Survival analyses of different treatment modalities and clinical stage for hypopharyngeal carcinoma

Objective We investigated the effects of different treatment modalities and clinical stage for hypopharyngeal carcinoma (HPC) patients. Methods Between February 2004 and December 2012, 167 HPC patients were reviewed. We calculated overall survival (OS), progression-free survival (PFS), local failure-free survival (LFFS), regional failure-free survival (RFFS), and distant metastasis failure-free survival (DMFFS) using the Kaplan–Meier method and compared various survival outcomes between definitive chemoradiotherapy (CRT) and surgery-based therapy (SBT). Results There were no significant differences in baseline characteristics between SBT (n = 102) and definitive CRT (n = 65) groups. The 5-year rates of OS (59.7% vs. 24.0%, p < 0.0001) and PFS (49.9% vs. 22.6%, p = 0.0002) were significantly better in patients who received SBT than in those who received definitive CRT. The SBT group also obtained better LFFS (p < 0.0001), RFFS (p = 0.0479), and DMFFS (p = 0.0110). We did similar analyses by different T-classification (T1–2, T3, and T4) and found that SBT had better OS (p < 0.0001 and p = 0.0020), PFS (p < 0.0001 and p = 0.0513), LFFS (p = 0.0002 and p = 0.0075), RFFS (p = 0.1949 and p = 0.0826), and DMFFS (p = 0.0248 and p = 0.0436) in the T4 and T1–2 subgroups but similar OS (p = 0.9598), PFS (p = 0.5052), RFFS (p = 0.9648), and DMFFS (p = 0.8239) in T3 patients. Analyses by different overall stages revealed no differences between definitive CRT and SBT for stage III patients but significantly better results for stage IV patients who received SBT. Conclusions SBT can obtain significant survival benefits when compared with definitive CRT for the whole cohort of patients. Definitive CRT has similar survival outcomes compared with SBT only for T3 tumors or overall stage III disease.


Introduction
Squamous cell carcinoma of the head and neck (SCCHN) develops from the mucosal epithelium in the oral cavity, oropharynx, hypopharynx, and larynx. SCCHN is the most common malignancy that arises in the head and neck region. It ranks the sixth most common cancer worldwide, with 890,000 new cases and 450,000 deaths in 2018 (1,2). Among different subsites of SCCHN, hypopharyngeal carcinoma (HPC) has the worst prognosis. The reported 5-year survival rates for stage III and IV HPC are only 36% and 24%, respectively (3). Because of a relatively low incidence rate, literature regarding treatment outcomes for HPC patients is not common, and prospective clinical trials that focused exclusively on HPC are very rare. Most clinical studies enrolled all SCCHN patients and HPC patients accounted for only a small proportion. In general, there are two different treatment approaches for advanced HPC, either surgery-based therapy or definitive chemoradiotherapy (CRT). So far, no consensus has been established on which treatment modality should be the standard of care. In this study, we reported the long-term treatment results of HPC patients and investigated the survival impacts of different treatment modalities, clinical T-classification, and overall stage as a reference for future therapy.

Patient selection
The major inclusion criteria for this retrospective study were patients with previously untreated, biopsy-proven squamous cell carcinoma of the hypopharynx who received curative treatment in the Taipei Veterans General Hospital, Taipei, Taiwan, during the period from February 2004 through December 2012. The staging criterion was according to the 7th edition American Joint Committee on Cancer TNM staging system (4). The exclusion criteria were patients with stage I, initial distant metastasis, nonsquamous cell carcinoma histology, synchronous multiple primary cancers, incomplete medical charts, and loss of follow-up. This study was approved by the Institutional Review Board of our hospital, and informed consent was waived because of the retrospective chart review without any intervention and contact with patients.
According to our institutional guideline, we offered organ preservation treatment, either transoral laser microsurgery or definitive (chemo)radiotherapy, to clinical T1 and T2 hypopharyngeal carcinoma patients. For clinical T3 tumors with laryngeal dysfunction, such as airway compromise, vocal fold fixation, or post-cricoid tumor with esophageal inlet or interarytenoid invasion, we recommended pharyngectomy with total laryngectomy; for those without laryngeal dysfunction, we applied organ preservation treatment, either transoral laser microsurgery or definitive CRT. For the majority of T4 tumors, pharyngectomy with total laryngectomy was advised for operable ones and otherwise patients underwent definitive CRT. Thanks to the National Health Insurance in Taiwan covering nearly 100% of the population, our patients had similar accessibility to definitive CRT and surgerybased therapy with affordable payment. The decision on which treatment strategy to receive depended on not only tumor factors including surgical risks and our institutional guidelines but also patients' preferences and performance status.
A total of 167 patients entered the final analysis. Table 1 demonstrates patient characteristics. There were 161 men and 6 women. Their age ranged from 33 to 91 (median 57) years. The primary origin of most patients arose from the pyriform sinus (n = 148) and the remaining from the posterior hypopharyngeal wall or post-cricoid area (n = 19). The clinical stage distribution revealed 16 patients with stage II, 35 with stage III, and 116 with stage IV. Sixtyfive patients received definitive CRT, and 102 patients received surgery-based therapy.
The most frequently used regimen of induction chemotherapy was triweekly cisplatin and 5-fluorouracil (PF) before 2009. It was shifted to TPF (taxotere+PF) afterward. The concurrent chemotherapy used in these years was triweekly high-dose cisplatin (80-100 mg/m 2 ) or weekly low-dose cisplatin 30-40 mg/ m 2 . BioRT represented a loading dose of cetuximab 400 mg/m 2 followed by weekly 250 mg/m 2 concurrently with radiotherapy (RT). No adjuvant chemotherapy was applied in these patients.
The RT was administered by intensity-modulated radiotherapy (IMRT) technique with a total dose of 70 Gy/35 fractions to the primary tumor site and metastatic regional nodal area plus a 2-4mm margin via a conventional fractionation schedule (a daily 2 Gy, 5 days per week).

Surgery-based treatment
Regarding the surgical treatment techniques for the primary site, we offered radical open surgery including total laryngectomy for cases with difficult endoscopic tumor exposure, arytenoid fixation, cartilage destruction, neck soft tissue invasion, or poor pulmonary function and organ preservation therapy with transoral laser microsurgery for the others. As for the neck, ipsilateral or bilateral neck dissection was advised for all patients. Postoperative adjuvant RT/CCRT/BioRT was applied for those with pathological Surgery-based therapy included surgery alone (n = 26), surgery +adjuvant RT (n = 8), surgery+adjuvant CCRT (n = 55), and surgery+BioRT (n = 13). Among these, five patients also received induction chemotherapy.

Survival analyses
Various survival curves were calculated according to the Kaplan-Meier method. Overall survival (OS) was defined as the date from the first day of curative treatment to death of any cause or the date of the last follow-up visit. Progression-free survival (PFS) was defined as the time from the first day of curative treatment to the time of disease progression or death. Local failure-free survival (LFFS), regional failure-free survival (RFFS), and distant metastasis failure-free survival (DMFFS) were calculated from the first day of curative treatment until the day of the primary, neck, or distant relapse or the date of the last follow-up visit. Survival differences between different subgroups were analyzed using the log-rank test. Patient characteristics and other variables were compared, as follows. The Mann-Whitney test was used for age, the continuous variable, of the two groups. The chi-square test was used for categorical or ordinal variables. Fisher's exact test was used when a small sample size existed. All statistical tests were two-sided, and a p-value of less than 0.05 is considered statistically significant. Analyses were performed by using MedCalc Statistical Software version 20.014 (MedCalc Software Ltd, Ostend, Belgium). The detailed causes of death analysis revealed that 48.7% (57/ 117) patients died of uncontrolled HPC and 51.3% died of other causes. Secondary malignancy (29/117 = 24.8%) and treatmentinduced complications (13/117 = 11.1%) were two major causes of non-HPC deaths, followed by intercurrent diseases (11/117 = 9.4%) and unknown (7/117 = 6.0%). The second primary malignancies consisted of cancers of the esophagus (9), lung (9), head and neck (7), hepatocellular carcinoma (1), urothelial carcinoma (1), lymphoma (1), mucoepidermoid carcinoma (1), and neuroendocrine carcinoma (1). Table 1 showed baseline characteristics between two different treatment approaches. There were no statistically significant differences in terms of age, gender, primary subsite, percentage of severe airway obstruction, T-classification, N-classification, and overall stage between the patients who received surgery-based therapy and definitive CRT.

Comparison of survival outcomes between surgery-based therapy and chemoradiotherapy
Thirty-eight of 65 (58.5%) patients in the definitive CRT group and 32 of 102 (31.4%) patients in the surgery-based therapy group developed tumor relapse (p = 0.0006). Table 2 demonstrated the detailed patterns of failure between the two groups. There were higher rates of local (32.3% vs. 5.9%, p < 0.0001), regional (23.1% vs. 13.7%, p = 0.1210), and distant recurrences (32.3% vs. 20.6%, p = 0.0897) in patients who received definitive CRT compared with those who received surgery-based therapy.

Discussion
Treatment strategies for HPC have changed over the past few decades. The traditional treatment approach for locally advanced HPC has been pharyngo-laryngectomy with adjuvant RT ± chemotherapy (5). However, radical surgery will compromise speech and swallowing functions and accordingly result in poor quality of life (6). In 1991, the Veterans Administration Laryngeal Cancer Study Group (VALCSG) published a randomized trial investigating the role of induction chemotherapy (IndCT) with a PF regimen followed by RT in order to spare patients a total laryngectomy (7). This study recruited patients with advanced laryngeal cancer only and showed that 64% of patients in the IndCT arm were able to preserve their larynx and avoid surgical resection. In addition, the 2-year OS rates were the same (68%) for both IndCT and surgical arms (p = 0.9846). The IndCT arm had fewer distant failures (p = 0.016) but more local recurrences (p = 0.0005) and no difference in the overall recurrence rate. Results of the VALCSG trial established the role of IndCT followed by RT and the concept of laryngeal preservation in the treatment of advanced laryngeal cancer. A phase III randomized trial (EORTC-24891) enrolled 202 HPC patients and reported no survival difference between IndCT+RT and surgery+adjuvant RT (8). Long-term follow-up of the EORTC-24891 trial confirmed that the initial larynx-preservation strategy did not compromise disease control or survival (9). Since the publication of both trials, the larynxpreservation approach has become one of the standard treatments for both HPC and laryngeal cancer in clinical practice for approximately three decades.
Many researchers further investigated different larynxpreservation protocols other than IndCT with a PF regimen followed by RT alone. The RTOG-9101 trial allocated patients with laryngeal cancer into three organ-preservation arms, IndCT followed by RT, CCRT, or RT alone, and they found that both chemotherapy-containing arms had significantly better larynxpreservation and locoregional control rates (10). The 2-year rates of larynx-preservation and locoregional control for CCRT vs. IndCT+RT vs. RT alone were 87% vs. 75% vs. 70%, and 78% vs. 61% vs. 52%. However, the OS rates had no significant difference. In the EORTC-24954 trial, Levebvre et al. designed a new schedule of alternating chemoradiotherapy versus the standard IndCT of PF followed by RT for both laryngeal cancer and HPC patients. They observed similar larynx preservation, PFS, and OS, as well as acute and late toxic effects (11). Except for studies focusing on the optimal sequence of combined chemoradiotherapy, a more effective chemotherapy regimen has been explored. 57.5%, p = 0.03) and disease-free survival (58% vs. 44%, p = 0.11), favoring the TPF arm (14). However, the OS was similar (60% vs. 60%, p = 0.57). Now, larynx-preservation strategy with definitive CRT (either CCRT alone, IndCT followed by RT/CCRT or alternating chemoradiotherapy) has become more popular than surgery-based therapy in patients with HPC and laryngeal cancer for more than 20 years. The best treatment strategy for advanced HPC is surgery first or organ preservation approach? It is still a matter of discussion to this day. To the best of our knowledge, there were only three prospective randomized trials focusing on HPC patients (8,9,15,16). As mentioned above, the EORTC-24891 trial demonstrated similar outcomes between surgery-based therapy and definitive CRT (8,9). The second randomized trial recruited a relatively small sample size (n = 92) and compared two different interventions, IndCT+RT vs. Comparison of overall survival and progression-free survival between surgery-based therapy (SBT) and definitive chemoradiotherapy (CRT) for the T1-2, T3, and T4 subgroups.
IndCT+total laryngectomy+RT (15). Obviously, this is an unbalance study design, three-combination therapy (chemotherapy, surgery, and RT) vs. two-combination therapy (chemotherapy and RT). The author reported a significantly better OS and local control favoring the threecombination therapy. However, the three-combination therapy arm consisted of total laryngectomy, not meeting the expectation of organ preservation by most patients. The third randomized trial enrolled 71 patients with T3M0 HPC and compared two different organ preservation strategies, CCRT vs. IndCT+RT (16). Results showed no differences in terms of OS, event-free survival, and local control, but CCRT had a higher 2-year larynx-preservation rate (92% vs. 68%, p = 0.016) than IndCT+RT. Based on these limited prospective randomized trials, we could not make a solid conclusion regarding "which treatment approach is better" and may say similar efficacy between primary surgery-based therapy and definitive CRT for HPC patients. There were several database studies to compare treatment outcomes between surgery-based therapy and definitive CRT. Comparison of overall survival and progression-free survival between surgery-based therapy (SBT) and definitive chemoradiotherapy (CRT) for the stage II, III, and IV subgroups. and Keelung) and also illustrated better OS (p < 0.001) and diseasefree survival (p = 0.003) for the surgery-based therapy arm (22). Based on the above database studies, we may conclude that surgerybased therapy is better than definitive CRT for HPC patients. Because of the inconsistent results of limited prospective randomized trials mentioned above and database studies, we further did a literature review and selected several observational studies with relatively large patient numbers (Table 4) (23-31). Three studies showed significant survival benefits favoring surgerybased therapy. Another six studies reported no significant difference between surgery-based therapy and definitive CRT. Results of the current study showed significantly better survival outcomes favoring surgery-based therapy compared with definitive CRT for all 167 patients. However, subgroup analyses obtained an important but rarely reported point of view-a definitive CRT approach can achieve the same survival outcome as surgery-based therapy only for patients with T3 tumor or overall stage III disease. This new argument may apply to the shared decision-making with the patients in our daily practice. When counseling a patient with a clinical T3 tumor or stage III disease, we could reference the current study and state that both surgery-based therapy and definitive CRT are good treatment options with the same oncologic efficacy. If the patient presented with an earlier or more advanced disease (T1-2, T4, stage II, or stage IV), the non-surgical approach tends to result in a significantly worse overall survival. Of note, our study has similar limitations as other series, such as a retrospective nature, small sample size, and no functional outcome.

Conclusion
In summary, there is no consensus on whether surgery-based therapy or definitive CRT should be the standard treatment approach to advanced HPC. Our new argument-definitive CRT approach-can achieve the same survival outcome as surgery-based therapy only for patients with T3 tumor or overall stage III disease, but needs a prospective phase III randomized trial with a large sample size to confirm.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Ethics statement
This study was reviewed and approved by the Institutional Review Board of Taipei Veterans General Hospital. The committee waived the requirement of written informed consent for participation.