Clinical features of lenvatinib for unresectable hepatocellular carcinoma in real‐world conditions: Multicenter analysis

Abstract Background/Aim Presently, there are no therapeutic options for unresectable hepatocellular carcinoma (u‐HCC) patients who are intolerant to sorafenib or regorafenib failure. There have been no reports with detailed clinical findings of lenvatinib (LEN), a newly developed first‐line tyrosine kinase inhibitor (TKI), obtained in real‐world practice. We aimed to elucidate the therapeutic efficacy of LEN. Materials/Methods From March to August 2018, 105 u‐HCC patients were treated with LEN. Following exclusion of those who started with a reduced LEN dose and/or had a short observation period (<2 weeks), 77 patients (72.0 ± 8.9 years, 59 males, 8 mg/12 mg = 49/28, Liver Cancer Study Group of Japan 6th [LCSGJ]‐TNM stage II/III/IVa/IVb = 8/28/4/37, and American Joint Committee on Cancer/Union for International Cancer Control 8th [AJCC/UICC]‐TNM stage IB:II:IIIA:IIIB:IVA:IVB = 2:27:6:5:9:28) were divided into two groups (TKI naïve [n = 33] and TKI experienced [n = 44], including 11 with regorafenib history). Therapeutic response was evaluated using mRECIST. Clinical data were retrospectively evaluated. Results There were significant differences in age (74.6 ± 11.2 vs 70.0 ± 5.9 years, P = 0.040), LCSGJ‐TNM (II:III:IVa:IVb = 8:12:1:12 vs 0:16:3:25, P = 0.006), and AJCC/UICC‐TNM (IB:II:IIIA:IIIB:IVA:IVB = 2:17:1:1:4:8 vs 0:10:5:4:5:20, P = 0.028), while hepatic reserve function, adverse event (AE) profiles, and progression‐free survival (89.7%/80.4% vs 90.5%/80.1%, P = 0.499) and overall survival (96.7%/96.7% vs 100%/92.3%, P = 0.769) after 4 and 12 weeks were not significantly different between the TKI‐naïve and TKI‐experienced groups. Overall response rate and disease control rate at 4 weeks (n = 52) were 38.5% and 80.8%, respectively, and 32.4% and 70.3%, respectively, at 12 weeks (n = 37). A significant decline in log10 AFP from the baseline to 4 weeks after introducing LEN was observed in patients with PR and SD (2.047 ± 1.148 vs 1.796 ± 1.179, P < 0.001). Conclusion Regardless of past TKI therapy, therapeutic response and AEs after introducing LEN were similar. LEN may be an important treatment for the present unmet need regarding TKI treatment against u‐HCC.


| MATERIALS AND METHODS
. We retrospectively examined the records of those patients and collected clinical data obtained at the introduction of LEN, as well as after 2 weeks and every 4 weeks thereafter. Following exclusion of those who started with a reduced LEN dose and/or had a short observation period (<2 weeks), 77 patients were enrolled and divided into two groups, TKI naïve (n = 33) and TKI experienced (n = 44), according to their past history with TKI treatments. Flow diagram of enrolled patients is shown in Figure 1. Clinical characteristics, therapeutic response stage II/III/IVa/IVb = 8/28/4/37, and American Joint Committee on Cancer/Union for International Cancer Control 8th [AJCC/UICC]-TNM stage IB:II:IIIA:IIIB:IVA: IVB = 2:27:6:5:9:28) were divided into two groups (TKI naïve [n = 33] and TKI experienced [n = 44], including 11 with regorafenib history). Therapeutic response was evaluated using mRECIST. Clinical data were retrospectively evaluated. Results: There were significant differences in age (74.6 ± 11.2 vs 70.0 ± 5.9 years,  including progression-free survival rate (PFSR) and overall survival rate (OSR), and AEs were analyzed in a retrospective manner.
Patients positive for anti-hepatitis C virus (HCV) were judged to have HCC due to HCV, while those positive for hepatitis B virus surface antigen (HBsAg) were judged to have HCC due to hepatitis B virus (HBV).

| Assessment of hepatic reserve function and prognosis.
Child-Pugh classification 8 and ALBI grade were used to assess hepatic reserve function. ALBI grade was calculated based on serum albumin and total bilirubin values using the following formula: [ALBI score = (log10 bilirubin (µmol/L) × 0.66) + (albumin (g/L) × −0.085)], and defined by the following scores: ≤−2.60 = Grade 1,>−2.60 to ≤−1.39 = Grade 2,>−1.39 = Grade 3. 9,10 For a more detailed evaluation of the middle grade of ALBI (grade 2), we used a modified ALBI (mALBI) grade by creating 4 grades, including sub-grading for the middle grade (2a and 2b), with an ALBI score of −2.270, which was reported as the cutoff value for indocyanine green retention 15 minutes 30%, used as the value for dividing 2a and 2b. 12,13

| Assessment of muscle volume
Muscle volume loss (MVL) was determined using the following previously reported index: [psoas index (PI): bilateral psoas muscle area of middle L3 level (cm 2 )/height (m) 2 ; cutoff values for MVL of males = 4.24 cm 2 /m 2 , females = 2.50 cm 2 /m 2 ], based on CT findings. 14 PI was manually calculated using psoas muscle area at the middle L3 level in CT findings with personal computer software (Centricity Web DX ver.3.7.3.6417: GE Healthcare Japan, Tokyo, Japan, or OsiriX DICOM Viewer MD 9.5: https://www.osirixviewer.com). Muscle volume was calculated using computed tomography (CT) findings obtained at the start of LEN and again after 4 and 12 weeks in patients for whom CT imaging data were available.

| Diagnosis and treatment of HCC
HCC was diagnosed based on an increasing course of alpha-fetoprotein (AFP), as well as dynamic CT, 15 magnetic resonance imaging (MRI), 16,17 contrast-enhanced ultrasonography (CEUS) with perflubutane (Sonazoid ® , Daiichi Sankyo Co., Ltd. Tokyo, Japan), 18,19 and/or pathological findings. American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) tumor node metastasis (TNM) stage was used for evaluation of tumor progression as well as TNM stage, which was determined as previously reported in studies for staging of HCC conducted by the Liver Cancer Study Group of Japan (LCSGJ). 20 The present study protocol was approved by the Institutional Ethics Committee of Ehime Prefectural Central Hospital (No. 29-75).

| LEN treatment and assessment of AEs
After obtaining written informed consent, LEN was orally administered at 8 mg/d to patients weighing <60 kg and 12 mg/d to those ≥60 kg, and discontinued when any unacceptable or serious AE or clinical tumor progression was observed. According to the guidelines for administration of LEN, the drug dose should be reduced or treatment interrupted when a patient develops any grade 3 or more severe AE or if any unacceptable grade 2 drug-related AE occurs. AEs were assessed according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. 21 The worst grade for each AE during the present observation period was recorded. If a drug-related AE occurred, dose reduction or temporary interruption was maintained until the symptom was resolved to grade 1 or 2, according to the guidelines provided by the manufacturer.

| Evaluation of therapeutic response
Local physicians at each institution evaluated tumors using enhanced CT or MRI results obtained at 4, 8, or 12 weeks after introducing LEN, in accordance with the modified RECIST guidelines. 22
After starting LEN treatment, patients with Child-Pugh class A had worsened to Child-Pugh B or C in 23.4% of patients at 4 (P < 0.001) and in 23.7% at 12 weeks (P = 0.002) (Figure 4). Furthermore, a significant decline in ALBI score from the baseline (−2.32 ± 0.49) was observed at 4 (−2.18 ± 0.54 [n = 67], P < 0.001) and 12 (−2.10 ± 0.55 [n = 43], P < 0.001) weeks after the start of LEN treatment. AEs that occurred following the start of LEN in these patients are presented in Table 2. HFSR was the most common in the present cohort (all grades: n = 31 [ an AE during the present observation period, LEN was abandoned in 17 (PD, n = 5; AEs, n = 12). The median period to initial dose reduction or pause of LEN was 40 days. Of 8 with an HFSR with a grade of 3 in previous treatments with a TKI, 87.5% also had an HFSR (grade 3, 2, 1; n = 2, 3, 2, respectively) with LEN treatment, while 44.4% of 18 with an HFSR with a grade of 2 during previous TKI treatments had an HFSR with LEN (grade 3, 2, 1; n = 3, 3, 2, respectively) ( Figure 5). The data of CT imaging findings at introducing LEN were obtained for 51 patients, of whom 18 (35.3%) had MVL. As a sub-analysis, relative changes in muscle volume at 4 and 12 weeks after introducing LEN were evaluated in 41 and 25 patients, respectively, whose data of CT imaging findings obtained at the start of and approximately 4 and 12 weeks after beginning treatment. In those, the relative change in PI with total psoas muscle area at L3 level in the 41 patients after 4 weeks was −0.210 ± 0.315 cm 2 /m 2 , while that in the 25 patients after 12 weeks was −0.275 ± 0.372 cm 2 /m 2 .

| DISCUSSION
Development of TKI treatment for u-HCC has improved the prognosis of affected patients. Following the introduction of SOR, global trials of sunitinib, 26 brivanib, 27 linifanib, 28 and erlotinib plus SOR 29 were conducted, though none of those drugs showed superiority and each of the trials ended in failure. Thus, there was no additional first-or second-line TKI treatment for u-HCC developed until the introductions of REG 3,4 and LEN. 6 Recently, REG was developed as a second-line option for SOR in patients who met the RESORCE trial criteria, 3 after which LEN became available as a firstline oral TKI targeting VEGF receptors 1-4, PDGF receptor α, RET, and KIT. 6 In the present study, early therapeutic response with LEN was favorable and patients obtained disease control within 4 weeks, as follow-up imaging showed a significant decline in log10 AFP from the baseline. Kuzuya et al 30 have reported that a decreased AFP ratio from the baseline at 4 weeks after introduction of SOR was a predictor of good therapeutic response (PR, SD). Thus, a decline of log10 AFP from the baseline within the first 4 weeks might also have predictive value for therapeutic response in patients receiving LEN treatment. In addition, a previous study noted that fever within 2 weeks after the start of SOR therapy may be useful as a predictor of favorable treatment response in patients with advanced HCC and receiving SOR treatment. 31 Three of our patients developed a fever within 2 weeks after the start of LEN (grade 1, 2, 3: each 1 patient), in whom therapeutic response was PD in 2 and SD in 1. It would be important to examine the clinical importance of fever in patients receiving LEN therapy as a useful predictor of a favorable treatment response in a future study with a larger number of patients.
Not all patients treated with SOR in clinical practice meet the RESORCE trial criteria; thus, the frequency of u-HCC patients indicated for REG has been reported to only range from 30.6% to 37.0% of those confirmed by radiological In the REFLECT trial reported by Kudo, 6 the frequency of HFSR was less in patients that received SOR (all grades: 26.9% vs 52.4%), whereas all grades of HFSR were observed more frequently in the present analysis (40.3%). Although there was no statistical difference in the frequency of HFSR between our TKI-naïve and TKI-experienced groups (33.3% vs 45.5%, P = 0.350), past history of high-grade HFSR might be an important risk factor of it developing during LEN therapy. Thus, HFSR should be kept in mind as a commonly occurring AE in patients with a history of that condition during previous SOR and/or REG treatments. HFSR in association with SOR has been reported as a factor predicting a better therapeutic effect of SOR for HCC. 36,37 Based on these findings, the relationship between HFSR and therapeutic response with LEN treatment should be analyzed in the future. Nevertheless, countermeasures against HFSR are most important for patients with a past history of HFSR in order to maintain TKI adherence for enhancing the anti-cancer effect.
In the present cohort, Child-Pugh class and ALBI score became worse at 4 and 12 weeks after starting LEN treatment. In addition to HFSR, general fatigue and appetite loss were the second and third most frequent AEs seen in the present analysis, and it is thought that LEN and these AEs have direct effects on liver function and nutritional status. To maintain adherence to LEN therapy, it is important not only to monitor and treat as carefully as possible AEs but also to keep in mind to introduce LEN, as well as SOR, as possible as in better liver function, if necessary for LEN treatment.
HCC is often seen in patients with chronic liver disease (CLD). Additionally, MVL is not rare in affected patients 14 and its frequency was shown to increase in association with the progression of CLD stage. 38 Moreover, it has been reported that MVL is a more significant prognostic factor than portal hypertension in liver cirrhosis (LC) patients (Child-Pugh class A/B), 39 and also an important prognostic factor following not only surgical resection 40 but also SOR treatment 41 in patients with HCC. Recently, a metaanalysis study reported that MVL was a prognostic factor regardless of therapeutic modality given for HCC by Chang et al. 42 Furthermore, in addition to LC and HCC, muscle volume has been reported to be an important prognostic factor in patients receiving treatments for other types of F I G U R E 5 Hand-foot skin reaction occurrence in tyrosine kinase inhibitor-experienced group. Of 8 patients who had undergone past tyrosine kinase inhibitor (TKI) treatments and developed a grade 3 hand-foot skin reaction (HFSR), 87.5% had an HFSR with lenvatinib (LEN) therapy (G3, 2, 1; n = 2, 3, 2, respectively). Furthermore of 18 with a grade 2 HFSR in past TKI treatments, 44.4% had an HFSR with LEN (G3, 2, 1; n = 3, 3, 2, respectively). Of 4 with a grade 1 HFSR in previous TKI treatments, grade 2 occurred in 1 and no HFSR was seen in 3 with LEN treatment. In 14 without an HFSR in past TKI treatments, 28.6% developed HFSR with LEN treatment (G3, 2, 1; n = 3, 3, 2, respectively) cancers. 43 In the present study, a fair percentage of our patients (35.3%) had MVL and a relative decline in muscle volume at 4 and 12 weeks after starting LEN was observed in patients who had data of CT imaging findings available, similar to our past study of patients treated with SOR. 41 It will be important to analyze the relationships among therapeutic response, OSR, and muscle volume in patients undergoing LEN therapy in a future study. Although some trials of intervention with nutrition and exercise CLD patients have been reported, 44,45 effective intervention methods have yet to be reported. Establishment of strategies for preventing and improving muscle wasting in CLD patients with and without HCC is needed for improving prognosis.
The present study has some limitations, including its retrospective nature. Additionally, though this was a multicenter study, the number of analyzed patients was not large and the observation period was limited. In a future study, we hope to examine the relationship between prognosis and clinical features including AEs and muscle volume with a larger number of patients receiving LEN therapy.
In summary, regardless of previous TKI treatments, therapeutic response and AE occurrence following introduction of LEN treatments were similar. We consider that LEN might complement the unmet need for TKI therapy for patients with u-HCC.