PII S0360-3016(98)00041-8

● Clinical Investigation

HIGH DOSE RATE VERSUS LOW DOSE RATE INTERSTITIALRADIOTHERAPY FOR CARCINOMA OF THE FLOOR OF MOUTH

TAKEHIRO INOUE, M.D.,* TOSHIHIKO INOUE, M.D.,* HIDEYA YAMAZAKI , M.D.,*MASAHIKO KOIZUMI, M.D.,* KAZUFUMI KAGAWA, M.D.,* KEN YOSHIDA, M.D.,*HIROYA SHIOMI, M.D.,* ATSUSHI IMAI , M.D.,* K IMISHIGE SHIMIZUTANI , D.D.S.,†

EICHII TANAKA , M.D.,† TAKAYUKI NOSE, M.D.,† TERUKI TESHIMA, M.D.,‡

SOUHEI FURUKAWA, D.D.S.,§ AND HAJIME FUCHIHATA, D.D.S.§

*Department of Radiation Oncology, Biomedical Research Center,†Department of Radiology,‡Department of Medical Engineering,Osaka University Medical School, and§Department of Oral and Maxillofacial Radiology,

Osaka University Dental School, Osaka, Japan

Purpose: Patients with cancer of the floor of mouth are treated with radiation because of functional and cosmeticreasons. We evaluate the treatment results of high dose rate (HDR) and low dose rate (LDR) interstitial radiationfor cancer of the floor of mouth.Methods and Materials: From January 1980 through March 1996, 41 patients with cancer of the floor of mouthwere treated with LDR interstitial radiation using 198Au grains, and from April 1992 through March 1996 16patients with HDR interstitial radiation. There were 26 T1 tumors, 30 T2 tumors, and 1 T3 tumor. For 21 patientstreated with interstitial radiation alone, a total radiation dose of interstitial therapy was 60 Gy/10 fractions/6–7days in HDR and 85 Gy within 1 week in LDR. For 36 patients treated with a combination therapy, a total doseof 30 to 40 Gy of external radiation and a total dose of 48 Gy/8 fractions/5–6 days in HDR or 65 Gy within 1 weekin LDR were delivered.Results: Two- and 5-year local control rates of patients treated with HDR interstitial radiation were 94% and94%, and those with LDR were 75% and 69%, respectively. Local control rate of patients treated with HDRbrachytherapy was slightly higher than that with 198Au grains (p 5 0.113). For late complication, bone exposureor ulcer occurred in 6 of 16 (38%) patients treated with HDR and 13 of 41 (32%) patients treated with LDR.Conclusion: HDR fractionated interstitial brachytherapy can be an alternative to LDR brachytherapy for cancerof the floor of mouth and eliminate radiation exposure for the medical staff. © 1998 Elsevier Science Inc.

Brachytherapy, Carcinoma of the floor of mouth, Dose rate.

INTRODUCTION

Some patients with cancer of the floor of mouth are treatedwith radiation because of functional and cosmetic reasons.At the Osaka University Hospital, more than half of patientswith cancer of the floor of mouth had been treated withinterstitial radiotherapy (1). Before 1980, patients weremainly treated with192Ir hair pins or Ra needles. For192Irhair pins or Ra needles, it is difficult to avoid the radiationfor normal tissue of the tongue because of the long activelength of these sources. From 1980,198Au grains wereavailable at our department. For198Au grains, it is easy toimplant under local anesthesia for patients with poor risk oraged patients (2).

From July 1991, we carried out high dose rate (HDR)interstitial radiation for head and neck carcinomas (3). HDR

brachytherapy can eliminate radiation exposure for the med-ical staff. According to the randomized trial of comparisonfor local control rates of tongue carcinoma patients treatedwith HDR and low dose rate (LDR) interstitial radiation,local control rate and late complication of the HDR groupwere the same as those of the LDR group (4). In this paper,we evaluate the treatment results of HDR and LDR inter-stitial radiation for carcinoma of the floor of mouth.

METHODS AND MATERIALS

From January 1980 through March 1996, 41 patients withcarcinoma of the floor of mouth were treated with LDRbrachytherapy using198Au grains. From 1992 to March1996, 16 patients were treated with HDR192Ir microsource.

Reprint requests to: Takehiro Inoue, M.D., Department of Ra-diation Oncology, Biomedical Research Center, Osaka UniversityMedical School, 2-2 Yamadaoka, Suita, Osaka 565, Japan.Acknowledgment—This study was supported in part by the Grant-in-Aid for Cancer Research from the Ministry of Health and

Welfare (7-33, 9-27) and by the Grant-in-Aid for Scientific Re-search (A) from the Ministry of Education, Science, Sports andCulture (06304032) of the Japanese Government.

Accepted for publication 24 November 1997.

Int. J. Radiation Oncology Biol. Phys., Vol. 41, No. 1, pp. 53–58, 1998Copyright © 1998 Elsevier Science Inc.Printed in the USA. All rights reserved

0360-3016/98 $19.001 .00

53

We compared the treatment results of patients treated withLDR brachytherapy using198Au grains and HDR brachy-therapy using192Ir microsource. Table 1 shows the age andsex distribution and TN classification of 41 patients treatedwith LDR brachytherapy and 16 patients with HDR brachy-therapy. The median ages of LDR and HDR groups were 62(range, 38–83) and 60 (range, 32–80) years old. There were36 male patients and 5 female patients in the LDR group; allpatients were male in the HDR group. There are no statis-tically significant differences in age or sex distributionsbetween the two groups. Of 41 patients of the LDR group,22 (54%) were classified as T1 and 19 (46%) as T2, and of16 patients of the HDR group, 4 (25%) were classified asT1, 11 (69%) as T2, and one (6%) as T3 (p 5 0.0569).Thirty-eight patients (93%) were classified as N0 and 3(7%) as N1 in the LDR group, and 14 (88%) were classifiedas N0, one (6%) each as N1 and N2 in the HDR group (notsignificant).

The radiation methods for individual patients were de-cided by more than two doctors in charge who had morethan 10 years’ experience of the brachytherapy. Table 2shows the tumor size and treatment method for carcinomaof the floor of mouth. For the LDR group, 15 patients weretreated with interstitial radiation alone, 11 patients with acombination of external radiation and interstitial radiation,and 15 patients with concurrent chemotherapy during ex-ternal radiation and interstitial radiation. We used 45 mg of

peplomycin (range, 25–60 mg) or 60 mg of bleomycin(range, 45–135) as the concurrent chemotherapy. The cor-responding figures for the HDR group were 6, 8, and 2patients for each treatment, respectively. Twenty-one of 43patients (49%) with a tumor of 30 mm or less were treatedwith interstitial radiation alone; all 14 patients with a tumormore than 30 mm or with nodal metastasis were treated witha combined treatment of external and interstitial radiationwith or without concurrent chemotherapy (p 5 0.003).

For the patients treated with interstitial radiation alone,the median dose of the LDR group was 85 Gy (range,58–132 Gy) within 1 week. In the HDR group, all patientswere treated with a total dose of 60 Gy/10 fractions/7 days(Table 3). For the patients treated with a combination of30–40 Gy of external radiation and interstitial radiation, themedian dose of LDR brachytherapy was 65 Gy (range,42–129 Gy) within 1 week. In the HDR group, all patientswere treated with 48 Gy/8 fractions/6 days after the externalradiation. The median interval between external radiationand LDR brachytherapy was 28 days (range, 5–76 days),and that for the HDR group was 20 days (1–29 days).

Method of HDR interstitial radiotherapy was reported inprevious papers (3, 4). Before the insertion of guide tubesfor carcinoma of the floor of mouth, an individualizedtemplate was made because of the parallel implantation andideal arrangement of applicators (Fig. 1). The pattern of theimplantation of applicators such as triangles or squares wasdecided according to the size and location of the tumors.Eight patients were treated with double-plane implantation,and eight patients with volume implant. Applicators for the192Ir microsource were inserted from the mouth floor to theskin surface of the submental or submandibular regionthrough the above mentioned template. In some patientswith tongue invasion, applicators were inserted from thedorsal surface of the tongue to the skin surface. Dwelllengths of the192Ir microsource were 10 to 35 mm with2.5-mm steps.

HDR interstitial radiation was carried out twice a dayusing microSelectron-HDR (Nucletron-Oldelft, The Neth-erlands). Time interval between two treatments was morethan 6 hours (3, 4). Reference point of dose calculation of

Table 1. Characteristics

LDR HDR

No. cases 41 16Age Range 38–83 32–80

Median 62 60Sex Male 36 88% 16 100%

Female 5 12% 0 0%T1 22 54% 4 25%T2 19 46% 11 69%T3 0 0% 1 6%N0 38 93% 14 88%N1 3 7% 1 6%N2 0 0% 1 6%

Table 2. Tumor size and treatment method

Tumorsize (mm)

LDR HDR

Int. alone Ext.1 Int. Ext. 1 Che.1 Int. Total Int. alone Ext.1 Int. Ext. 1 Che.1 Int. Total

N0¶ 20 11 8 3 22 2 0 1 3¶ 30 4 2 4 10 4 4 0 8¶ 40 0 1 5 6 0 1 1 2. 40 0 0 0 0 0 1 0 1

N1–20 0 3 3 0 2 0 2

Total 15 11 15 41 6 8 2 16

Int. 5 interstitial radiation; Ext.5 external radiation; Che.5 chemotherapy.

54 I. J. Radiation Oncology● Biology ● Physics Volume 41, Number 1, 1998

interstitial radiation was 5 mm from the source. Dose rate atthe reference point changed according to the activity of the192Ir microsource. Average dose rates (single fraction dose/overall treatment time in the first fraction) at the refer-ence point ranged from 0.79 to 6.86 Gy/min (median: 2.5Gy/min).

Forty-one patients of the LDR group were followed atleast 27 months or until their deaths. The maximum fol-low-up period was 136 months and the median was 65months in the LDR group. One patient in the HDR groupwas lost to follow-up at 11 months after treatment becauseof the Hanshin Earthquake in January 1995. Fifteen patientsof the HDR group were followed at least 19 months or untiltheir deaths. The maximum was 61 months and the medianfollow-up period was 30 months in the HDR group.

Student’s t test and chi-square test were used for thecomparison of patients’ characteristics. Local control ratesand cause-specific survival rates were calculated with theKaplan–Meier method. Statistical significance was tested bymeans of log rank test.

RESULTS

Of 16 patients treated with HDR interstitial radiation, onedied with nodal metastasis without local recurrence 7months after radiation (Table 4). Two patients died withsecondary esophageal or rectal cancer. Another patient died

with cerebral infarction 19 months after radiation. Onepatient was lost to follow-up at 11 months after treatment.The remaining nine are alive with no evidence of disease.Only one T3 tumor recurred 6 months after a combinationtherapy of 32 Gy of external radiation and 48 Gy/8 fractionsof HDR interstitial radiation. This patient was treated withsurgery for local recurrence, and died with rectal carcinoma20 months after radiation as mentioned above. All 13 pa-tients with T1 and T2 tumors were locally controlled.

Of 41 patients treated with LDR, 12 developed localrecurrence. Seven patients died with local tumor and twowith nodal metastases. Two patients died with lung cancer,one with esophageal cancer, and one with another oralcancer.

Two- and 5-year local control rates of patients treatedwith HDR interstitial radiation were 94% (Fig. 2). On theother hand, two- and 5-year local control rates of patientstreated with198Au grains were 75% (T1, 81%; T2, 67%)and 69% (T1, 76%; T2, 60%), respectively. This differencewas not statistically significant (p 5 0.1130).

Two- and 5-year cause-specific survival rates of patientstreated with HDR interstitial radiation were 94% (Fig. 3).On the other hand, 2- and 5-year local control rates ofpatients treated with198Au grains were 82% (T1, 91%; T2,72%) and 76% (T1, 85%; T2, 67%), respectively. Thisdifference was not statistically significant (p 5 0.2891).

Table 5 shows local control of carcinoma of the floor ofmouth according to the tumor size. For the LDR group, 17of 22 tumors (77%) not more than 20 mm and 10 of 12(83%) not more than 30 mm were locally controlled, andonly 2 of 7 tumors (29%) more than 30 mm were controlledwith radiation (p 5 0.0253). For the HDR group, all tumorsnot more than 40 mm were locally controlled.

Table 6 shows local control according to the treatment

Table 3. Radiation dose

External radiation

Interstitial radiation

LDR* HDR†

No. patients Median dose (Min.–Max.) No. patients Dose

0 15 85 Gy/7 d (58–132) 6 60 Gy/10 Fr/7 d30–40 Gy 26 65 Gy/7d (42–129) 10 48 Gy/8 Fr/6 d

* Dose delivered within 1 week. 0.833 permanent dose of198Au.† All patients were treated with 60 Gy or 48 Gy.

Table 4. Treatment results

LDR HDR

No. cases 41 16Local recurrence 12 29% 1 6%Death with primary tumor 7 17% 0 0%Death with regional node 2 5% 1 6%Death with double cancer 4 10% 2 13%Bone exposure and/or ulcer 13 32% 6 38%Fig. 1. Individualized template for carcinoma of the floor of mouth.

55HDR brachytherapy for mouth floor cancer● T. INOUE et al.

method and interval between external and interstitial radiation.For the LDR group, 13 of 15 tumors (87%) treated withinterstitial radiation alone were locally controlled, and 16 of 26tumors (62%) treated with a combination of external andinterstitial radiation with or without chemotherapy were alsolocally controlled. Twenty-six tumors treated with a combina-tion of external and interstitial radiation were divided into twogroups according to interval between external and interstitialradiation. Twelve of 15 tumors (80%) with interval not morethan 28 days were controlled, and only 4 of 11 tumors (36%)with interval more than 28 days were locally controlled (p 50.0641). For the HDR group, all 6 tumors treated with inter-stitial radiation alone were locally controlled, and 9 of 10patients treated with a combination of external and interstitialradiation were controlled.

As for grades 2 and 3 of objective score for late complica-tion according to Late Effects Normal Tissues (LENT) Sub-jective Objective Management and Analytic (SOMA) Tables(5), bone exposure or ulcer of the mucosa of the floor of mouthoccurred in 6 of 16 (38%) patients in the HDR group, and 13of 41 (32%) patients in the LDR group (Table 7). The median

delay between brachytherapy and complication in the HDRgroup was 9 months (range, 2–20 months), and that in the LDRgroup was 17 months (range, 5–38 months). One of 19 patientswith complication received surgery for bone exposure, andanother 18 patients were treated with non-narcotic medicationand/or antibiotics. In the HDR group, 3 of 5 patients withgingival invasion and 3 of 11 without gingival invasionshowed late complication. Among 11 patients without gingivalinvasion, 6 were treated with dwell length of 10 or 15 mm and5 patients with dwell length of 20 to 35 mm. No late compli-cation occurred in the patients treated with dwell length of 10or 15 mm. Three of the 5 patients treated with dwell length of20 to 35 mm developed late complication. In the LDR group,3 of 6 patients with gingival invasion showed late complica-tion, and 10 of 35 patients without gingival invasion showedlate complication.

For the LDR group, late complication occurred in 3 of 15patients (20%) treated with interstitial radiation alone, andin 10 of 26 patients (38%) with a combination therapy ofexternal and interstitial radiation (p 5 0.3815). For the HDRgroup, late complication occurred in 3 of 6 patients treatedwith interstitial radiation alone, and in 3 of 10 patientstreated with a combination therapy of external and intersti-tial radiation (p 5 0.7897).

Table 8 shows bone exposure and/or soft tissue ulcerby tumor size of patients without lower gum invasion.For the HDR group, bone exposure and/or soft tissueulcer occurred in 3 of 9 tumors not more than 30 mm.There were only 2 tumors more than 30 mm without guminvasion, and one of them had local recurrence and

Table 5. Local control by tumor size

Tumorsize (mm)

LDR HDR

No.patients

Localcontrol %

No.patients

Localcontrol %

¶ 20 22 17 77 4 4 100¶ 30 12 10 83 9 9 100¶ 40 7 2 29 2 2 100. 40 — — — 1 0 0Total 41 29 71 16 15 94

¶ 30 versus. 30 in LDR; p 5 0.0253.

Table 6. Local control by treatment method and intervalbetween external and interstitial radiation

LDR HDR

Interval (days)No.

patientsLocalcontrol %

No.patients

Localcontrol %

Interstitial alone 15 13 87 6 6 100Ext. 1 Int.

¶ 28 15 12 80 8 7 88Ä 29 11 4 36 2 2 100

Total 41 29 71 16 15 94

¶ 28 versusÄ 29 in LDR; p 5 0.0641.

Fig. 2. Local control rates of patients treated with HDR brachy-therapy using192Ir microsource and LDR brachytherapy using198Au grains.

Fig. 3. Cause-specific survival rates of patients treated with HDRbrachytherapy using192Ir microsource and LDR brachytherapyusing198Au grains.

56 I. J. Radiation Oncology● Biology ● Physics Volume 41, Number 1, 1998

another tumor was controlled without complication. Forthe LDR group, late complication occurred in 4 of 20tumors (20%) not more than 20 mm, and in 6 of 15tumors (40%) more than 20 mm (p 5 0.3586).

DISCUSSION

Local control rates of carcinoma of the floor of mouthtreated with radiation in Osaka University Hospital werereported by Ikedaet al in 1985 (1). In this paper, localcontrol rates of T1, T2, and T3 carcinoma of the mouthfloor were reported as 80%, 56%, and 17%, respectively.More than half of the patients were treated with intersti-tial radiation. Forty patients were treated with Ra needlesor 192Ir hair pins and six patients with Rn seeds and198Augrains. However, from January 1985 through April 1993,26 patients were treated with198Au grains and only threepatients with Ra needles,137Cs needles, and192Ir hairpins. Shimizutaniet al reported the results of 45 patientswith carcinoma of the oral cavity treated with198Augrains in Osaka University Hospital (2). Local controlrate for patients with T1 and T2 oral cancers includingthe carcinoma of the floor of mouth were 80% and 59%,respectively. Matsumotoet al (6) reported the local con-trol rates of T1, T2a (¶3 cm), and T2b (.3 cm) lesionstreated with198Au grains as 89%, 76%, and 56%. Ourresults also showed lower local control rates in tumorsmore than 3 cm. Shimizutaniet al reported a lower localcontrol rate in the patients with longer interval betweenexternal and interstitial radiation. For the patients withcarcinoma of the floor of mouth, a longer interval re-sulted in a lower control rate (p 5 0.0641).

In recent years, we mainly used198Au grains for LDRbrachytherapy because of small irradiated volume of normaltissue. According to the present study, local control rates ofHDR brachytherapy were better than those of198Au grains.Delclos et al stated that gold grain implants were moresubject to cold spots than radium or iridium implants (7). Itis difficult to get the homogeneous dose distribution with198Au grains because of the technical problem. Anotherdemerit of 198Au grain is the expulsion or dislocation ofgold grains during high activity of the radioisotope. Thisproblem also resulted in the dose inhomogeneity. For HDRbrachytherapy, it is easy to implant the applicators in idealarrangement using template. Even for the irregular arrange-

ments of applicators, we can get the homogeneous dosedistribution using dose optimization of a stepping sourcedosimetry system (4).

Mazeronet al (8) reported the local control rates of T1N0and T2N0 carcinoma of the floor of mouth treated with192Irhair pins as 93.5% and 74.5%, respectively. Grades 1, 2, and3 complications of soft tissue were also reported as 15 of 95(16%), 8 of 95 (8%), and 8 of 95 (8%), respectively. Ikedaet al (1) reported that 22 of 46 (48%) patients showed boneexposure, and these complications occurred in the patientstreated with Ra needles or192Ir hair pins. Shimizutaniet al(2) reported that bone exposure occurred in 8 of 26 (31%)patients treated with198Au grains. According to the presentdata, 6 of 16 (38%) patients in the HDR group and 13 of 41(32%) patients in the LDR group showed bone exposureand/or ulcer. There were no significant differences betweencomplication rates of the patients treated with HDR andLDR interstitial radiation.

For the HDR group, there were three patients with boneexposures of mandible in five with lower gum invasion, andfor the LDR group 3 of 6 patients with lower gum invasionshowed bone exposure. In these patients, lower gum wasincluded in the treatment volume. It was difficult to avoidbone exposure of the mandible for these patients with gin-gival invasion. For the patients without lower gum invasion,dwell length of the stepping source was the important factorfor bone exposure in the HDR interstitial radiation. Carefulfollow-up for bone exposure or ulcer in patients treated withdwell length of 20 mm or more is needed.

Matsumotoet al (6) reported lower local control rate(55%) in patients with lower gum invasion than that inpatients without gum invasion (82%). Our results suggesteda high complication rate in patients with lower gum inva-sion. Tumors with lower gum invasion treated with inter-stitial radiation have lower local control and higher compli-cation rates. Hickset al (9) reported high local control rate(87%) of surgical treatment, regardless of T stage, in pa-tients with negative margins. In our hospital, patients withlower gum invasion were mainly treated with surgery, andinterstitial radiation is applied for the patients who refusedsurgical treatment.

According to the phase III trial for early tongue carci-noma treated with HDR and LDR interstitial radiation,

Table 7. Bone exposure and/or ulcer by dwell length

SourceLower gum

invasionDwell

length (mm)No.

patientsBone exposure

and/or ulcer

HDR 1 5 32 ¶ 15 6 0

Ä 20 5 3Total 16 6 (38%)

LDR 1 6 32 35 10

Total 41 13 (32%)

Table 8. Bone exposure and/or ulcer by tumor size of patientswithout lower gum invasion

Tumorsize (mm)

No.patients

Bone exposureand/or ulcer

HDR ¶ 20 4 1¶ 30 5 2¶ 40 1 0. 40 1 0

11 3 (27%)LDR ¶ 20 20 4

¶ 30 10 4¶ 40 5 2

35 10 (29%)

57HDR brachytherapy for mouth floor cancer● T. INOUE et al.

HDR interstitial radiation had almost the same localcontrol rate and the same complication compared withLDR interstitial radiation (4). We found the same resultin a retrospective study for carcinoma of the floor ofmouth. HDR brachytherapy for carcinoma of the floor ofmouth had slightly higher local control (p 5 0.1130) andthe same complication compared with LDR brachythe-

rapy. HDR fractionated interstitial radiation may becomean alternative to LDR interstitial radiation for carcinomaof the floor of mouth. However, in this retrospectivestudy, follow-up periods in patients treated with HDRbrachytherapy were short compared with the LDR group,and longer follow-up is needed for local control andcomplication.

REFERENCES

1. Ikeda, H.; Nishiyama, K.; Masaki, N.; Shigematsu, Y.; Inoue,Ta.; Nakamura, M.; Kubo, K.; Fuchihata, H.; Shimizutani, K.;Kawasaki, Y.; Tanaka, Y. Squamous cell carcinoma of thefloor of the mouth treated by radiotherapy. Nippon ActaRadiol. 45:877–893; 1985 (in Japanese).

2. Shimizutani, K.; Koseki, Y.; Inoue, To.; Teshima, T.; Fu-rukawa, S.; Kubo, K.; Fuchihata, H.; Masaki, N.; Ikeda, H.;Tanaka, Y. Application of198Au grain for carcinoma of oralcavity. Strahlenther. Onkol. 171:29–34; 1995.

3. Teshima, T.; Inoue, T.; Ikeda, H.; Murayama, S.; Furukawa,S.; Shimizutani K. Phase I/II study of high dose rate interstitialradiotherapy for head and neck cancer. Strahlenther. Onkol.168:617–621; 1992.

4. Inoue, Ta.; Inoue, To.; Teshima, T.; Murayama, S.; Shimizutani,K.; Fuchihata, H.; Furukawa, S. Phase III clinical trial of high andlow dose rate interstitial radiotherapy for early tongue cancer. Int.J. Radiat. Oncol. Biol. Phys. 36:1201–1204; 1996.

5. LENT SOMA Tables. Radiother. Oncol. 35:17–60; 1995.6. Matsumoto, S.; Takeda, M.; Shibuya, H.; Suzuki, S. T1 and

T2 Squamous cell carcinoms of the floor of the mouth: Resultsof brachytherapy mainly using198Au grains. Int. J. Radiat.Oncol. Biol. Phys. 34:833–841; 1996.

7. Delclos, L.; Lindberg, R. D.; Fletcher, G. H. Squamous cellcarcinoma of the oral tongue and floor of mouth. Am. J.Roentgenol. 126:223–228; 1976.

8. Mazeron, J. J.; Grimard, L.; Raynal, M.; Haddad, E., Piedbois,P.; Martin, M.; Marinello, G.; Nair, R. C.; Le Bourgeois, J. P.;Pierquin, B. Iridium-192 curietherapy for T1 and T2 epider-moid carcinomas of the floor of the mouth. Int. J. Radiat.Oncol. Biol. Phys. 18:1299–1306; 1990.

9. Hicks, W. L.; Loree, T. R.; Garcia, R. I.; Maamoun, S.;Marshall, D.; Orner, J. B.; Bakamjian, V. Y.; Shedd, D. P.Squamous cell carcinoma of the floor of mouth: A 20-yearreview. Head Neck 19:400–405; 1997.

58 I. J. Radiation Oncology● Biology ● Physics Volume 41, Number 1, 1998