High dose rate brachytherapy for carcinoma of the oral tongue.

PURPOSE: The purpose of this study is to assess the feasibility of treating early-staged tongue cancer with high dose rate (HDR) remote afterloading technique. Furthermore, a new figure of merit, the Geometry Index (GI), is introduced to quantify the quality of the implants. METHODS AND MATERIALS: Between 1994 and 1995, eight patients with carcinoma of the oral tongue were treated solely with interstitial implant using the HDR remote afterloading technique. Five patients had T1 N0 disease and the remaining three had T2 N0 disease. Elective neck treatment was withheld. The male-to-female ratio was 1:1, and the mean age 60 years (range: 32-72 years). The median follow-up time was 26 months (range: 6-30 months). The afterloading catheters were positioned through the submandibular approach with the assistance of templates. Six patients had single planar implant and the remaining two had double planar implant. The median number of catheters inserted was 5 (range: 4-9). The median dose given was 60 Gy in 10 fractions over 6 days. The interfraction interval was 7 h. Mandibular and maxillary shields were inserted prior to treatment. Thomadsen et al. introduced the use of Implant Quality Index (QI). We introduce a new parameter, GI, which is defined as ratio of the QI of the nonoptimized executed implant to the corresponding QI value of the nonoptimized idealized implant.
RESULTS: The mucositis lasted for 6 to 20 weeks (median: 10 weeks). There was no local failure up to a median follow-up of 26 months. Two patients developed ipsilateral neck node metastases at 2 and 4 months following implant, respectively. One patient had involvement at level II and the other failed at level I to III. Both patients were salvaged by neck node dissection and regionally remained in control. One patient with multiple nodal metastases and extracapsular spread developed biopsy-proven liver metastases and succumbed 6 months following implant. One patient treated with double planar implant developed Grade 3 necrosis of the soft tissue and bone. This complication is largely preventable now, as we have acquired more technical expertise. The mean GI values for the single and double planar implants were 0.88 (range: 0.84-0.91) and 0.8, respectively. This correlates with our practical experience that it is more difficult to maintain a good geometry as double planar implant is required. The GI gives a better view of the geometry of implant as it compares the nonoptimized QI of the executed implant with its ideal counterpart. The failure to achieve a high GI in double planar implants is presumed to relate to technical difficulties rather than variation in individual performance.
CONCLUSION: Our preliminary experience in treating early-staged tongue cancer with the HDR remote afterloading technique is inspiring, as it gives a local control rate of 100% with acceptable morbidity. Further studies are eagerly awaited to delineate the optimum schedule for this modality of treatment. It is hoped that the GI values, which represents the skills of insertion, could be routinely reported so that treatment results between different centers could be compared in a more precise manner.