The objective of this retrospective cohort study was to define the efficacy and safety of fractionated radiotherapy (FRT) and stereotactic radiosurgery (SRS) for the treatment of patients with pituitary adenoma.
The objective of this retrospective cohort study was to define the efficacy and safety of fractionated radiotherapy (FRT) and stereotactic radiosurgery (SRS) for the treatment of patients with pituitary adenoma.
Between January 1995 and April 2006, 125 consecutive patients with pituitary adenomas (54 hormone-secreting adenomas and 71 nonsecretory adenomas) received FRT or underwent SRS. Sixty-four patients received FRT, for which the mean total dose was 50.4 grays (Gy) (range, 48–54 Gy), and 61 patients underwent gamma-knife SRS with mean marginal dose of 25.1 Gy (range, 9–30 Gy).
After mean follow up of 36.7 months, the tumor volume was increased in only 4 patients (3.2%). The overall actuarial progression-free survival rate was 99% at 2 years and 97% at 4 years. No difference was observed between the FRT group and the SRS group in the control of tumor growth. Based on the endocrinologic results in the patients who had secretory adenomas, the overall hormone complete remission rate was 26.2% at 2 years and 76.3% at 4 years. The median time to complete remission was 26 months in the SRS group and 63 months in the FRT group (P = .0068). Hypopituitarism developed as a delayed complication in 11.5% of patients at a median of 84 months.
Both FRT and SRS were efficient treatment modalities for the control of tumor growth in patients with pituitary adenomas. The current results indicated that single-dose radiosurgery more promptly produces an effect on the hypersecretion of pituitary hormones and may be recommended over FRT for suitable patients. Cancer 2007. © 2007 American Cancer Society.
Pituitary adenomas are relatively common tumors that account for 10% to 20% of all primary central nervous system tumors.1 Control of tumor growth and preservation of vision are the major objectives of treatment for nonsecretory adenomas, whereas the correction of endocrinopathies is a more important target in hormone-secretory adenomas.2–7 Therefore, surgical removal is the treatment of choice for most patients who have nonsecretory adenomas with mass effect. However, the usual treatment for patients who have secretory adenomas is surgery and/or medical treatment, depending on their hormone status and response to treatment. Radiotherapy is recommended for patients who have tumors that are not controlled successfully with surgical or medical treatment.8–10 Traditionally, radiotherapy (RT) has been given in a fractionated manner at standard doses from 1.8 grays (Gy) to 2 Gy per daily fraction. In contrast, stereotactic radiosurgery (SRS) is a new method that delivers a single dose of irradiation with high conformity and selectivity. There are well-established guidelines limiting the total dose of fractionated RT (FRT) or the maximal dose to the optic apparatus in SRS to prevent radiation injury to the optic nerve.9, 11, 12 However, to date, no accepted method has been demonstrated that prevents the deterioration of normal endocrine function. In adequately selected patients, highly conformal and selective irradiation by SRS enables the delivery of a high dose of irradiation to the tumor and simultaneously reduces irradiation of the adjacent, radiosensitive structures to a level that is enough to guarantee safety. This property of SRS may provide a theoretical advantage not only for the preservation of vision but also for the prevention of hypopituitarism by reducing irradiation to the hypothalamus and pituitary stalk to a much lower level compared with the biologic equivalent dose delivered by FRT.13 However, it remains to be clarified in a randomized, controlled study whether SRS is superior to FRT.
For this report, we studied a retrospective cohort of 125 patients with pituitary adenomas who underwent FRT or SRS. To evaluate the efficacy and safety of each modality of irradiation, we investigated the clinical data in terms of tumor growth control and normalization of excessive hormone secretion. With regard to the safety issues of radiation treatment, we focused on the problems of visual function and endocrine outcomes.
Between January 1995 and April 2006, 125 consecutive patients with pituitary adenomas were treated with either FRT (64 patients) or SRS (61 patients) in a single institution. Among them, 117 patients received irradiation after surgical removal, which comprised 13.3% of all 881 patients who underwent transsphenoidal tumor removal during the same period. The mean age of patients was 41.3 years (range, 14–73 years). 65 patients were women (52%), and 60 patients were men (48%). With regard to endocrine status, 71 patients had nonfunctioning adenomas, and 54 patients had hormone-secreting tumors (8 corticotrophin [ACTH]-secreting tumors, 30 growth hormone [GH]-secreting tumors, and 16 prolactin-secreting tumors). With regard to size, 6 tumors that measured ≥3 cm in greatest dimension were included in among nonsecretory adenomas. Preirradiation hormone assays demonstrated that 20 patients already had complete or partial hypopituitarism that required hormone-replacement therapy, mostly as a result of surgical complications.
RT (either FRT or SRS) was given to patients for their pituitary adenoma if surgical and medical treatment had failed to remove the tumor completely or normalize hypersecretion of hormone. Eight patients in whom surgery was not feasible for medical reasons also were included. From January 1995 to December 2001 all patients received FRT, because SRS was not available until 2002, when the gamma knife was installed at our institute. Since the introduction of the gamma knife, patients underwent SRS if they met the following inclusion criteria: 1) a maximum tumor dimension <30 mm, and 2) a distance ≥2 mm between the tumor and the optic apparatus. Conversely, patients received FRT for tumors of any size that were located in close proximity to the optic chiasm. The dose of irradiation was variable according to each treatment modality and the characteristics of the tumor. FRT was delivered with 3-dimensional, conformal dose plan. The total dose delivered by linear accelerator (Varian Medial Systems, Palo Alto, Calif) for FRT was 50.4 Gy (range, 48–54 Gy) with daily dose of 2 Gy. Almost equal doses were applied regardless of the functional type of adenoma. In contrast, the gamma-knife radiosurgical dose varied from 9 Gy to 30 Gy at 50% isodose (median dose, 25.1 Gy), which usually was determined by a few important parameters, such as tumor volume, dose to the optic apparatus, and endocrine status (Table 1). Source blocking was applied frequently to obtain a sharper dose fall-off toward the optic nerves and chiasm. The maximum dose applied to the optic nerves and chiasm was ≤8 Gy, and the lens doses were ≤0.6 Gy. A lower radiation dose (16.3 Gy) was applied in nonsecretory adenomas, because they tended to be larger compared with secretory adenomas, and tumor growth could be arrested by a low radiation dose.2, 13–16 In secretory adenomas, a higher radiation dose (25.2 Gy) was administered for a relatively smaller volume with the objective of promoting early and complete control of hormone hypersecretion.
|Characteristic||No. of patients|
|Tumor volume, mm3|
|Tumor size, cm|
|Method||Linear accelerator||Gamma knife|
Data for the current analysis were taken from medical records and magnetic resonance images (MRIs) from our archives. MRI scans and pituitary hormone assessments with dynamic tests were performed annually at regular intervals. To evaluate the efficacy of each mode of RT, tumor growth was investigated in nonfunctioning adenomas. In addition to control of tumor growth, endocrine normalization was assessed in the secretory adenomas. The development of hypopituitarism was used as a parameter of long-term safety, and the onset of clinically significant hypopituitarism was determined according to when hormone replacement became mandatory.
Disease progression was assessed by quantitative analysis of the extent of tumor on T1-weighted, enhanced, coronal MRIs. The time to tumor progression was calculated from the first day of FRT or from the day of SRS. A complete response was defined as a reduction in tumor size >25% compared with baseline measurements. A partial response was defined as a reduction in tumor size <25%. Tumors were considered stable if any measured change in size was <10%. Tumor control was defined as the absence of radiologic tumor progression at the follow-up evaluation.
Criteria for remission of acromegaly were defined by basal GH levels <2.5 ng/mL or glucose-suppressed GH levels <1 ng/mL and normal insulin-like growth factor values. Prolactinoma was defined as prolactin basal levels <20 ng/mL. In Cushing diseases, remission was defined as normal cortisol levels, urinary free cortisol levels in the normal range, and resolution of clinical stigmata. Complete remission was defined as a remission state that fulfilled the criteria described above without requiring medications to suppress hormone secretion from the tumor.
Toxicities that originated from radiation injury were evaluated from the viewpoints of visual function and the development of hypopituitarism. Data on the subjective assessment of vision were retrieved from outpatient medical records. However, objective assessments of visual field and acuity were performed only for patients who had a complaint of visual deterioration. The occurrence of hypopituitarism was assessed regularly in clinical examinations and hormone assays supplemented with combined pituitary function tests. The development of hypopituitarism that required hormone-replacement therapy >3 months after irradiation was regarded as postirradiation hypopituitarism.
The time to tumor progression and the hormone remission rates were calculated using the Kaplan-Meier method and were analyzed using the log-rank test comparing the FRT group and the SRS groups. The independent variables that affected hormone outcomes were determined by logistic regression analysis. Cox proportional-hazard regression analysis was used to define better the potential interrelations among the related factors. Statistical software (SPSS version 10.0; SPSS Inc., Chicago, Ill) was used for the analyses, and P values <.05 were considered significant for all tests.
Only 4 of 125 patients had tumor progression at a mean follow-up of 36.8 months (range, 2–140 months). One patient with acromegaly died 15 months after SRS from cardiovascular complications, and 1 patient with prolactinoma died 21 months after FRT from lung cancer. The actuarial progression-free survival rate was 99% at 2 years and 97% at 4 years. In terms of tumor growth, the objective response rate (complete and partial tumor shrinkage) was 39.5% at 2 years and 81.8% at 4 years. Volumetric analysis did not reveal a significant difference in the pattern of posttreatment volume change between the FRT group and the SRS group (Fig. 1), although the mean follow-up in the FRT group was longer than that in the SRS group (46.4 months vs 25.4 months).
Based on the endocrinologic results in the patients with secretory adenomas, the overall complete remission rate was 26.2% at 2 years and 76.3% at 4 years (Fig. 2). In an attempt to evaluate the prognostic factors that affected hormone outcomes, multiple factors were analyzed using log-rank tests. All secretory tumors that received radiation treatment measured <3 cm in greatest dimension.
The univariate analysis revealed that the treatment modality and the type of secretory adenoma were correlated with better outcomes. Age, sex, and tumor size were not associated with endocrinologic outcomes. Complete remission was achieved by 14 patients (43.8%) in the SRS group and by 8 patients (36.4%) in the FRT group. The median time to complete remission was 26 months in the SRS group and 63 months in the FRT group (P = .0068) (Fig. 3). ACTH-producing adenomas were the most sensitive to both types of radiation treatment, resulting in an overall complete remission rate of 87.5% at 2 years and 100% at 4 years. In contrast, the complete remission rate was only 4.4% at 2 years and 46.9% at 4 years in 30 patients with GH-producing adenomas. In 16 patients with prolactin-producing adenomas, the complete remission rate was 19.2% at 2 years and 51.6% at 4 years (Fig. 4). In all patients with secretory adenomas, 18 of 30 patients with acromegaly and out of 16 patients with prolactinoma had received hormone-suppressing agents (bromocriptine or octreotide) during the period of RT (SRS or FRT). However, there was no difference in the remission rate between the patients who did and did not receive medication during the treatment period, a finding that was not consistent with previously reported results.17, 18 In the multivariate analysis, the treatment modality (SRS vs FRT; P = .023) and the type of secretory adenoma (ACTH-secreting vs GH- or prolactin-secreting; P = .005) were statistically significant factors related to outcome (Table 2).
|Univariate analysis||Multivariate analysis|
|Type of tumor (ACTH, GH, prolactin)||<.05||.005|
|Treatment modality (SRS, FRT)||<.05||.023|
No patients had complained of progressive visual deterioration at the time of radiation in either treatment group. After irradiation, 5 patients with Cushing disease or acromegaly complained of visual dimness. However, they were diagnosed with cataracts or glaucoma as a result of hypertension or diabetes, not necessarily indicating complications directly related to radiation. No other patients complained of visual deterioration after irradiation, although objective testing was not done in the majority of patients. Hypopituitarism that required hormone-replacement therapy was reported in 11 of 95 patients (11.6%) as a delayed complication, excluding 20 patients who had pituitary hypofunction prior to RT (Fig. 5). The median time to the development of hypopituitarism was 84 months (range, 14–84 months), and the cumulative risk for developing hypopituitarism was 5.7% at 3 years and 27.3% at 5 years. Only 1 of 11 patients who developed hypopituitarism belonged to the SRS group (median, 14 months). Secretory or nonsecretory adenoma, treatment modality, patient age, and sex were factors that did not affect the development of hypopituitarism, but it was noted that patients who had complete remission at the last follow-up had a longer median time to the development of hypopituitarism (53 months vs 84 months; P = .018).
All patients who had secretory adenomas achieved suboptimal control of hormone secretion by medical treatment only before irradiation. Among them, 32 patients did not reach complete remission after FRT or SRS at the last follow-up; therefore, medical treatment had been continued in 28 patients. At the last follow-up, 12 additional patients (42.9%) demonstrated target hormone levels in normal ranges, although they still needed medication, such as bromocriptine or octreotide. Radiation treatment provided additional remission with the assistance of medication in 33.5% of patients at 2 years and 38.3% of patients at 4 years (median, 64 months). Taken together, hormone hypersecretion was controlled well in 34 patients (63.0%) with or without supplementary medication after radiation treatment. Considering both complete remission and remission with medication for GH and prolactin, the median time to endocrine normalization was significantly shorter in the SRS group compared with that in the FRT group for GH (23 months vs 63 months) and prolactin (15 months vs 26 months; P = .0088).
In patients with pituitary adenomas, long-term tumor control rates for microsurgery alone vary from 50% to 90%.10, 19, 20 Despite advances in microsurgical techniques, patients may develop postoperative recurrences because of tumor invasion into surrounding structures, such as the cavernous sinuses. Recurrence rates after surgery reportedly have been 20% at 5 years postsurgery and 40% at 10 years postsurgery.14 Subsequently, either conventional FRT or SRS needs to be used as an adjunct to surgery to inhibit tumor growth or to treat symptomatic residual tumors.2 FRT has been accepted widely as second-line therapy after microsurgery. FRT may reduce recurrence rates to 6% after 10 years and 12% after 20 years.15, 21–23 Despite much improved control of tumor growth, there are concerns about potential late toxicity of RT and the delay in achieving hormone control in secreting adenomas.24–27 Selectively high radiation doses to tumor and low doses to a minimal volume of normal tissues are expected to reduce complications and enhance efficacy. In theory, SRS has potential advantages for those purposes in properly selected patients; however, sufficient evidence from a randomized trial has been absent.
The current study had major limitations in its ability to reach a decisive conclusion about a comparison of the 2 treatment modalities. First, this study was not a randomized, controlled trial, and a direct comparison of the results from each modality was impossible. Between both groups, there were significant differences in tumor size and length of follow-up. Furthermore, the results from this study cannot be generalized to newer radiation methods, such as intensity-modulated RT (IMRT). However, the results from this study still provide some information that enables a comparison of the 2 modalities, because the choice of treatment modality depends largely on the chronology of treatment. All patients received FRT without exception before the introduction of the gamma knife at our institute. Conversely, since the introduction of the gamma knife, the majority of patients, except those who had very large tumors in close contact with the optic apparatus, underwent SRS. If a small proportion of patients with very large, nonsecretory tumors (that is, >3 cm) were excluded, then there would not be significant selection bias of patients with small tumors who were eligible for SRS and FRT. Therefore, the current study suggests several significant findings despite some drawbacks. First, the control of tumor growth was excellent with both treatment modalities, which is in agreement with other published data.1 Although the length of follow-up for the SRS group was too short to draw final a conclusion, the trend toward volume change after SRS over a shorter follow-up was comparable to that after FRT over a longer follow-up, suggesting that SRS using an adequate marginal dose is not worse than FRT for the control of tumor growth. Second, in view of the length of time to endocrine normalization in patients with secretory adenomas, our results correspond to those reported by other authors, who suggested that SRS offers faster correction of hormone hypersecretion than FRT.28–32 In our series, the median time to remission was significantly longer in the FRT group than in the SRS group (46 months vs 23 months). Even if final remission rates are in the same range between both groups (36.4–43.8%), there is a possibility that the final outcome will be better in the SRS group, because follow-up for the SRS group was much shorter than that for the FRT group. It seems reasonable to choose SRS with priority for qualified patients, although this will require clarification in long-term follow-up. Taken together, a more prompt hormone effect can be expected with SRS, at least in patients who have tumors within the current eligibility criteria for SRS. Regarding radiation-induced complications, the current techniques of FRT or SRS within the generally recommended dose-range guidelines for the optic apparatus reportedly have a reliably low probability of visual complications. Therefore, the risk of optic neuropathy is not a major consideration as long as dose planning conforms to those guidelines. Radiation-induced damage to the residual normal pituitary function is a more important indicator of safety.32–34 Several studies demonstrated that SRS resulted in a lower incidence of hypopituitarism by injury to the hypothalamohypophyseal axis, as we also observed in the current study. However, most reports from the literature regarding the adverse effects of SRS were based on a shorter follow-up than the reports on FRT.33 In the current study, we had the same limitations of a shorter follow-up and relatively smaller tumors in the SRS group; however, the development of hypopituitarism probably does not occur earlier with SRS, unlike the normalization of hormone hypersecretion.
From this series, several tentative conclusions can be drawn, although they need to be confirmed in randomized, controlled trials. Both FRT and SRS are efficient treatment modalities for the control of tumor growth in patients with pituitary adenomas, and SRS results in more prompt normalization of hormone hypersecretion. The risk of radiation-induced hypopituitarism is not greater with SRS than with FRT in the short term. In patients who have small tumors with enough distance from the optic apparatus, SRS is more suitable than FRT in the period before endocrine normalization.