Long-term follow-up of radiotherapy for pituitary adenoma: The absence of late recurrence after ≥ 4500 cGy

Abstract

Recent literature has suggested that late recurrence of pituitary adenoma after radiotherapy is common. We hypothesized that late failures might be a result of inadequate dose (< 4500 cGy). To investigate, we analyzed 105 patients treated at our institution between 1965 and 1986 (analysis, 2/89). The minimum observation time was ≥ 5 years in 58% and ≥ 10 years in 30% of the patients. All patients received megavoltage radiotherapy (range, 4200–5500 cGy; mean, 4821 cGy) at a mean dose per fraction of 172 cGy; 100 patients received ≥ 4500 cGy tumor dose. Twenty-nine patients received radiotherapy alone, and 76 had postoperative radiotherapy after frontal craniotomy (20 patients) or transsphenoidal hypophysectomy (56 patients). At presentation, 71 % of patients had extrasellar disease, 57% had visual field deficits, and 50% had endocrinopathy. Of patients treated postoperatively, 74% had gross residual disease. Four local failures occurred at 13, 16, 57, and 64 months after postoperative radiotherapy, all within the irradiated volume (tumor doses of 4700, 4715, 5000, and 5100 cGy). All four patients had presented with moderate to extensive extrasellar disease with visual field defects. Two of the four remain free of second recurrence at 7 and 13 years after salvage transsphenoidal hypophysectomy. The local control rate with radiotherapy (product-limit method) at 10 years was 100% in the radiotherapy-alone group and 92% in the postoperative radiotherapy group (95% for all patients). To prevent bias, seven patients who received bromocriptine, none of whom demonstrated a recurrence, were censored from the local control analysis at the initiation of the drug. No patient in this study suffered recurrence > 64 months after radiotherapy, with 31 patients (none with bromocriptine) observed 10 to 21 years. We conclude that treatment of pituitary adenoma with ≥ 4500 cGy in 25 fractions can result in a high (≥ 90%) probability of stable long-term control.