The median follow-up period was 38.4 months (range 2.9–217.1 months). There were 49 males and 9 females among the subjects. The median age of the subjects was 68 years (range 38 to 87 years). The patients' characteristics are summarized in detail in Table 1. All of the patients had no nodal involvement. All patients had squamous cell carcinomas. The median body surface area (BSA) at the beginning of RADPLAT was 1.578 m2 (range 1.192 to 2.094 m2).
Table 1
Patient characteristics (n = 58)
| ≦ 60 Gy (n = 13) | 66 Gy (n = 27) | 70 Gy (n = 18) | p value |
Age (y) median (range) | 68 (48–87) | 68 (38–87) | 71 (51–84) | 0.264 |
Gender (male/female) | 10/3 | 23/4 | 16/2 | 0.514 |
T Stage (n) | T2 | 1 (7.7%) | 1 (3.7%) | 0 | 0.841 |
T3 | 2 (15.4%) | 3 (11.1%) | 6 (33.3%) |
T4a | 7 (53.8%) | 17 (63.0%) | 10 (55.6%) |
T4b | 3 (23.1%) | 6 (22.2%) | 2 (11.1%) |
Radiotherapy (3DCRT/IMRT) | 11/2 | 22/5 | 1/17 | 0.08 |
intra-arterial infusion (cycle) median (range) | 6 (4–10) | 6 (2–9) | 6 (1–7) | 0.975 |
The median prescribed dose to the local lesion was 66 Gy (range 50 to 70 Gy) in the 3DCRT group and 70 Gy (range 60 to 70 Gy) in the IMRT group. One patient in the 3DCRT group was irradiated with only 50 Gy because of a difficulty in avoiding the optic chiasm and the contralateral optic nerve. In this study, we analyzed three groups defined according to the total radiation dose ( ≦ 60 Gy, 66 Gy and 70 Gy).
CDDP was performed with a median of 6 cycles once a week (range 1 to 10 cycles). The median total CDDP dosage was 1015 mg (range 190 to 1820 mg). There were no significant differences in BSA ( ≦ 60 Gy group: median 1.560 m2, 66 Gy group: median 1.570 m2, 70 Gy: median 1.612 m2, p = 0.163). There were also no significant differences in the total dose of CDDP among the three groups ( ≦ 60 Gy group: median 960 mg, 66 Gy group: median 987 mg, 70 Gy: median 1080 mg, p = 0.778).
The 5-year local control rate was 69.9% for all patients (n = 58) (Fig. 1), 80.8% for patients with T2-3 disease (n = 13), 68.4% for those with T4a disease (n = 34), and 61.4% for those with T4b disease (n = 11) (Fig. 2). The 5-year local control rates were 86.2% in the IMRT group and 64.0% in the 3DCRT group (p = 0.130).
The 5-year overall survival rates were 72.2% for all patients (n = 58) (Fig. 1), 92.3% for patients with T2-3 disease (n = 13), 71.6% for T4a (n = 34), and 51.9% for T4b (n = 11). The 5-year overall survival rate was 88.5% in patients treated with 70 Gy(n = 18), 66.6% in patients treated with 66 Gy(n = 27), and 72.7% in patients treated with 60 Gy or lower(n = 13) (p = 0.627) (Fig. 3).
The 5-year local control rate was 87.7% in patients treated with 70 Gy(n = 18), 72.1% in patients treated with 66 Gy(n = 27), and 41.0% in patients treated with 60 Gy or lower(n = 13) (p = 0.035) (Fig. 4). In particular, the local control rate in the 70 Gy group was significantly higher than that in the ≦ 60 Gy group (p = 0.011).
Acute toxicities were manageable in all patients. Grade 1, 2, and 3 dermatitis developed in 17, 29, and 3 patients, respectively. Grade 1, 2, and 3 mucositis developed in 6, 31, and 13 patients, respectively. No patients showed renal toxicity of grade 3 or worse. Two patients had mild brain infarction caused by the intra-arterial cisplatin procedure.
No late grade 3 or higher eye disorders except for cataracts developed in the IMRT group. In the 3DCRT group, grade 4 toxicity (blindness) occurred in 3 patients treated with 60 Gy, one patient treated with 66 Gy. In two of these patients, drug infusions were administered via the ophthalmic artery based on the judgement of an interventional radiologist because it was difficult to control tumors that had invaded the optic nerves.
Other late toxicities were recognized as follows: grade 2 maxillary osteonecrosis in 3 patients, oculomotor nerve paralysis in one patient, and carotid cavernous fistula in one patient. One patient had a skin perforation due to the shrinkage of the tumor.
The sites of recurrence and residual disease were identified wherever possible. Local recurrence occurred in 15 patients (25.9%). Of these, seven patients underwent salvage surgery, and six of these seven patients were successfully salvaged. Two patients received stereotactic radiotherapy, and one of those patients was salvaged. Neck recurrence occurred in two patients. Of these, one patient was treated successfully by salvage neck dissection. Distant metastasis occurred in eight patients (lung 2, brain 1, lung + bone 1, lung + brain 2, lung and liver 1, brain + adrenal grand 1). Of these, three patients with lung metastasis were successfully salvaged (surgery 1, stereotactic radiotherapy 2). One patient died due to acute subdural hematoma, which was not considered a treatment-related death.