Retrospective analysis of the clinical efficacy of definitive chemoradiotherapy for patients with hypopharyngeal cancer

Abstract

Objective

A retrospective analysis was performed to evaluate the clinical efficacy of definitive chemoradiotherapy including intensity-modulated radiotherapy for patients with hypopharyngeal cancer.

Methods

Previously untreated 204 patients with hypopharyngeal cancer were treated with definitive chemoradiotherapy. Of note, 66–70 Gy was delivered to the primary and involved nodes and 36–54 Gy was delivered to the prophylactic lymph node using standard fractionated radiotherapy. One hundred and forty-six patients received induction chemotherapy as a larynx preservation strategy, followed by definitive radiotherapy with or without concurrent chemotherapy. Intensity-modulated radiotherapy was also performed after 2006.

Results

The median follow-up time of this cohort was 43.4 months (range; 6.9–151.0). The 3-year overall survival, progression-free survival and larynx preservation survival rates were 78.8% (95% confidence interval; 73.0–85.0), 58.4% (95% confidence interval; 51.8–65.9) and 67.5% (95% confidence interval; 61.0–74.7), respectively. Multivariate analyses identified the following significant prognostic factors: an advanced age, the T category and N category for overall survival, the T category and N category for progression-free survival and the T category for larynx preservation survival. Acute toxicities of Grade 3 or higher were observed in 47 patients (23.0%). Two patients (1.0%) had Grade 4 pharyngeal edema. Suspicious treatment-related death due to lethal pharyngeal hemorrhage occurred in 1 (0.4%) patient. The rates of Grade 2 xerostomia in patients treated with intensity-modulated radiotherapy were 28.1, 17.4 and 9.5% at 6 months, 1 and 2 years after the completion of radiotherapy, respectively.

Conclusions

The efficacy and safety of definitive chemoradiotherapy are considered feasible with sufficient laryngeal preservation.

Introduction

Surgery with or without postoperative radiotherapy (RT) has historically been considered as the standard treatment for resectable advanced hypopharyngeal cancer (HPC); however, laryngectomy severely deteriorates the quality of life of patients. Thus, laryngeal preservation approaches have been developed. Clinical trials conducted in the 1990s revealed better locoregional control and laryngeal preservation rates with chemoradiotherapy (CRT) than with RT alone. CRT is also regarded as the standard treatment for HPC patients who want laryngeal preservation (1–3).

Induction chemotherapy (IC) followed by definitive RT (IC-RT) represents an alternative laryngeal preservation approach for locally advanced HPC. Although IC-RT is not as efficacious as CCRT, it suppresses distant metastases more than RT alone or CCRT (4–6). Tumor responses to IC may be related to locoregional control and provide predictive information for determining whether a larynx preservation approach is feasible (5).

The standard treatment for advanced head and neck cancer is currently CCRT. Although CCRT is considered the most promising treatment for tumor control, it is associated with severe late toxicity events such as dysphagia (7, 8). Furthermore, the incidence rates of severe toxicity events with salvage surgery for locoregional recurrence after CCRT were previously reported to be high (9).

We have performed intensity-modulated radiotherapy (IMRT) for HPC since 2006. IMRT delivers a highly conformal dose to a target volume while sparing the surrounding normal tissue. IMRT has been clinically used since the 1990s and significantly decreases the incidence of late toxicity events such as xerostomia (10). However, the clinical findings of IMRT for HPC patients are more limited than those for nasopharyngeal cancer (NPC) or oropharyngeal cancer (OPC) (11).

In this retrospective study, we analyzed the clinical efficacy of definitive CRT for HPC patients, including IMRT series in our institution.

Patients and methods

Patient and tumor characteristics

We analyzed 204 consecutive patients with non-metastatic HPC treated with definitive CRT at Aichi Cancer Center Hospital between 1995 and 2013. Eligibility criteria were as follows: previously untreated and histologically proven squamous cell carcinoma without distant metastasis, and an Eastern Cooperative Oncology Group performance status (PS) of 0–2. Patients with a treatment history of RT to the head and neck region were excluded, while those with a second primary cancer were included in this analysis. Patients were diagnosed according to the standard International Union Against Cancer (UICC) 6th edition. Pretreatment evaluations were as follows: a physical examination, laryngeal endoscopy, gastrointestinal endoscopy, chest radiography, computed tomography (CT) of the neck and chest and magnetic resonance imaging (MRI) of the primary site and involved neck nodes. ¹⁸F-fluorodeoxyglucose-positron emission tomography (¹⁸F-FDG PET) or PET/CT was also performed after 2001.

Radiotherapy

One hundred and forty-eight patients (72.5%) received three-dimensional conformal RT (3DRT), while 45 (22.1%) received definitive IMRT during the entire treatment period and 11 (5.4%) were treated with 3DRT followed by an IMRT boost. IMRT was delivered using Helical TomoTherapy.

In the 3DRT group, 36–40 Gy was delivered to the whole neck, with an additional boost of 26–30 Gy being delivered to the gross tumor and involved lymph node using the cone down technique in order to spare the spinal cord. The details of target volume delineation were as reported previously (12).

In the IMRT group, 70 Gy was delivered over 7 weeks to the PTV primary and PTV nodes, while 54 Gy was delivered to the PTV initial using the simultaneous integrated boost (SIB) technique. In the combined group, 40 Gy was prescribed to the whole neck by 3DRT, and 30 Gy was then delivered to the PTV primary and PTV nodes using an IMRT boost. The prescribed dose of IMRT was calculated on the basis of D50 (50% of the target volume received prescribed dose) for PTV primary and PTV node, while the dose for the PTV initial node was calculated at D95 (95% of the target volume received the prescribed dose).

RT was administered at 1.8–2.0 Gy per fraction once daily using 6 MV photon. The treatment was planned using an X-ray simulator or radiation planning system for 3DRT. A radiation planning system (Tomoprovider^®) was used to plan IMRT.

Chemotherapy

We basically used IC as a larynx preservation approach. IC was administered to 146 patients (71.6%). Two courses of high-dose cisplatin and 5-fluorouracil (5-FU) were administered, and responders to IC received definitive RT with or without concurrent chemotherapy, which consisted of weekly or triweekly cisplatin. Fifty-eight patients (28.4%) who refused laryngectomy or were diagnosed with unresectable disease were treated with concurrent or alternating CRT. Thirty-four patients (16.7%) received concurrent regimen consisted of triweekly cisplatin and 24 patients (11.8%) patients received alternating regimen consisted of 5-FU and cisplatin. The details of our strategy have already been described in a previous study (12).

Surgery

Planned neck dissection was considered for patients with initially bulky nodes or suspicious residual nodal disease 1–2 months after the completion of CRT. Salvage surgery was indicated in cases of residual or the progression of either the primary or neck lesion.

Follow-up

Six weeks after the completion of CRT, initial treatment responses were assessed using laryngeal endoscopy, MRI and CT. The indication for planned neck dissection was generally determined at this point for patients with residual lymph nodes. We followed patients every month for the first 6 months to 1 year after the completion of RT and then every 3–6 months thereafter. In each follow-up, a physical examination, and laryngoscopy were performed, in addition to CT or MRI scanning of the head and neck. ¹⁸F-FDG PET or PET/CT was also performed at least annually during the follow-up after 2011.

Acute toxicity was scored according to the Common Terminology Criteria of Adverse Events version 3.0 (13). Late toxicity was scored according to the Radiation Therapy Oncology Group mobility grading scale (14).

Statistical analysis

Overall survival (OS) was defined as the time from the start of any treatment to death from any cause. Progression-free survival (PFS) was calculated from the start of any treatment to death by any cause, local/regional progression or distant metastasis. Larynx preservation survival (LPS) was defined as the time from the start of any treatment to death from any cause, salvage laryngectomy, severe dysphagia or laryngeal stenosis. OS, PFS and LPS rates were estimated using the Kaplan–Meier method (15). The Cox proportional hazard model was used to estimate the hazard ratio (16). P < 0.05 was considered significant.

Results

Patient characteristics

Between October 1995 and December 2013, 216 patients with HPC received definitive RT at Aichi Cancer Center hospital. Six patients who received an insufficient total radiation dose (<60 Gy) were excluded from the analyses. No follow-up data were available for six patients. Patient and tumor characteristics were summarized in Table 1.

Table 1.

Patient and tumor characteristics

Characteristics
Age (years) median (range)	64 (37–87)
Sex—N (%)
Male	189 (92.6)
Female	15 (7.4)
T category
1	23
2	95
3	64
4a	22
4b	0
N category
0	59
1	31
2a	10
2b	51
2c	42
3	11
Clinical Stage
I	8
II	32
III	41
IVA	112
IVB	11
Dose (Gy) median (range)	69.3 (61.2–76.9)
Modality–N (%)
3DRT	148 (72.5)
IMRT	56 (27.5)
Chemotherapy regimen—N (%)
IC followed by CCRT	146 (71.6)
CCRT	34 (16.7)
AltCRT	24 (11.8)
Treatment duration—d median (range)
IC followed by CCRT	89 (44–201)
CCRT	51 (44–63)
AltCRT	69 (57–81)

3DRT, three-dimensional radiotherapy; IMRT, intensity-modulated radiotherapy; IC, induction chemotherapy; CCRT, concurrent chemoradiotherapy; AltCRT, alternating chemoradiotherapy.

The median age was 64 years old (range; 37–87). One hundred and twenty-three patients (60.3%) were diagnosed with Stage IVA–IVB disease. The median dose to the primary tumor and node was 69.3 Gy (range; 61.2–76.9 Gy).

Treatment outcomes

In the last follow-up, 130 patients (63.7%) were alive without disease, 8 (3.9%) were alive with disease, 53 (26.0%) died from the disease, 13 (6.4%) died from other reasons. The OS curve for all cohorts was shown in Figure 1. The 3 years of OS (3yOS), 3 years of PFS (3yPFS) and 3 years of LPS (3yLPS) rates were 78.8% [95% confidence interval (CI); 73.0–85.0], 58.4% (51.8–65.9) and 67.5% (95%CI; 61.0–74.7), respectively.

Figure 1.

Overall survival rate of the entire cohort.

Based on the initial responses of the primary lesion, 190 patients (93.1%) were assessed as achieving complete response (CR), 12 (5.9%) partial response (PR) and 2 (1.0%) progressive disease (PD), respectively. Regarding initial responses for neck disease, 108 (74.5%) achieving a CR, 29 (20.0%) PR and 3 (2.0%) PD.

Seventy-seven patients (37.8%) developed treatment failures; 41 (20.1%) for local progression, 30 (14.7%) for regional and 30 (14.7%) for distant metastasis.

Regarding additional treatments for local recurrence, 34 (16.7%) were indicated for salvage laryngectomy and 1 (0.5%) for endoscopic laryngo-pharyngeal surgery. Twenty-eight patients (13.7%) underwent neck dissection for regional recurrence, while 20 (9.8%) were indicated for planned neck dissection (Fig. 2).

Figure 2.

Larynx preservation survival rates of groups divided by the T category.

Second primary cancer

Synchronous or metachronous second primary cancer developed in 87 patients (42.6%). Esophageal cancer was the most common (57 patients), followed by stomach (15 patients) and lung cancers (6 patients). Among the 57 patients with esophageal cancer, 10 were diagnosed prior to the treatment of HPC and 47 were diagnosed simultaneously or developed later.

Univariate and multivariate analyses

The results of univariate analyses were shown in Table 2.

Table 2.

Univariate analysis

Prognostic factor	N	3-year survival rate (%)
		OS	P	PFS	P	LPS	P
Age (years)
≧64	102	73.6	0.012*	54.9	0.259	60.8	0.012*
<64	102	83.7		62.0		74.1
Sex
Male	189	78.5	0.738	58.2	0.808	68.0	0.758
Female	15	82.5		60.9		60.9
Subsite
PS	167	79.2	0.664	59.4	0.829	70.0	0.487
Others	37	77.0		54.5		57.0
T category
T1–2	87	88.4	<0.001*	68.2	<0.001*	76.4	<0.001*
T3–4	117	65.5		45.4		55.1
N category
N0–1	90	88.5	<0.001*	67.8	0.005*	74.4	0.002*
N2a–3	114	71.0		51.1		61.9
Clinical Stage
I–II	40	96.9	0.002*	72.9	0.034*	80.4	0.004*
III–IV	164	74.4		54.9		64.3
Synchronous SPC*
Yes	45	76.2	0.441	57.5	0.323	65.1	0.620
No	159	79.4		58.8		68.2
Dose (Gy)
>66	131	81.6	0.391	55.1	0.166	63.4	0.367
≦66	103	76.9		63.9		74.4
Modality
3DRT	148	78.0	0.872	62.6	0.070	70.5	0.676
IMRT	56	80.6		39.7		53.7
IC
Yes	146	76.7	0.307	55.0	0.144	64.3	0.285
No	58	84.3		67.5		76.1

OS, overall survival; PFS, progression-free survival; LPS, larynx preservation survival PS, pyriform sinus; SPC, second primary cancer.

*P < 0.05.

The 3yOS was significantly lower in the group with an advanced age (≧64 years old) than in the group with a younger age (<64 years old) (73.6 vs. 83.7%, P = 0.012). The group with the T3–4 category had a significantly worse OS rate than that of the T1–2 group (3yOS: 65.5 vs. 88.4%, P < 0.001). The group with N2–3 had a significantly lower OS rate than that in the group with N0–1 (3yOS: 71.0 vs. 88.5%, P < 0.001). The group with clinical Stage III–IVA had a significantly lower OS rate than that in the group with clinical Stage I–II (3yOS: 74.4 vs. 96.9%, P = 0.002). No significant differences were observed between the group that received IC and the non-IC group (3yOS: 84.3 vs. 76.1%, P = 0.290).

The 3yPFS rate was significantly worse in the group with T3–4 than in the group with T1–2 (55.1 vs. 76.4%, P < 0.001). The group with N2a–3 had a significantly lower PFS rate than that in the group with N0–1 (3yPFS: 61.9 vs. 74.4%, P = 0.002). The group with clinical Stage III–IVA had a significantly lower PFS rate than that in the group with clinical Stage I–II (3yPFS: 54.9 vs. 72.9%, P = 0.034). The group that received IC had a slightly lower PFS rate than that in the non-IC group (3yPFS: 67.5 vs. 54.6%, P = 0.142).

The 3yLPS rate was significantly more unfavorable in the group with an advanced age than in the group with a younger age (3yLPS: 74.1 vs. 60.8%, P = 0.012). The group with T3–4 had a significantly worse LPS rate than that in the group with T1–2 (3yLPS: 55.1 vs. 76.4%, P < 0.001). The 3yLPS rate of the group with N2–3 was significantly lower than in the group that with N0–1 (3yLPS: 61.9 vs. 74.4%, P = 0.002). The group with clinical Stage III–IVA had a significantly lower LPS rate than that in the group with clinical Stage I–II (3yLPS: 64.3 vs. 80.4%, P = 0.004). No significant differences were observed between the group who received IC and the non-IC group (3yLPS: 76.1 vs. 63.9%, P = 0.259).

Among the patients who received IC, the IMRT group tend to had a slightly longer overall treatment time than that of the 3DRT group (OTT: 101.4 days vs. 95.8 days, P = 0.07).

The results of multivariate analyses were shown in Table 3. An advanced age (HR 1.80 95%CI; 1.04–3.14, P = 0.038), the T category (HR 2.06, 95%CI; 1.19–3.57, P = 0.010) and N category (HR 2.63, 95%CI; 1.46–4.71, P = 0.001) were identified as significant prognostic factors for OS, those for PFS were the T category (HR 1.90, 95%CI; 1.25–2.88, P = 0.003) and N category (HR 1.70, 95%CI; 1.10–2.63, P = 0.018); while that for LPS was the T category only (HR 2.00, 95%CI; 1.26–3.19, P = 0.003).

Table 3.

Multivariate analysis

Prognostic factor	OS		PFS		LPS
	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
Age
>64	1.80 (1.04–3.14)	0.038*	1.23 (0.81–1.87)	0.333	1.55 (0.97–2.46)	0.063
T category
T3–4	2.06 (1.19–3.57)	0.010*	1.90 (1.25–2.88)	0.003*	2.00 (1.26–3.19)	0.003*
N category
N2a–3	2.63 (1.46–4.71)	0.001*	1.70 (1.10–2.63)	0.018*	1.58 (0.99–2.53)	0.055
Modality
IMRT	NA	NA	1.41 (0.88–2.25)	0.156	NA	NA
IC
Yes	NA	NA	1.21 (0.73–2.01)	0.450	NA	NA

CI, confidence interval; NA, not applicable.

*P < 0.05.

Toxicities

Acute toxicities of Grade 3 or higher were observed in 47 patients (23%) (Table 4). The most common hematological toxicity event of Grade 3 or 4 was neutrocytopenia in 24 patients (11.8%). Nine patients (4.4%) developed Grade 3 or 4 dermatitis. The rates of Grade 2 xerostomia in patients treated with IMRT were 28.1, 17.4 and 9.5% at 6 months, 1 and 2 years after the completion of RT, respectively.

Table 4.

Incidence of Grade 3 or 4 toxicities

Toxicity	N
	Grade 3	Grade 4
Acute toxicity
Neutrocytopenia	15	9
Thrombocytopenia	11	6
Dysphagia	16	1
Mucositis	12	0
Dermatitis	8	1
Late toxicity
Laryngeal stenosis	3	0

Regarding late toxicity events, Grade 4 pharyngeal edema developed in 2 patients and Grade 2 hypothyroidism in 7. Four patients developed Grade 2 peripheral neuropathy. Suspicious treatment-related death due to lethal pharyngeal hemorrhage occurred in 1 patient because exact records from another hospital could not be obtained.

Discussion

We reported the efficacy and safety of CRT for patients with HPC. The median follow-up period was 43.4 months. The 3yOS, 3yPFS and 3yLPS rates were 78.8% (95%CI; 73.0–85.0), 58.4% (95%CI; 51.8–65.9) and 67.5% (95%CI; 61.0–74.7), respectively. In this analysis, 146 patients (71.6%) were treated with IC followed by RT, while 58 (28.4%) were treated with CCRT or alternating CRT.

Lefebvre et al. previously reported the findings of randomized phase III trial that compared a surgery arm with adjuvant RT with an IC-RT arm for patients with Stage II–IV HPC in the 1990s. The 3yOS, 3yPFS and 3yLPS in the IC-RT arm were 57, 42 and 43%, respectively. Thus, IC-RT was considered to be promising from the viewpoint of organ preservation without sacrificing survival rates (3yOS: IC-RT 57% vs. Surgery 43%) (3, 17).

The clinical results of organ preservation using CRT for HPC patients were shown in Table 5. Most of these trials also included patients with laryngeal cancer, and only locally advanced cases with Stage III–IV were analyzed. However, our results also showed similar efficacy even in the subset of patients with Stage III–IV disease (3yOS, 3yPFS and 3yLPS were 74, 55 and 64%, respectively).

Table 5.

Results of treatment outcomes of definitive chemoradiotherapy for patients with hypopharyngeal cancer

Trial	N	Site	Stage	Treatment	3-year survival rate (%)
					OS	PFS	LPS
EORTC 24891	100	Hypopharynx	II–IV	FP -> RT vs. (Surgery -> RT)	57	43	42
GORTEC 2000-01	213	Hypopharynx Larynx	III/IV	FP -> RT vs. TPF -> RT	60 60	NA	57 70
EORTC 24954	450	Hypopharynx Larynx	III/IV	FP -> RT vs. Alternating FP + RT	62 65	50 50	40 45
TREMPLIN	116	Hypopharynx Larynx	III/IV	TPF -> CRT vs. TPF -> Cetuximab + RT	75 73	NA	NA
Prades	71	Pyriform sinus	III/IV	PF -> RT/Surgery CCRT	NA NA	36 41	68 92
This study	204	Hypopharynx	I–IV (III/IV)	FP -> RT/CRT CCRT Alternating FP/FN + RT	79 (74)	58 (55)	68 (64)

RT, radiotherapy; FP, 5-fluorouracil + cisplatin; FN, 5-fluoroucacil + nedaplatin; TPF, docetaxel + 5-fluorouracil + cisplatin.

Concurrent CRT is regarded as the standard treatment for patients with locally advanced HPC based on the findings of RTOG 91–11 and meta-analyses (2, 8, 18). In the present study, IC was administered to 146 patients (71.6%). The IC group had a significantly lower PFS rate in the univariate analysis (IC vs. non-IC: 55.0 vs. 67.5%, P = 0.144); however, no significant difference was detected in the multivariate analysis (HR 1.21, 95%CI; 0.73–2.01, P = 0.450).

Posner et al. previously reported the findings of a randomized Phase III trial for the treatment of squamous cell carcinoma of the head and neck, in which with IC using docetaxel plus cisplatin and fluorouracil (TPF) was compared with cisplatin and fluorouracil (FP) followed by CRT. The TPF group showed significantly more favorable survival than that of the FP group (HR 0.70, 95%CI; 0.54–0.90; P = 0.006) (6). FP was mainly used in our series; therefore we will consider using TPF regimens for IC strategies in the future.

The IMRT group had a slightly poorer PFS rate in the univariate analysis; however, no significant differences were observed in the multivariate analysis. No significant differences were observed in OS, PFS or LPS rates between the IMRT group and the 3DRT group.

The overall treatment time of the IMRT group was slightly longer due to its longer planning time. One of the reasons of prolongation was that preparation of IMRT became longer compared with 3DCRT. In the case of IC, it should be arrange to minimize delay of consultation to radiation oncologist, because it might affect efficacy of definitive radiotherapy. This difference may have had a potentially negative impact on treatment outcomes.

CRT has recently been reported to be associated with a higher incidence of late toxicity events such as dysphagia (8). It was only 4.9% in the present study. In the long-term results of RTOG 91–11, the IC group had a slightly better OS rate than that of the CCRT group in spite of a lower locoregional control rate. The reason for this discrepancy was considered to be increases in the morbidity of non-cancer causes in the CCRT group. The lower incidence of late toxicity events in our cohort may have been due to the use of IC.

It was difficult to evaluate late toxicity events in the IMRT group due to the short-term follow-up. But the rate of Grade 2 xerostomia gradually decreased after CRT in the IMRT group. The advantage of parotid sparing by IMRT has been reported in cases of NPC or OPC, but less so in HPC patients in the reported series. However, the same benefit was also observed in the present series for HPC. In the present analysis, only four cases of Grade 2 hypothyroidism and 4 of Grade 2 peripheral neuropathy were observed as late toxicity events. We performed additional evaluations in relation to laryngeal edema with radiation methods, and are now preparing another study.

This present study had several limitations. Since it was a single-institution, retrospective study, there were several biases due to the variable methods of chemotherapy and RT, which had a marked impact on the data. Thus, it was difficult to determine the optimal treatment for HPC in the present study. However, to the best of our knowledge, this is the first study on CRT for HPC that included more than 200 patients.

Our clinical results were consistent with those of previous series. In spite of the smaller sample size with a shorter follow-up period, we also considered IMRT for HPC to be efficacious and safe in our cohort. Further improvements need to be considered in prospective multi-institutional trials in order to acquire reliable evidence for Japanese patients.

Funding

This research was partially supported by the Research for Innovative Cancer Control from the Japan Agency for Medical Research and Development, AMED (15ck0106055h0102 and 15ck0106093h0102) and The National Cancer Center Research and Development Fund (25-B-2 and 26-A-4).

Conflict of interest statement

None declared.