Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study

Radiotherapy and Oncology xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Radiotherapy and Oncology journal homepage: www.thegreenjournal.com

Original article

Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study Jeong Il Yu a,1, Jae Won Park b,1, Hee Chul Park a,c,⇑,2, Sang Min Yoon b,⇑,2, Do Hoon Lim a, Joon Hyeok Lee d, Han Chu Lee e, Seon Woo Kim f, Jong Hoon Kim b a Department of Radiation Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine; b Department of Radiation Oncology, Asan Liver Center, Asan Medical Center, University of Ulsan College of Medicine; c Department of Medical Device Management and Research, SAIHST, Sungkyunkwan University; d Department of Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine; e Department of Gastroenterology, Asan Liver Center, Asan Medical Center, University of Ulsan College of Medicine; and f Biostatistics Team, Samsung Biomedical Research Institute, Seoul, Republic of Korea

a r t i c l e

i n f o

Article history: Received 30 June 2015 Received in revised form 26 October 2015 Accepted 18 November 2015 Available online xxxx Keywords: Hepatocellular carcinoma Portal vein tumor thrombosis Radiotherapy Transarterial chemoembolization Validation

a b s t r a c t Purpose: To evaluate the relationship between portal vein tumor thrombosis (PVTT) response and clinical outcomes in patients with hepatocellular carcinoma (HCC) treated with transarterial chemoembolization followed by radiotherapy (TACE-RT). Materials and methods: The study enrolled 329 patients in the training set and 179 patients in the validation set. All patients who were treated with TACE-RT from 2002 to 2008 and satisfied the inclusion criteria were enrolled in the study. The median follow-up period was 11.7 months (range, 1.6-108.6) in the training set and 11.9 months (range, 1.7–105.1) in the validation set. Results: After TACE-RT, PVTT response was complete or partial in 32 (9.7%) and 134 (40.7%) patients of the training set and in 18 (10.1%) and 96 (53.6%) patients in the validation set, respectively. Failure to obtain PVTT response was significantly related with elevated post-treatment Child-Pugh score (P < 0.001). Furthermore, progression-free survival was significantly related with PVTT response (P < 0.001, hazard ratio 0.33, 95% confidence interval 0.25–0.42) in multivariate analysis. In receiveroperating characteristics analysis of 1-year progression prediction, the PVTT response showed an area under the curve of 0.74. Most of the findings were successfully reproduced in the independent external validation set. Conclusions: Positive PVTT response was closely associated with favorable clinical outcomes. The PVTT response to TACE-RT reduces metastasis and makes it possible to maintain normal liver function and achieve longer survival. Ó 2015 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology xxx (2016) xxx–xxx

Primary liver cancer is the second leading cause of cancerrelated death, and hepatocellular carcinoma (HCC) accounts for approximately 70–90% of primary liver cancers occurring worldwide [1]. Portal vein tumor thrombosis (PVTT) is recognized as a

⇑ Corresponding authors at: Department of Radiation Oncology, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnamgu, Seoul 135-710, Republic of Korea (H. C. Park). Department of Radiation Oncology, Asan Liver Center, Asan Medical Center, University of Ulsan College of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, Republic of Korea (S. M. Yoon). E-mail addresses: [email protected] (H.C. Park), [email protected]. kr (S.M. Yoon). 1 Jeong Il Yu and Jae Won Park contributed equally to this work as first authors. 2 Hee Chul Park and Sang Min Yoon contributed equally to this work as corresponding authors.

common accompanying manifestation, with reported rates of approximately 30–40% in patients with advanced HCC [2]. Several reports have described its role as one of the most important dismal prognostic indicators in these patients [3,4]. Although sorafenib is generally accepted as a standard of care for the treatment of advanced HCC, the objective response rate is somewhat disappointing at 2–5% [5,6]. Furthermore, median time to progression was only 2.8 months in a study of the Asia–Pacific region. There is no widely accepted local modality as standard treatment; trans-arterial chemoembolization (TACE) followed by radiotherapy (RT) has shown encouraging survival as well as a favorable local response rate [7–9]. In many studies significant survival prolongation was reported in responders after TACE followed by RT (TACE-RT) compared with non-responders [7–12].

http://dx.doi.org/10.1016/j.radonc.2015.11.019 0167-8140/Ó 2015 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

2

Clinical impact of TACE-RT for HCC with PVTT

A positive response after treatment is one of the most important and best-known determinants of prognosis in many oncologic fields, including HCC. In particular, considering the fact that PVTT is the main prognostic factor, a PVTT response might be the most important factor for improving the clinical outcomes of advanced HCC. However, the effect of the PVTT response on survival prolongation is poorly understood. Comparison of clinical outcomes according to the PVTT response after TACE-RT may provide an explanation for this relationship. Increasing RT application using higher conformal RT techniques, like stereotactic ablative RT (SABR) or particle beam RT, may be further emphasized in the present situation. Moreover, further treatment might be modified according to the response pattern based on more precise information on the effect of TACE-RT on PVTT. Based on this background, the authors conducted the present analysis with an independent external validation study to evaluate the relationship between PVTT response and clinical outcomes in HCC patients with PVTT treated with TACE-RT. Materials and methods Patients and methods The present study was conducted in HCC patients with main and/or first branch PVTT who were treated with TACE followed by three-dimensional conformal RT (3D-CRT) from August 2002 to September 2008 at Asan Medical Center. The diagnosis of HCC was based on guidelines proposed by the American Association for the Study of Liver Diseases (AASLD), and PVTT was confirmed by using dynamic contrast enhanced multi-phase computed tomography (CT) scans, magnetic resonance imaging (MRI) scans, or angiography. On the contrast-enhanced CT or MRI, PVTT was identified by the presence of intraluminal filling defect of portal vein adjacent to the primary tumor on portal phase with contrast enhancement of inner side of the filling defect on arterial phase. We categorized the thrombi as main PVTT if thrombi were located in the bilateral first branches and the main trunk. Tumor size was defined as the length of the longest diameter of the primary tumor with PVTT. Patients with PVTT below the first branch were excluded to maintain homogeneity of the study because other local treatment modalities can be considered when PVTT is confined to only small branches. Other exclusion criteria were as follows: (1) distant metastasis before initiation of RT, (2) Eastern Cooperative Oncology Group performance status P3, (3) poor liver function of Child-Pugh class C, (4) uncontrollable ascites or hepatic encephalopathy, (5) combined treatment with intra-arterial chemotherapy, (6) presence of double primary malignancies, and (7) planned RT was not completed. Patients for whom it was not possible to evaluate the response at 4–12 weeks after the completion of RT were also excluded from the analysis. The detailed TACE and RT procedures were described in our previous study [8]. RT was started at 2–3 weeks after TACE or TACI (transarterial chemoinfusion), and every effort was made to encompass the tumor invading the portal vein as much as possible; if this could not be achieved, partial tumor and PVTT with margin was determined as the target. The clinical target volume (CTV) was regarded the same as gross tumor volume (GTV) defined by dynamic enhanced CT or MRI. For daily set-up of variation and respiratory movement of the liver, the planning target volume was determined using 1–2 cm margin from CTV. To reduce the uncertainty of respiratory motion, fluoroscopy as well as planning CT was checked in simulation with shallow-breathing in both sets. GTVs were determined by dynamic enhanced lesion, and they tried to encompass the entire tumor and PVTT as possible. The dose per fraction was 2 to 5 Gy at 5 fractions per week. The total dose was determined by the volume of normal liver receiving >50% of the

prescription dose is maintained under 50%, and the maximum dose to the stomach or duodenum was limited to 36 Gy with 3 Gy or 44 Gy with 2 Gy fraction. The next session of TACE was routinely planned for 6–8 weeks after completion of RT, and TACE continued until the viable intrahepatic tumor completely disappeared in patients with preserved hepatic function. Patients were examined at least once a week during RT and then followed up 1 month after completion of RT. Thereafter, follow up was continued at 2- to 3-month intervals. The RT response was determined with dynamic contrast enhanced multi-phase liver CT scans 4–12 weeks after completion of treatment. To evaluate the effect of RT on PVTT with respect to overall clinical outcome, evaluation of RT response focused only on PVTT. In the evaluation of PVTT response, the greatest perpendicular diameter of the tumor thrombus was calculated and compared with the initial value as described previously [8]. Objective response was calculated as the combined number of patients with complete response and partial response. Adverse events were scored using the Common Terminology Criteria for Adverse Events (CTCAE; version 3.0). For external validation of the findings we used an independent external cohort of HCC patients with PVTT who satisfied the present inclusion criteria and received TACE-RT at Samsung Medical Center during the same period as the training set from August 2002 to September 2008. The procedure of TACE and RT used in the validation set was reported in another published article [11], and additional treatment, mostly TACE, was added on demand after TACE-RT. The present study was approved with permission by the institutional review boards of both Asan Medical Center (IRB No. 20150264) and Samsung Medical Center (IRB No. 2015-03-073). Informed consent was waived because of the retrospective nature of the study.

Statistical analysis Intrahepatic metastasis-free survival (IHMFS), distant metastasis-free survival (DMFS), progression-free survival (PFS), and overall survival (OS) were estimated using the Kaplan–Meier method. OS, IHMFS, DMFS, and PFS were measured from the date of the start of RT to the date of a patient’s death, the last followup examination, or the date that an event developed, respectively. In univariate survival analyses, the log-rank test was used to evaluate differences. Potential prognostic variables that showed statistical significance in univariate analysis were used to perform multivariate analysis with the Cox proportional hazards model using the Schoenfeld residuals method. Correlation between PVTT response and other variables was evaluated using the chi-square test or Fisher’s exact test. Receiver-operating characteristic (ROC) curves were drawn to evaluate the status of diseases whether they progressed or not at 1 year from the date of starting RT according to the prognostic factors that had statistical significance in multivariate analysis, and area under the curve (AUC) with 95% confidence interval (CI) was used to evaluate the power of prognostic factors. ROC curves were drawn and evaluated with the external validation set. All statistical analysis was performed using SPSS Statistics version 21.0 (Asan Medical Center) and version 22.0 (Samsung Medical Center) software for Windows (IBM, Armonk, NY, USA), and P < 0.05 was considered statistically significant. Results During the study period, 440 patients with HCC who had PVTT were treated with TACE-RT at Asan Medical Center. Among them, a

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

3

J.I. Yu et al. / Radiotherapy and Oncology xxx (2016) xxx–xxx

total of 73 patients were excluded for the following reasons: planned RT was not completed in 20 patients (3: refusal of further treatment, 9: cirrhosis-related complications, 3: TACE-RT related toxicities, 5: tumor progressions during TACE-RT), previous RT history in 3, combined treatment with intra-arterial chemotherapy in 3, double primary malignancies in 2, initial Child-Pugh class C in 3, initially diagnosed distant metastasis before TACE-RT in 41, and RT for a site other than PVTT in 1. After completion of RT, response evaluation of PVTT at 4–12 weeks was not possible in 38 patients, therefore they were excluded from the present analysis. Because the survival outcomes between a total of 440 patients and the enrolled 329 patients showed broadly similar patterns (Supplementary Fig. 1), 329 patients were finally enrolled in the present study as a training set to achieve the purpose of the present study, and an additional 179 patients were enrolled as an external validation set at Samsung Medical Center. Diagrams of patient cohorts are shown in Supplementary Fig. 2. The median follow-up period was 11.7 months (range, 1.6–108.6) in the training set and 11.9 months (range, 1.7–105.1) in the validation set. Table 1 summarizes the comparison of patient characteristics between the training and validation sets. The median age was 53 in the training set and 55 in the validation set. In comparisons between the training and validation sets, the proportion of ChildPugh class showed the greatest difference. Main PVTT was observed in 142 patients (43.2%) in the training set and 75 patients (41.9%) in the validation set. The median value of alpha-fetoprotein (AFP) was 563 ng/mL (range, 1–98,200) in the training set, and 400 ng/mL (range, 1–82,640) in the validation set. The median size of the primary tumor was 9.0 cm (range, 2–21.0) in the training set and 6.0 cm (range, 2–19.0) in the validation set. TACE was used before TACE-RT in 314 (95.4%), and 166 (92.7%) patients in the training set and the validation set, respectively, and the median number was two in both groups. The median interval of TACE-RT was 22 days (range, 3–54) in the training set, and 23 days (range, 10–59) in the validation set. There was no case presenting PVTT without primary tumor. PVTT without encompassing primary tumor was targeted in 267 (81.2%) patients in the training set, and 65 (36.3%) patients in the validation set. Median administered dose was 54.6 Gy10 in the training set and 50.4 Gy10 in the validation set as a biologic effective dose (BED) with a/b = 10. In all patients enrolled in the present study, 3DCRT was used based on the dose–volume histogram analysis. Among all patients, 36 patients (10.9%) of the training set and 103 patients (57.5%) of the validation set had no further treatment after RT completion. The remaining 293 patients (89.1%) of the training set and 76 (42.5%) of the validation set received at least one additional treatment (Table 1). The median number of additional TACE after TACE-RT was two (range, 1–21) in the training set, and two (range, 1–9) in the validation set. Elevation of liver enzymes during and within 3 months after treatment was common, but it was mainly confined to grade 2 or less, and grade 3 or more was developed in 103 (31.3%) patients in the training set, and 48 (26.8%) patients in the validation set. Liver dysfunction without evidence of intrahepatic tumor progression was developed in 37 (11.2%) patients in the training set, and 12 (6.7%) patients in the validation set. Grade II or III gastroduodenal toxicities were observed in 14 (4.3%) patients in the training set, and 6 (3.4%) patients in the validation set. In the training and validation sets, the median survival was 11.7 months (range, 1.5–108.6) and 11.0 months (range, 2.0– 104.0), and 1-year OS was 48.6% and 43.7%, respectively. The median PFS was 8.7 months (range, 1.0–108.6) and 4.9 months (range, 1.0–101.9), respectively (Supplementary Fig. 3). On assessment of PVTT response, 32 patients (9.7%) in the training set were classified as complete response (CR), 134 (40.7%) as partial response (PR), 129 (39.2%) as stable disease (SD), and 34

Table 1 Baseline characteristics of patients in the training and the validation sets. Variables Age (years) Gender ECOG performance status

Cause of hepatitis

Child-Pugh class alpha-fetoprotein (ng/ml) Tumor size (cm) Multiplicity of viable tumor Type of tumor

Stage (mUICC)

Main PVTT Previous treatment (repeated measure)

No of TACE before TACE-RT TACE-RT interval

Total RT dose (BED, a/ b = 10) RT field

RT dose per fraction Treatment after TACE-RT (repeated measure)

No. of TACE after TACE-RT

Median Range Male Female 0 1 2 HBV HCV Alcohol Others A B Median Range median range Solitary Multiple Nodular Infiltrative Massive II III IVA Involve Not involve Surgery RFA or PEIT TACE or TACI None Median Range Median (days) Range (days) Median (Gy) Range (Gy) PVTT and Entire HCC PVTT with margin Median (Gy) Range (Gy) Surgery RFA or PEIT TACE or TACI Re-RT Sorafenib None Median Range

Training set (%) n = 329

Validation set (%) n = 179

53 24–81 287 (87.2) 42 (12.8) 68 (20.7) 231 (70.2) 30 (9.1) 278 (84.5) 23 (7.0) 17 (5.2) 11 (3.3) 219 (66.6) 110 (33.3) 563 1–98200 9.0 2.0–21.0 164 (49.8) 165 (50.2) 72 (21.9) 253 (76.9) 4 (1.2) 0 (0.0) 70 (21.3) 259 (78.7) 142 (43.2) 187 (56.8) 18 (5.5) 32 (9.7) 314 (95.4)

55 28–80 160 (89.4) 19 (10.6) 30 (16.8) 141 (78.8) 8 (4.5) 165 (92.2) 9 (5.0) 4 (2.2) 1 (0.6) 157 (87.7) 22 (12.3) 400 1–82640 6.0 2.0–19.0 83 (46.4) 96 (53.6) 72 (40.2) 58 (32.4) 49 (27.4) 3 (1.7) 66 (36.9) 110 (61.5) 75 (41.9) 104 (58.1) 19 (10.6) 26 (14.5) 166 (92.7)

13 (4.0) 2 0–18 22

6 (3.4) 2 0–9 23

3–54

10–59

54.6 27.3–78.0 62 (18.8)

50.4 15.6–71.5 110 (61.5)

267 (81.2)

69 (38.5)

3.0 2.0–5.0 11 (3.3) 14 (4.3) 288 (87.5)

3.0 1.8–4.5 2 (1.1) 3 (1.7) 65 (36.3)

45 (13.7) 1 (0.3) 36 (10.9) 2 1–21

8 (4.5) 4 (2.2) 103 (57.5) 2 1–9

Abbreviations: ECOG, Eastern Cooperative Oncology Group; HBV, hepatitis B virus; HCV, hepatitis C virus; mUICC, modified International Union Against Cancer; PVTT, portal vein tumor thrombosis; RFA, radiofrequency ablation; PEIT, percutaneous ethanol injection therapy; TACE, transarterial chemoembolization; TACI, transarterial chemoinfusion; RT, radiotherapy; BED, biologically equivalent dose; Gy, Gray.

(10.4%) as progressive disease (PD). In the validation set, 18 patients (10.1%) were classified as CR, 96 (53.6%) as PR, 42 (23.5%) as SD, and 23 (12.8%) as PD. Typical cases showing positive response of PVTT after TACE-RT can be found in Supplementary Fig. 4. Table 2 summarizes the results of correlation analysis between potentially related factors and objective PVTT response after TACERT in patients of the training set. As pretreatment factors, ChildPugh class (P = 0.005), total dose of RT (P = 0.03), and AFP level P400 ng/ml (P = 0.006) were significantly related to PVTT

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

4

Clinical impact of TACE-RT for HCC with PVTT

Table 2 Potential factors related to PVTT response after TACE-RT in the training set. Variables

Pre-treatment

Gender Age (years) ECOG performance status Child-Pugh class Child-Pugh score

Tumor size Viable tumor Level of PVTT Previous treatment Total dose (BED with a/b = 10) PreAFP P 400 ng/ml Post-treatment

AFP decrement Child-Pugh class

Child-Pugh score elevation Further treatment

N

Objective PVTT response

P value

Responder (%)

Non-responder (%)

Male Female <55 P55 0–1 2 A B 5 6 7 8 9 <5 cm P5 cm Single Multiple Main Other Yes No <55 Gy P55 Gy Yes No

287 42 180 149 299 30 219 110 99 120 64 23 23 50 279 164 165 142 187 316 13 165 164 178 151

149 (51.9) 17 (40.5) 85 (47.2) 81 (54.5) 152 (50.8) 14 (46.7) 123 (56.2) 43 (39.1) 61 (61.6) 62 (51.7) 28 (43.8) 10 (43.5) 5 (21.7) 31 (62.0) 135 (48.4) 89 (54.3) 77 (46.7) 67 (47.2) 99 (52.9) 161 (50.9) 5 (38.5) 73 (44.2) 93 (56.7) 77 (43.3) 89 (58.9)

138 (48.1) 25 (59.5) 95 (52.8) 68 (45.6) 147 (49.2) 16 (53.3) 96 (43.8) 67 (60.9) 38 (38.4) 58 (48.3) 36 (56.3) 13 (56.5) 18 (78.3) 19 (38.0) 144 (51.6) 75 (45.7) 88 (53.3) 75 (52.8) 88 (47.1) 155 (49.1) 8 (61.5) 92 (55.8) 71 (43.3) 101 (56.7) 62 (41.1)

Yes No A B C Yes No Yes No

205 124 171 126 32 162 167 293 36

128 (62.4) 38 (30.6) 114 (66.7) 43 (34.1) 9 (28.1) 65 (40.1) 101 (60.5) 155 (52.9) 11 (30.6)

77 (37.6) 86 (69.4) 57 (33.3) 83 (65.9) 23 (71.9) 97 (59.9) 66 (39.5) 138 (47.1) 25 (69.4)

0.19 0.22 0.71 0.005 0.006

0.09 0.19 0.32 0.41 0.03 0.006 <0.001 <0.001

<0.001 0.01

Abbreviations: PVTT, portal vein tumor thrombosis; ECOG, Eastern Cooperative Oncology Group; BED, biologically equivalent dose; Gy, Gray; AFP, alpha-fetoprotein.

response. Especially, there was a negative trend of PVTT response and Child-Pugh scores (P = 0.006). After TACE-RT, a decrease in AFP level (P < 0.001) and elevation of Child-Pugh score and class (both P < 0.001) at 4–12 weeks were closely related with PVTT response. The Child-Pugh score decreased in 48 patients (14.6%), and did not change in 119 patients (36.2%). Among those 167 patients, 101 patients (60.5%) patients showed PVTT response. Further treatment after TACE-RT was also significantly associated with the PVTT response (P = 0.01). Although the findings were slightly different in the validation set, the total dose of RT (P = 0.02) as a pretreatment factor, and decreased AFP (P = 0.005) and increased Child-Pugh score (P = 0.003) and class (P = 0.006) as post-treatment factors were significantly related to PVTT response, as in the training set (Supplementary Table 1). The Child-Pugh score decreased in 11 patients (6.1%), and did not change in 91 patients (50.8%). Among those 102 patients, 75 patients (73.5%) showed PVTT response. The median PFS of the training set was 8.7 months. In univariate analysis of potential prognostic variables, age (P = 0.05), nodular morphology of primary tumor (P = 0.001), single viable tumors (P = 0.003), tumor size (P = 0.02), stage (P < 0.001), pretreatment AFP (P = 0.002), and PVTT response (P < 0.001) were significantly associated with PFS (Table 3). On multivariate analysis, the independent favorable prognostic factor for PFS were nodular morphology of primary tumor (P = 0.04, hazard ratio [HR] 0.71, 95% CI 0.52–0.98), single viable tumor (P = 0.03, HR 0.75, 95% CI 0.58–0.97), and pretreatment AFP < 400 ng/ml (P = 0.03, HR 0.77, 95% CI 0.61–0.98). In particular, PVTT response (P < 0.001, HR 0.33, 95% CI 0.25–0.42) was the most important favorable prognostic indicator of PFS (Table 3, Fig. 1).

In the validation set, morphology of primary tumor (P < 0.001), single viable tumors (P = 0.003), and PVTT response (P < 0.001) were significant favorable prognostic factors whereas pretreatment AFP (P = 0.08) showed a marginal significance for PFS in univariate analysis (Supplementary Fig. 5). On multivariate analysis, nodular morphology of primary tumor (P = 0.02, HR 0.67, 95% CI 0.48–0.94) and PVTT response (P < 0.001, HR 0.32, 95% CI 0.22– 0.47) were significant favorable factors (Supplementary Table 2 and Fig. 5). The PVTT response was the most significant prognostic factor in both univariate and multivariate analyses in the validation as well as the training set in terms of OS (Supplementary Tables 3 and 4). In the ROC analysis of 1-year progression prediction using significant prognostic factors on multivariate analysis, the PVTT response after TACE-RT showed the highest AUC value of 0.74 (95% CI 0.68–0.79, Supplementary Fig. 6A), followed by tumor morphology at 0.57 (95% CI 0.5–0.63), number of viable tumors at 0.54 (95% CI 0.47–0.60), and pretreatment AFP at 0.57 (95% CI 0.50– 0.63). Consistent findings were obtained in the validation set, and the AUC value of PVTT response in this set was 0.71 (95% CI 0.63–0.79, Supplementary Fig. 6B). The IHMFS, DMFS, and OS were also significantly affected by PVTT response in the training set (all P < 0.001, Fig. 2). The median IHMFS, DMFS, and OS were 17.6 months, 30.4 months, and 19.3 months in PVTT responders, compared with 4.3 months, 8.3 months, and 8.0 months in non-responders. Similar survival differences were observed in the validation set (Supplementary Fig. 7). In the ROC analysis of 1-year intrahepatic, distant progression, and survival prediction according to the PVTT response after

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

5

J.I. Yu et al. / Radiotherapy and Oncology xxx (2016) xxx–xxx Table 3 Univariate and multivariate analyses of potential prognostic factors for PFS in the training set. Variables

1-yr PFS (%)

P value UVA

HR

95% CI

0.04

0.71

0.52–0.98

0.75

0.58–0.97

0.03

0.77

0.61–0.98

<0.001

0.33

0.25–0.42

MVA

Gender

Male Female

39.1 41.5

0.74

Age (years)

<55 P55

35.7 47.9

0.05

ECOG performance status

0–1 2

41.7 34.8

0.61

Morphology of primary tumor

Nodular Other

53.9 37.6

0.001

Child-Pugh class

A B

43.1 37.3

0.30

Number of viable tumor

Single Multiple

44.7 37.7

0.003

0.03

Tumor size

65 cm >5 cm

55.7 38.6

0.02

0.16

Main PVTT

Not involve Involve

44.1 37.2

0.11

Pretreatment AFP (ng/ml)

6400 >400

48.3 35.1

0.002

Total RT dose (BED, a/b = 10)

655 Gy >55 Gy

37.8 44.5

0.12

TACE-RT response of PVTT

No Yes

19.1 62.3

<0.001

0.08

Abbreviations: UVA, univariate analysis; MVA, multivariate analysis; HR, hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; PVTT, portal vein tumor thrombosis; AFP, alpha-fetoprotein; RT, radiotherapy; BED, biologically equivalent dose; Gy, Gray; TACE, transarterial chemoembolization.

Fig. 1. Kaplan–Meier curves of progression-free survival. In the training set, PFS was significantly affected by nodular morphology (A), solitary viable tumor (B), pretreatment AFP <400 ng/ml (C), and PVTT response (D) in both univariate and multivariate analyses.

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

6

Clinical impact of TACE-RT for HCC with PVTT

Fig. 2. Kaplan–Meier survival curves. The intrahepatic metastasis-free survival (IHMFS, A), distant metastasis-free survival (DMFS, B) and overall survival (OS, C) were significantly affected by PVTT response in the training set (all P < 0.001).

TACE-RT, showed AUCs of 0.75 (95% CI 0.69–0.80), 0.69 (95% CI 0.62–0.75), and 0.75 (95% CI 0.69–0.80) in the training set, and 0.71 (95% CI 0.62–0.79), 0.69 (95% CI 0.60–0.79), and 0.73 (95% CI 0.66–0.80) in the validation set, respectively (Supplementary Fig. 8).

Discussion The present study evaluating the correlation between clinical outcomes and PVTT response after TACE-RT showed clear significance of a PVTT response with respect to delaying disease progression as well as improving patients’ survival outcomes. The PVTT response was the most reliable progression predictor within one year. The AUC values of PVTT response after TACE-RT were 0.69– 0.75 in survival prediction. These values were quite impressive to be considered as clinical prognostic factors. Additionally, nonresponsiveness of PVTT was closely related to increase posttreatment Child-Pugh class and/or Child-Pugh score. Those findings were successfully reproduced in an independent external validation set. PVTT is well known as the most important prognostic indicator in HCC [13–15], and the median OS of patients with HCC with PVTT is reported to be less than 3 months at best without appropriate treatment [16]. PVTT can lead to liver function deterioration thus theoretically directly causing ischemic liver injury, and can increase portal pressure thus promoting variceal bleeding or rupture [17–19]. PVTT may promote intrahepatic and distant metastasis and the generation of collateral vessels, such as an arterioportal shunt, which can interfere with successful TACE [20–23].

Recently, numerous reports were published reporting favorable clinical outcomes of RT using a 3D technique for HCC [7–11]. In many reports, RT was frequently combined with TACE to enhance the RT effect [7–9,24]. In fact, favorable outcomes of responders among patients with HCC with PVTT who were treated with RT with or without TACE have been reported in many articles [7,9–11]. Our group also reported the results of TACE-RT on HCC with PVTT, with a median OS among responders of 19.4 months [8]. Although sorafenib is recommended as the sole treatment of choice for advanced HCC based on the results of randomized phase III trials, the response rate was disappointingly limited to 2–5% of all patients [5,6]. Given the importance of an early response in patients with advanced HCC, more reliable local modalities with high response rates should be considered first. In this regard, TACE-RT could be the most optimal modality in HCC with PVTT for the following reasons. First, PVTT itself may cause deterioration of liver function [25]. Additionally, disease progression can accelerate this process. This unfavorable role of PVTT was also observed in our study; an elevation in Child-Pugh score after treatment was more frequent in the non-responder group. Conversely, a reduction in PVTT after TACERT was clearly related to optimal liver function maintenance. Also, in the report of HCC patients with PVTT by Zhang et al., the group treated with TACE-RT to enhance portal vein patency showed longer survival than the group not treated with TACE-RT, although the study did not assess liver function maintenance [26]. The PVTT response after TACE-RT was the most significant prognostic factor for PFS, IHMFS, DMFS, as well as OS in the present study. In the prospective study by Han et al., both PFS and OS were significantly affected by response to chemoradiotherapy in HCC

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

J.I. Yu et al. / Radiotherapy and Oncology xxx (2016) xxx–xxx

patients with PVTT [27]. A similar prolongation of time to progression was reported in responders treated with TACE with sorafenib [28]. Although these studies only focused on the overall response rather than PVTT response with different treatment modalities, the results suggest that the role of PVTT as a source of metastasis could be efficiently controlled by suppressing PVTT as shown in our study. Lastly, PVTT itself is an obstacle to the treatment of HCC with respect to liver function or efficacy of treatment, as discussed above. Kim et al. reported that subsequent treatment including repeated TACE was significantly more frequent among patients who received RT and that prolonged survival was achieved in these patients as a consequence [29]. As in the present study, additional treatments were significantly more frequent among patients with CR or PR in the training set. Among these, 24 patients (7.3%) received salvage surgery and/or radiofrequency ablation/ percutaneous ethanol injection, with a median OS of 41.6 months (range, 9.6–100.7). One clinical importance of the PVTT response is that it could provide an opportunity to receive further treatment, including curative modalities. More advanced RT techniques, like SABR or particle beam RT, provide the possibility of more conformal and higher dose delivery, therefore more rapid and higher response could be expected on PVTT in HCC. In fact, more favorable clinical outcomes with comparable toxicities of those techniques for HCC were reported in several published articles [30–33], though the relation between this conformal higher dose RT and PVTT response should be studied further. This study has several limitations that should be noted. First, this is a retrospective study, and is therefore inevitably subject to selection bias. Therefore, the significance of baseline liver function, like Child-Pugh score which could affect and be affected by multiple factors (including indication of the study, tumor size, degree of PVTT, dose of RT, etc.) might not be easily interpreted in the present study. Second, there are some differences in baseline characteristics between the training and validation sets that are related to different institutional management protocols. This may affect direct generalization of results of the present study. Third, we investigated the PVTT response and related factors, such as Child-Pugh score, and decreased AFP at specific times, but the causality between the response and these factors might be obscure. Nevertheless, the present study provides valuable information about the PVTT response after TACE-RT. The effect of selection bias might be minimized because the study had a relatively large number of cases that were homogenously treated according to a consistent protocol. Despite some discrepancies in the characteristics between the training and validation sets, most of the results were successfully validated in a considerable number of cases in an external validation set sharing the same inclusion criteria. Therefore, this improves the reliability of the clinical importance of the PVTT response in patients with HCC combined with PVTT who were treated with TACE-RT. Additionally, the PVTT response after TACE-RT have reversed the clinical outcome according to the common course of the disease. Therefore, it might be possible to accept the causality between PVTT response and improved clinical outcomes. In conclusion, PVTT response after TACE-RT was clearly related with favorable clinical outcomes in the training set and this finding was successfully validated in the independent validation set. This response makes it possible to maintain normal liver function, reduces intrahepatic and/or distant metastases, and achieves longer survival with or without additional treatment. Considering the major contribution of PVTT to a dismal prognosis, achievement of an early response using TACE-RT should be considered a priority.

7

Further prospective studies are needed to verify the clinical importance of PVTT. Conflicts of interest statement We have no conflicts of interest to declare. Acknowledgments This research was supported by a Samsung Medical Center Grant (GF01130081) and by a Grant (2015-7214) from the Asan Institute for Life Sciences, Asan Medical Center, Seoul, Republic of Korea. This research was also supported by a grant from Marine Biotechnology Program (20150220) funded by the Ministry of Oceans and Fisheries, Republic of Korea. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.radonc.2015.11. 019. References [1] Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin 2012;62:10–29. [2] Ogren M, Bergqvist D, Bjorck M, Acosta S, Eriksson H, Sternby NH. Portal vein thrombosis: prevalence, patient characteristics and lifetime risk: a population study based on 23,796 consecutive autopsies. World J Gastroenterol 2006;12:2115–9. [3] Hiraoka A, Horiike N, Yamashita Y, et al. Risk factors for death in 224 cases of hepatocellular carcinoma after transcatheter arterial chemoembolization. Hepatogastroenterology 2009;56:213–7. [4] Zhou L, Rui JA, Wang SB, et al. Outcomes and prognostic factors of cirrhotic patients with hepatocellular carcinoma after radical major hepatectomy. World J Surg 2007;31:1782–7. [5] Cheng AL, Kang YK, Chen Z, et al. Efficacy and safety of sorafenib in patients in the Asia-Pacific region with advanced hepatocellular carcinoma: a phase III randomised, doubleblind, placebo-controlled trial. Lancet Oncol 2009;10:25–34. [6] Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008;359:378–90. [7] Kim SW, Oh D, Park HC, et al. Transcatheter arterial chemoembolization and radiation therapy for treatment-naive patients with locally advanced hepatocellular carcinoma. Radiat Oncol J 2014;32:14–22. [8] Yoon SM, Lim YS, Won HJ, et al. Radiotherapy plus transarterial chemoembolization for hepatocellular carcinoma invading the portal vein: long-term patient outcomes. Int J Radiat Oncol Biol Phys 2012;82:2004–11. [9] Yu JI, Park HC, Lim do H, et al. Scheduled interval trans-catheter arterial chemoembolization followed by radiation therapy in patients with unresectable hepatocellular carcinoma. J Korean Med Sci 2012;27:736–43. [10] Kim DY, Park W, Lim DH, et al. Three-dimensional conformal radiotherapy for portal vein thrombosis of hepatocellular carcinoma. Cancer 2005;103:2419–26. [11] Yu JI, Park HC, Lim do H, et al. Prognostic index for portal vein tumor thrombosis in patients with hepatocellular carcinoma treated with radiation therapy. J Korean Med Sci 2011;26:1014–22. [12] Xi M, Zhang L, Zhao L, et al. Effectiveness of stereotactic body radiotherapy for hepatocellular carcinoma with portal vein and/or inferior vena cava tumor thrombosis. PLoS ONE 2013;8:e63864. [13] Prospective validation of the CLIP score: a new prognostic system for patients with cirrhosis and hepatocellular carcinoma. The Cancer of the Liver Italian Program (CLIP) Investigators. Hepatology 2000; 31:840–5. [14] Chevret S, Trinchet JC, Mathieu D, Rached AA, Beaugrand M, Chastang C. A new prognostic classification for predicting survival in patients with hepatocellular carcinoma. Groupe d’Etude et de Traitement du Carcinome Hepatocellulaire. J Hepatol 1999;31:133–41. [15] Pons F, Varela M, Llovet JM. Staging Systems in Hepatocellular Carcinoma, 7. HPB (Oxford); 2005. pp. 35–41. [16] Llovet JM, Bustamante J, Castells A, et al. Natural history of untreated nonsurgical hepatocellular carcinoma: rationale for the design and evaluation of therapeutic trials. Hepatology 1999;29:62–7. [17] Fimognari FL, Violi F. Portal vein thrombosis in liver cirrhosis. Intern Emerg Med 2008;3:213–8. [18] Kadouchi K, Higuchi K, Shiba M, et al. What are the risk factors for aggravation of esophageal varices in patients with hepatocellular carcinoma? J Gastroenterol Hepatol 2007;22:240–6.

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019

8

Clinical impact of TACE-RT for HCC with PVTT

[19] Ng WD, Chan YT, Ho KK, Kong CK. Injection sclerotherapy for bleeding esophageal varices in cirrhotic patients with hepatocellular carcinoma. Gastrointest Endosc 1989;35:69–70. [20] Toyosaka A, Okamoto E, Mitsunobu M, Oriyama T, Nakao N, Miura K. Pathologic and radiographic studies of intrahepatic metastasis in hepatocellular carcinoma; the role of efferent vessels. HPB Surg 1996;10:97–103 [discussion 103–4]. [21] Addario L, Tritto G, Cavaglia E, Amodio F, Giannelli E, Di Costanzo GG. Preserved liver function, portal thrombosis and absence of oesophageal varices are risk factors for metastasis of hepatocellular carcinoma. Dig Liver Dis 2011;43:319–34. [22] Morse SS, Sniderman KW, Galloway S, Rapoport S, Ross GR, Glickman MG. Hepatoma, arterioportal shunting, and hyperkinetic portal hypertension: therapeutic embolization. Radiology 1985;155:77–82. [23] Oh D, Shin SW, Park HC, Cho SK, Lim do H, Paik SW. Changes in arterioportal shunts in hepatocellular carcinoma patients with portal vein thrombosis who were treated with chemoembolization followed by radiotherapy. Cancer Res Treat 2015;47:251–8. [24] Kim GA, Shim JH, Yoon SM, et al. Comparison of chemoembolization with and without radiation therapy and sorafenib for advanced hepatocellular carcinoma with portal vein tumor thrombosis: a propensity score analysis. J Vasc Interv Radiol 2015;26:e326. [25] Schwartz ME, Emre S, Mor E, Katz E, Miller CM. Acute portal vein thrombosis: a cause of rapid deterioration of liver function, treatable by thrombectomy and transplantation. Transplant Proc 1993;25:1972. [26] Zhang XB, Wang JH, Yan ZP, Qian S, Du SS, Zeng ZC. Hepatocellular carcinoma with main portal vein tumor thrombus: treatment with 3-dimensional

[27]

[28]

[29]

[30]

[31]

[32]

[33]

conformal radiotherapy after portal vein stenting and transarterial chemoembolization. Cancer 2009;115:1245–52. Han KH, Seong J, Kim JK, Ahn SH, Lee do Y, Chon CY. Pilot clinical trial of localized concurrent chemoradiation therapy for locally advanced hepatocellular carcinoma with portal vein thrombosis. Cancer 2008;113:995–1003. Zhu K, Chen J, Lai L, et al. Hepatocellular carcinoma with portal vein tumor thrombus: treatment with transarterial chemoembolization combined with sorafenib–a retrospective controlled study. Radiology 2014;272:284–93. Kim KM, Kim JH, Park IS, et al. Reappraisal of repeated transarterial chemoembolization in the treatment of hepatocellular carcinoma with portal vein invasion. J Gastroenterol Hepatol 2009 May;24:806–14. Bujold A, Massey CA, Kim JJ, et al. Sequential phase I and II trials of stereotactic body radiotherapy for locally advanced hepatocellular carcinoma. J Clin Oncol 2013;31:1631–9. Huertas A, Baumann AS, Saunier-Kubs F, et al. Stereotactic body radiation therapy as an ablative treatment for inoperable hepatocellular carcinoma. Radiother Oncol 2015;115:211–6. Lee SU, Park JW, Kim TH, et al. Effectiveness and safety of proton beam therapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis. Strahlenther Onkol 2014;190:806–14. Qi WX, Fu S, Zhang Q, Guo XM. Charged particle therapy versus photon therapy for patients with hepatocellular carcinoma: a systematic review and meta-analysis. Radiother Oncol 2015;114:289–95.

Please cite this article in press as: Yu JI et al. Clinical impact of combined transarterial chemoembolization and radiotherapy for advanced hepatocellular carcinoma with portal vein tumor thrombosis: An external validation study. Radiother Oncol (2016), http://dx.doi.org/10.1016/j.radonc.2015.11.019