A phase 2 multicenter study of stereotactic body radiotherapy for hepatocellular carcinoma: Safety and efficacy
Abstract
Background
Although several prospective studies have reported the efficacy of stereotactic body radiotherapy (SBRT) for hepatocellular carcinoma (HCC), treatment-related toxicity varies and has not been determined. Therefore, the authors evaluated the safety and efficacy of SBRT for patients with HCC in a hepatitis B virus-endemic area.
Methods
This multicenter phase 2 trial enrolled patients with unresectable HCC. Patients received SBRT with 45 to 60 Gy in 3 fractions. To evaluate gastroduodenal toxicity, esophagogastroduodenoscopy (EGD) was performed before and 2 months after SBRT. The primary endpoint was treatment-related severe toxicity at 1 year after SBRT. The secondary endpoints were the 2-year local control, progression-free survival, and overall survival rates.
Results
In total, 74 patients were enrolled between January 2012 and April 2015, and 65 eligible patients were analyzed. One patient experienced radiation-induced liver disease with acute grade ≥3 toxicity 1 month after SBRT. In addition, 1 patient had a grade 3 esophageal ulcer with stenosis 5 months after SBRT. The actuarial rate of treatment-related severe toxicity at 1 year was 3%. The pre-SBRT and post-SBRT EGD findings were not significantly different among the 57 evaluable patients who underwent EGD. The 2-year and 3-year local control rates were 97% and 95%, respectively. The progression-free and overall survival rates were 48% and 84% at 2 years, respectively, and 36% and 76% at 3 years, respectively.
Conclusions
With a median follow-up of 41 months, this prospective multicenter study demonstrated that SBRT for patients with HCC is well tolerated and is an effective treatment modality.
Introduction
Stereotactic body radiotherapy (SBRT) delivers a high dose of radiation to the target, uses a small number of fractions with a high degree of precision within the body, and applies new radiobiology.1, 2 Although many studies on SBRT for hepatocellular carcinoma (HCC) have demonstrated its efficacy, with 1-year local control (LC) rates from 75% to 100%, most studies were retrospective case series, and a only few were level II evidence studies.3-13 A previous phase 2 study on SBRT after incomplete transarterial chemoembolization (TACE) from our institution reported promising treatment results, with an LC rate of 95% and an overall survival (OS) rate of 69% at 2 years.6 However, 5 of 47 patients in that study (11%) experienced severe gastrointestinal toxicity. Subsequent analysis of the patients who underwent SBRT for abdominopelvic malignancies, including these patients, showed that the dose to the gastroduodenum (GD) and a history of ulcer before SBRT were the best predictors of severe GD toxicity.14 On the basis of our single-institutional, phase 2 study and retrospective toxicity study, we launched the current phase 2, multicenter study to reduce treatment-related severe toxicity while sustaining the efficacy of SBRT. We used our own GD constraint and prospectively conducted esophagogastroduodenoscopy (EGD).
The objective of the current study was to evaluate the safety and efficacy of SBRT in patients with unresectable HCC who have an incomplete response after TACE. This study is registered at clinicaltrials.gov (NCT01850667).
Materials and Methods
Study Design and Participants
This was a single-arm, multi-institutional, open-label, phase 2 study. Patients were prospectively enrolled in a single cohort approved by the institutional review boards of each of the 6 participating institutions. The trial protocol is provided in the Supporting Information and is available online at https://clinicaltrials.gov/ct2/show/NCT01850667.
Eligible patients were aged ≥20 years who were diagnosed with primary or recurrent HCC.15 Patients were identified as having unresectable HCC and as ineligible for local ablative treatment after a review with surgeons and interventional radiologists at the institutional liver conference. Only a Child-Pugh (CP) class of A or B7 was permitted. Patients who had an Eastern Cooperative Oncology Group performance status of 0 to 1 were included. Tumor diameters were <10 cm, and the normal liver volume (NLV) (the total liver volume minus the cumulative gross tumor volumes [cGTV]) had to be ≥700 mL. Single or multinodular tumors, including portal vein tumor thrombosis (PVTT), which can be covered in the radiation field with 1 session or consecutive sessions of SBRT, were permitted. All patients received 1 to 5 TACE sessions. Patients were required to have no evidence of an uncontrolled lesion at any other site, uncontrolled complications of cirrhosis, and uncontrolled intercurrent illness. Patients were excluded if they had a previous history of radiotherapy (RT) to the abdomen or direct invasion to the esophagus, stomach, or colon by HCC. Written informed consent was obtained from all patients before enrollment.
SBRT Simulation
After TACE, patients were evaluated for their eligibility for SBRT. The patients underwent contrast-enhanced computed tomography (CT) simulation. The cGTV was defined as enhancing lesions, including lipiodol-laden areas, on CT and/or magnetic resonance imaging (MRI) studies. The tumor volumes used during planning were larger than the cGTV according to the discretion of the treating radiation oncologist to identify the borders of the tumor on simulation CT images and were referred to as internal target volumes (ITVs). The planning target volume (PTV) was defined as the ITVs + 4 mm in the longitudinal direction and the ITVs + 2 mm in all other directions. Forced shallow breathing with abdominal compression and respiratory gating using Real-Time Position Management (Varian Medical Systems) or the Synchrony Respiratory Tracking System (Accuray, Inc) were mandatory to control liver motion.
SBRT Dose Prescription and Normal Tissue Constraints
The SBRT dose reached a maximum of 60 Gy in 3 fractions. However, the total doses could be reduced to 45 Gy by 1 Gy per fraction until normal tissue constraints were allowed. In this procedure, at least 90% of the prescription dose should cover the PTV. The following liver constraints were used: at least 700 mL or 70% of the NLV should not receive a total dose >17 Gy (reverse V17Gy [rV17Gy], ≥700 mL). The minimum dose of 2 mL (D2mL) to the GD should not exceed 35 Gy, or 28 Gy in patients without ulcers on EGD before SBRT and in those with ulcers, according to our retrospective study.14 Although the study reported that the maximum point dose (Dmax) to the GD was the best dosimetric predictor, we think a near-point maximum dose volume of 2 mL was both clinically realistic and comparable when calculated across different planning systems in a multi-institutional study. For the spinal cord, the maximum dose should not exceed 22 Gy, and and ≤0.25 mL of the irradiated volume of the spinal cord receiving >18 Gy.
SBRT Treatment
All treatments were delivered using 3 fractions within 14 days, with at least 48-hour intervals between fractions, at the Korea Institute of Radiological and Medical Sciences, Dongnam Institute of Radiological and Medical Sciences, Soonchunhyang University Seoul Hospital, Soonchunhyang University Cheonan Hospital, Haeundae Paik Hospital, and Inha University Hospital. Daily imaging for localization was required. To ensure the availability of basic SBRT quality, all 6 participating institutions submitted a facility questionnaire before the initiation of the study. In addition, a regular workshop was convened 4 times a year during the study. Representative cases were reviewed to provide protocol requirements, structure delineation, and dose prescription and to minimize protocol deviations. After the end of SBRT, clinical data were verified through a cross-check of patient files and RT charts, including Digital Imaging and Communications in Medicine (DICOM) RT format files.
Toxicity Evaluation
Patients had regular follow-up visits at 2, 4, 6, 9, and 12 months and every 4 to 6 months thereafter until disease progression or death. If a patient presented any symptom of treatment-related toxicity between visits, they could come in for an evaluation at any time. Follow-up visits entailed history taking, physical examination, complete blood cell counts, chemistry panels, and CT or MRI.
To evaluate GD toxicity, we prospectively performed the first EGD within 6 months before SBRT and the second EGD at 2 months after SBRT. However, EGD could be omitted if the asymptomatic patient refused EGD or the patient had a high risk of bleeding by underlying cirrhosis. Hepatic toxicity was scored according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 4.0. In addition, we assessed the occurrence of classic radiation-induced liver disease (RILD) (ie, anicteric hepatomegaly, ascites, or elevated alkaline phosphatase more than twice the upper limit of normal value), and nonclassic RILD (ie, elevation of liver transaminases more than 5 times the upper limit of normal level or a worsening of the CP score ≥2 points) within 3 months after SBRT. Other toxicities were scored according to the CTCAE, and treatment-related severe toxicity was defined as grade ≥3 adverse events or classic/nonclassic RILD.
Statistical Analysis
The primary endpoint was 1-year treatment-related severe toxicity, defined as grade ≥3 adverse events according to the CTCAE or classic/nonclassic RILD. The null hypothesis was a 1-year, treatment-related, severe toxicity-free rate of 85%, corresponding to a low risk of hepatic toxicity (≤5%) from several studies16 and 11% gastrointestinal toxicity from our retrospective study.6 An improvement in severe toxicity-free rate from 85% to 95% with application of our own constraints was considered promising. A Simon optimal 2-stage design was used with a type I error of 5% and power of 80%. With an anticipated drop-out rate of 10%, 71 individuals were required for the safety analysis. At the end of the first stage, the study would be continued if there was a minimum of 11 patients without severe toxicity among 13 patients. At the second stage, 52 additional patients would be recruited, and ≥59 patients without severe toxicity would be required to demonstrate safety. The secondary endpoints were the LC, progression-free survival (PFS), and OS rates at 2 years. LC was defined as the absence of local failure or the absence of tumor growth or regrowth in any direction beyond that on pre-SBRT images of the treated lesions using the Modified Response Evaluation Criteria in Solid Tumors. OS and PFS were calculated from the start date of SBRT to the date of death from any cause and until tumor progression at any site or death. Survival was estimated using the Kaplan-Meier method and was compared between groups in univariate analysis using the log-rank test. The McNemar test was used to analyze differences in the prevalence of EGD findings before and after SBRT. Binary logistic regression analysis was used to identify the association between changes in EGD findings and the D2mL to the GD. All statistical analyses were performed using Statistical Package for the Social Sciences software, version 23.0 (IBM Corporation), and a 2-sided P < .05 was considered statistically significant.
Results
Study Population
Seventy-four patients were enrolled in this study between January 2012 and April 2015; of these, 65 patients were included the outcome analysis (Fig. 1). The characteristics of the patients are summarized in Table 1. The median age was 61 years (range, 44-84 years). Hepatitis B virus (HBV) infection was the predominant cause of liver disease (64%). All patients received TACE, and 71% received TACE ≤2 times. Except for 1 patient, all 64 patients had CP class A disease. The median tumor size was 2.4 cm, and most patients (88%) had a single lesion. PVTT was present in 4 patients (6%). The median SBRT dose was 60 Gy in 3 fractions (range, 45-60 Gy). Dose reductions are shown in Supporting Table 1.
| Characteristics | No. of Patients |
|---|---|
| Age: Median (range), y | 61 (44-84) |
| Sex | |
| Women | 16 |
| Men | 49 |
| ECOG performance status | |
| 0 | 29 |
| 1 | 36 |
| Hepatitis | |
| No | 4 |
| HBV | 41 |
| HCV | 14 |
| Alcoholic | 6 |
| Cirrhosis | |
| No | 12 |
| Yes | 53 |
| Initial CP score | |
| 5 | 47 |
| 6 | 17 |
| 7 | 1 |
| Diagnosis | |
| Initial | 37 |
| Recurrent | 28 |
| No. of previous TACE procedures | |
| 1 | 32 |
| 2 | 14 |
| 3 | 13 |
| 4 | 5 |
| 5 | 1 |
| Tumor size: Median (range), cm | 2.4 (1.0-9.9) |
| No. of tumors | |
| 1 | 57 |
| 2 | 8 |
| Tumor location | |
| One lobe | 60 |
| Both lobes | 5 |
| Portal vein tumor thrombosis | |
| No | 61 |
| Yes | 4 |
| AFP: Median (range), IU/mL | 7.9 (1.1-1192.0) |
| Milan criteria | |
| Within | 54 |
| Beyond | 11 |
| BCLC stage | |
| 0 | 25 |
| A | 32 |
| B | 4 |
| C | 4 |
| mUICC | |
| 1 | 29 |
| 2 | 25 |
| 3 | 11 |
| cGTV: Median (range), mL | 9.21-318.4 |
| PTV: Median (range), mL | 17.2 (3.3-699.2) |
| SBRT dose: Median (range), Gy | 60 (45-60) |
- Abbreviations: AFP, α-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; cGTV, cumulative gross tumor volume; CP score, Child-Turcotte-Pugh score; ECOG, Eastern Cooperative Oncology Group; HBV, hepatitis B virus; HCV, hepatitis C virus; mUICC, the modified International Union Against Cancer stage; PTV, planning target volume; SBRT, stereotactic body radiotherapy; TACE, transarterial chemoembolization.
Safety
Acute toxicity within 3 months after the end of SBRT is detailed in Table 2. Most patients had grade ≤2 toxicity that resolved spontaneously or after medication. Nonclassic RILD occurred in 1 patient who had a 4.5-cm HCC and an α-fetoprotein level of 960.9 IU/mL. He received SBRT at 60 Gy. His NLV and rV17Gy were 973 mL and 712 mL, respectively. He experienced grade 4 hyperbilirubinemia, grade 3 hepatic failure, and a worsening CP score from 6 points to 10 points. Multiple intrahepatic tumors developed, and his α-fetoprotein level increased to 116,610 IU/mL at 1 month. He died at 4 months after SBRT. Another patient had a 7-cm HCC and received SBRT at 51 Gy. The Dmax to esophagus was 42 Gy. He experienced abdominal pain at 2 weeks after SBRT, and EGD showed grade 2 radiation-induced esophagitis. The symptom worsened despite oral medication, and follow-up EGD showed an ulcer at the esophagogastric junction 2 months after SBRT and an esophageal ulcer with stenosis at 5 months after SBRT. His symptom was relieved with an esophageal stent. New intrahepatic tumors outside of the SBRT field developed at 29 months, and he died at 39 months. No additional treatment-related severe toxicities occurred within 1 year in the remaining patients. The actuarial treatment-related severe toxicity rate at 1 year was 3% (95% CI, 0%-7.1%).
| CTCAE Term | No. of Patients | |||
|---|---|---|---|---|
| Grade 1 | Grade 2 | ≥Grade 3 | ||
| CTCAE Category | ||||
| General disorder | Edema | 1 | 1 | 0 |
| Fever | 0 | 1 | 0 | |
| Fatigue | 9 | 0 | 0 | |
| Malaise | 1 | 0 | 0 | |
| GI disorder | Nausea | 8 | 7 | 0 |
| Abdominal discomfort/pain | 3 | 7 | 0 | |
| Diarrhea | 0 | 1 | 0 | |
| Vomiting | 1 | 0 | 0 | |
| Constipation | 0 | 0 | 0 | |
| Diarrhea | 0 | 1 | 0 | |
| Investigation | Bilirubin increased | 7 | 9 | 1a |
| Hypoalbuminemia | 8 | 5 | 0 | |
| AST increased | 28 | 0 | 1a | |
| ALP increased | 18 | 1 | 0 | |
| ALT increased | 11 | 0 | 0 | |
| INR increased | 11 | 0 | 0 | |
| Hepatobiliary disorder | Hepatic failure | 0 | 0 | 1a |
| Respiratory disorder | Pneumonia | 2 | 1 | 0 |
| Type of RILD | ||||
| Classic RILD | 0 | |||
| Nonclassic RILD | 1a | |||
- Abbreviations: ALP, alkaline phosphatase; ALT, alanine transaminase; AST, aspartate transaminase; CTCAE, National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0; GI, gastrointestinal; INR, international normalized ratio; RILD, radiation-induced liver disease.
- a The same patient experienced grade ≥3 toxicity according to the CTCAE and nonclassic RILD with intrahepatic progression.
Endoscopic Findings Before and After SBRT
Both pre-SBRT and post-SBRT EGDs were prospectively performed in 57 patients, as shown Table 3. Most patients had gastritis. Abnormal EGD findings were single in 32 versus 28 patients on pre-EGD versus post-EGD studies, and multiple in 24 versus 28 patients, respectively. Medication based on pre-SBRT EGD studies was prescribed for 11 patients with gastritis, 1 patient with a gastric ulcer, 1 patient with a duodenal ulcer, and 1 patient with candidiasis. After SBRT, 22 new abnormalities developed in 18 patients (32%); however, no patients required medication, except for the above-mentioned patient with esophageal toxicity. The prevalence of EGD findings between pre-SBRT and post-SBRT was not statistically different on the McNemar test (see Supporting Table 2). Varix or portal hypertensive gastropathy did not significantly increase after SBRT. On binary logistic regression analysis, newly developed gastritis and duodenitis after SBRT had a statistically significant association with the D2mL to the GD (Fig. 2; see Supporting Table 2).
| Variable | No. of Patients (%) | ||
|---|---|---|---|
| Findings Before SBRT | Disappearance After SBRT | New Abnormalities After SBRT | |
| Normal | 1 (2) | 1 (2) | 1 (2) |
| Portal hypertensive gastropathy | 5 (9) | 2 (4) | 4 (7) |
| Esophageal varix | 14 (25) | 2 (4) | 1 (2) |
| Gastric varix | 12 (21) | 5 (9) | 1 (2) |
| Esophagitis | 7 (12) | 4 (7) | 5 (9) |
| Duodenitis | 1 (2) | 1 (2) | 5 (9) |
| Gastritis | 48 (84) | 2 (4) | 3 (5) |
| Gastric ulcer | 1 (2) | 1 (2) | 2 (4) |
| Duodenal ulcer | 1 (2) | 0 (0) | 0 (0) |
| Esophageal candidiasis | 1 (2) | 0 (0) | 2 (4) |
- Abbreviation: SBRT, stereotactic body radiotherapy.
Efficacy
The median follow-up was 41 months (range, 4-69 months) for the entire cohort and 45 months (range, 12-69 months) for the survivors. The 2-year and 3-year LC rates were 97% and 95%, respectively. The 2-year and 3-year PFS rates were 48% and 36%, respectively. The median OS had not been reached, and the 2-year and 3-year OS rates were 84% and 76%, respectively. The survival curves are presented in Figure 3. On univariate analysis, there was no significant prognostic factor for LC (see Supporting Table 3). The number of TACE sessions before SBRT was the only significant factor for PFS (see Supporting Table 4). Tumor size <3 cm, a PTV ≤17 mL, and disease within the Milan criteria were significantly favorable prognostic factors for OS (see Supporting Table 5).
Discussion
The main finding in this prospective, multicenter study was the low treatment-related severe toxicity with a sustained high rate of tumor control. Because HBV infection is endemic in Korea and almost 90% of patients with HCC have underlying cirrhosis or chronic HBV infection, both tumor control and the prevention of hepatic decompensation are essential requirements for the treatment of HCC. On the basis of a previous phase 2 study, prospectively performed EGD in the current study demonstrated that SBRT does not increase variceal bleeding or RT-induced bleeding. The satisfaction of NLV constraints preserved liver function after SBRT, and the rate of treatment-related severe toxicity was quite low at 3%. On the basis of these findings, as well as the high 3-year LC and OS rates of 95% and 76%, respectively, SBRT is hereby established as a safe and effective local modality for the treatment of HCC.
To the best of our knowledge, this study is the first prospective SBRT trial concerning EGD-based toxicity evaluation. Few studies have performed EGD for toxicity evaluation after RT.17-19 Yu et al19 reported that grade 3 GD toxicity developed in 15% of patients at 2 months after 3-dimensional conformal RT (3DCRT) for perihilar HCC, and an elevated CP score at 3 months after 3DCRT was the most significant factor. Another study showed that the prevalence of esophageal varix, gastric varix, and gastric ulcers was significantly higher after concurrent chemoradiotherapy than before chemoradiotherapy for advanced HCC.17 Meanwhile, the current study showed that the prevalence of EGD findings before and after SBRT was not statistically different, although abnormalities were newly developed in 32% of patients. Moreover, EGD findings after SBRT were not aggravated, and there was no severe GD toxicity during the follow-up period. There are several possible reasons for this difference. First, historical RT-induced gastritis was defined as a red lesion, usually >2 cm, that was sharply demarcated from the surrounding normal mucosa.20 3DCRT includes a substantial volume of the GD even in high-dose regions, whereas SBRT minimizes the irradiated volume by a steep dose gradient. The physical difference in GD dose distribution might cause the change in EGD findings. Second, SBRT is considered for highly selected patients. Although 82% of patients in our study had cirrhosis, only 2% had gastric ulcers on pre-SBRT EGD. This suggested that patients had well preserved hepatic function, considering that portal hypertension after cirrhosis induces an impaired gastric mucosal defense mechanism and may contribute to the increased risk of gastric ulcer in such patients compared with the general population (20% vs 2%).21 Our previous studies suggested that a history of gastric/duodenal ulcer and GD dose were risk factors for the development of severe GD toxicity.6, 14 However, the risk of the presence of an active ulcer before SBRT and the consequent severe GD toxicity after SBRT might be minimized if patients are carefully selected who have preserved liver function with a CP score <7 and the D2mL to the GD is kept under 28 Gy.
The Princess Margaret Hospital conducted a prospective phase 1/2 SBRT study between 2004 and 2010, when liver SBRT was in the early phase of development and optimization.7 In that study, the 2-year LC and OS rates were 74% and 34%, respectively. After SBRT, 29% of patients had CP class deterioration. Subsequently, Takeda et al10 reported improved treatment outcomes from a phase 2 SBRT study performed between 2007 and 2012, with 3-year LC and OS rates of 96% and 67%, respectively. Nonclassic RILD occurred in only 9% of those patients. The current study, in which we used the most recent technologies developed between 2012 and 2015 and is the first to be conducted in a multicenter setting, confirmed the efficacy of SBRT, with 3-year LC and OS rates of 95% and 76%, respectively, in a sufficiently long median follow-up of 41 months. This is comparable to the findings from studies that used charged-particle therapy listed in Table 4.7, 10, 22-25 Small-sized HCC treated with a sufficient SBRT dose in patients with preserved liver function might have contributed to the highest OS among published studies regardless of the modality used (ie, charged particles vs photons).
| Reference (Study Period) | Study Design | Beam | No. of Patients | CP Class A/B, % | PVTT, % | Median Tumor Size [range], cm | Dose | Median Follow-Up, mo | LCR, % | OS, % | Toxicity |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Bujold 20137 (2004-2010) | SP 1/2 | SBRT | 102 | 100/0 | 55 | 7.2 [1.4-23.1] | 24-54 Gy/6 Fx | 31 | 87 at 1 y | 55 at 1 y, 34 at 2 y | CP class deterioration in 29%; grade ≥3 toxicity in 30% |
| Takeda 201610 (007-2012) | SP 2 | SBRT | 90 | 91/9 | NS | ≤4 | 35-40 Gy/5 Fx | 42 | 96 at 3 y | 67 at 3 y | Nonclassic RILD in 9%; grade 3 transaminase ↑ in 2% |
| Kawashima 200522 (1999-2003) | SP 2 | Proton | 30 | 67/33 | 40 | 4.5 [2.5-8.2] | 76 GyE/20 Fx | 31 | 96 at 2 y | 77 at 1 y, 66 at 2 y, 62 at 3 y | Grade 5 hepatic toxicity in 13% |
| Kim 201523 (2007-2010) | SP 1 | Proton | 27 | 89/11 | 15 | 1.37 | 60 GyE/20 Fx, 66 GyE/22 Fx, 72 GyE/24 Fx | 31 | 80 at 3 y, 64 at 5 y | 56 at 3 y, 42 at 5 y | No grade 3 toxicity |
| Hong 201624 (2009-2015) | MP 2 | Proton | 44 | 79/21 | 30 | 5 [1.9-12.0] | 15-67.5 GyE/15 Fx | 20 | 95 at 2 y | 63 at 2 y | CP class deterioration in 4% |
| Kasuya 201725 (1997-2003) | SP 1/2 and 2 | Carbon | 126 | 77/23 | 17 | 4 [1.0-12.0] | 48-69.6 GyE/412 Fx | 27 | 95 at 1 y, 91 at 3 y, 90 at 5 y | 90 at 1 y, 50 at 3 y, 25 at 5 y | Nonclassic RILD in 8%; grade 3 skin toxicity in 4% |
| Current study (2012-2015) | MP 2 | SBRT | 65 | 98/2 | 6 | 2.4 [1.0-9.9] | 45-60 Gy/3 Fx | 41 | 97 at 2 y, 95 at 3 y | 84 at 2 y, 76 at 3 y | Nonclassic RILD in 2%; grade 3 esophageal stenosis in 2% |
- Abbreviations: ↑, elevation; CP score, Child-Turcotte-Pugh score; Fx, fractions; GyE, gray equivalent; LCR, local control rate; MP, multi-center prospective study; OS, overall survival; PVTT, portal vein tumor thrombosis; RILD, radiation-induced liver disease; SBRT, stereotactic body radiotherapy; SP, single-center prospective study.
We acknowledge that our study has several limitations. First, we did not conduct a prospective quality-assurance program. Prospective feedback during patient accrual promotes the quality and uniformity of SBRT treatment in the trial but has the possible disadvantage of rendering the trial less representative for clinical practice.26 After a retrospective review of DICOM RT format files, we found a major deviation of the study protocol in 1 patient. She had 5.5-cm, 2.5-cm, and 0.5-cm HCCs and received SBRT at a dose of 51 Gy. Her NLV and rV17Gy were 1181 mL and 483 mL, respectively. Nonclassic RILD and rapid tumor progression, including multiple lung metastases, occurred at 3 months after SBRT, and she died at 7 months after SBRT. We excluded this patient from the treatment outcome analysis. In addition, 1 patient experienced severe esophageal toxicity at an unconstrained structure on our protocol. Although unexpected deviations from the protocol do not affect the primary endpoint, this emphasizes that a prospective and continuous quality-assurance program should be considered for a multicenter study. Second, all patients received a D2mL to the GD ≤28 Gy as our own constraint, although we recommended a different D2mL to the GD, depending on the presence of ulcer. This was because we reduced 1 Gy per fraction and 3 Gy per total dose. A sharp dose fall-off to achieve optimal sparing leads to a significantly inhomogeneous dose deposition within the GD volume, which results in large variations between the Dmax and the doses encompassing a finite hot-spot isodose volume such as D2mL.27 Therefore, a reduction of 3 Gy at a time would induce a more pronounced change in the D2mL to the GD. The use of a D2mL to the GD ≤28 Gy induced no severe GD toxicity in the current study. However, this is a higher value than the general recommendation. Because the total number of patients was relatively small to confirm safety, further prospective studies on larger study populations are needed to validate our own constraints.
In conclusion, this prospective study demonstrated that SBRT for HCC was well tolerated and was an effective treatment modality after long-term median follow-up of 41 months. Our study encourages a randomized comparison of SBRT versus other treatment modalities for HCC.
Funding Support
This work was supported by a grant of the Korea Institute of Radiological and Medical Sciences (KIRAMS), funded by Ministry of Science and ICT (MSIT), Republic of Korea. (no. 50543-2018; Won Il Jang). This work also was supported by the Soonchunhyang University Research Fund (no. 20190003; Sun Hyun Bae).
Conflict of Interest Disclosures
The authors made no disclosures.
Author Contributions
Won Il Jang: Methodology, investigation, data curation, funding acquisition, software formal analysis, writing–original draft, writing–review and editing, and visualization. Sun Hyun Bae: Conceptualization, methodology, investigation, funding acquisition, software formal analysis, writing–original draft, writing–review and editing, and visualization. Mi-Sook Kim: Conceptualization, project administration, supervision, funding acquisition, and writing–review and editing. Chul Ju Han: Conceptualization, methodology, and writing–review and editing. Su Cheol Park: Conceptualization, methodology, and writing–review and editing. Sang Bum Kim: Investigation and writing–review and editing. Eung-Ho Cho: Writing–review and editing. Chul Won Choi: Investigation and writing–review and editing. Kyung Su Kim: Writing–review and editing. Sangyoun Hwang: Investigation and writing–review and editing. Jin Ho Kim: Methodology, investigation, and writing–review and editing. A Ram Chang: Methodology, investigation, and writing–review and editing. Younghee Park: Writing–review and editing. Eun Seog Kim: Methodology, investigation, and writing–review and editing. Woo Chul Kim: Conceptualization, methodology, and writing–review and editing. Sunmi Jo: Investigation and writing–review and editing. Hae Jin Park: Investigation and writing–review and editing. All authors had access to the study data and reviewed and approved the final article.
