Safety and efficacy profile of lenvatinib in cancer therapy: a systematic review and meta-analysis

To systematically review the safety and efficacy of lenvatinib in the treatment of patients, we retrieved all the relevant clinical trials on the adverse events (AEs) and survival outcomes of lenvatinib through PubMed, Medline, Embase, Web of Science and Cochrane Collaboration's Central register of controlled trial. Fourteen eligible studies involving a total of 978 patients were included in our analysis. The most common all-grade AEs observed in patients treated with lenvatinib were hematuria (56.6%), fatigue (52.2%) and decreased appetite (50.5%). The most frequently observed grade ≥3 AEs were thrombocytopenia (25.4%), hypertension (17.7%) and edema peripheral (15.5%). The incidences of both all-grade and high-grade hypertension were significantly increased. Meanwhile, the controlled trial suggested that progression free survival (PFS) was significantly longer in the lenvatinib group than the placebo group. Subgroup analyses showed that mean PFS for renal cell carcinoma was 10.933±1.828 months (95% CI 7.350-14.515, p < 0.001), and that for thyroid cancer was 18.344±0.083 months (95% CI 18.181-18.506, p < 0.001). In conclusion, lenvatinib is an effective agent in thyroid cancer. Early monitoring and effective management of side effects are crucial for the safe use of this drug.


INTRODUCTION
Angiogenesis is critical for the local invasion and progression of tumor cells [1]. The aberrant formation and proliferation of blood vessels is due to an imbalance in pro-and anti-angiogenic factors, with the first weighing more [2]. Vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF), platelet-derived growth factor (PDGF) and epidermal growth factor (EGF) are several positive regulators of angiogenesis [3]. Over the last decade, multi-targeted tyrosine kinase inhibitors (TKIs) have been developed and approved in clinical oncology practice [4].
Lenvatinib has brought clinical benefits for patients, but adverse events (AEs) are inevitable such as hypertension, fatigue, proteinuria, nausea, decreased weight and abdominal pain, which may decrease the quality of life of patients and influence their acceptance of treatment [11,12]. Therefore, we conducted a meta-analysis to estimate various AEs and clinical benefits of lenvatinib.

Literature search results
We ran an initial broad search that yielded 422 unique articles after deletion of duplicates. After title and abstract screening, 344 were excluded since they were narrative review articles or interviews, or completely not associated with clinical assessment of lenvatinib. Forty were further excluded for they were conference abstracts based on published clinical trials, leaving 38 potentially relevant studies for full review. After estimating the full texts of these articles, 24 articles were ruled out for insufficient information. Ultimately, 14 eligible studies [13][14][15][16][17][18][19][20][21][22][23][24][25][26] involving a total of 978 patients met our metaanalysis criteria. Two articles [22,26] with the same first author which had different study designs were both included in our study, one was a phase II trial, and the other was a phase III, randomized multicenter study. No additional unpublished trials were added to the literature search results. A flow diagram of the trial selection process is provided in Figure 1.

Study characteristics
Of the studies that were included in the final analysis, 3 studies were based on thyroid cancer patients, 5 evaluated advanced solid tumors, 1 evaluated non-small-cell lung cancer, 1 was based on melanoma, 2 were performed on metastatic renal cell carcinoma, 1 was on advanced hepatocellular carcinoma and 1 on healthy adults. Schlumberger M [22] compared lenvatinib with placebo in radioiodine-refractory thyroid cancer patients, and Motzer RJ [19] used lenvatinibeither in combination with everolimus or as a single agent in patients with metastatic renal cell carcinoma. The characteristics of each trial are summarized in Table 1.

Survival outcomes and subgroup analysis
The efficacy analysis of lenvatinib was mainly based on the controlled trial of lenvatinib in patients with thyroid cancer [22]. The median progression-free survival was 18.3 months in the lenvatinib group and 3.6 months in the placebo group (hazard ratio for progression or death 0.21, 99% CI 0.14-0.31, P < 0.001). In addition, Motzer RJ [19] reported that median PFS was 7.4 months (95% CI 5.6-10.2) for single-agent lenvatinib in patients with metastatic renal cell carcinoma and 5.5 months (95% CI 3.5-7.1) for single-agent everolimus, representing the significantly prolonged PFS of lenvatinib compared with everolimus alone (HR 0.61, 95% CI 0.38-0.98, p = 0.048). Seven trials [13,15,16,19,22,25,26] reported encouraging response rates, median time to response, or PFS observed in patients with different types of tumors, demonstrating the antitumour efficacy of lenvatinib ( MTC: medullary thyroid cancer; RR-DTC: radioiodine-refractory, differentiated thyroid cancer Schlumberger M 1 and 2 : the former was a single-arm trial, while the latter was a controlled trial The all-grade adverse event rates and 95% CIs using a random-effects model; c. The grade ≥ 3 adverse event rates and 95% CIs using a fixed-effects model; d. The grade ≥ 3 adverse event rates and 95% CIs using a random-effects model.

Risk of bias and quality assessment
The risk of bias and quality assessments of the included studies are outlined in Figure 3a, 3b. Overall, the quality of the studies was satisfactory.

DISCUSSION
To the best of our knowledge, this is the first study to evaluate both the safety and efficacy of the novel antitumor agent lenvantinib in different types of tumors systematically. The adverse events of lenvatinib were tyrosine kinase inhibitor-related and were also seen in other TKIs. In one meta-analysis [27], the VEGFR-TKIs group (cediranib and axitinib) was associated with higher rates of diarrhea, fatigue, hypertension and thrombocytopenia compared with bevacizumab. Vandetanib [28], a dual VEGFR and EGFR inhibitor, yielded an improvement in PFS but more frequent grade 3 or greater hypertension. Although the incidence of hematuria was high, most people experienced low grade (grade 0) hematuria.
It should be noted that lenvatinib was associated with a significantly increased risk in all-grade (47.0%) and high-grade (17.7%) hypertension. The mechanism of lenvatinib-associated hypertension has not been clarified, and may be due to a possible perturbation of endothelial cell function in patients treated with VEGF-targeting agents [29]. It has been documented upon administration of bevacuzimab and cediranib, and several other inhibitors of the VEGF signalling pathway [30][31][32]. All of these suggest that patients who were administered lenvatinib should be monitored for high blood pressure, and managed with antihypertensive drugs or dose reductions when necessary.
Grade ≥ 3 thrombocytopenia was experienced in about a quarter of patients. Through binding to PDGFR, PDGF promotes the recovery of platelets and the formation of bone marrow colony-forming unit-megakaryocyte [33,34], thus the inhibition of PDGFR by lenvatinib might cause thrombocytopenia. Hematopoietic growth factors and transfusions [35] could be used to deal with persistent toxicities on platelets, but the effects of them on tumor cells remain to be explored.
In February 2015, US FDA has approved lenvatinib for the treatment of radioiodine-refractory thyroid cancer [9] based on the randomized controlled trial [22] included in our analysis. We find a similar mean PFS (18.344±0.083 months, 95% CI 18.181-18.506, p < 0.001) for thyroid cancer in our pooled analysis. However, our results of adverse events (Figure 4a, 4b) are different, since the relatively larger sample size may allow us to better determine the AE values. Survival outcomes of other tumors are mainly based on phase I and phase II trials, and more subsequent randomized, controlled phase III trials are needed.
The dose of lenvatinib administered in patients with solid tumors varied in different situations, but in 8 [15, 16, 18-20, 22, 23, 26] of the 14 included studies, patients received lenvatinib at a daily dose of 24 mg per day in 28-day cycles, and two studies [18,20] demonstrated that the 24-mg QD dose of lenvatinib was determined to be   tolerable with encouraging anti-tumor activity in patients with solid tumors. The heterogeneity in our analysis could arise from different tumor types, the very heterogeneous study population with pre-treated disease and the ethnicity difference. In addition, there are several limitations of our study. Firstly, because lenvatinib is a relatively new drug, reports about it are few and are mostly phase I and II studies. Secondly, only one study provided the overall survival data, so prolonged follow-ups are needed. Thirdly, we did not perform subgroup analysis of melanoma and non-small-cell lung cancer because of lack of enough information.
In conclusion, lenvatinib has clinically meaningful benefits in survival outcomes of patients with thyroid cancer. The pooled analyses suggest that patients should be monitored for potential thrombocytopenia and increases in blood pressure, and dose reductions or delays or antihypertensive drugs are needed accordingly. Correct estimates of treatment-related toxicities and the efficacy of lenvatinib are fundamental to provide appropriate guidance and to conduct ongoing trials.

Search strategy
We performed a literature search of PubMed, Medline, Embase, Web of Science and The Cochrane Library for all the relevant clinical trials on the safety and efficacy of lenvatinib (until April 26, 2016, 201). In order to ensure the completeness of the results, we expanded the search scope by using the search terms "lenvatinib" or "E7080" or "lenvima". We also carried out further searches for relevant unpublished trials in the clinical trial registry (http://www.clinicaltrials.gov). Papers in all languages were sought and translated where appropriate to reduce the chances of bias.

Inclusion and exclusion criteria
To be included in the analysis, patients must be diagnosed with histologically confirmed tumors, survival outcomes and toxicities were mandatory to be reported. All phase clinical trials were eligible for inclusion if they evaluated the side effects and efficacy of lenvatinib. Studies were excluded if they did not provide enough data for toxicities and survival outcomes. They were also excluded for which full-text reports were not available.

Selection process and data extraction
Two reviewers selected studies independently. Any disagreements were resolved through discussion with another author. We excluded those studies that clearly did not meet the inclusion criteria, and made efforts to rule out duplicated studies by comparing author lists, publication year, and the main contents if necessary. Articles with the same author(s) or medical center(s) were carefully reviewed and discussed for eligibility.
Data extracted from all eligible articles included the first author, year of publication, sample size, study phase, tumor type, treatment regime, progression-free survival (PFS), hazard ratio (HR) and adverse events. ADRs were graded using the National Cancer Institute (Washington, DC, USA) Common Toxicity Criteria, version 3.0.

Data analysis
We used patients' all-grade and grade ≥ 3 Common Toxicity Criteria Adverse Events (CTCAE) counts to calculate the incidence rates of AEs and the corresponding 95% confidence intervals (CIs). I-squared was calculated to test heterogeneity of the studies, and I 2 > 50% and P ≤ 0.1 indicated strong heterogeneity between the studies. All the analysis was carried out using the software Comprehensive Meta-Analysis (CMA) program 2 (Biostat, Englewood, NJ) and Review manager 5.3 (Copenhagen, Sweden).

Risk of bias and quality assessment
To evaluate the risk of bias and quality of the studies, QUADAS-2 was used as a systematic review assessment method, which consisted of four key domains: patient selection, index test, reference standard and flow and timing [36]. Risk of bias was rated as high/low/ unclear. The assessment was measured using Review Manager 5.3 (Copenhagen, Sweden).