A new intra-arterial chemoradiation regimen that involves infusing low-dose cisplatin in combination with definitive irradiation was used in 36 patients diagnosed with squamous cell carcinoma of the maxillary sinus. The safety and therapeutic efficacy of this regimen were reviewed retrospectively. An intra-arterial catheter was inserted in a retrograde manner into the target artery via the superficial temporal artery or occipital artery. Intra-arterial infusion was performed using cisplatin at a dose of 20–50 mg/m 2 per week for 6–8 weeks. At the same time, sodium thiosulphate was infused as a neutralizing agent. Irradiation was performed at 60 Gy in 30 fractions. All 36 patients completed treatment. Grade 3 adverse events occurred in only seven patients (19.4%) and no grade 4 events were noted. As a primary therapy, the complete response rate was 83.3%, the partial response rate was 16.7%, and the overall response rate was 100%. The 2-year local control rate was 63.0%, and the 2-year overall survival rate was 75.5%. The 2-year preservation rate of the hard palate was 97.1%, that of the eyeball was 97.2%, and that of visual function was 94.4%. This treatment regimen can contribute to improving the quality of life of patients without reducing the curability of the therapy.
Cancer of the nasal cavity and paranasal sinus is relatively rare, accounting for 3–5% of all head and neck cancers and less than 1% of all malignancies in Japan and the USA. Cancer of the maxillary sinus most commonly occurs in the paranasal sinuses, and squamous cell carcinoma (SCC) is the most common histopathological classification in Japan, Europe, and the USA. Interestingly, the incidence of such cancers is higher in Japan (77%) than in Europe and the USA (55%).
Because of the lack of large-scale clinical trials, there is no clear evidence on the optimal treatment strategies for maxillary sinus cancer. In Europe and the USA, surgical resection is the main treatment strategy regardless of cancer tissue-type, and postoperative irradiation (radiotherapy alone or concurrent chemoradiation therapy) is used as a standard therapy depending on the invasion of the cancer. However, in the past 30 years, no clear prognostic improvement has been seen, and studies indicate that additional treatment may be necessary.
In Japan, the treatment of tissue-type cancers other than SCC has been performed using a strategy similar to that used in Europe and the USA. However, in the 1960s, a multimodality therapy was introduced for the treatment of maxillary sinus SCC; this was called ‘tri-modality therapy’ and comprised the intra-arterial infusion of, for example 5-fluorouracil, through a catheter placed in the external carotid artery via the superficial temporal artery, in combination with preoperative irradiation, followed by a partial or total maxillectomy. This is because selective intra-arterial infusion is easy through the maxillary artery, which is the main supply to the maxillary sinus cancer. In addition, SCC is more sensitive to chemotherapy and radiation than glandular cancer.
Maxillary sinus cancer is asymptomatic until it invades the surrounding tissues, and it is advanced at the time of initial presentation. Therefore, it is often necessary to perform a total maxillectomy or extended total maxillectomy in these patients. These surgeries result in significant changes to the facial appearance and the functions of articulation and eating. In order to improve these effects, tri-modality therapy was frequently applied. However, because the catheter position in these cases was checked only by infusion of a dye without the use of computed tomography angiography (CTA), the therapeutic effects of retrograde intra-arterial chemotherapy were unstable.
In the 1990s, Robbins et al. introduced RADPLAT (RADiation and intra-arterial cisPLATin), which is the simultaneous therapy of super-selective intra-arterial chemotherapy with high-dose cisplatin by the Seldinger technique from the femoral artery in combination with radiation therapy. Since the development of a technique for selective catheterization into the target artery under fluoroscopic guidance, good therapeutic results have been achieved by simultaneously infusing high-dose cisplatin and sodium thiosulphate (STS) as a neutralizing agent. In Japan, intra-arterial chemotherapy is practiced widely because CTA with intra-arterial infusion to the lesion has shown clinical benefits.
Intra-arterial chemoradiation therapy has been performed previously in our department, by placing a catheter in the external carotid artery via the superficial temporal artery as preoperative therapy, followed by radical resection. On the basis of previous reports on the development of techniques for the super-selective placement of the catheter into the target artery, a new intra-arterial chemoradiation regimen is presented herein. This regimen involves infusing weekly low-dose cisplatin in combination with definitive irradiation to treat SCC of the maxillary sinus, through retrograde catheterization via the superficial temporal artery or occipital artery. The aim of this treatment is to preserve function and aesthetics, as well as to cure patients who refuse surgery and those who are inoperable.
Materials and methods
The clinical data of patients with SCC of the maxillary sinus who underwent initial therapy via intra-arterial chemoradiation between January 2009 and January 2014 in the Head and Neck Oncology Centre of the International University of Health and Welfare Mita Hospital were reviewed retrospectively. Eligible patients had an Eastern Cooperative Oncology Group (ECOG) performance status of 0–2 and no severe disorders of the vital organs such as the bone marrow, liver, and kidney.
Surgical method for catheter placement
In accordance with the reports of Fuwa et al. and Hasegawa et al., retrograde catheterization into the target artery was done via the superficial temporal artery or occipital artery. The branches of the external carotid artery were examined preoperatively using three-dimensional CTA (3D CTA). The operation was performed with the patient under general anaesthesia.
The superficial temporal artery or occipital artery was exposed and a Supercath catheter (22-gauge; Medikit Co., Ltd, Tokyo, Japan) was inserted in a retrograde manner. The guide wire (Radifocus Guide wire M, angle 45°, diameter 0.406 mm, length 120 cm; Terumo Corporation, Tokyo, Japan) was then inserted through the external cylinder of the Supercath. After removal of the outer cylinder of the puncture needle, a Medikit Angiographic Catheter (NECK-type M, diameter 4.0 Fr, length 40 cm; Medikit Co., Ltd) was inserted along the guide wire into a common carotid artery, and the guide wire then removed. While pulling the catheter gradually during infusion of the contrast agent, the catheter was hooked to a branch of the maxillary artery. After inserting the guide wire into the peripheral side of the maxillary artery through the angiographic catheter once again, the angiographic catheter was removed, but the guide wire left inserted. Finally, an Anthron P-U catheter (Tapering Type, proximal portion 5.0 Fr, distal portion 2.7 Fr; Toray Medical Co., Ltd, Tokyo, Japan) was inserted along the guide wire and fixed in place. The catheter was inserted into the maxillary artery on the peripheral side of the middle meningeal artery ( Fig. 1 A). In cases of tumours invading the buccal mucosa or facial skin, the catheter was inserted into the external carotid artery on the central side of the facial artery because there is a expected flow from the transverse facial artery or the facial artery ( Fig. 1 B and C). Heparinized saline was administered continuously into the catheter via an infusion pump (Baxter Infusor, 7-days type; Baxter, Chicago, IL, USA). The day after surgery, it was checked that the tumour was entirely enhanced by performing contrast-enhanced magnetic resonance imaging (MRI) through the catheter ( Fig. 1 D and E). It was also verified that there was no occlusion of the catheter or change in arterial infusion area by performing staining with indigo carmine (Daiichi Sankyo, Tokyo, Japan).
Chemotherapy was administered in the form of weekly cisplatin by intra-arterial infusion for 5 h per day. The cisplatin dose given was determined to be 20–30 mg/m 2 in patients who underwent super-selective placement into the maxillary artery, and 40–50 mg/m 2 in patients who underwent selective placement into the external carotid artery. Intra-arterial chemotherapy was administered one or two times before starting irradiation therapy, and it was continued weekly until the end of irradiation ( Fig. 2 A). To prevent vascular pain and vasculitis, 1 ml of 7% sodium hydrogen carbonate was added to the intra-arterial infusion of 10 mg of cisplatin. In addition, to neutralize cisplatin, 2000 mg of STS was infused intravenously for every 5 mg of cisplatin.
Radiotherapy was performed for the primary tumour in cases without metastasis, and it was performed for both the primary tumour and the diseased side of the neck in cases showing cervical lymph node metastasis. The total irradiation dose was 60 Gy in 30 fractions. Once 30 Gy in 15 fractions had been administered, the therapeutic efficacy was evaluated using MRI and it was determined whether any changes were needed to the arterial infusion area. If necessary, the position of the catheter was adjusted ( Fig. 2 B and C).
Evaluation of anti-tumour effects and adverse events
The therapeutic efficacy was evaluated according to the Response Evaluation Criteria in Solid Tumours (RECIST) version 1.1 criteria. Evaluation of the therapeutic efficacy was based on the best response observed using computed tomography (CT), MRI, or fluorodeoxyglucose positron emission tomography ( 18 F-FDG-PET) every 3 months after treatment. For both the primary tumour and any cervical metastases, pathologically therapeutic evaluation via biopsy or planned surgery was not performed, as would conventionally be done for tri-modality therapy. These examinations were performed only for patients in whom residual tumour or recurrence was suspected on the basis of endoscopic findings or diagnostic images. All toxicities caused by the treatment were evaluated according to the Common Terminology Criteria for Adverse Events version 3.0.
The Kaplan–Meier method was used to calculate the overall survival (OS) rate (date of cancer diagnosis to death from all causes) and the local control (LC) rate (date of cancer diagnosis to the date of local failure, including disease progression or tumour relapse). Univariate analysis was performed using the log-rank test-to-test associations with the prognosis and orbital invasion. Statistical analyses were conducted using StatMate software (ATMS Co., Ltd, Tokyo, Japan), and differences were considered significant at P < 0.05.
This study included 36 patients with maxillary sinus SCC who were first treated at the Head and Neck Oncology Centre, International University of Health and Welfare Mita Hospital (Tokyo, Japan) from January 2009 to January 2014. Of these 36 patients, 16 had an ECOG performance status of 0, 15 had a performance status of 1, and five had a performance status of 2. There were 27 men and nine women. The median age was 63 years (range 41–80 years). By T and N classification (Union for International Cancer Control (2005)), three patients had T2 tumours, four had T3 tumours, 24 had T4a tumours, and five had T4b tumours; one had an N1 tumour and two had N2b tumours. Furthermore, 24 patients showed tumour invasion of the orbit, of whom 11 showed destruction of the orbital inferomedial wall bone, eight showed infiltration beyond the orbital periosteum, and five showed invasion of the orbital apex ( Table 1 ). The median observation period was 30.1 months (range 11–62.2 months).
|Age, years, median (range)||63 (41–80)|
|ECOG performance status|
Catheters were inserted in the maxillary artery in 15 patients and in the external carotid artery in 21 patients. Catheterization was successful in all patients who underwent super-selective catheterization into the maxillary artery. The catheter was first inserted via the superficial temporal artery in all patients, except in one patient who developed a catheter-induced infection; for this patient, the catheter was re-inserted via the occipital artery for another round of therapy after removal of the catheter associated with the infection. Intra-arterial infusions were single-sided in 34 patients, but were double-sided in two patients because of tumour invasion across the anatomical midline. The median number of intra-arterial infusions was 8 (range 6–8). The median cisplatin dose was 30 mg/m 2 (range 20–50 mg/m 2 ) and the median total cisplatin dose was 265 mg/m 2 (range 180–370 mg/m 2 ). The median indwelling period of the catheter was 51.3 days (range 28–69 days).
The catheter position was adjusted during therapy in 10 cases. Intra-arterial infusions were performed through a catheter that was placed as planned in 30 of the 34 cases. During therapy, the catheter was removed in four patients because of fever thought to be caused by a catheter infection. The catheter was removed after three infusions in one patient, but a total of eight infusions were completed after replacement of the catheter via the occipital artery. The catheter was removed after seven infusions in three patients. All eight infusions were performed via the Seldinger technique in one patient, and intra-arterial infusion was ceased in two patients because of evaluation of a complete response. All patients completed radiotherapy as planned.
Eight adverse events of grade 3 occurred in seven of the 36 patients (19.4%). Of the grade 3 events associated with chemotherapy and radiotherapy, leukopenia occurred in three patients (8.3%), anaemia occurred in one patient (2.8%), hyponatremia occurred in two patients (5.6%), hyperkalemia occurred in one patient (2.8%), and mucositis occurred in one patient (2.8%) ( Table 2 ). There were no adverse events of grade 4 or more. There was no cerebrovascular disease associated with catheter placement. Grade 1 or 2 mucositis was seen in almost all patients, in combination with anorexia, as it was difficult to ingest food orally. As a result, a gastrostomy feeding tube was required in four patients, but it was removed within 2 months after treatment. Three patients showed cranial nerve paralysis. One patient showed cranial nerve II, III, and V paralysis after one cisplatin infusion. One patient showed cranial nerve VII paralysis after three cisplatin infusions. One patient showed cranial nerve X paralysis after the completion of therapy. These cases of paralysis persisted for 33 months, 30 months, and 44 months, respectively, in each patient, but there were no functional problems. Trismus occurred in five patients; four developed trismus during treatment and one developed it after the completion of treatment. This improved in two patients after opening training for approximately 1 year, but the trismus persisted in one patient for 8 months after onset. The remaining two patients died. Visual dysfunction occurred in one patient after treatment and persisted for 10 months, showing no improvement at the time of last follow-up.