Weekly chemotherapy as first line treatment in frail head and neck cancer patients in the immunotherapy era

Objective First-line therapy for metastatic squamous cell carcinoma of the head and neck (R/M HNSCC) has been revolutionized by the introduction of anti-checkpoint monoclonal antibodies, which have shown a significant improvement in overall survival (OS) gaining approval in a first line setting. Efficacy and safety of first-line weekly chemotherapy, compared to 3-weeks treatment, was retrospectively evaluated in a frail patient population with R/M HNSCC with the aim to evaluate its role as part of a personalized first-line approach. Methods A total of 124 patients with locally incurable R/M HNSCC receiving weekly (21) or three-weekly (103) chemotherapy plus cetuximab in a first line setting from December 2010 to September 2020 were retrospectively reviewed. Treatment outcomes in terms of objective response rate (ORR), progression-free survival (PFS), overall survival (OS) and toxicities were analysed. Results Patients in the three-week subgroup were ECOG PS 0 (39) and 1 (64) while patients in weekly group (21) were all PS 2. No significant differences were reported in terms of age, sex, smoking and previous alcohol abuse considering the two distinct subgroups. Moreover, no statistically significant difference was found in PFS and OS between the two treatment subgroups. The response rate was 35% (36 patients) and 34% (7 patients) in three-week and weekly treatment group, respectively. Seventy patients (68%) in the three-week group experienced chemotherapy-related toxicities, predominantly G3. In the weekly group a predominantly low-grade toxicity was found in a lower number of patients (52%). Conclusion The weekly schedule appears to be an active and safe strategy in frail patients with R/M HNSCC. Based on these data, a weekly schedule could be considered as a first line treatment in all frail patients excluded from pembrolizumab treatment and a study on the combination of weekly chemotherapy and immunotherapy should be performed.


Introduction
Head and neck squamous cell carcinoma (HNSCC) represent a heterogeneous spectrum of diseases originating predominantly from oral cavity, oropharynx, hypopharynx, and larynx 1 . HNSCC is globally the sixth most common type of cancer with 830,000 new cases and around 430,000 deaths each year worldwide 2 . Recurrent/metastatic head and neck squamous cell carcinoma (R/M-HNSCC), locally incurable, is associated with a poor prognosis 3 . The old standard treatment for R/M HNSCC, incurable with local therapies, was platinum-based chemotherapy plus cetuximab according to the therapeutic regimen EXTREME 4 . In this phase III trial, Vermorken et al. demonstrated that the addiction of cetuximab to platinum-5Fluoruracil (5FU) chemotherapy prolonged the median OS and PFS (from 7.4 months to 10.1 months and from 3.3 to 5.6 months, respectively). There were no signi cant differences in terms of Grade 3 and 4 toxicities between the two arms (82% and 76% in cetuximab and chemotherapy alone arm, respectively); however, in cetuximab arm there were signi cantly more cases of sepsis (p = 0.02) and hypomagnesemia (0.05) compared to platinum-based chemotherapy alone group 4 .
Subsequently, a large randomized trial con rmed the good survival outcomes and response rate of the taxane-based TPE-x regimen observed in the phase II GORTEC study in rst line R/M HNSCC 5 . Despite the lack of signi cant increase of overall survival (OS) compared to EXTREME, TPEx required a shorter duration (4 vs 6 cycles) with a lower toxicity pro le resulting in improvement of quality of life 6 . Furthermore, recent evidence have shown that docetaxel was able to act on the tumour microenvironment favouring the priming of immune response. This mechanism supports the hypothesis that taxanes act synergistically with immunotherapy, highlighting the importance of their use in clinical practice 7,8,9,10 .
Nevertheless, a large proportion of patients with R / M HNSCC have poor clinical conditions, weight loss and symptoms, such as pain or obstruction related to the extent of the cancer, comorbidities and impaired organ function but are not suitable for EXTREME treatment due to the high risk of developing high grade toxicity. Conversely, a weekly chemotherapy regimen plus cetuximab was considered a reliable therapeutic strategy in frail cancer patients with R/M-HNSCC 11 . Weekly taxane-based chemotherapy has proved to control cancer growth and related symptoms, with reduced toxicity and with an appropriate safety pro le 12 .
Currently, rst-line therapy for R/M-HNSCC has been revolutionized by the introduction of immunecheckpoint monoclonal antibodies inhibitors (ICIs), a class of drugs targeting the inhibitory immunecheckpoint receptors. The open label randomized phase 3 study KEYNOTE-048 results led to approval of the anti PD-1 pembrolizumab in rst line setting alone or in combination with cisplatin / 5 uorouracilbased chemotherapy 13 . The study evidenced a signi cantly prolonged OS vs EXTREME in patients with PD-L1 combined positive score (CPS) >/=20 and CPS >/= 1. Toxicity pro le was favorable for pembrolizumab vs EXTREME and comparable for pembrolizumab plus chemotherapy. The results were con rmed at a long-term 48-months follow-up 14 .
Nevertheless, some questions are still to be de ned. The KEYNOTE-048 study enrolled patients in good general condition and ECOG Performance Status (PS) = 0/1, although R/M HNSCC with poor clinical condition (PS = 2) and symptoms, were excluded from this clinical changing study, making its results di cult to be extend to the frail patients' population.
In this context the combination strategy in frail patients still remains un unmet need.
This study is carried out to retrospectively observe the clinical outcomes and toxicities of patients with baseline poor clinical condition treated in rst line with a weekly taxane-based chemotherapy, compared to patients, with good PS, treated EXTREME like chemotherapy.

Patients
This is a retrospective study, including patients with R/M HNSCC, who received rst line chemotherapy in association with cetuximab, in our center from December 2010 to September 2020.
Patients were clinically staged with contrast enhanced computed tomography (CT) scan and magnetic resonance imaging (MRI) before starting rst line chemotherapy. All patients were discussed and judged not eligible for locoregional treatments by the multidisciplinary team of the Policlinico Umberto I, Sapienza University of Rome.
Data including age, sex, ECOG PS, comorbidities, history of tobacco smoking and alcohol abuse, primary tumor sites, site of relapse (locoregional vs metastatic) and chemotherapy treatment schedule were retrospectively collected. On the basis of PS, related symptoms, age, nutritional status and comorbidities, patients were judged frail or clinically t and therefore scheduled for the two different treatments.
Patients who had comorbidities contraindicating platinum-based chemotherapy and/or cetuximab were not included in the analysis as well as all patients deemed unsuitable for chemotherapy who have undergone best supportive care.

Treatment and Assessments
Three weekly chemotherapy according to the Extreme schedule cisplatin at a dose of 100 mg per square meter of body-surface area on day 1, or carboplatin at an area under the curve (AUC) of 5 mg per milliliter/ minute on day 1, plus uorouracil (5-FU) at a dose of 1000 mg per square meter/day for 4 days, plus cetuximab at a dose of 400 mg per square meter in loading dose, then 250 mg per square meter, per week every 21 days for a maximum of 6 cycles, were administered to patients considered t at the baseline clinical evaluation. Cetuximab maintenance was performed in all patients who achieved at least stable disease (SD) as best response.
Weekly chemotherapy according to the modi ed schedule paclitaxel 80 mg per square meter day 1,8,15 carboplatin AUC 2 day 1,8, 15 and cetuximab 400 mg loading dose followed by 250 mg per square meter weekly every 21 days (PCC), were administered intravenously to those patients judged frail and un t for Extreme schedule.
Tumor response was assessed every 12 weeks using Response Evaluation Criteria in Solid Tumors (RECIST) and classi ed as complete response (CR), partial response (PR), SD, and progressive disease (PD). Toxicities were recorded at day 1 of every cycle and classi ed according to the National Cancer Institute Common Terminology Criteria for Adverse Events (version 4.0). Progression-free survival (PFS) was de ned as the time from patient's rst administration of treatment until the rst progression or treatment death. The OS was de ned as the time from patient registration to death from any cause.

Statistical Analysys
In the descriptive analysis, quantitative variables were described as median and range, while qualitative variables were reported as number and percentage. The association between each clinicopathological characteristic and outcomes was evaluated using the univariate and multivariable logistic regression model. Statistical signi cance was set at p < 0.05. All analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

Patients
A total of one hundred twenty-four patients with locally incurable R/M HNSCC treated with rst line platinum-based chemotherapy in association with cetuximab were included in this retrospective study. The baseline clinicopathological characteristics are reported in Table 1.
Patients who continue to smoke during treatment as well as previous alcohol abuse were reported in medical history in 26 (21%) and 33 (27 %) patients, respectively. The site of relapse was locoregional, metastatic or both in 50 (40 %), 59 (48%) and 15 (12 %) patients, respectively.
Overall, 103 patients (83%) received standard rst line platinum-based chemotherapy, according to the phase III EXTREME trial, while 21 patients (17%) received rst line PCC weekly chemotherapy.
As it is shown in Table 1, no signi cant difference was reported in terms of age, sex, smoking and previous alcohol abuse considering the two distinct subgroups of patients based on the received chemotherapy schedule. Instead, patients in the EXTREME subgroup were ECOG PS 0 (39) and 1 (64) while patients (21) in PCC scheme group were all PS 2.

Outcomes
PFS was 4 months (range 1-20) in the overall population; median PFS was 4 and 5 months in EXTREME and PCC schedule subgroup, respectively. No statistically signi cant difference was found in PFS between the two treatment subgroups (p = 0.275, Fig. 1). On the univariate analysis (Table 2), no clinical features, including the type of treatment regimen, were associated with the PFS.
OS was 12 months (range 1-50) in the overall population; median OS was 12 months both in Extreme and PCC schedule subgroups. No statistically signi cant difference was found in OS between the two treatment subgroups (p = 0.400, Fig. 2). On the univariate analysis (Table 3), none of the clinical features, including the type of treatment regimen, were found to be predictive of survival.
The response rate was 35 % in EXTREME group (36 patients) and 34 % in PCC group (7 patients) with a relevant palliative effect. None of the patients of both groups achieved a CR.

Discussion
In our retrospective study, frail patients treated with PCC chemotherapy regimen had similar PFS compared to t patients treated with EXTREME schedule (p = 0-275) in rst line setting. Therefore, our data support that modulating the chemotherapy schedule based on the patient's general condition is a reasonable and safe approach that does not compromise disease control in frail patients.
In daily clinical practise, the coexistence of several factors often makes patients with locally incurable R/M HNSCC fragile and di cult to manage: they often have an ECOG PS > 1, advanced age, severe symptoms and relevant comorbidities (such as diabetes mellitus type 2, arterial hypertension and cardiopathies) that often limit the use of chemotherapy. Moreover, patients with R/M HNSCC are characterized by frequent malnutrition, poor habits and social status 15 . Consequently, there is the urgent need to personalize the therapeutic approach in order to decrease toxicities and morbidities without compromising the oncological outcomes.
The treatment of recurrent and metastatic setting should be exible and should be easily adapted to the patient's clinical conditions, disease extension and treatment purpose.
In contrast to the PS 0-1 population included in randomized clinical trials, in clinical practice EXTREME is contraindicated in PS 2 patients for the high risk of infectious diseases and high-grade toxicities. Therefore, in literature cetuximab was evaluated in combination with different platinum-based regimens 16,17 . Several retrospective and prospective studies have evaluated rst-line chemotherapy based on the weekly combination of paclitaxel and cetuximab in R/M-HNSCC showing promising activity. The schedule has proven to be a relevant option in the treatment of patients with poor prognosis considered unsuitable for the EXTREME regimen 18 .
In a retrospective study, the combination of paclitaxel and cetuximab signi cantly prolonged the PFS compared to EXTREME regimen, above all in older male patients and in presence of tracheostomy 19 .
Pêtre A. and colleagues evaluated the activity and safety of the weekly carboplatin and paclitaxel chemotherapy in patients with R/M HNSCC un t for cisplatin due to comorbidities and poor baseline clinical condition. The weekly chemotherapy was e cient and well tolerated in this subgroup of particularly fragile patients 11 .
In our study the PCC chemotherapy regimen, used in a population with poor baseline clinical conditions, showed a non-inferior e cacy pro le with an excellent tolerability compared to the EXTREME regimen administered in PS 0-1 t patients. Furthermore, the safety pro le was optimal with no high-grade toxicity in our study population compared to the heavy toxicity pro le reported in the EXTREME group.
The e cacy is probably due to this optimal tolerability pro le, demonstrated by the low degree of toxicity, that allowed an optimal dose intensity even in this group of frail patients. The tolerability can be correlated to the exibility of the weekly treatment that allows delays and / or reductions in the dosage of drugs, highlighting the accumulation of high-grade toxicity with the subsequent delays for recovery. Our data are in line with Naverson et al. who have proved that the use of PCC, in patients affected by R/M-HNSCC with PS 0-2, appears to be a pro table therapeutic strategy with low toxicity. This retrospective analysis highlighted that 41 % of patients showed PR or CR and 34 % of patients showed SD 20 . Similarly, in our experience, the PCC chemotherapy regimen obtained 34% of ORR, demonstrating a relevant cytoreductive effect compared with EXTREME ORR.
Recently, the phase III Keynote 048 13 led to the approval of pembrolizumab alone or in combination with platinum/5FU chemotherapy in rst line in R/M HNSCC expressing PD-L1, showing an advantage in terms of OS, compared to standard EXTREME regimen in patients with tumours expressing combined proportion score (CPS) of 1 or more. Pembrolizumab in monotherapy could be the rst choice in frail patients' population, since they will not be able to receive chemo/immunotherapy due to the high toxicity rate of the combination. However, Keynote 048 includes a population of PS 0-1 patients with no comorbidities. Consequently, patients with poor clinical conditions and comorbidities are not represented in this clinical trial and unfortunately, there is no data regarding the association of pembrolizumab with weekly platinum-based chemotherapy schedule.
Therefore, the PCC schedule, for its optimal e cacy and toxicity pro le could be considered in rst line in all frail patients excluded from pembrolizumab treatment, in relation to negative CPS or due to the presence of clinical contraindications to immunotherapy treatment. A third group of patients which could potentially be candidates to the PCC treatment are those with frail condition requiring a rapid tumour response as HNSCC that is an aggressive disease often associated to rapid worsening symptoms, such as pain, breathing and feeding di culties (Fig. 3).
Our study has some limitations due to its monocentric-retrospective nature. On the contrary, the strength of our study is the evaluation of a frail population usually excluded from randomized clinical trials.
The challenge is to de ne personalized rst-line approach considering the patient's preference, socioeconomic status and the availability of caregivers to all head and neck cancer patients without excluding the frail patients' population.
Each treatment should be individualized on patient characteristics including PS and comorbidity, symptoms, and CPS, the need of rapid response and risk of complications, site of relapse, necessarily in the context of a multidisciplinary team evaluation. A personalized therapeutic approach in the complex frail patients' population can lead to excellent oncological outcomes gaining disease control without compromising the patient's quality of life.
In conclusion, our study con rms that a PCC schedule appears indicated in frail patients and could currently be considered in all frail patients excluded from immunotherapy treatment (CPS < 1) or when immunotherapy treatment has contraindications as well as in all cases where a prompt and deep response is required for disease control in critical sites. It appears advisable in the future to explore the combination of check point inhibitors with modulated chemotherapy regimens in frail patients' population.
In frail patients with R/M HNSCC a personalized rst line chemotherapy exible, low-toxic and effective may represent a therapeutic option when immunotherapy is contraindicated or a rapid cytoreductive response is required, allowing a personalization of rst-line treatment critical primarily in frail patients' population.

Declarations
All patients provided a written informed consent, and the protocol approval of Local Ethics Committee was obtained [CE 5618].
Consent for publication: "Not applicable".
Availability of data and materials: All data generated or analysed during this study are included in this published article [and its supplementary information les].

Figure 1
Kaplan Meyer curves. No statistically signi cant difference between the EXTREME and PCC schedule subgroups in terms of PFS was reported (p value 0.427).

Figure 2
Kaplan Meyer curves. No statistically signi cant differences between the EXTREME and PTC schedule subgroups in terms of OS was reported in our patient population (p value 0.400).