Outcomes of lenvatinib therapy in poorly differentiated thyroid carcinoma

Background and objective Lenvatinib showed promising results in a subgroup of patients with poorly differentiated thyroid carcinoma (PDTC) in the SELECT trial. Our aim was to report the effectiveness and tolerability of lenvatinib in our series of PDTC patients. Methods Medical records of eight consecutive patients with PDTC treated with lenvatinib in a single center between January 2019 and October 2022 were retrospectively reviewed. Inclusion criteria were PDTC diagnosis based on Turin criteria and evidence of disease progression in the previous 6 months. Results Eight PDTC patients received an average dose of lenvatinib of 18.1 mg for a median duration of treatment of 10.3 months. The baseline Eastern Cooperative Oncology Group performance status was ≥2 in 50% of patients. Two patients had unresectable primary tumor. Seven patients showed extrathyroidal disease, particularly mediastinal lymph nodes (85.7%), lung (71.4%), and bone (71.4%). The disease control rate was 100%, with partial response and stable disease in 12.5 and 87.5%, respectively. The median time to best overall response was 3 months, and the median duration of response was 7.5 months. Median progression-free survival was 12 months and median overall survival was not reached. At 6, 12, and 18 months, overall survival was 87.5, 71.4, and 57.1%, respectively. All patients experienced drug-related adverse effects (AEs). Four (50%) had dose reductions and two (25%) had temporary treatment interruptions. Lenvatinib was stopped in two patients due to grade ≥3 AEs. Conclusion Lenvatinib is an effective treatment for real-world PDTC patients. Adequate management of comorbidities and AEs increases treatment tolerability and minimizes dose reductions.


Introduction
The term poorly differentiated thyroid carcinoma (PDTC) was first introduced by Granner and colleagues in 1963, although it was only recognized as a distinct entity by the World Health Organization in 2004 (1,2). thyroid carcinoma (DTC) (1,4). At presentation, the majority of patients with PDTC have compressive symptoms with radiological evidence of extrathyroidal neck disease (1,4). Distant metastasis eventually emerges during follow-up in up to 85% of cases, being the cause of most disease-related deaths (1). While DTC diagnosis can be usually established by fine-needle aspiration cytology, PDTC usually requires histological examination (4). Survival rates range between 50-85, 34-50, and 0% at 5, 10, and 15 years after diagnosis, respectively (1,4,5). Higher mortality rates have been associated with older age, primary tumor >4 cm, pT3/pT4 staging, and presence of distant metastasis (5). Therefore, more effective treatment alternatives will be of utmost importance for patients with PDTC. The available treatment modalities are the same as radioiodine (RAI)-refractory DTC, despite the absence of specific trials focusing on PDTC patients. Total thyroidectomy with lymph node dissection is the preferred treatment modality, with a significant impact on disease progression if all gross disease is removed (1,4). Adjuvant treatment includes RAI, external beam radiotherapy (EBRT), and systemic chemotherapy. High RAI activities are recommended for patients with avid lesions. However, RAI avidity is quite variable in patients with PDTC, mainly due to the presence of less differentiated areas which decrease iodine-131 uptake and retention (1,4). EBRT can be considered after R2 surgery, loco-regionally recurrent disease, or bone metastasis. Nevertheless, survival benefit has not been established (1,6). Systemic cytotoxic chemotherapy has been used in a small number of cases without significant impact on the outcomes and with important side effects (1, 7  (11,12,13,14,15,16,17,18). Based on these premises, we have retrospectively reviewed the effectiveness and tolerability of lenvatinib in patients with PDTC followed at our center.

Patients
We retrospectively analyzed the medical records of all adult patients with the diagnosis of PDTC treated with lenvatinib in the Endocrinology department of Instituto Português de Oncologia de Lisboa Francisco Gentil Informed consent for the treatment was obtained from all patients. Institutional treatment protocols were strictly followed, and the study was conducted in compliance with the Helsinki Declaration and was approved by IPOLFG Ethics Committee.

Treatment and safety
Decision to start lenvatinib was made by a multidisciplinary team and followed evidence of progression in the last 6 months, either after surgery and adjuvant RAI or in progressive unresectable disease. Starting doses were defined according to patients' characteristics, particularly age, Eastern Cooperative Oncology Group (ECOG) performance status scale (https://ecog-acrin.org/resources/ecog-performancestatus/), and comorbidities. Whenever possible, lenvatinib was increased to the maximal tolerated dose. Also, dose reductions were performed based on tolerability and according to the manufacturer's recommendations. Interruption or definitive suspension of treatment was considered when grade 3+ AEs occurred (according to Common Terminology Criteria for Adverse Events 5.0 (https://ctep.cancer.gov/protocoldevelopment/ electronic_applications/ctc.htm) or in cases of persistent progression. Treatment of comorbidities was optimized before the initiation of lenvatinib. Clinical and biochemical follow-up was performed every month. All patients had a phone connection to the clinic for a prompt report of AEs and treatment.

Effectiveness measures
Treatment effectiveness was assessed by computed tomography (CT), according to RECIST criteria 1.1 (19). CT scans were performed every 3 months. In selected cases, 18 fluorodeoxyglucose positron emission/CT ( 18 FDG-PET/ CT) and magnetic resonance imaging were also used to evaluate progression. PFS, disease-specific survival (DSS), overall survival (OS), best overall response (BOR), disease control rate (DCR), and duration of response (DoR) were established as effectiveness outcomes. PFS was defined as the time between lenvatinib start and disease progression, death or last follow-up. OS and DSS were defined as the time between lenvatinib start and death by any cause or related to the disease, respectively. DoR was defined as the time between first evidence of response (stable disease, partial response, or complete response) and evidence of progression, death, or last follow-up.

Statistical analysis
Continuous variables were analyzed through descriptive statistics and respective results were presented as mean with s.d. and median with interquartile range. Categorical variables were expressed as absolute numbers or percentages. Kaplan-Meier curves were used to calculate PFS, DSS, and OS. Statistical analyses were performed using SPSS ® (v.23) software.

Patients' characteristics at baseline
Baseline characteristics of the eight patients are summarized in Table 2. The median age at diagnosis was 59 years old and only one patient (12.5%) had more than 65 years old, with a female predominance. ECOG performance status was zero in one (12.5%), one in three (37.5%), and two in four (50%) patients. RAI treatment was performed in all six patients who were previously submitted to surgery, and the median total activity was 150 mCi. Four patients were also submitted to local EBRT and one to chemotherapy with paclitaxel and carboplatin. EBRT was also performed in bone metastatic lesions in three patients and surgical excision of metastasis was undertaken in two patients due to compressive symptoms. Treatment with lenvatinib was initiated either due to evidence of progression after previous treatments (six patients, 75%) or in the neoadjuvant setting in patients with progressive inoperable disease (two patients, 25%). Loco-regional metastasis was present in four (50%) and distant metastasis was present in seven (87.5%) patients at lenvatinib start, particularly in mediastinal lymph nodes (six patients), lung (five patients), and bone (five patients).

Treatment outcomes
Details of lenvatinib treatment and effectiveness outcomes are listed in Table 3. The median duration of treatment was 10.3 months, and the median follow-up from lenvatinib start was 16.0 months. DCR was 100%, with a BOR of partial response in one patient (12.5%) and stable disease in seven patients (87.5%). As stable disease e230003 J Roque, T Nunes Silva et al.

12:2
includes a wide range of tumor responses, Fig. 1 details BOR according to the variation of summed longest diameter of target lesions in each patient. Two patients have been monitored with PET-FDG, so they could not be included in this analysis. The median time to BOR was 3 months, and the median DoR was 7.5 months. The median PFS was 12.0 months (Fig. 2). The median OS was not reached since more than half of the patients were alive by the end of the study. OS at 3, 6, 12, and 18 months was 87.5, 87.5, 71.4, and 57.1%, respectively. Two patients continued lenvatinib beyond the evidence of tumor progression. However, subsequent evaluations would confirm further progression of the disease, and treatment was suspended 3 and 6 months later, respectively. The two patients with unresectable disease at lenvatinib start could not be submitted to surgery later, one due to rapid progression of the disease and the second due to unaltered dimensions of the primary tumor. By study end, three patients have died from disease-related causes.
As such, in this study, DSS overlaps with OS. Five patients were still alive, one with partial response, three with stable disease, and one with disease progression.

Lenvatinib dosage, adverse effects, and tolerability
The median initial dose was 17 mg/day and only three patients have started with 24 mg/day due to ECOG performance status, age, and comorbidities. The median average dose of lenvatinib was 18.1 mg/day, confirming the possibility to increment drug dose in some patients. Reduction of dose and temporary interruption of treatment were necessary in four (50%) and two (25%) patients, respectively, due to grade 1-2 AEs. Of those, two patients needed multiple reductions and temporary suspensions due to weight loss and hand-foot syndrome, respectively. The median time to first treatment reduction was 6.5 months. Lenvatinib-related AEs were experienced by 100% of patients, and a total of 11 different AEs were reported. The median time to the first AE was 37 days. The most frequent were fatigue (6 patients, 75.0%), weight loss (5 patients, 62.5%), anorexia (5 patients, 62.5%), hypertension (4 patients, 50.0%), and diarrhea (four patients, 50%). Fatigue was mainly managed with more time for rest and sleep, and avoidance of heavy physical activities. Adrenal insufficiency was investigated in selected cases and was not confirmed in any patient. Grade 3 AEs were reported in two patients: tracheoesophageal fistula and leg cellulitis. The former occurred in a 34-year-old woman who was first submitted to a first total thyroidectomy with lymphadenectomy in another hospital, and then submitted to new cervical surgery, 150 mCi RAI and neck EBRT at our institution 2 years later. Considering the rapid progression, a low dose of 14 mg/day was initiated. However, tracheoesophageal fistula would develop 2 months after lenvatinib start,    leading to treatment suspension. A new progression of disease was evident shortly after and the patient eventually died 9 months after this event. The second patient was a 52-year-old man who was submitted to total thyroidectomy and lymphadenectomy and 150 mCi RAI 5 years before, has progressed over sorafenib, and showed stable disease for 18 months under lenvatinib. At this time, a diffuse itching cutaneous rash eventually emerged, evolving into right thigh cellulitis. Treatment with flucloxacillin was successful, but lenvatinib suspension was decided. He was under dexamethasone treatment for bone metastasis edema control, which may have facilitated the development of this complication. This patient was still alive at study-end, 9 months after this event. All reported AEs are detailed in Table 4.

Discussion
This study aimed to evaluate the real-world effectiveness and tolerability of lenvatinib in patients with PDTC. This represents a subgroup of RAI-refractory thyroid carcinomas with aggressive behavior and limited therapeutic options. As noted earlier, TKIs have been the most studied drugs in this setting, but specific results in PDTC patients are scarce. In the SELECT trial, lenvatinib elicited a PFS of 14.8 months (vs 2.1 with placebo) in PDTC patients (10). Since then, multiple real-world studies have also shown interesting results in RAI-refractory thyroid carcinoma patients, with PFS ranging from 10.0 to 35.3 months for both DTC and the minority of PDTC (11,12,13,14,15,16,17,18). Our study has reached a PFS of 12 months, which is significant in a population with confirmed progressive disease and a worse expected prognosis compared to overall RAI-refractory thyroid carcinomas. However, comparisons of results between studies should be made carefully, due to different cohort characteristics. Our cohort had evidence of active disease progression, with a progression rate in the last 6 months of 100%. This is different from other real-world studies. Also, in the SELECT trial, patients were included if radiologic evidence of disease progression was evident within the last 13 months, eventually selecting a cohort with less aggressive disease (10). Concerning ECOG performance status, our population had considerably unfavorable ratios, with 50% with ECOG grade 2, compared to less than 5% in SELECT and 0-18% in other real-world studies (10,11,12,13,14,15,16). Compared to the mentioned studies, the median initial dose was lower, which could have had a negative impact on clinical results. This was decided mainly due to pre-treatment performance status and in an attempt to increase treatment tolerability. Noteworthy, the median average dose was slightly higher than the initial dose, which suggests good tolerability. Our study has also included a significant proportion of TKI pre-treated patients (37.5%), which suggests the presence of more advanced and drug-resistant diseases. All of our patients experienced at least one AEs, which is similar to previous reports. Similarly, fatigue, weight loss, anorexia, hypertension, and diarrhea were the most frequent AEs reported. The majority of this grade 1-2 AEs were managed without lenvatinib dose reductions or suspensions. One patient with significant weight loss had transient dose reductions and could then sustain stable weight under megestrol acetate. Also, two patients with very frequent diarrhea episodes were successfully managed with loperamide and intermittent reductions of lenvatinib dose. Finally, one patient with hand-foot syndrome had multiple dose reductions and interruptions every time symptoms recurred. Limitations to our study include the small cohort, which is in line with the rarity of the disease. The retrospective design could also impact our results. However, all patients were treated by the same multidisciplinary team and according to the same protocols, which limits variability. In conclusion, this study is the first to specifically confirm the realworld effectiveness of lenvatinib in PDTC patients. This applies to patients previously treated with conventional therapy and TKIs, and, also, to patients treated in the neoadjuvant setting.

Declaration of interest
There is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.