Predictive value of indocyanine green retention rate with respect to complications of radiofrequency ablation in 878 patients with hepatocellular carcinoma

Abstract

Background and aims: Radiofrequency ablation (RFA) is a minimally invasive technique used for the treatment of hepatocellular carcinoma (HCC). It may produce complications. The indocyanine green (ICG) retention rate at 15 min (ICGR15) has been used to predict complications after hepatectomy. In this study, the prediction of the value of ICGR15 for complications of RFA to the patients with HCC was evaluated.

Methods: Some 878 cases of HCC treated between June 2009 and June 2013 were evaluated. All patients were treated by percutaneous radiofrequency ablation. Patients were divided into two groups: a complication group (85 cases) and a complication-free group (793 cases). ICGR15 and other baseline characteristics of the two groups were compared. A logistic regression model was used to analyse the merits of assessing liver reserve to predict complications post-RFA.

Results: Complications such as intra-abdominal haemorrhage and pleural effusion occurred in 85 (9.68%) patients after RFA. Patients in the two groups did not differ with regard to baseline parameters. Patients in the two groups did differ significantly in ICGR15 and tumour site (p < 0.05). Tumour site was found to have a significant impact on the rate of complications post-RFA. There was no significant difference in ICGR15 values among patients with the same Child-Pugh scores or in the same tumour site.

Conclusions: The present results demonstrated that RFA is minimally invasive and suitable for the treatment of HCC. They also showed that ICGR15 did not independently predict for liklihood of post-RFA complications, after controlling for tumour site. Patients with tumours located subcapsularly or near the porta hepatis were found to have significantly higher rates of post-operative complications after RFA than to patients with tumours in the liver parenchyma.

• Complications
• hepatocellular carcinoma (HCC)
• indocyanine green retention rate at 15 min (ICGR15)
• liver function reserve
• radiofrequency ablation (RFA)

Introduction

Hepatocellular carcinoma (HCC) is one of the most common malignant tumours worldwide [1]. It is characterised by a high degree of morbidity and significant mortality. A large number of patients (80–90%) with HCC have concomitant cirrhosis. Liver transplantation, hepatectomy and ablation are the three curative modalities; radiofrequency ablation (RFA) is currently the most common liver tumour ablation technology [2,3]. Studies have shown that, for liver tumours less than 3 cm in diameter, the effectiveness of RFA treatment is comparable to that of hepatectomy [4]. A recent study by this group showed that for patients with liver tumours <4 cm in diameter, the 3-year survival rate after RFA is no different from the survival rate of hepatectomy patients [5]. Although RFA is a minimally invasive and its safety is well recognised, many types of complications still occur, and some may even be life threatening [6–8]. As a result, the development or discovery of a sensitive indicator that can satisfactorily predict post-operative complications after treatment of HCC with RFA remains an important research topic in the field of RFA treatment for HCC.

The hepatic functional reserve test which involves measurement of the rate of retention of indocyanine green (ICG) retention rate at 15 min after administration (ICGR15) is a comprehensive indicator of liver function. It reflects dynamic hepatic functional reserve [9,10]. The usefulness of ICGR15 in the prognostic evaluation of liver cancer patients, especially in the prediction of complications after hepatectomy, has been confirmed [11–13]. Studies have demonstrated that among Child-Pugh grade A patients, if the ICGR15 is greater than 14%, the surgical risk for hepatectomy increases and the rate of liver dysfunction after surgery significantly increases. If the ICGR15 is greater than 20%, large-scale hepatectomy involving more than the two hepatic segments is contraindicated [14]. Some investigators have proposed a safety assessment protocol for hepatectomy based on ICGR15 [15].

So far no report has been published regarding the predictive value of ICGR15 for complications of RFA in patients with HCC. Only one paper has shown that the differences between the pre-operative ICGR15 and the post-operative ICGR15 values of RFA are not significantly different [16]. Because testing hepatic functional reserve with ICGR15 is clearly useful in the prediction of post-operative complications, especially liver dysfunction after hepatectomy, it was here determined whether it would also be useful in the prediction of complications after RFA.

Materials and methods

Patients

Hepatic functional reserve testing with ICG was started on 1 June 2009. A total of 2,693 HCC patients were treated in our hospital between 1 June 2009 and 30 June 2013. Among these patients, 1,054 were treated with RFA and 878 of these cases met the inclusion criteria and were subjected to analysis. These 878 patients included 722 men and 156 women, giving a male to female ratio of 4.6:1. The mean age of patients was (55 ± 12.15) years. Of the 878 cases, 572 were HCC patients who had experienced a recurrence after hepatectomy and were subsequently treated with RFA, while the other 308 patients had been diagnosed with HCC for the first time. Of these, 264 cases were confirmed with biopsy, and the remaining 42 were diagnosed using clinical diagnostic criteria for HCC [17].

The inclusion criteria were as follows:

1. Patients had been diagnosed with HCC using pathological or clinical diagnostic criteria and were treated with RFA. These included newly diagnosed HCC cases and HCC recurrence after hepatectomy.

2. Patients had fewer than three liver tumours and the diameters of all tumours were ≤5 cm.

3. Patients had liver function of Child-Pugh grade A or B.

4. There was hepatitis B infection and the hepatitis B virus DNA copy number was less than 105/mL.

5. An ICG15 test was performed pre-operatively.

6. All treatments were performed by percutaneous radiofrequency ablation.

The exclusion criteria were as follows:

1. Patients with a history of variceal upper gastrointestinal bleeding and platelet counts less than 30,000, or platelet counts less than 30,000 after blood transfusions; patients with a prothrombin time (PT) longer than 5 s.

2. Extrahepatic metastasis of the tumour.

3. Tumour thrombus found in hepatic vasculature, tumour invasion of the portal vein, hepatic vein, and bile duct trunk or branch.

4. Other accompanying malignant tumours existing before or at the same time as HCC.

5. RFA was performed during hepatectomy simultaneously.

ICGR15 measurement

ICG (Dandong Yichuang Pharmaceutical, Dandong City, Liaoning Province, China) was injected at a dose of 0.5 mg/kg body weight via the cubital vein within a 10 s period. A DDG-3300K liver reserve function analyser (Nihon Kohden, Tokyo, Japan) was used for measurement. An optical sensor was clipped onto the two sides of the ala of one nostril, and ICG concentration was measured continuously both in vitro and in vivo and ICGR15 was calculated automatically. ICG is a deep blue-green dye that enters the blood via intravenous injection and binds to serum proteins. It is selectively absorbed by the liver and then released into the bile in free form. It is non-toxic, does not participate in the intestine liver circulation, and is not excreted by the kidneys. The instrument used to take ICG measurements involves a pulsatile pigment concentration method, i.e. it uses infrared to continuously monitor the blood concentration of a light-absorbing substance. Because ICG absorbs light well, its concentration in the blood can be determined accurately [18].

RFA

An LDR F-120 S RFA system (Lead Electron, Mianyang, China) was used to perform percutaneous transhepatic RFA for HCC treatment [19]. After the rotary cluster electrode was fully opened, it had the shape of a sphere with a diameter of nearly 5 cm. In this way, with a single electrode, lesions about 5 cm in diameter could be ablated and tumour lesions with diameters <4 cm would definitely be fully ablated. Under ultrasound guidance a radiofrequency electrode was placed at the bottom of the tumour and the cluster electrode was opened until it was larger than the largest tumour by at least 1 cm in diameter. The initial power was set at 50 W, and the power was increased by 10 W/min until it reached a maximum of 120 W. The power remained at this level until impedance increased to a maximum, at which point the output power automatically dropped; this completed one round of treatment. After the cluster electrode was switched off it was turned clockwise by 15° from the original position, then turned on again for another round of ablation in order to eliminate the maximum number of tumour cells in the treatment area. On the day of treatment ECG was continuously monitored after the procedure. On the first day after treatment, fasting routine blood tests and liver function tests were performed. Hepatoprotective therapy was implemented and prophylactic antibiotics were administered for 1–2 days.

Methods of analysis

The 878 patients who met the inclusion criteria were divided into two groups according to whether or not complications occurred: the complication group (85 cases, 9.68%) and the complication-free group (793 cases, 90.32%) (Table 1). The alpha-fetoprotein (AFP), haemoglobin, platelet, alanine aminotransferase, aspartate aminotransferase, total bilirubin (TBil), direct bilirubin (DBil), gamma-glutamyl transpeptidase, alkaline phosphatase, albumin, prothrombin time (PT), prothrombin time international normalised ratio (PTINR), tumour diameter, ICGR15, Child-Pugh grade and tumour site of patients in the two groups were analysed. The tumour site was classified as one of the following three sites: subcapsular (closest tumour edge within 1 cm away from the liver capsule), near the porta hepatis (closest tumour edge within 2 cm of the left and right portal veins at the first porta hepatis, or the left, the right hepatic veins or inferior vena cava at the second porta hepatis), and inner liver parenchyma (cases other than the previous two). Complications were defined according to the literature [20–22].

Table 1. Complications of 85 patients.

	n (cases)	Incidence (%)	Outcome
Bleeding	11	12.9	Cure
Bile leakage	9	10.6	Cure
Fever	26	30.6	Cure
Pleural effusion	11	12.9	Cure
Seroperitoneal effusion	14	16.5	Cure
Liver dysfunction	8	9.4	Cure
Liver abscess	6	7.1	Cure

The present study conformed to the Declaration of Helsinki, and all procedures were approved by the Southwest Hospital’s ethics committee. All patients signed an informed consent forms before surgery.

Data collection

The main purpose of the present study was to determine whether ICGR15 could be used to predict the rate of occurrence of complications after RFA and to compare different pre-operative examination indicators between the two groups. These indicators included AFP, haemoglobin, platelet count, alanine aminotransferase, Child-Pugh grade, and tumour site. Raw data from the enrolled patients was collected by an assistant and taken to the clinical research centre in our institute and entered into the liver cancer database management system.

In order to determine whether post-operative complications occurred, vital signs were monitored after RFA, and physical examinations were performed. Liver imaging tests (mainly contrast-enhanced ultrasound or enhanced CT) and blood tests (including routine blood tests, liver function tests, AFP and AFP heterogeneity) were also performed after RFA. Once a complication was confirmed, appropriate treatments were provided in accordance with the specific condition of the patient. Only one patient needed laparotomy to stop bleeding. The other 84 were successfully treated using more conservative treatments. After the patients were discharged, active follow-up exams were conducted. Within 6 months of RFA, AFP and AFP heterogeneity and liver imaging tests (contrast-enhanced ultrasound or enhanced CT) were performed. After 6 months these tests were performed once every 3 months. If tumour recurrence was confirmed, appropriate treatment measures were taken based on the specific conditions of the patients, including another resection, RFA, percutaneous ethanol injection, transcatheter arterial chemoembolisation, radiotherapy, chemotherapy, immune adoptive therapy, or liver transplantation [5].

Statistical analysis

Continuous variables are here presented as mean ± standard deviation (), whereas categorical variables were presented as percentages. Comparisons of continuous variables between two groups were analysed with Student's t-test, whereas comparisons of categorical variables between groups were analysed using a chi-square test. A logistic regression model was used to fit all variables, and a forward stepwise method was adopted for regression. The level of significance was set at p < 0.05. All analyses were carried out using SPSS 17.0 statistical software.

Results

Analysis of the clinical data from patients in the complication and complication-free groups showed that the two groups did not differ significantly in age, sex, AFP, haemoglobin, platelet count, alanine aminotransferase, aspartate aminotransferase, TBil, DBil, gamma-glutamyl transpeptidase, alkaline phosphatase, albumin, PT, PTINR, tumour diameter, or Child-Pugh grade (Table 2), but there were significant differences in ICGR15 and tumour site between the two groups (p < 0.05). Fitting all clinical data as variables with a logistic regression model showed that only tumour site had a significant impact on the rate of occurrence of complications after RFA. Patients with tumours located subcapsularly or near the porta hepatis had a significantly higher incidence of post-operative complications after RFA than patients with tumours inside the liver parenchyma. There was no significant difference between the two groups with respect to Child-Pugh grade, but the ICGR15 values of patients in the complication group were higher than those in the complication-free group (Child-Pugh A 13.41 ± 14.94 vs. 10.48 ± 11.45; Child-Pugh B 13.49 ± 10.47 vs. 10.91 ± 11.69) (Tables 3 and 4). However, within the same tumour site there was no significant difference between the two groups (subcapsular 11.26 ± 12.21 vs. 13.00 ± 13.57; near porta hepatis 9.51 ± 10.93 vs. 20.60 ± 17.41; inside liver parenchyma 11.12 ± 11.75 vs. 8.24 ± 3.47) (Table 5).

Table 2. Clinical data of the two groups.

	Complication group (n = 85 cases)	Complication-free group (n = 793 cases)	P Values
Age (years)	55.74 ± 13.35	54.92 ± 12.02	0.553
Sex (male:female) (cases)	73:12	649:144	0.354
Alpha-fetoprotein (ug/L)	1361.16 ± 4133.98	3104.07 ± 53818.91	0.766
Haemoglobin (g/L)	11.76 ± 1.41	11.93 ± 1.47	0.324
Platelet count ( × 10^9/L)	197.31 ± 63.03	202.00 ± 61.41	0.504
Alanine aminotransferase (IU/L)	45.38 ± 41.62	48.83 ± 60.79	0.610
Aspartate aminotransferase (IU/L)	50.22 ± 63.32	46.53 ± 49.58	0.527
Total bilirubin (μmol/L)	20.26 ± 7.09	20.83 ± 11.44	0.508
Direct bilirubin (μmol/L)	8.90 ± 3.73	9.12 ± 5.80	0.736
Alkaline phosphatase (IU/L)	104.93 ± 42.68	100.53 ± 39.35	0.332
Gamma-glutamyl transpeptidase (IU/L)	71.11 ± 81.73	70.96 ± 74.69	0.986
Albumin (g/L)	39.00 ± 6.70	39.17 ± 5.92	0.806
PT (s)	12.72 ± 2.10	12.55 ± 1.69	0.319
PTINR	1.04 ± 0.09	1.05 ± 0.13	0.788
Diameter (cm)	2.98 ± 0.95	2.95 ± 0.86	0.790
ICGR15 (%)	13.44 ± 13.72	10.59 ± 11.50	0.034*
Child-Pugh (A:B) (cases)	60:25	599:194	0.316
Site
Subcapsular (cases)	72	25	<0.001*
Near porta hepatis (cases)	8	266	<0.001**
Inside liver parenchyma (cases)	5	502

ICGR15, indocyanine green retention rate at 15 min; PT, Prothrombin time; PTINR, Prothrombin time international normalised ratio.

*Compared to inside liver parenchyma, p < 0.05; **Compared to inside liver parenchyma, p < 0.05.

Table 3. Logistic regression results.

	B	SE	Wald-χ^2	P value	Exp (B)	Exp (B) 95%CI
ICGR15	0.024	0.013	3.243	0.072	1.024	0.998–1.051
Diameter	−0.083	0.204	0.166	0.683	0.920	0.616–1.373
Child-Pugh	0.241	0.429	0.317	0.574	1.273	0.549–2.952
Site			162.470	<0.001
Site (subcapsular)	6.082	0.561	117.610	<0.001	437.810	145.858–1314.134
Site (near porta hepatis)	1.207	0.587	4.231	0.040	3.344	1.059–10.566

ICGR15, indocyanine green retention rate at 15 min.

B, regression coefficient. SE, standard error.

Table 4. Comparison of ICGR15 values of patients with the same Child-Pugh grade between the two groups.

	ICGR15 values (%) Complication group	ICGR15 values (%) No complication group	P Value
Child-Pugh A	13.41 ± 14.94	10.48 ± 11.45	0.067
Child-Pugh B	13.49 ± 10.47	10.91 ± 11.69	0.295

ICGR15, indocyanine green retention rate at 15 min.

Table 5. Comparison of ICGR15 values of patients in the same tumour site between the two groups.

	ICGR15 values (%) Complication group	ICGR15 values (%) No complication group	P Value
Subcapsular	11.26 ± 12.21	13.00 ± 13.57	0.573
Near porta hepatis	9.51 ± 10.93	20.60 ± 17.41	0.116
Inside liver parenchyma	11.12 ± 11.75	8.24 ± 3.47	0.584

ICGR15, indocyanine green retention rate at 15 min.

Discussion

In the present study the overall rate of occurrence of complications after HCC treatment with RFA was 9.68%, and there was only one severe case that needed surgical intervention (laparotomy was performed to stop massive bleeding). No surgical deaths occurred and even when ICGR15 value reached 45%, RFA treatment could still be performed. These results demonstrate that treating HCC with RFA is safe and reliable as well as being minimally invasive. Statistical results from this study also indicated that patients in the complication group and complication-free group did not differ significantly in basic clinical indicators including age, sex, tumour diameter, Child-Pugh grade, AFP, haemoglobin, platelet count, indicators of liver function, PT or PTINR, but they differed significantly in ICG15 and tumour site (p < 0.05). Fitting all of the above clinical indicators within a logistic regression model revealed that only the tumour site had a significant impact on the rate of occurrence of complications after RFA. Patients with subcapsular tumours or tumours near the porta hepatis had a higher incidence of complications after RFA, such as bleeding and bile leakage, than patients with tumours inside liver parenchyma. ICG15 did not add any predictive value for complications beyond that provided by the tumour site.

The present study is the first to carry out large sample analysis to determine the usefulness of hepatic functional reserve testing with ICGR15 in predicting complications after treating HCC with RFA. The results demonstrate that RFA treatment for HCC is a minimally invasive, safe method with a low complication rate even for patients with poor liver functional reserve, especially those with ICGR15 >20%, who are at high risk for surgical resection, or cannot tolerate surgical resection at all. However, although ICG15 can predict the incidence of post-operative complications of hepatectomy and is more sensitive in evaluating hepatic functional reserve than the Child-Pugh grade, it is not correlated with the incidence of complications after RFA. A high ICGR15 value is not a contraindication against RFA treatment.

In placing the electrode via percutaneous transhepatic puncture, the surgeon's level of operating experience may affect the rate of occurrence of complications after RFA. Kasugai et al. [23] reported that surgeons who performed RFA three or four times per month had a significantly fewer incidence than patients whose with surgeons who had less experience. However, a study conducted in Italy showed that the incidence rate of complications and mortality after RFA was not correlated with the surgeon’s experience [7]. The different conclusions of these two studies may result from different levels of experience at the two medical centres. In the current study our hospital started RFA treatment for HCC patients in June 1999, and has so far treated 3,211 patients. In recent years this hospital has treated approximately 500 HCC cases annually. In the present study all RFA was done by skilled physicians who had performed successful RFA treatments in more than 300 previous cases. In this way, complications due to insufficient operating experience were substantially reduced.

In general, the rate of occurrence of complications after RFA is low. A previous study showed that the incidence of major complications after RFA was 3.54%, and the mortality rate was 0.038% [20]. Complications caused by RFA can be divided into two categories: complications caused by the operation of electrode insertion, and complications attributable to thermal damage. Complications caused by the former include infection, bleeding, tumour spreading and pneumothorax, and those caused by the latter include damage to nearby organs and local burns caused by the electrode plate [20].

There have also been studies that have shown that when the tumour is located subcapsularly or near the liver centre, heat damage often occurs. With subcapsular tumours thermal damage may easily occur and can involve the stomach, peritoneal wall, diaphragm and/or the hepatic flexure of the colon. It has been shown that the probe must penetrate through a sufficient amount of normal liver parenchyma in order to prevent electrical burning of the needle track when retracting, otherwise bleeding, tumour rupture and spread may occur [21]. Tumours near the porta hepatis are close to the bile duct and blood vessels. Therefore, if the operator wants to completely ablate the tumour, the range of the ablated area may need to exceed the safe zone, which is 5–10 mm from the edge of the tumour [24]. This may cause damage to the bile duct and blood vessels [21,22]. In the present study, patients with subcapsular tumours or tumours near the porta hepatis had a higher incidence of complications such as bleeding and bile leakage after RFA than patients with tumours inside the liver parenchyma. Multivariate analysis demonstrated that ICGR15 did not predict for complications post-RFA, independently of the tumour site.

ICGR15 is a sensitive indicator of liver function, and it can reflect the hepatic functional reserve of patients with chronic liver disease. ICGR15 evaluation of hepatic functional reserve before hepatectomy has been widely adopted [25]. ICGR15 plays an important role in predicting the incidence of post-operative complications, especially liver dysfunction, after hepatectomy.

The present results suggest that for Child-Pugh A patients, the distribution of ICGR15 values was relatively discrete (0–47.2%). Comparison of ICGR15 values between the two groups in patients with the same Child-Pugh grade showed there to be no significant difference, yet the mean ICGR15 value of patients in the complication group was higher than that of patients in the complication-free group. This suggests that the majority of HCC patients have abnormal ICG excretion test results.In nearly half of the Child-Pugh grade A patients, the abnormality in ICGR15 values manifested before the abnormality in conventional liver functions. Compared to the conventional Child-Pugh grade, the ICGR15 excretion test is superior in reflecting hepatic functional reserve of patients with decompensated liver cancer. In this way, the hepatic functional reserve test with ICGR15 is more sensitive than Child-Pugh grade with respect to the evaluation of liver function.

Limitations of this study

Firstly, among the 878 patients included, 42 refused a biopsy test before RFA due to the risk of tumour metastasis in needle tracking during tumour puncture. These were only diagnosed with HCC with clinical diagnostic criteria. Secondly, the ethics committee allowed patients to choose treatment methods voluntarily. In this way, some HCC patients with RFA indications chose liver transplantation or hepatectomy and were not included in the study.

Conclusion

RFA is a minimally invasive procedure with a small impact on liver function. It can be performed even when the hepatic functional reserve is poor. However, ICGR15 cannot predict the incidence of complications after RFA treatment. The tumour site is an independent risk factor for incidence of complications after RFA. Patients with subcapsular tumours or tumours near the porta hepatis are more prone to complications such as bleeding and bile leakage after RFA than patients with tumours inside the liver parenchyma.

Declaration of interest

This work was supported by the National Natural Science Foundation of China (grant no. 30972894 and no. 81272688). The authors alone are responsible for the content and writing of the paper.