Proposal of a new grading system for evaluation of tongue ...

Radiotherapy and Oncology 61 (2001) 87–92 www.elsevier.com/locate/radonline

Proposal of a new grading system for evaluation of tongue hemiatrophy as a late effect of brachytherapy for oral tongue cancer Yasuo Yoshioka a,*, Ken Yoshida a, Kimishige Shimizutani b, Souhei Furukawa c, Takehiro Inoue a, Teruki Teshima d, Hideya Yamazaki a, Eiichi Tanaka b, Toshihiko Inoue a a

Department of Multidisciplinary Radiotherapy, Osaka University Graduate School of Medicine (D10), 2-2 Yamada-oka Suita, Osaka 565-0871, Japan b Department of Diagnostic Medicine, Osaka University Graduate School of Medicine, Osaka, Japan c Department of Oral and Maxillofacial Radiology, Osaka University Graduate School of Dentistry, Osaka, Japan d Department of Medical Engineering, Faculty of Medicine, Osaka University, Osaka, Japan Received 3 November 2000; received in revised form 19 April 2001; accepted 22 May 2001

Abstract Purpose: To evaluate tongue hemiatrophy as a late effect of brachytherapy, a new grading system was designed and applied to patients who had received low dose rate (LDR) or high dose rate (HDR) brachytherapy for early tongue cancer. Methods and materials: Between December 1998 and April 1999, 49 patients who had received brachytherapy for early tongue cancer (T1/ T2 ¼ 22:27) at Osaka University Hospital were investigated. All patients had undergone either LDR or HDR brachytherapy with Ir-192 (LDR/HDR ¼ 30:19) between 1980 and 1998. Atrophic changes in their tongue were classified into four categories (G0–G3): G3, not able to protrude the tongue beyond incisors; G2, hemiatrophy is seen on the irradiated side in the resting position of the tongue; G1, deviation of the tip of the tongue to the irradiated side is seen when protruded; and G0, none of these signs. The relationship between tongue hemiatrophy and tumor factors, treatment factors, and patients’ functional impairment was then investigated. The median time from treatment to assessment was 75 months (range 8–219 months). Volume index was defined as the number of needles that were implanted vertically into the tongue. Results: Fourteen patients were classified as G0, 29 as G1, five as G2, and one as G3. None of the G0 patients showed any speech or swallowing dysfunction, pain or contracted feeling, or general dissatisfaction with post-treatment tongue status. There was a tendency for such problems to increase with the tongue hemiatrophy grade. The frequency of T2 and non-superficial type tumors also tended to increase with the tongue hemiatrophy grade. The volume index of the G2–3 hemiatrophy group was significantly larger than that of the G0–1 group ðP ¼ 0:041Þ. Conclusion: This new grading system makes evaluation of atrophic changes in the tongue after brachytherapy easy and effective. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Tongue cancer; Hemiatrophy; Late effect; Brachytherapy; High dose rate; Grading system

1. Introduction Since the aim of brachytherapy is to reduce treatment related adverse effects (while maintaining or improving local control), it is important to evaluate any adverse effect as comprehensively as possible. Tongue hemiatrophy after brachytherapy is considered to be relatively common and an easily observable late adverse effect. However, the focus of examination has been on soft tissue necrosis in this adverse effect, while no systematic and objective scoring system for tongue hemiatrophy has been published so far. As a result, the degree of tongue

* Corresponding author.

hemiatrophy can usually not be adequately described, especially not for a careful and systematic study of the adverse effects of radiotherapy. In this paper a scoring system for tongue hemiatrophy after brachytherapy is introduced, and its correlation with dysfunction, patients’ subjective feelings, tumor factors, and treatment factors is investigated.

2. Materials and methods 2.1. Patient characteristics and selection Between December 1998 and April 1999, a total of 54 patients, who had received brachytherapy for early tongue cancer at Osaka University Hospital, came to the hospital to

0167-8140/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0167-814 0(01)00394-2

88

Y. Yoshioka et al. / Radiotherapy and Oncology 61 (2001) 87–92

see a doctor for follow-up. Forty-nine (32 males and 17 females) of them met our study entry criteria and are the subject of this report. Protocol eligibility was extended to (a) patients who had had clinical TNM stage T1–2N0M0 oral tongue cancer at the lateral border of the mobile tongue, (b) patients who had received low dose rate (LDR) or high dose rate (HDR) brachytherapy with Ir-192 for tongue cancer, and (c) patients whose informed consent had been obtained. Four patients were eliminated because the tumors were located on the ventral or dorsal subsite of the tongue, and one patient because she had been treated with a radium source. The median age during treatment was 53 years (range 24– 78 years). The UICC TNM classification of 1997 identified 22 patients as T1 and 27 as T2. Thirty patients had undergone LDR and 19 HDR brachytherapy between 1980 and 1998. The median time from treatment to assessment was 75 months (range 8–219 months), 100 months (range 32–219 months) for the LDR group, and 38 months (range 8–75 months) for the HDR group. Tumor types were divided into superficial ðn ¼ 22Þ, exophytic ðn ¼ 8Þ, indurative or nodular ðn ¼ 9Þ, and ulcerative type ðn ¼ 7Þ. The tumor type was unidentified in three cases.

2.2. Treatment The LDR implant was accomplished with the aid of the guide gutter technique using Ir-192 hairpins with or without an Ir-192 single pin. In the LDR group, 24 patients were treated with brachytherapy alone with a total dose of 65– 85 Gy (median 70 Gy), and six with 50–63.5 Gy of brachytherapy (median 59 Gy) combined with 18–40 Gy of external beam irradiation (EBI) (median 27.5 Gy). EBI was performed through opposing lateral portals or an ipsilateral portal with 2 Gy per daily fraction for five consecutive days per week. The HDR implant was achieved with flexible applicators followed by irradiation with a remote afterloading system and a dose optimization program. In the HDR group, 14 patients were treated with brachytherapy alone with a total dose of 54–60 Gy (median 60 Gy), and five received 42– 50 Gy of brachytherapy (median 48 Gy) combined with 30– 40 Gy of EBI (median 40 Gy). HDR brachytherapy irradiation was performed with a fraction size of 4–6 Gy twice daily with time intervals of more than 6 h on consecutive days. EBI was performed in the same way as for LDR cases. In both LDR and HDR brachytherapy, dose prescription was based on the Manchester method [12]. Dose prescription was at a 5-mm distance from the radioactive source plane, with 10 mm of constant intervals between the sources. Double-plane implant was adopted when the tumor thickness exceeded 10 mm. No tumor exceeded 20 mm thickness. Four patients were treated with doubleplane implants and all the others (45 patients) with singleplane implants. Dose rates at the dose prescribed points for

LDR and HDR were 0.33–0.98 (median 0.58) and 62.4– 167.7 Gy/h (median 107.3), respectively. 2.3. Grading The definitions of the grades of tongue hemiatrophy are shown in Table 1. The patients who were not able to protrude the tongue beyond their incisors were classified as G3. The other patients were classified as G0–2 when (a) deviation of the tip of the tongue to the irradiated side was seen when protruded or (b) tongue hemiatrophy on the irradiated side was seen even when the tongue was in the resting position. The patients with both (a) and (b) were classified as G2, those with (a) but without (b) as G1, and those without either (a) or (b) as G0. Representative samples of each grade are shown in Fig. 1. The results of the grading were analyzed in relation to the tumor factors, treatment factors, and the answers to questions listed below. 2.4. Questions to patients All the patients were asked if they had any kind of difficulty in speaking or swallowing, if they felt pain or a sense of contraction, and if they were satisfied with their posttreatment tongue status. Answers to the last question were not obtained from eight patients (two patients in the G0 group and six in G1). 2.5. Volume analysis To investigate the relationship between tongue hemiatrophy and target volume, an index was needed that could be used for the target volume for both LDR and HDR brachytherapy. We termed this index ‘volume index’ and defined it as representing the number of needles that were implanted vertically into the tongue, with all the needles spaced at identical 10-mm intervals for both LDR and HDR. Volume index ¼ ðnumber of single pinsÞ 1ðnumber of hairpinsÞ £ 2 ðLDRÞ ¼ ðnumber of applicatorsÞ ðHDRÞ 2.6. Statistical methods The chi-square test was used to compare subgroups. The Table 1 Definition of grades of tongue hemiatrophy Hemiatrophy on the Deviation of the tip to the irradiated side (tongue in irradiated side (tongue in the resting position) the protruded position) G0 G1 G2 G3

(2) (2) (2) (1) (1) (1) Not able to protrude the tongue beyond the incisors


89

Fig. 1. Examples of grades of tongue hemiatrophy: (a) G0, the tip does not deviate to the irradiated side during protrusion; (b) G1, the tip deviates to the irradiated side during protrusion; (c) G2, hemiatrophy of the irradiated side is observed in the resting position, but the patient is able to protrude the tongue beyond the incisors; (d) G3, the patient is not able to protrude the tongue beyond the incisors due to tongue hemiatrophy.

Mann–Whitney U-test was used as a non-parametric test for comparing volume index or the duration of follow-up period and G2–3 hemiatrophy rate. P-values of less than 0.05 were considered to be significant. 3. Results

3.2. Correspondence between hemiatrophy grade and patients’ problem (Table 2) None of the G0 patients showed any speech or swallowing dysfunction, pain or contracted feeling, or general dissatisfaction with their post-treatment tongue status. There was a tendency for such problems to increase with the hemiatrophy grade.

3.1. Results of grading As mentioned earlier, classification into G0–3 resulted in 14 patients being classified as G0, 29 as G1, five as G2, and one as G3.

3.3. Hemiatrophy grade and tumor factors (Table 3) There were 22 T1 and 27 T2 patients. One of the T1

Table 2 Hemiatrophy grade and patients’ problem

G0 ðn ¼ 14Þ G1 ðn ¼ 29Þ G2 ðn ¼ 5Þ G3 ðn ¼ 1Þ a b

Speech dysfunction

Swallowing dysfunction

Pain or contracted feeling

Dissatisfaction a

0 10 (34%) 3 (60%) 1 (100%)

0 0 0 1 (100%)

0 2 (7%) 1 (20%) 1 (100%)

0b 6 b (21%) 3 (60%) 1 (100%)

Patients’ general satisfaction (dissatisfaction) with their post-treatment tongue status. Eight (two G0 and six G1) of 49 patients did not answer the question.

90


Table 3 Hemiatrophy grade and tumor or treatment factor

G0 ðn ¼ 14Þ G1 ðn ¼ 29Þ G2 ðn ¼ 5Þ G3 ðn ¼ 1Þ a b c

T1/T2

Superficial/other type a

RND (1)/(2)

EBI b (1)/(2)

8/6 13/16 1/4 0/1

6/7 15/12 1/4 0/1

3/11 4 c/25 0/5 1 c/0

2/12 7/22 2/3 0/1

Includes exophytic, indurative or nodular, and ulcerative type. Tumor type was not classifiable in three cases. External beam irradiation to the tongue. Includes one case that received postoperative EBI to the neck lymph node area.

patients (5%) showed G2–3 hemiatrophy as did five T2 patients (19%) ðP ¼ 0:29Þ. The number of patients whose tumor types had been classified macroscopically as superficial was 22, while that of patients with other types of tumors was 24 (not classifiable in three cases). The ratio of G2–3 atrophy was 5% for superficial tumors and 20% for other types ðP ¼ 0:23Þ. 3.4. Hemiatrophy grade and treatment factors The volume index for the G2–3 group ranged from 4 to 6 with a median of 5 and a mean of 5.00 ðn ¼ 6Þ, that for the G0–1 group from 3 to 8 with a median of 4 and a mean of 4.23 ðn ¼ 43Þ, with that for the G2–3 group significantly larger than for the G0–1 group ðP ¼ 0:041Þ (Table 4). The double-plane implant group ðn ¼ 4Þ comprised one patient in each classification grade. Radical neck dissection (RND) was performed in eight cases. One case showed G1 hemiatrophy immediately after RND (2 months after brachytherapy), while the other seven showed atrophy, if any, at least 3 years after RND. Eleven cases received EBI to the tongue at a dose of not more than 40 Gy. Twelve cases received mild chemotherapy. There was no statistical correlation between RND, EBI, or chemotherapy and the hemiatrophy grades. 3.5. Hemiatrophy grade and duration of follow-up The mean and standard deviation of the duration of follow-up for the G0–1 group was 78.2 ^ 50.5 months ðn ¼ 43Þ, and that for the G2–3 group 78.5 ^ 50.3 months ðn ¼ 6Þ ðP ¼ 0:85Þ. Note that the duration of follow-up is Table 4 Hemiatrophy grade and volume index a

G0 ðn ¼ 14Þ G1 ðn ¼ 29Þ G2 ðn ¼ 5Þ G3 ðn ¼ 1Þ a

3

4

5

6

7

8

1 1 0 0

12 21 2 0

0 6 2 0

1 0 1 1

0 0 0 0

0 1 0 0

Volume index for the G2–3 group was significantly larger than for the G0–1 group ðP ¼ 0:0410Þ.

not determined by the date of occurrence of hemiatrophy but the date of last follow-up. 3.6. Analysis of G2–3 hemiatrophy cases Of the six G2–3 cases, five were T2 cases, and the remaining case had a T2-equivalent target volume because of the additional area of leukoplakia. Five of these cases had tumor types other than superficial. The G3 case did not receive EBI to the tongue but received 50 Gy of postoperative (post-RND) EBI to the neck lymph node area. Four of the G2–3 cases showed ulceration, including three with both ulceration and bone exposure. One G2 patient continued drinking heavily even after brachytherapy, and another G2 patient had received excisional biopsy before brachytherapy. 4. Discussion Oral tongue cancer can generally be cured with interstitial brachytherapy [6–9,15,16]. Clinicians often observe atrophic changes in the tongue of patients who have received radiotherapy for oral tongue cancer. It is obviously one of the late adverse effects of radiotherapy and related to dysfunction of the tongue because tongue atrophy causes difficulty in speaking and swallowing. Tongue hemiatrophy, which is rather more typical of interstitial brachytherapy than of external beam radiotherapy, is easier to identify than whole tongue atrophy because we can compare ipsilateral hemiatrophy with the contralateral side. However, no systematic and objective evaluation of tongue hemiatrophy has been published so far. Tongue hemiatrophy has often been associated with hypoglossal nerve damage after high dose EBI with or without neck dissection [1,2]. In our study, three cases appear to fit this description. One is a G1 case who showed tongue hemiatrophy immediately after RND (2 months after brachytherapy) and the other two are G1 and G3 cases that received both RND and 50 Gy of EBI to the neck postoperatively. No other patients received more than 40 Gy of EBI, which is not enough to cause cranial nerve atrophy as shown by previously published data (62.5–100 Gy, 6000– 8000 R) [1,2]. Seventy-three percent of our patients (36 out of 49) were treated with brachytherapy alone, so that their


tongue hemiatrophy was undoubtedly caused by brachytherapy. We attempted to objectify the degree of tongue hemiatrophy with definite criteria and to clarify the correspondence between tongue hemiatrophy and patients’ troubles. As shown in Table 1, hemiatrophy grades were determined by three simple criteria: (1) whether patients could protrude the tongue beyond the incisors, (2) whether hemiatrophy was observed with the tongue in the resting position, and (3) whether hemiatrophy was observed in the protruded position. Our finding that no G0 patient complained of functional or sensorial trouble suggests that G0 status corresponds to freedom from any adverse effect. The tendency for the higher hemiatrophy grades to correspond to more frequent patient complaints indicates that the grades reflect patients’ functional and subjective problems. G2–3 tongue hemiatrophy was more closely related to T2 than T1 tumors and other than superficial types, although statistical significance was not obtained. The target volume seemed to be significantly related to tongue hemiatrophy provided that our definition of ‘volume index’ is valid. No difference was seen between the HDR and LDR groups in terms of frequency of tongue hemiatrophy. However, the median time from treatment to assessment for the HDR group (38 months) was shorter than for the LDR group (100 months), so the possibility remains that some of the patients of the HDR group may show tongue hemiatrophy during a longer follow-up, although there seemed to be no significant relationship between the duration of follow-up and the frequency of hemiatrophy. The analysis of G2–3 hemiatrophy cases suggested some risk factors such as ulcer formation, bone exposure, continuation of heavy drinking after treatment, and excisional biopsy. The three cases that displayed both ulcer formation and bone exposure may have been treated with overdose irradiation. EORTC/RTOG has established standard criteria to describe the late effects of radiotherapy in the form of the LENT/SOMA scale [3,10,13,14]. This scale is widely used and reliable, but it is mainly applicable to external beam radiotherapy rather than brachytherapy, especially in the case of the tongue. The LENT/SOMA scale refers to mucosal reactions including ‘patchy’ or ‘diffuse’ atrophy of the tongue, but pays no special consideration to ‘hemi-’ atrophy or speech dysfunction (although swallowing dysfunction is mentioned). The aim of our grading system is to classify the degree of hemiatrophy of the tongue and to describe, in detail, one of the late effects definitely related to the field of brachytherapy. Evaluation of the late effects of HDR brachytherapy for tongue cancer is of special interest, because it has been reported that HDR brachytherapy used as monotherapy results in local control and acute reaction equivalent to that obtained with LDR, while no studies have as yet been published about the late effects of HDR brachytherapy [8]. It has been suggested that HDR brachytherapy has a higher rate of late adverse effects than LDR because of the ‘dose

91

rate effect’. However, this effect has been discussed only in relation to LDR continuous brachytherapy [4,5,11]. An objective and rational method is therefore needed to describe and compare late adverse effects of LDR and HDR brachytherapy, and our grading system is thought to contribute to such a method. Our findings also add to the little information now available regarding the frequency of tongue hemiatrophy following brachytherapy for the border of the tongue.

Acknowledgements This work was supported in part by the Grant-in-Aid for Cancer Research from the Ministry of Health and Welfare of the Japanese Government (10-17).

References [1] Berger PS, Bataini JP. Radiation-induced cranial nerve palsy. Cancer 1977;40:152–155. [2] Cheng VST, Schulz MD. Unilateral hypoglossal nerve atrophy as a late complication of radiation therapy of head and neck carcinoma: a report of four cases and a review of the literature on peripheral and cranial nerve damages after radiation therapy. Cancer 1975;35:1537– 1544. [3] Cox JD, Stetz J, Pajak TF. Toxicity criteria of the radiation therapy oncology group (RTOG) and the European organization for research and treatment of cancer (EORTC). Int J Radiat Oncol Biol Phys 1995;31:1341–1346. [4] Dutreix J. Expression of the dose rate effect in clinical curietherapy. Radiother Oncol 1989;15:25–37. [5] Fontanesi J, Hetzler D, Ross J. Effect of dose rate on local control and complications in the reirradiation of head and neck tumors with interstitial iridium-192. Int J Radiat Oncol Biol Phys 1989;17:365–369. [6] Hareyama M, Nishio M, Saito A, et al. Results of cesium needle interstitial implantation for carcinoma of the oral tongue. Int J Radiat Oncol Biol Phys 1992;25:29–34. [7] Horiuchi J, Okuyama T, Shibuya H, Takeda M. Results of brachytherapy for cancer of the tongue with special emphasis on local prognosis. Int J Radiat Oncol Biol Phys 1982;8:829–835. [8] Inoue Ta, Inoue To, Teshima T, et al. Phase III trial of high and low dose rate interstitial radiotherapy for early oral tongue cancer. Int J Radiat Oncol Biol Phys 1996;36:1201–1204. [9] Kondo M, Hashimoto S, Dokiya T, et al. Local control of squamous cell carcinoma of the mobile tongue: an experience of different modalities. Int J Radiat Oncol Biol Phys 1986;12:755–760. [10] LENT SOMA tables: table of contents 1995. Radiother Oncol 1995;35:17–60. [11] Mazeron JJ, Simon JM, Le Pechoux C, et al. Effect of dose rate on local control and complications in definitive irradiation of T1–2 squamous cell carcinomas of mobile tongue and floor of mouth with interstitial iridium-192. Radiother Oncol 1991;21:39–47. [12] Paterson R. The radium dosage system. In: Paterson R, editor. The treatment of malignant disease by radiotherapy, 2nd ed. London: Edward Arnold (Publishers) Ltd., 1963. pp. 121–138. [13] Pavy JJ, Denekamp J, Letscher J, et al. EORTC late effects working group. Late effects toxicity scoring: the SOMA scale. Radiother Oncol 1995;35:11–15. [14] Rubin P, Constine LS, Fajardo LF, Phillips TL, Wasserman TH. RTOG late effects working group. Overview of late effects normal tissues (LENT) scoring system. Radiother Oncol 1995;35:9–10.

92


[15] Shibuya H, Hoshina M, Takeda M, Matsumoto S, Suzuki S, Okada N. Brachytherapy for stage I and II oral tongue cancer: an analysis of past cases focusing on control and complications. Int J Radiat Oncol Biol Phys 1993;26:51–58.

[16] Yamazaki H, Inoue T, Koizumi M, et al. Comparison of the long term results of brachytherapy for T1–2N0 oral tongue cancer treated with Ir-192 and Ra-226. Anticancer Res 1997;17:2819–2822.