381 Radiation Therapy for Glottic Cancer Using 6-MV Photons Robert L. Foote, M.D.’ Gordon L. Grado, M.D! Steven J. Buskirk, M.D? Kerry D. Olsen, M.D? James A. Bonner, M.D.’ John D. Earle, M.D.’ Jan L. Kasperbauer, M.D? Thomas V. McCaffrey, M.D., Ph.D.2 Vera J. Suman, P ~ . D . ~ ’ Division of Radiation Oncology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Department of Otorhinolaryngology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Section of Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Department of Radiation Oncology, Mayo Clinic Scottsdale, Scottsdale, Arizona. Department of Radiation Oncology, Mayo Clinic Jacksonville, Jacksonville, Florida. BACKGROUND. It has been recommended that cobalt-60 or 4-MV photons be used when treating glottic cancer with radiation therapy. Underdosing may occur when using higher energy photons, particularly when the anterior commissure is involved. The authors report their experience using higher energy photons (6 MV) for the treatment of glottic cancer. METHODS. Between January 1975 and July 1991, 73 patients with Tis, T1, ‘r2, or T3 glottic tumors underwent radiation therapy with curative intent. Cobalt-60 or 4MV photons were used to treat the cancers of 30 patients, and 6-MV photons were used to treat 43 patients. Clinical records were reviewed retrospectively to determine patterns of treatment failure, cancer deaths, and local tumor control in the 43 patients receiving treatment with 6-MV photons. Patients were followed until death or for a minimum of two years. RESULTS. Treatment failures were: local recurrence, one patient; local recurrence and distant metastasis, one patient; delayed neck metastasis, two patients; and delayed neck metastasis with distant metastasis, one patient. Three patients who had treatment failure are alive and free of cancer after salvage treatment. Two patients died of neck and distant metastases. The 2-year initial local tumor control rate was 94.8%. CONCLUSIONS. Glottic cancer can be treated successfully with 6-MV photons. Local tumor control is similar to that reported using cobalt-60 or 4-MV photons. Cancer 1996; 77:381-6. 0 1996 American Cancer Society. KEYWORDS: glottic cancer, radiation therapy, higher energy photon beam, cobalt-60. R Presented at the 2nd World Congress on Laryngeal Cancer, Sydney, Australia, February 20 to 24, 1994. Address for reprints: Robert L. Foote, M.D., Mayo Clinic, 200 First Street SW, Rochester, MN 55905. Received June 15, 1995; revision received September 8, 1995; accepted September 8, 1995. 8 0 1996 American Cancer Society adiation therapy for early glottic cancer is a highly successful form of treatment. Tumor control rates in large series of patients from major cancer treatment centers have been reported to be 88% to 96% for Stage T1 tumors and 52% to 80% for Stage T2 tumors after treatment with c0balt-60.’-~It has been suggested that the optimal beam energy is cobalt60 or 4-MV photons.” Underdosing may occur when higher energy photons are used, particularly when the anterior commissure is involved, because of inadequate buildup of electrons in the thin wedge-shaped anterior neck and at the surface of the vocal cord-air interface. However, the number of cobalt-60 machines in radiation therapy facilities in the United States has decreased from 970 in 1975 to 504 in 1990. During this same time, the number of linear accelerators has increased from 407 to 1893.’’ Therefore, fewer radiation facilities have cobalt-60 machines available to treat early glottic cancer. If higher energy photons are found to be less effective, perhaps patients with early stage glottic cancer should be referred to treatment centers with cobalt-60 machines. The cobalt-60 machine at our institution was decommissioned in 1982. Since that time, 382 CANCER January 15, 1996 / Volume 77 I Number 2 the majority of patients have received treatment with 6MV photons. Our initial review revealed 5-year local tumor control rates of 77% for cobalt-60, 75% for 4-MV photons, and 100% (16 patients) for 6-MV photons when treating early (Tis, T1, T2) glottic cancer.” The purpose of the present review is to determine the pattern of treatment failure, cause of death, and local tumor control rate when using higher energy photons (6 Mv) for the treatment of glottic cancer. MATERIALS AND METHODS Between January 1975 and December 1991, 73 patients with Stage Tis, T1, T2, or T3 squamous cell carcinoma of the true vocal cord underwent treatment with curative intent with radiation therapy at the Mayo Clinic in Rochester, Minnesota, Scottsdale, Arizona, and Jacksonville, Florida. Of these 73 patients, 30 had treatment with cobalt-60 or 4-MV photons and 43 with 6-MV photons. Clinical records were reviewed retrospectively to determine patterns of treatment failure and cancer-related deaths in the 43 patients who underwent treatment with 6-MV photons. Patients were followed until death or for a minimum of two years. The tumors were staged according to the 1992 American Joint Committee on Cancer staging ~ y s t e m . Stage ’~ T1 tumors were stratified as follows: Tla, involvement of one true vocal cord with or without involvement of the anterior commissure; and Tlb, involvement of both true vocal cords. Stage T2 tumors were stratified according to the mobility of the vocal cord(s): T2a, normal mobility, and T2b, reduced mobility. The 43 patients consisted of 41 males and 2 females (median age, 68 years; range, 36 to 88 years). Three patients had Stage Tis tumors, 27 had Stage T1 tumors (15 patients, Tla and 12 patients, Tlb), 10 had Stage T2 tumors (6 patients, T2a and 4 patients, TZb), and 3 patients had stage T3 tumors. The anterior commissure was involved with tumor in 26 patients. All patients underwent endoscopic biopsy before radiation therapy. All radiotherapy was administered with a linear accelerator using a 6-MV photon beam. In four patients, a three-field treatment technique was used that incorporated an anterior electron beam during the last few treatments. All patients underwent simulation and treatment in the supine position except for one patient who was treated in the lateral decubitus position. Since 1988, a mask has been fabricated for immobilization. In addition to weekly port films, the treatment fields were verified each day on the treatment machine by a physician using the anatomical landmarks of the thyroid and cricoid cartilages. For Tis, T1, and T2 tumors involving the entire true vocal cord, the superior and inferior borders of the field extended from the middle of the thyroid notch to the bottom of the cricoid cartilage; the posterior border was 1 to 1.5 cm behind the posterior aspect of the thyroid cartilage, and there was a 1- to 2-cm fall off anteriorly. Planned reductions in the field size during treatment were rarely performed. The field borders were modified for each patient, depending on the anatomical extent of the particular tumor. The median field length was 5 cm, and the median field width was 6.5 cm. For Stage T3 tumors, the entire neck was initially treated before reducing the field size. Opposed lateral treatment fields were used in the treatment of 33 patients. Six patients received treatment with a three-field technique and four patients with oblique fields. Multiple ( 2 2 ) fields were treated each day in 40 patients; 1 field was treated each day in 3 patients. Wedge filters were used in 36 patients to improve the dose homogeneity. Cerrobend blocks were fabricated for 15 patients. One patient with a narrow anterior neck received treatment with bolus. The lower cervical nodes were treated with an anterior low neck field for T3 tumors. The tumor dose was specified to the isodose volume that encompassed the tumor, which was determined by computer-generated dosimetry. The median total dose delivered was 65.25 Gy (range, 27 to 74.4 Gy). The patient receiving treatment with 27 Gy died of an unrelated intercurrent medical illness during radiation therapy. The dose per fraction ranged from 1.2 to 2.3 Gy (median, 2 Gy). Beginning in 1988, the doses have been standardized to 63 Gy, using 2.25-Gy fractions, for T1 tumors and to 74.4 to 76.8 Gy, using 1.2 Gy per fraction twice a day, for T2 and T3 tumors. There were 41 patients who had treatment with once-daily fractionation, and 2 patients (1with T2 tumor and 1with T3 tumor) who had treatment with twice-daily fractionation. Seventeen patients received continuous course treatment. In the 26 other patients, treatment was interrupted for a median of 3 days (range, 1 to 86 days). The median number of days from the start of radiation therapy to the end of radiation therapy was 45 (range, 16 to 141 days). All patients were followed until death or for a minimum of two years. The median length of follow-up was 3.4 years. All three patients with Stage Tis tumors were followed up for two or more years. Of the 15 patients with Stage T l a tumors, 12 were followed for 2 or more years, as were 10 of the 12 patients with Stage T l b tumors, 5 of the 6 patients with Stage T2a tumors, 2 of the 4 patients with Stage T2b tumors, and 1 of the 3 patients with Stage T3 tumors. Ten patients died before two years of followup. Therefore, the majority of patients were followed for two or more years. Most of the local recurrences (290%) have been reported to occur during the first two years of follow-up. The Kaplan-Meier method was used to estimate time to initial local failure, time to any initial failure (local, nodal, or distant), and survival to death from any cause.I4The log rank test was used to compare survival and recurrence times. Initial local control was defined as Radiation Therapy for Glottic Cancer/Foote et al. 383 TABLE 1 Initial Local Control According to T Stage for 43 Patients Stage No. of patients Tis 3 Tla 15 12 6 4 Tlb T?a* T?b' T3 All 3 43 2-year control (%) SE (%) 100 100 100 83 75 100 94.8 I SE: standard error. 0 1 2 3 4 5 Years from diagnosis 'lwo-year initial local control for id1 T2 IT?a and T2b) tiimors, 78.8%ISE. 13%). FIGURE 1. Relapse-free and overall survival. complete and continuous disappearance of disease at the primary site. Variables evaluated for potential prognostic impact on local control included tumor stage and anterior commissure involvement. RESULTS Local Control Initial local control according to T Stage is outlined in Table 1. For all patients, the rate of initial local tumor control at 2 years was 94.8% (standard error [SE] = 3.6%). There were two local failures (one at 4 months and the other at 16 months after diagnosis),which were surgically managed with total laryngectomy. One of the two patients initially presented with a Stage T2a tumor that would have required partial vertical laryngectomy. The second patient presented with a Stage T2b tumor that would have required total laryngectomy. Pathologically, this tumor was a Grade 4 spindle cell variant of squamous cell carcinoma. In both patients, the anterior commissure was initially involved and the anterior commissure and true vocal cords were involved with recurrent tumor. When the anterior commissure was involved, the 2-year initial local tumor control rate was 91.6% (SE = 5.7%). When the anterior commissure was not involved, the 2-year initial local tumor control rate was 100%. This difference was not statistically significant ( P = 0.26). Patterns of Treatment Failure and Relapse Free Survival At latest follow-up evaluation, 38 patients had not experienced treatment failure (local recurrence, nodal metastasis, distant metastasis). Of the five other patients, cervical node metastases had developed in two, both nodal and distant metastases in one, local recurrence in one, and both local recurrence and distant metastases in one. The 2-year relapse free survival was 92.3% (SE = 4.3%; Fig. 1). Overall Survival At latest follow-up evaluation, 21 patients were still alive and free of cancer. Of the 18 patients who had died with evidence of glottic cancer, 14 died of an intercurrent medical illness and 4 of second malignancies. One patient is alive with glottic cancer. Of the three patients who died of glottic cancer, two died of uncontrolled nodal or distant metastasis (or both) and the other died of an intercurrent illness during the course of radiation therapy. The 2-year overall survival was 76.7% (SE = 6.4%; Fig. 1). Complications Chondronecrosis did not develop in any of the patients. Grade 3 laryngeal edema (according to the Radiation Therapy Oncology Group (RTOG)/European Organization for Research and Treatment of Cancer (EORTC) late radiation morbidity scoring schema) developed in one patient (Stage T2b tumor). This required placement of a permanent tracheostomy. DISCUSSION It has been recommended that the optimal beam energy for treatment of early glottic cancer is cobalt-60 or 4-MV photons." The concern is that with higher energy photon beams underdosing may occur along the air-tumor tissue interface because of the air cavity within the larynx, resulting in inadequate buildup of electrons at the surface of the vocal cord cancer. Also, in the relatively thin wedgeshaped anterior neck, inadequate buildup of electrons in the region of the anterior commissure may occur witn higher energy photon beams. Dosimetry studies designed to measure the degree of underdosing have yielded conflicting (written communication, Dr. Michael D. Sombeck, November, 1993). The relative merits of 6 MV compared with cobalt-60 for treatment of early vocal cord cancer are still being investigated. Sombeck and associates (written communication, November, 1993) have reported a comparison of dose values in an anthropomorphic phantom made of methyl methacrylate (Lucite) for 364 CANCER January 15,1996 I Volume 77 I Number 2 TABLE 2 Radiation Therapy for Glottic Cancer Using 6-MV Photons Literature Review-Local Control for T1 Glottic Cancer Orthovoltage Institution Washington University Columbia-Presbyterian Shinshu University National Cancer Center Hospital, Tokyo Thomas Jefferson University Hospital Northwestern University Gunma University Maya Clinic Cobalt 60 Length of follow-up (yr) No. of patients 5-yr control rate (W) No. of patients 5-yr control rate (96) 3 ND 60 13 78 100 54 16 17 85 87 88 19 46 15 84 85 68 ND 5 2 2 Nn 2 ND: no data. a specific irradiation geometry (wedged parallel opposed fields) and technique (prescription to an isodose surface). Whether their observation of lower dose (16% to 35%) at the skin surface and for 1 point 3 mm below the anterior neck surface with 6 MV photons compared with cobalt60 is clinically relevant or universally applicable is still a question. Doses measured along the surface of the vocal cords and at the anterior commissure were essentially identical. Initial reports of phantom dosimetric studies suggested that the dose absorbed at the anterior commissure may be decreased by 12% with 6-MV photons and by 18%with 10-MVphotons when compared with cobalt60.’* This is a practical concern because most cobalt-60 units in the United States are being replaced by higher energy linear accelerators.” However, very limited clinical information is available to evaluate the efficacy of 6MV or higher photon beams in the treatment of early glottic cancer. No prospective randomized clinical trials exist. The following is a review of retrospective studies. At the Mallinckrodt Institute of Radiologyin St. Louis, 281 patients with T1 glottic cancer received the following treatment: 60 patients with 250-kilovolt peak (kVp) photons, 54 with cobalt-60 gamma rays, 144 with 4-MV photons, and 23 with 6-MV photons. Most patients received 6000 cGy in 200-cGy daily fractions. All patients were followed for a minimum of three years. The 5-year local tumor control was reported to be 78% for patients receiving treatment with 250-kVp photons, 85% for those receiving cobalt-60,84% for those receiving 4-MV photons, and 70% for those receiving 6-MV photon^.'^ Sung et al. reported 5-year local tumor control rates for T1 and T2 glottic tumors of 88% for orthovoltage (26 patients), 84% for cobalt-60 (19 patients), 65% for 6-MV photons (20 patients), and 76% for 22-MV photons (97 patients).’” The differences in tumor control rates were significant ( P = 0.05) in favor of cobalt-60 versus 6-MV photons for Stage T1 tumors (87% versus 60%, respectively). The differences in tumor control rates were also significant (P= 0.03 and P = 0.01) in favor of orthovoltage versus 6-MV and 22-MV photons for Stage T1 tumors (100%versus 60% and 74%, respectively). No wedge filter or bolus was used with the 22-MV photon beam. Doses varied from 5000 to 6600 cGy in 200-cGy daily fractions.” Izuno et al. reported the results of treatment of 53 patients who had TlNOMO glottic tumors.2’ Of these 53 patients, 17 underwent treatment with cobalt-60, 25 with 8- or 10-MV photons, and 11 with 4-MV photons. The mean total dose was 6330 cGy in the cobalt-60 group, 6590 cGy in the 8- or 10-MV group, and 6380 cGy in the 4-MV group, with 180- or 200-cGy daily fractions. Wedge filters were used for most treatments in the 8- to 10-MV and 4-MV groups but not in the cobalt-60 group. The 5year local control rate was 88% (15 of 17) for cobalt-60 and 60% (15 of 25) for 8- or 10-MV photons (P = 0.05). The follow-up period for the 4-MV photon group was limited, but the 3-year local control rate was 91% (10 of 11). These retrospective studies suggest that the estimated 5-year local tumor control rate may be lower with 6-Mv or higher energy photons when compared with cobalt-60 gamma rays or 4-MV photons. However, other patient-, tumor-, or treatment-related factors maybe responsible for the differences noticed because these are nonrandomized clinical data. Akine et al. described the results for 154 patients with Stage T1 tumors treated with 6-MV photons (8 of 151 patients receiving 6 MV had 6 MV and cobalt-60) or cobalt-60 (3 patients).” In 110 of the patients receiving treatment with 6-MV photons, no bolus or wedge filter was used. Doses varied from 5750 to 7250 cGy, with most patients having 200-cGy daily fractions. The minimum follow-up period was five years. The 5-year local control Radiation Therapy for Glottic Cancer/Foote et al. 2Mv No. of patients 5-yr control rate (%) No. of patients 5-yr control rate (96) 144 84 No. of patients 23 15 4 100 11 91 53 77 90 41 8/10 Mv 6MV 4Mv 5-yr control rate (W) 5-yr control rate (%) 25 60 33 89 70 60 151 15 14 86 92 27 100 rate was 89%, comparable with the 88% to 96% rate reported by others using ~ o b a l t - 6 0 . ’ - ~However, ~’~ for the 27 patients with anterior commissure involvement, the 5-year local control rate was 81%, in contrast to 91% for the 127 patients without anterior commissure involvement. This was of borderline significance univariately ( P = 0.061, and the analysis did not control for other important treatment factors, such as total dose, fraction size, and overall treatment time.2z Nevertheless, a trend was suggested and is of clinical concern, In contrast, a review of approximately 2200 Stage TI vocal cord tumors treated with cobalt-60 showed no increase in local failure with anterior commissure involvement.I0 However, Rudoltz et al. reported the results of using various photon energies to treat Stage TI glottic tumors in 91 patient^.'^ The minimum follow-up period was two years. Wedge filters were used in 50 patients. The median dose was 6400 cGy. The 5-year local tumor control rate for cobalt-60 was 84% (19 patients); for 2-MV photons, 100% (4 patients); for 4-MV photons, 77% (53 patients); and for 6-MV photons, 86% (15 patients). None of these differences were statistically significant with multivariate analysis, including evaluation of anterior commissure involvement ( P = 0.58).23 Small et al. reported the results of using various photon energies to treat Stage TI glottic tumors in 101 pat i e n t ~The . ~ ~minimum follow-up period was two years. Wedge filters were used in 25 patients. The mean total dose and daily fraction dose were 6444 cGy and 228 cGy, respectively. The 5-year recurrence rates reported for cobalt-60 were 15% (46 patients); for 4-MV photons, 10% (41 patients); and for 6-MV photons, 8% (14 patients). No significant differences in outcome were observed among the groups.z4 Finally, Hayakawa et al. reported a 5-year local tumor control rate of 68% for 15 patients with Stage TI glottic No. of patients 385 22 Mv No. of patients 5-yrcontrol rate (%) 75 74 89 tumors treated with cobalt-60 and 45-degree wedge filters compared with 85% in 27 patients having treatment with 10-MV photons without wedge filters and 89% in 33 patients receiving 10-MV photons without wedge filters either alone or in combination with cobalt-60.25Total doses ranged from 6000 to 7000 cGy, with either 300 cGy per fraction 3 days a week or 200 to 250 cGy per fraction 5 days a week. These retrospective studies suggest that the estimated 5-year local tumor control rates with 6-, 8-, or 10MV photons may be similar to those achieved with cobalt60 or 4-MV photons. However, these nonrandomized series with a relatively small number of patients have the same limitations as the studies that suggested inferior results with the use of 6 - M v or higher energy photons (Table 2). We report a 2-year initial local tumor control rate of 100% for Stage T1 glottic tumors and 78.8% (SE = 13.4%) for Stage T2 glottic tumors treated with 6-MV photons. No significant difference was noted between the tumors that involved the anterior commissure and those that did not (2-year initial local tumor control 91.6% compared with 100%; caution should be used when interpreting this because of the small sample size). These results are comparable with the 88% to 96% local tumor control rates reported for TI glottic tumors and 52% to 80% local tumor control rates reported for T2 glottic tumors treated with c0balt-60.*-~ We have previously reported 5-year local tumor control rates of 77% for cobalt-60, 75% for 4-MV photons, and 100% €or 6-MV photons when treating Stage Tis, TI, or T2 tumors with radiation therapy.” In that review, we found that total dose (263 Gy) was a significant predictor of tumor control but photon energy was not ( P = 0.007 and P = 0.082, respectively). Patients who had treatment with cobalt-60 or 4-MV photons were more likely to receive doses of less than 63 Gy. Therefore, on 386 CANCER January 15,1996 I Volume 77 I Number 2 the basis of our experience and that reported by others, we believe that early glottic cancer can be treated safely and effectively with 6-MV photons and that results very similar to those reported using cobalt-60 can be achieved. For patients with a very thin anterior neck and cancer involving the anterior commissure, beam spoiling or bolus could be considered to aid in increasing the dose to the anterior commissure. REFERENCES 1. Chacko DC, Hendrickson FR, Fisher A. Definitive irradiation of T,-T.,N,, larynx cancer. Cancer 1983;51:994- 1000. 2. Fein DA, Mendenhall WM, Parsons JT, Million RR. Tl-TZ squamous cell carcinoma of the glottic larynx treated with radiotherapy: a multivariate analysis of variables potentially influencing local control. Int J Radiat Oncol Biol Phys 1993;25:605- 11. 3. Fein DA, Mendenhall WM, Parsons JT, Stringer SP, Cassisi NJ, Million RR. Carcinoma in situ of the glottic larynx: the role of radiotherapy. Int J Radiat Oncol Biol Phys 1993;27:379-84. 4. Harwood AR, Beale FA, Curnmings BJ, Keane TJ, Rider WD. T2 glottic cancer: an analysis of dose-time-volume factors. Znt/ Rudiat Oncol Biol Phys 1981;7:1501-5. 5. Harwood AR, Hawkins NV, Keane T, Cummings B, Beale FA, Rider WD, et al. Radiotherapy of early glottic cancer. Laryngoscope 1980;90:465-70. 6. Howell-Burke D, Peters LJ, Goepfert H, Oswald MJ.T2 glottic cancer: recurrence, salvage, and survival after definitive radiotherapy. Arch Otolaryngol Head Neck Surg 1990;116:8305. 7. Mittal B, Rao DV, Marks JE, Perez CA. Role of radiation in the management of early vocal cord carcinoma. Int J Radiat Oncol Biol Phys 1983;9:997- 1002. 8. Wang CC. Cancer of the larynx: radiation therapy. CA Cancer J Clin 1976;26:212-8. 9. Woodhouse RJ, Quivey JM, Fu KK, Sien PS, Dedo HH, Phillips TL. Treatment of carcinoma of the vocal cord: a review of 20 years experience. Laryngoscope 1981;91:1155-62. 10. Million RR. The larynx . . . s o t o speak: everything I wanted to h o w about laryngeal cancer I learned in the last 32 years. Int J Radiat Oncol Biol Phys 1992;23:691-704. 11. Owen JB, Coia LR, Hanks GE. Recent patterns of growth in radiation therapy facilities in the United States: a patterns of care study report. lnt J Radiat Oncol Biol Phys 1992;24:983-6. 12. Foote RL, Olsen KD, Kunselman SJ, Schaid DJ, Buskirk SJ, Grado GL, et al. Early-stage squamous cell carcinoma of the glottic larynx managed with radiation therapy. Muyo Clin Proc 1992;67:629-36. 13. American Joint Committee on Cancer. In: Beahrs OH, Henson DE, Hutter RW, Kennedy BJ, editors. Manual for staging of cancer. 4th edition. Philadelphia: JB Lippincott Company, 1992. 14. Kaplan EL, Meier P. Nonparametric estimation from incornplete observations. J A m Stat Assoc 1958;53:457-81. 15. Epp ER, Boyer AL, Doppke KP. Underdosing of lesions resulting from lack of electronic equilibrium in upper respiratory air cavities irradiated by 10 MVx-ray beams. IntJRadiat Oncol Biol Phys 1977;2:613-9. 16. Klein EE, Chin LM, Rice RK, Mijnheer BJ. The influence of air cavities on interface doses for photon beams. IntJ Radiat Oncol Biol Phys 1993;27:419-27. 17. Nilsson B, Schnell P-0. Build-up effects at air cavities measured with thin thermoluminescent dosimeters. Acta Radio/ 1976;151427-32. 18. Shimizu R. An evaluation of absorbed doses of high energy electromagnetic radiation in radiotherapy of laryngeal cancer. Nippon Jibiinkoka Gakkai Kaiho 1990;93:796-807. 19. Devineni VR, King K, Perez C, Mittal B, Sirnpson J, Emami B. Early glottic carcinoma treated with radiotherapy: impact of treatment energy on success rate (abstract). Int J Rndiar Oncol Biol Phys 1992;24 (Suppl 1):186-7. 20. Sung DI, Chang CH, Harisiadis L, Rosenstein LM. Primary radiotherapy for carcinoma in situ and early invasive carcinoma of the glottic larynx. Int J Rudiar Oncol Biol Phys 1979;3467 -72. 21. Izuno 1, Sone S, Oguchi M, Kiyono K, Takei K. Treatment of early vocal cord carcinoma with “Co gamma rays, 8/10 M V x-rays, or 4 MV x-rays-are the results different? Actu Oncol 1990;29:637-9. 22. Akine Y, Tokita N, Ogino T, Tsukiyama I, Egawa S, Saikawa M, et al. Radiotherapy of T1 glottic cancer with 6 MeV x rays. IntJ Radiat Oncol Biol Phys 1991;20:1215-8. 23. Rudoltz MS, Benammar A, Mohiuddin M. Prognostic factors for local control and survival in T1 squamous cell carcinoma of the glottis. lnt 1 Radiat Oncol Biol Phys 1993;26:767-72. 24. Small W Jr, Mittal BB, Brand WN, Shetty RM, Rademaker AW, Beck GG, et al. Results of radiation therapy in early glottic carcinoma: multivariate analysis of prognostic and radiation therapy variables. Radiology 1992;183:789-94. 25. Hayakawa K, Mitsuhashi N, Tamaki Y, Takahashi M, Honjo J, Kamei T, et al. Radiation therapy for stage I and I1 laryngeal cancer using 10 MV x-rays or cobalt-60 gamma rays. Radiat Med 1992;10:199-205.