follow-up. Triphasic liver CT was performed every 3 months in the first and second year, and then every 6 months.

 

OS was calculated by the Kaplan-Meier curve. Log rank test was used to compare outcomes among survival curves for potential prognostic factors. Significant factors in the univariate analyses (p < 0.1) were included in the multivariable regression analyses using the Cox proportional hazards regression model. Factors with p < 0.05 were considered to be significant in the Cox model.

 

A total of 174 patients undergoing IHRT for HCC were included in this study. Patients with ALBI grade A3 were excluded as there were only three of them. The majority (86%) of the patients were male, and a majority (79%) of them were hepatitis B carriers. Most (74%) patients had Barcelona clinic liver cancer stage C disease. The median size of the tumour was 9.8 cm (range, 2.3-25.7 cm) and median GTV was approximately 440 mL (range, 9.2-4009 mL).

 

All patients in this study had CP grade A liver reserve; 63% had CP grade A5 and 37% had CP grade A6. The median ALBI score was -2.39 (range, -3.61 to -1.41). According to the ALBI grading system, 34.5% were classified as grade A1, 65.5% were grade A2.

 

Table 1 shows the correlations between CP and ALBI grades. Of patients with ALBI grade A1, 95% had CP grade A5. Of patients with ALBI grade A2, 46% had CP grade A5 and 54% had CP grade A6. Among patients with CP grade A5, 53% were stratified into ALBI grade A1 and 47% into ALBI grade A2. For patients with CP grade A6, 95% had ALBI grade A2.

 

Table 1. Correlations between Child-Pugh and albumin-bilirubin grade.

 

With a median follow-up of 23.5 months (range, 2.8-136.3 months), the median OS of the entire cohort was 24.1 months (95% confidence interval [CI]=19.7-28.4 months). The 1-year and 2-year OS were 74.6% (95% CI=71.3-77.9%) and 50.2% (95% CI=46.4%-54.0%), respectively.

 

Figures 1 and 2 show the OS of our patients in terms of ALBI and CP grades, respectively. Pretreatment ALBI grade stratified patients into distinct survival cohorts with a median OS of 34.7 months (95% CI=25.6-43.7 months) for grade A1 versus 19.2 months (95% CI=13.9-24.3 months) for grade A2 (p = 0.015). For CP classified patients, the median OS for CP grade A5 was 25.4 months (95% CI=20.4-30.4 months) and for CP grade A6 was 18.2 months (95% CI=14.0-24.2 months). The difference was not significantly different (p = 0.467).

 

Figure 1. Overall survival for albumin-bilirubin (ALBI) grade A1 versus A2 patients.

 

Figure 2. Overall survival for Child-Pugh (CP) grade A5 versus A6 patients.

 

In the univariate analysis for the OS, the ALBI grade, tumour size, number of lesions, presence of portal vein thrombosis, presence of metastases, and the total biological effective dose with α/β = 10 were significantly associated with OS (Table 2). CP score was not a significant factor.

 

Table 2. Univariate analysis of the factors for predicting overall survival.

 

In the multivariable analysis for OS, pretreatment ALBI score (p = 0.007) remained a significant independent factor in predicting OS. Other significant factors included: size of lesion (p = 0.014), number of lesions (p = 0.005), presence of portal vein thrombosis (p = 0.023), presence of metastasis (p = 0.002), and total dose of radiotherapy (BED with α/β = 10, p = 0.014; Table 3).

 

Table 3. Variables for multivariable analysis.

 

In this study, we investigated the CP versus ALBI grade in patients with HCC treated with IHRT. Our data showed that patients with CP grade A5 were stratified almost equally by ALBI into grades A1 and A2. For patients with CP grade A6, the majority were classified as ALBI grade A2. In contrast, nearly all patients with ALBI grade A1 had CP grade A5. Patients with ALBI grade A2 were roughly distributed equally between CP grades A5 and A6.

 

In the OS analysis, this study demonstrated that the median OS of patients with ALBI grade A1 (35 months) almost doubled that of patients with ALBI grade A2 (19 months). Conversely, the median OS difference between CP grade A5 (25 months) and CP grade A6 (18 months) was not significant. In multivariable regression analysis, ALBI was shown to be an independent predictor of OS, whereas CP was not. This indicates that ALBI grade is a significantly better predictor of OS than CP.

 

Our findings that ALBI is superior to CP in predicting survival are consistent with the current evidence of ALBI grade as a predictor of survival for other HCC treatments, including surgical resection,[3] [4] TACE, yttrium-90 radioembolisation,5 sorafenib,[6] [7] and SBAR.[8] In the study by Wang et al[4] that evaluated the prognostic performance of ALBI versus CP after liver resection and the study by Lo et al[8] on SBAR, both demonstrated no significant differences in OS between CP grades A5 and A6. Their findings were consistent with our data, although the patient and tumour characteristics, as well as treatments, were different. The result of this study may help to stratify patients with CP grade A5 into two distinct prognostic groups using ALBI. A prospective study will likely give a more definitive result for ALBI to be used in daily practice.

 

The present study had some limitations. The retrospective design of this study is prone to selection bias. Also, the different doses may have an effect on OS. We have strived to make this study a valuable one by including all consecutive eligible patients, which resulted in a relatively large sample size, a long follow-up period, and a long recruitment period of nearly a decade. Toxicities were not included in the analysis owing to the fact that grade ≥3 toxicities (prospectively collected) were very rare, and no radiation-induced liver disease occurred.

 

In conclusion, our findings suggest that ALBI grade is a better predictor of OS than the CP grade in patients with HCC who receive IHRT. A prospective study is recommended in order to validate the use of ALBI grade in daily practice.