Tumor growth rate as a prognostic factor for metastatic or recurrent adenoid cystic carcinoma of the head and neck patients treated with carboplatin plus paclitaxel

Large prospective studies of chemotherapy for metastatic or recurrent adenoid cystic carcinoma (ACC) of the head and neck are lacking due to the rarity of ACC. The aim of this study is to evaluate the efficacy of carboplatin plus paclitaxel toward ACC and perform an exploratory investigation of the prognostic factors to investigate the optimal strategy for metastatic or recurrent ACC. We retrospectively analyzed recurrent or metastatic ACC patients treated with carboplatin plus paclitaxel between April 2007 and September 2019 in our hospital. Objective response rate (ORR), disease control rate (DCR), progression-free survival (PFS), and overall survival (OS) were evaluated, and an exploratory analysis of the prognostic factors was conducted. A total of 26 ACC patients were enrolled. ORR and DCR were 11.5 and 76.9%; the median PFS and OS were 8.1 and 22.3 months, respectively. From the results of the multivariate analysis, higher (≥ 6%/month) tumor growth rate (TGR) was associated with worse PFS (hazard ratio [HR] 7.00, 95% CI 1.34–36.53, p = 0.02) and OS (HR 29.33, 95% CI 3.38–254.80, p < 0.01). The median PFS (10.6 vs. 6.6 months, log-rank p < 0.05) and OS (48.5 vs. 16.9 months, log-rank p < 0.01) were significantly shorter in patients with higher TGR. Carboplatin plus paclitaxel showed modest efficacy for recurrent or metastatic ACC patients. Watchful waiting may be optimal for ACC patients with lower TGR. Systemic chemotherapy should be considered when TGR increases during active surveillance.


Introduction
Adenoid cystic carcinoma (ACC) is a rare form of head and neck cancer. The yearly incidence of ACC is reported to be 3-4.5 cases per million, which represents approximately 1% of all head and neck cancers [1,2]. It mainly originates from major and minor salivary glands, accounting for about 10% of all salivary gland tumors [3]. The prognosis of ACC is reported to be indolent, with a 5-and 10-year overall survival (OS) rate of 90.3 and 79.9%, respectively [4]. However, it is difficult to cure ACC, with a recurrence rate of up to 100% 30-years after curative treatment [5]. Therefore, most ACC patients require salvage therapy, including chemotherapy, at some point in their clinical courses.
The evidence for the efficacy of salvage chemotherapy in recurrent or metastatic ACC patients is not well established. Several reports describe the modest efficacy of carboplatin plus paclitaxel in patients with salivary gland carcinoma (SGC), including patients with ACC and non-ACC histology [6][7][8]. Moreover, carboplatin plus paclitaxel has frequently been used as a therapeutic option for recurrent or metastatic ACC patients in real-world clinical practice, as it can be conveniently administered in an outpatient setting. However, due to the rarity of ACC, the number of patients included in previous studies is small. The optimal timing to initiate systemic chemotherapy for metastatic or recurrent ACC is controversial, as ACC is a slow-growing and 1 3 chemotherapy-resistant tumor. Some reports suggested that watchful waiting might be the appropriate strategy for asymptomatic and indolent SGC [9]. We, therefore, analyzed clinical outcomes of ACC patients treated with carboplatin plus paclitaxel in an effort to assess treatment efficacy and performed an exploratory analysis on prognostic factors of carboplatin plus paclitaxel in the treatment of ACC to investigate the optimal timing to initiate chemotherapy.

Patients and methods
We performed a retrospective analysis of patients with recurrent or metastatic ACC who were treated with carboplatin plus paclitaxel from April 2007 to September 2019 at the Department of Medical Oncology of the Cancer Institute Hospital of the Japanese Foundation for Cancer Research. All patients were administered a treatment consisting of carboplatin (area under the curve [AUC] = 6) and paclitaxel (200 mg/m 2 ) once every 3 weeks. Premedication with an H 1 and H 2 histamine antagonist plus dexamethasone was intravenously administered before each infusion. Intravenous administration of a serotonin-3 (5-HT 3 ) antagonist was provided for antiemetic prophylaxis. Dose modifications or delays during the treatment schedules were allowed according to the physicians' discretion.
Treatment responses were evaluated by computed tomography (CT) scans according to the Response Evaluation Criteria in Solid Tumors (RECIST) criteria (ver. 1.1). Progression-free survival (PFS) was defined as the time between treatment initiation and disease progression or death by any cause. We defined overall survival (OS) as the time between treatment initiation and death by any cause. Tumor growth rate (TGR) was defined as the percentage change in tumor volume over 1 month (%/month). TGR was calculated by TGR = 100 × (exp(TG)-1), where TG = 3 × log(D2/D1)/time (months) [10,11]. Tumor size (D) was determined using the sum of longest diameters of target lesions according to RECIST v1.1. Non-target and new lesions were excluded. D1 represents tumor sizes at the penultimate CT evaluation date (date 1), and D2 represents the CT evaluation date (date 2) immediately prior to starting carboplatin plus paclitaxel. Time (months) is defined as (date 2 − date 1 + 1) × 365/12.
We used EZR (R ver. 2.13.0) software to perform the statistical analyses [12]. PFS and OS were estimated by the Kaplan-Meier method and were compared using a logrank test. The survival results are expressed as the median value with a 95% confidence interval (CI). The Cox hazard regression model was used to analyze prognostic factors. A stepwise Cox proportional hazards model was used to evaluate independent prognostic factors for each variable that was significant in the univariate analysis. We selected final covariates using a backward stepdown selection process with the Akaike information criterion. The association between TGR and survival time was analyzed using Spearman's rank correlation coefficient. Statistical significance was defined as p < 0.05.

Patients and treatment delivery
Between April 2007 and September 2019, 26 ACC patients were treated with carboplatin plus paclitaxel at our hospital. Patient characteristics are shown in Table 1. The median age was 54 years (range 30-74), and 15 (58%) patients were female. The median body mass index (BMI) was 20.8 (range 14.1-27.7) kg/m 2 . The most common primary tumor site was the submandibular gland (34.6%), followed by the oral cavity (19.2%), and the parotid gland (15.4%). Approximately 58% of the patients had received surgery,

Discussion
A standard chemotherapeutic regimen for recurrent or metastatic ACC has not been previously established. This retrospective analysis evaluated the efficacy of carboplatin plus paclitaxel in a larger number of ACC patients than in previous reports. Moreover, this is the first report to evaluate the prognostic factors for PFS and OS in ACC patients treated with carboplatin plus paclitaxel. The results of our study indicate a modest efficacy of carboplatin plus paclitaxel for recurrent or metastatic ACC patients. There are a limited number of prospective studies investigating the effects of chemotherapy for ACC. A small prospective trial involving a regimen of cyclophosphamide, doxorubicin, and cisplatin (CAP) on 22 SGC patients, including 12 ACC patients, demonstrated a 25% ORR for the ACC patients [13,14]. However, this triplet regimen is difficult to administer in an outpatient setting, due to the amount of hydration required for cisplatin administration. Recently, several single-arm prospective phase 2 trials of multi-kinase inhibitors (MKIs), such as sunitinib, sorafenib, axitinib, dasatinib, and lenvatinib, for recurrent or metastatic ACC patients have been conducted [15][16][17][18][19][20]. In these trials, the number of patients enrolled were 14-40, the ORR were 0-15%, and the median PFS were 4.8-17.5 months. To date, these MKIs are not approved internationally for use in ACC patients. Pembrolizumab, a monoclonal antibody targeting programmed cell death 1 (PD-1), has been approved for metastatic or recurrent programmed cell death ligand 1 (PD-L1) positive cancers, regardless of their primary sites. However, PD-L1 positivity in ACC is reported to be low [21]. In a phase 1b trial of pembrolizumab for various PD-L1 positive solid tumors, an objective response was not observed in ACC patients (0/2), even when PD-L1 positive [22]. In several prospective phase 2 trials, the ORR was 8.6% (4/46) in nivolumab alone, and 6% (2/32) in nivolumab plus ipilimumab, suggesting the limited efficacy of immune checkpoint inhibitors for ACC [23,24]. Despite being retrospective data, the ORR, PFS, and OS of carboplatin plus  A large multicenter prospective study (N = 470) for ACC suggested that age, BMI, and lymph node invasion were significant prognostic factors for ACC [25]. Our exploratory analysis of prognostic factors indicated a link between higher TGR or poor ECOG PS and worse PFS and OS. BMI was also a prognostic factor for PFS in our results, which was consistent with the previous study. In the previous cohort, only 18% of the patients underwent chemotherapy and the detailed therapeutic data were excluded from the search for prognostic factors. Therefore, our results can be meaningful as all patients included in the cohort received chemotherapy with the same therapeutic regimen.
The ORR and DCR of our cohort were 11.5 and 76.9%, suggesting that the aim of a carboplatin plus paclitaxel regimen should be disease control, not tumor shrinkage. Patients with lower TGR demonstrated longer OS (median OS; 48.5 months). Therefore, watchful waiting and active surveillance can be an optimal strategy for patients with slowgrowing (lower TGR) ACC. Initiation of systemic chemotherapy should be considered when TGR increases up to 6%/month during active surveillance or the patient develops tumor-related symptoms affecting ECOG PS. Patients with massive tumor volumes, aggressive TGR or severe tumorrelated symptoms might not benefit from carboplatin plus paclitaxel. For these patients, palliative care would be preferable to intensive systemic chemotherapy, which accounts for the patients' quality of life.
Our study has several limitations. First, this was a retrospective analysis of a small number of patients at a single institute. Due to the rarity of ACC, not more than 40 ACC patients were enrolled in previous prospective studies investigating chemotherapy. Our results should be interpreted in the context of the difficulty of conducting large randomized phase 3 trials for ACC. Second, although the pathological specimens were diagnosed by a pathologist who specializes in head and neck cancers, genetic confirmation with MYB fusion genetic testing could not be performed in the majority of the patients. Third, some patients were excluded by the TGR assessments due to the absence of measurable target lesions. This may cause selection bias. Though we used a cutoff of 6%/month for TGR according to the median value of enrolled patients, the optimal cutoff value was not established. To resolve these limitations, we are now planning a prospective multicenter cohort study.

Conclusions
Carboplatin plus paclitaxel showed modest efficacy in patients with metastatic or recurrent head and neck ACC. This treatment regimen represents a valid therapeutic option for selected ACC patients in an outpatient setting. TGR and ECOG PS were independent prognostic factors for PFS and OS. During watchful waiting and active surveillance, when TGR increases up to 6%/month, initiation of chemotherapy should be considered. A prospective multicenter cohort study with a greater number of patients is required to establish firm evidence supporting this treatment regimen for this rare cancer.