ht. J. Radiarion Onco/ogy Biol. Phys.. Vol. 13. PP. 785-788 0360-3016/87 $3.00 + .OO

Printed in the U.S.A. All rights reserved. Copyright 8 1987 Pergamon Journals Ltd.

??Brief Communication

INTRAOPERATIVE RADIATION THERAPY FOR ADVANCED OR RECURRENT HEAD AND NECK CANCER

PETER GARRETT, M.D.,’ NEWELL PUGH, M.D.,’ DAVID Ross, M.D.,’ RON HAMAKER, M.D.2 AND MARK SINGER, M.D.2

‘Department of Radiation Therapy, Methodist Hospital of Indiana; and ‘Head and Neck Surgery Associates, Indianapolis, IN

Between 1982 and 1984, we treated 28 patients with advanced head and neck cancer with surgery, combined with intraoperative radiation. All patients had sqaamous cell carcinoma. One patient had two separate sites treated and a second patient was treated on two occasions, allowing 30 sites for analysis. The overall survival for all patieats treated was 67% at 1 year. Local failure occurred in 13% of those with close surgical margins and 25% of those with mhxoscopic residual disease. With gross residaal disease 100% had local failure. Patients with gross residual disease had significantly higher local failure than microscopic residual disease (p < 0.02) or close sargical margins (p < 0.01). Carotid blowout was the major treatment complication. We believe intraoperatlve radiition is an effective treatment for advanced or recurrent head and neck cancer when all gross disease has been resected.

Intraoperative radiition, Head and neck cancer.

INTRODUCTION

Patients with advanced nodal disease are difficult to con- trol with radiation alone.’ When recurrence takes place, surgery is difficult with fixed disease; long term prognosis is poor.4 After initial reports of encouraging results, most chemotherapeutic regimens have failed to dramatically alter the course of this advanced disease. ’ I

To treat advanced or marginally resectable disease in the abdomen, the use of radiation at the time of surgery has become po~ular.‘*~,~,“,‘~ Intraoperative Radiation Therapy (IORT) allows the radiation oncologist to give a high single dose of radiation to the tumor volume. The tumor can be visualized directly and critically normal structures can be moved from the field. A large single fraction of 1500 to 2000 cGy can overcome the initial shoulder of repair and produce tumor kill of sev- eral log. I2

We began a pilot project in 1982 to treat advanced or recurrent head and neck cancer with resection combined with intraoperative radiation. The head and neck is ide- ally suited for this type of treatment. The surgical field is easily visualized and normal tissue, such as skin and soft tissue can be moved out of the radiation field. The spinal cord can be avoided by using appropriate electron ener- gies. This paper reports the initial results of that project.

METHODS AND MATERIALS

Between May of 1982 and October of 1984 we treated 28 patients with advanced head and neck malignancies with IORT. One patient was treated on two separate oc- casions and a second had two separate sites treated con- currently. Thus there were 30 sites for analysis in 28 pa- tients.

An analysis was done of local tumor control and sur- vival. Comparisons between treatment groups were made using the log-rank method.’ The actuarial method was used to calculate survival3 and survival was calcu- lated from the time of IORT.

There were 2 1 males and 7 females in the study (ratio of 3: 1). The patients’ ages ranged from 44 to 8 1 years, with a median of 60 years. Eight patients were treated in 1982, seven in 1983 and 15 in 1984. The minimum fol- low-up is 14 months.

There were three indications for the use of IORT. The first was gross residual disease in seven sites. These pa- tients had the bulk of tumor resected but tumor persisted in the surgical field. The second indication was micro- scopic residual disease in eight sites. This group had all visible disease resected but there was tumor at the margin of the specimen on frozen section at the time of surgery. The final indication was close margins in 15 sites. An

Reprint requests to: Peter Garrett, M.D., Department of Ra- diation Therapy, Methodist Hospital of Indiana, P.O. Box 1367, 1701 N. Senate Blvd., Indianapolis, IN 46206.

Accepted for publication 8 December 1986.

785

786 I. J. Radiation Oncology 0 Biology 0 Physics May 1987, Volume 13, Number 5

acceptable margin of normal tissue could not be ob- tained around the tumor in patients in this group. This most often occurred with disease around the carotid ar- tery.

All patients had squamous cell carcinoma. At the same time we also treated 13 patients with malignant salivary gland tumors. An additional 24 patients were initially considered for IORT, but there was either inoperable dis- ease at the time of surgery or an acceptable margin of normal tissue was obtained.

The neck was the most common site treated with IORT ( 17 sites). All these patients had N3 disease. Table 1 summarizes the sites treated.

Seventeen sites were treated after failing conventional external beam radiation. The dose ranged from 5000 to 8200 cGy with a median of 6000 cGy. Ten patients re- ceived postoperative external beam radiation with a range from 4500 to 6000 cGy and a median of 5000 cGy. Two patients received a planned preoperative course of 4500 cGy followed by IORT. All treatments were given 180 to 200 cGy daily fractions.

All treatment was given using a linear accelerator. Twenty-three sites were treated using 4 Mev electrons (surface dose = 834, D,, at 0.5 cm). Six were treated using 7 Mev electrons (surface dose = 87% D, at 1.2 cm). One patient was treated using 11 Mev electrons (surface dose = 91%, D- at 1.8 cm). The patients with gross residual disease were treated with the higher ener- gies.

The most commonly prescribed dose was 2000 cGy ( 17 sites). Nine patients received 1500 cGy. Two patients were given 1000 cGy and two patients received 2500 and 10,000 cGy, mspectively. The patient given 10,000 cGy had disease invading the mandible. All doses were pre- scribed at D,.

We designed transparent circular lucite cones to facili- tate treatment. The surface of the cone abutted the sur- face of the treatment field. Subcutaneous tissue and skin edges were retracted out of the field. Other critical struc- tures, such as the esophagous, were blocked individually within the lead cone. The most frequent cone size used was 6 cm (12 sites). Eight were treated with 5 cm cones and 7 with 4 cm cones. Three patients were treated with 9 cm cones.

RESULTS

The results of treatment were analyzed in terms of lo- cal control by disease group. There were two failures within the surgical field in the 15 sites with close surgical margins. Of the eight patients treated for microscopic re- sidual disease, there were two local failures.

There were seven patients treated with gross residual disease. One patient died 2 days postop from airway ob- struction because of a mucous plug. This case is not in- cluded in local control analysis but is included in survival

Table I. Sites of disease treated with intraoperative radiation

Site treated

Fixed nodal disease Pterygoid Maxilla Floor of mouth Temporal re$on Parotid Tongue

Number

17 4 2

: 2 1

data. All six treated with gross residual disease failed within the surgical field.

There was a higher frequency of local control for close margins (87%) when compared to gross residual disease (0%). This superior local control is statistically significant (p < 0.02). There was no significant difference between close margins and microscopic residual.

Table 2 summarizes local control by surgical group compared to the use of external beam radiation. There were six failures in the 17 patients who failed previous external beam therapy (35% local failure). There were four failures in the ten patients with no prior radiation. These patients also received postoperative external beam therapy.

Table 3 compares local control by surgical group ver- sus anatomic site of disease. There were six failures in the 17 patients with nodal disease (35% local failure). There were four failures in the 12 other sites outside the neck.

The actuarial l-year survival for all patients treated was 67%. For close surgical margins the l-year survival was 76% and for microscopic disease was 86%. The actu- arial l-year survival for patients with gross residual dis- ease was 14%. None of the patients with gross residual disease survived.

The major morbidity associated with the treatment of advanced head and neck cancer was carotid blowout. This occurred in two patients and was fatal in both cases. The first case had received 7000 cGy followed by a laryn- gectomy for cancer of the larynx. This patient was re- ferred with a stomal recurrence. He had gross persistent disease invading the carotid. He received 2000 cGy at the time of surgery and died 1 month later of a carotid blowout. A postmortem examination was not obtained. The second patient had disease resected off the carotid and had a close surgical margin. This patient received

Table 2. Local tumor failure versus conventional external radiation

Margins Microscopic Gross Totals

Preop XRT Of2 O/O O/O O/2 Failures of XRT 219 216 212 6117 Postop XRT O/4 O/2 414 4110 Totals 2115 2/g 616 10129

Intraoperative RT for head and neck cancer 0 P. GARRETT et al. 787

Table 3. Local failure versus site of disease

Margins Microscopic Gross Totals

Neck Pterygoid Maxilla Floor of mouth Temporal

2112 l/2 313 6117 O/l O/3 O/O O/4 O/O l/l l/l 212 O/l O/l O/O O/2

region Parotid Totals

O/O O/O 212 212 O/l O/l O/O O/2 2115 218 616 IO/29

1500 cGy at the time of IORT. This was followed by postoperative radiation with an additional 5000 cGy in 24 treatments. Subsequent to this, an ulcer appeared in the pyriform sinus secondary to radiation. This pro- gressed to a sinus formation with infection and a carotid blowout occurred 9 months after the IORT. Postmortem showed no local tumor and radiation effect on the ca- rotid artery.

There was one other major complication of osteone- crosis of the mandible. This patient had a carcinoma of the floor of the mouth treated by a composite resection of the floor of the mouth, partial glossectomy, marginal mandibulectomy, and radical neck dissection. There was microscopic residual disease in the remaining mandible and this area received 10,000 cGy intraoperatively fol- lowed by 6000 cGy in 30 treatments postoperatively. Two months alter the IORT, a fistula developed and sub sequently, exposed bone was visible. The mandible was removed and focal necrotic debris found.

DISfXSSION

Patients with advanced or recurrent head and neck cancer have an extremely poor prognosis. Disease may be disfiguring and provide a painful and distressing de- mise, with ulceration of tumor nodules with associated infection and bleeding.

The results of treatment were excellent for the group with close margins or microscopic residual disease. Fail- ure in the operative field was uncommon and over 75%

were alive 1 year after treatment. Patients with gross dis- ease remain a problem. The survival is poor and all pa- tients at risk have failed locally. Thus it appears for IORT to be successful, all gross disease must be resected.

Patients who have failed conventional external beam radiation tolerated intraoperative radiation in the doses given. The results with local control were similar to those receiving IORT and postoperative radiation. It should be noted, the higher proportion of patients with gross dis- ease is the postoperative radiation group.

The major complication observed was carotid blow- out. The two cases occurred in the early part of the series. After these cases we took two precautions to try to pre- vent this from recurring. If disease was removed from the region of the carotid a flap with a myocutaneous graft was placed covering the surgical defect. This improved the vascular&y of the region and provided extra protec- tion to the carotid. Also, if gross disease was left on the carotid, IORT was not given. Shrinkage of the tumor may weaken the carotid wall or result in a direct ulcer- ation leading to a blowout. Since we have adopted these measures, we have had no more of these complications.

We did have one case of osteonecrosis when we used a massive dose of 10,000 cGy to the mandible. We had previously treated the mandible to a similar dose after removal with successful results.’ Disease invading the mandible remains a difficult problem; the tolerance level of the mandible to high single doses of IORT is not known. We have subsequently reduced our dose to the mandible to 2000 to 2500 cGy.

We are encouraged by these preliminary results with intraoperative radiation in head and neck cancer. Fixed disease in the neck has been controlled in 65% of pa- tients. Similar disease treated with external beam therapy has a control rate of 39% in Harwood’s series from the Princess Margaret Hospital.’ Further patient accrual is required to co&m these results in large patient num- bers. It will be necessary to study this treatment in a ran- domized multicentric trial in the future. The addition of radiosensitizers to IORT is an interesting possibility. By treating with IORT at peak serum levels, some of the dose related toxicity of radiosensitizers may be avoided.’

REFERENCES

Abe, M., Takahashi, M.: Intraoperative radiotherapy: The Japanese experience. Int. J. Radiat. Oncol. Biol. Phys. I: 863-868,198l. Abe, M., Takaha&i, M., Yabumoto, E., Ada&i, H., Yoshi, M., Mori, K.: Clinical experiences with intraoperative ra- diotherapy of locally advanced cancers. Cancer 45: 40-48, 1980. Cutler, S.J., Ederer, F.: Maximum utilization of the life ta- ble method in analyzing survival. J. Chron. Dis. 8: 699- 712,1958. Garrett, P., Beale, Fr., Cummings, B., Harwood, A., Keane, T., Payne, D., Rider, W.: Cancer of the tonsil: Re-

5.

6.

sults of radical radiation with surgery in reserve. Am. J. Surg. 146: 432-435,1983. Gray, A., Dische, S., Adams, G., Flockhart, I., Foster, J.: Clinical testing of the radiosensitizer Ro-O7-0582,1. Dose, tolerance, serum and tumor concentration. Clin. Radiol. 27: 151-157,1976. Gunderson, L., Coehn, A., Dosemtx, Dr., Shipley, W., Hedberg, S., Wood, W., Rodkey, G., Suit, H.: Residual un- resectable or recurrent colorectal can= External beam irradiation and intraoperative electron beam boost t/-re- section. Int. J. Radiat. Oncol. Biol. Phys. 9: 1597-1606, 1983.

788 I. J. Radiation Oncology 0 Biology 0 Physics May 1987, Volume 13, Number 5

7.

8.

9.

10.

Hamaker, R., Singer, M., Shockley, W., Pugh, N., Shidnia, H.: Irradiated mandibular autografts. Cancer 52: 10 17- 1021,1983. Harwood, A., Beale, F., Cummings, B., Keane, T., Payne, D., Rider, W., Rawlinson, E., Elhakim, T.: Sumaglottic laryngeal carcinoma: An analysis of dose-time-volume fac- tors in 410 patients. ht. J. Radiat. Oncol. Biol. Phys. 9: 311-319,1983. Peto, R., Poke, M., Armitage, P.: Design and analysis of randomized clinical trials requiring prolonged observation of each patient 11. Analysis and examples. Br. J. Cancer 35: l-39,1977. Shipley, W., Wood, W., Tepper, J., Warshaw, A., Orlow, E., Kaufman, S., Battit, G., Nardi, Gr.: Intraoperative elec-

11.

12*

13.

tron beam irradiation for patients with unresectable pan- creatic cancer. Ann. Surg. 200: 289-294, 1984. Vogel, S., Schoenfield, D., Kaplan, B., Lemer, H., Eng- Strom, P., Horton, J.: A randomized prospective compari- son of methotrexate with a combination of methodtrexate, bleomycin and cisplatin in head and neck cancer. Cancer 56: 432-442, 1985. Withers, J.R.: The dose survival relationship for irradia- tion of epithelial cells of mouse skin. Br. J. Radiol. 40: 187-194,1967. Wood, W., Shipley, W., Gunderson, L., Cohen, A., Nardi, G.: Intraoperative irradiation for unresectable pancreatic carcinoma. Cancer49: 1272-1275,1982.