Japanese Journal of Clinical Oncology

DISCUSSION

The pathogenesis of GO is believed to involve lymphocytic infiltration of the extraocular muscle along with fibroblast activation. Recent studies indicate that activated retrobulbar T-lymphocytes secrete various cytokines, which in a paracrine manner are thought to induce glycosaminoglycan (GAG) production by fibroblasts (6). Excessive secretion of the very hydrophic GAG molecules results in tissue edema, which, in addition to lymphocytic infiltration, causes marked swelling of the extraocular muscle in GO. How radiotherapy works is unknown. The leading theory is that X-rays decrease the function of, or kill, lymphocytes in the orbit (7). Therefore, radiation is thought to be effective in the acute, active phase of the disease (1).

The grade of GO is established mainly on the basis of the no sign, only sign, soft tissue involvement, extraocular muscle involvement, corneal involvement and sight loss (NOSPECS) classification system (5). This system was also used for evaluation of treatment results; however, it was difficult to use for quantitative analysis and especially for the eye movement. Therefore, we analyzed the effect of radiotherapy on GO by using newly developed quantitative indexes for the eye movements. Most patients have experienced an improvement of the eye-muscle motility by using the newly developed quantitative indexes in our series. However, Petersen et al. (8) noted only small improvements in extraocular muscle function. One reason for this is that the index which was used to evaluate extraocular muscle function was different. They used only three grades for abnormalities of extraocular muscle function.

We were able to show changes in impairment of motility of individual extraocular muscles by using the newly developed quantitative indexes. The response to irradiation was not different among extraocular muscles and the motility of all muscles improved within 2-3 weeks and stabilized. However, we found that extraocular muscles for elevation were impaired more severely and their functional impairment tended to remain after irradiation. There have been no similar reports as far as we are aware. In the NOSPECS system commonly used, there are only three grades for abnormalities of extraocular muscle involvement: (a) limitation of motion at extremes of gaze, (b) evident restriction of gaze and (c) fixation of a globe or globes. Such a rough classification cannot detect delicate changes of movement of extraocular muscles.

We used a radiation dose of 24-28 Gy in 2 Gy fractions. The results of low dosages of the order of 10 Gy are conflicting (9). Grauthoff et al. (10) reported fair results with 10 Gy retrobulbar irradiation without glucocorticoid therapy. Ravin et al. (11), however, reported poor results with only 10 Gy retrobulbar irradiation in the treatment of GO. Nakahara et al. also reported that response in the 24 Gy group was better than that in the 10 Gy group, in which nine out of 15 patients experienced persistent or recurrent symptoms of GO (12). A higher cumulative dose of 30 Gy did not provide further benefit. In a series of over 300 patients, Petersen et al. (8) showed that 2000 cGy in 10 fractions gave results identical with those for 3000 cGy in 15 fractions and 76% of treated patients responded positively to treatment. Judging from these reports, the appropriate total dose seems to be 20-30 Gy. Regarding fraction size, 1 Gy per fraction is not a sufficiently high dose (12), but a fraction larger than 2 Gy per fraction should not be used because of the danger of retinopathy.

Retrobulbar irradiation was tolerated very well and adverse reactions were minimal in our series. Prummel et al. (13) reported transient hair loss at the temple and a transient small increase in periorbital swelling, chemosis and redness of the eye were observed in 14% of patients, which is consistent with the adverse reactions in our results.

There may be a risk of radiation-induced retinopathy, cataracts and sarcoma. We have experienced no late complications in this follow-up period. Parsons et al. (14) reported that no radiation-induced cataracts or second malignancies were seen among their 311 patients. Because the lens is not included within the radiation field, the risk of cataracts is very low. Sandler et al. (15) reported that three patients had cataracts after radiotherapy. However, upon review, these were felt to be senile cataracts in all cases.

The risk of radiation-induced retinopathy would be very small with fractionated doses of 20-30 Gy (14). The few described cases of radiation-induced retinopathy appear to be caused largely by dosimetric errors resulting in fractionated doses greater than 2 Gy (16) or by the coexistence of diabetes mellitus (17). Because diabetes and chemotherapy are likely to potentiate the onset of radiation-induced retinopathy, these conditions may be contraindications for retrobulbar irradiation. Although there are reports of second malignancies arising after irradiation for benign disease (18,19), the risk in this setting is felt to be small. No therapy-related tumors were observed up to 21 years after irradiation in the Stanford series (8).

In our patients who were treated with steroids before radiotherapy, one patient suffered secondary diabetes mellitus. We have experienced no late complications of radiotherapy so far. When compared with steroid medication, radiotherapy seemed to have better tolerability. Prummel et al. performed a double-blind randomized trial, in which patients with GO were treated with predonisone or irradiation. Radiotherapy and oral predonisone appeared to be equally effective as initial treatment of GO. They concluded that radiotherapy should be considered the treatment of first choice in view of its better tolerability.

Spontaneous remission with improvement has been observed during the natural history of GO (20). Some authors advocated including a control group that receives neither steroids nor irradiation (21). However, it is difficult in practice to do a controlled study of GO treatment. Because some patients with severe symptoms suffer immotility of the extraocular muscle in an inactive end stage of inflammation, we should accomplish early arrest of active inflammation in the orbit of every patient. Wiersinga et al.'s (20) and Prummel et al.'s (13) summary of the role of radiotherapy is that patients in the early stage with active but still mild eye disease might have the greatest benefit from radiotherapy by preventing the development of severe eye disease, thus precluding the need for any surgery. For patients with active and severe eye disease, radiation-induced inactivation of the eye disease will permit eye surgery at an earlier point in time (20).

In summary, radiotherapy was effective in preventing exacerbations of active inflammatory ophthalmopathy in patients with Graves' disease with minimal morbidity and thus eliminated the adverse reactions associated with protracted corticosteroid use. However, proptosis did not improve or improved only slightly. We were able to show changes in impairment of motility of individual extraocular muscles by using the newly developed quantitative indexes. Extraocular muscles that work for elevation were impaired more severely than the other muscles and the impairment of motility tended to remain after irradiation.