| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Radiation Oncology (J.H.P., T.I.Y., J.J.C., M.A., M.B., J.S.L.), Neurosurgery (B.S., P.C.), and Medicine (Neuroendocrinology) (B.M.K.B., A.K.), Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts 02114
Address all correspondence and requests for reprints to: Jay S. Loeffler, M.D., Department of Radiation Oncology, Massachusetts General Hospital, 100 Blossom Street, Cox 347, Boston, Massachusetts 02114. E-mail: jloeffler{at}partners.org.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to evaluate the efficacy and safety of PSR for persistent Cushing’s disease (CD) and Nelson’s syndrome (NS).
Design: This was a retrospective review of 38 patients (33 with CD and five with NS) treated between 1992 and 2005.
Participants: All patients had transsphenoidal surgery without biochemical cure. Four had previous irradiation with photons. The patients with NS underwent bilateral adrenalectomy 29–228 months (median 40) before PSR.
Intervention: Single-fraction PSR was delivered at a median dose of 20 Cobalt Gray Equivalents (range 15–20) on 1 treatment day.
Main Outcome Measures: Complete response (CR) was defined as sustained (
3 months) normalization of urinary free cortisol off medical therapy. CR in NS was based on normalization of plasma corticotropin.
Results: At a median follow-up of 62 months (range 20–136), CR was achieved in five patients (100%) with NS and 17 (52%) patients with CD. Among all patients with CR, median time to CR was 18 months (range 5–49). No secondary tumors were noted on follow-up magnetic resonance imaging scans, and there was no clinical evidence of optic nerve damage, seizure, or brain injury. There were 17 patients (52%) who developed new pituitary deficits.
Conclusions: PSR is effective for patients with persistent corticotroph adenomas with low morbidity after a median follow-up of 62 months; longer follow-up is warranted for late radiation-related sequelae.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Stereotactic radiosurgery (SRS), in which patients can be treated at a single sitting, has gained widespread use over the past decade. SRS can be performed using either photons [such as Gamma Knife (Center for Image-Guided Neurosurgery, University of Pittsburgh Medical Center, Presbyterian Hospital, Pittsburgh, PA), CyberKnife (Accuracy, Sunnyvale, CA), or linear accelerator-based radiosurgery] or protons. The primary advantage of protons is their characteristic Bragg peak, which essentially eliminates exit dose to structures beyond the target (Fig. 1
). This allows the treatment of an irradiated volume that conforms more closely with the actual target volume (6). Therefore, proton stereotactic radiotherapy (PSR), often referred to as proton SRS, achieves more selective treatment of the tumor mass while minimizing dose to surrounding tissues when compared with SRS using photons.
|
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
From 1992–2005, 38 patients (six males and 32 females) were treated with PSRS for ACTH-producing pituitary adenomas that were not cured after TSS. All patients had undergone one to four prior pituitary surgeries with a median of two surgical attempts for cure. All surgeries before PSR were transsphenoidal resections, and no patient underwent craniotomy. Two patients had been previously treated with fractionated radiotherapy: one with Cushing’s disease (CD) was treated 5 yr before PSR, and one with Nelson’s syndrome (NS) was treated 8 yr before PSR. Two other patients with CD had previously undergone Gamma Knife radiosurgery (GKRS) (3.5 and 4.5 yr before PSR). Patient and treatment characteristics for CD and NS are summarized in Tables 1 and Table 2, respectively.
|
|
For the 33 patients with CD, a pituitary etiology was confirmed by ACTH-positive staining cells in a surgical specimen for 25 patients (76%). The remaining eight (24%) without positive pathology all had findings consistent with ACTH-dependent pituitary CD and positive inferior petrosal sinus sampling, defined as a pituitary to peripheral ACTH ratio of two or more before ovine CRH and/or three or more after CRH 1 µg/kg. At PSR, three patients were refractory to adrenal blockade medication, six were controlled on medication, and 24 were uncontrolled but had not yet received medication. Medication was not withdrawn at PSR for the nine patients receiving adrenal-suppressive medication.
NS
Five patients had undergone previous bilateral adrenalectomy between 29 and 228 months (median 40) before PSR and had developed NS. The patients with NS all had both elevated plasma ACTH levels and tumor growth demonstrated on imaging. None of these patients had neurological deficits at PSR.
Treatment parameters
All patients were treated with PSR using two to five convergent beams of 160 or 230 MEV protons. Treatment volumes were constructed using fused CT and MRI (7). The entire sella, medial walls of the cavernous sinuses, and inferior dura of the sella turcica were targeted in all patients. Additional residual tumor outside of the sella, visible on MRI, was also targeted in 12 patients (32%). Three fiducial markers were placed in the outer table of the skull before imaging to ensure accurate localization on the day of treatment (8). A stereotactic head frame was placed on the day of radiosurgery to obtain adequate immobilization.
The median radiation dose was 20 Cobalt Gray Equivalents (CGEs) (range 15–20). The dose was prescribed at the 90% isodose line in 37 patients and the 100% line in one patient. The dose to the optic chiasm was limited to less than eight CGEs in all cases, and less than four CGEs for patients with a history of prior radiation treatment. All patients were treated to 20 CGEs except for three patients with a history of prior irradiation, for whom the prescription dose was reduced. One patient had previously been treated with GKRS (12 Gy minimum target dose) 3.5 yr before PSR, and the prescription dose was reduced to 15 Gy to keep the optic chiasm dose under four CGEs. The other two patients had previously received fractionated radiotherapy as described previously, and the prescription dose was reduced to 18 CGEs.
Follow-up evaluation
The study was approved by the Massachusetts General Hospital Institutional Review Board. Retrospective chart review included follow-up obtained from referring endocrinologists, neurosurgeons, and primary care physicians. Some patients also continued to be seen regularly in the Radiation Oncology Clinic and/or the Neuroendocrine Clinical Center at the Massachusetts General Hospital. All patients were evaluated with serial urinary free cortisol (UFC) levels, or plasma ACTH in patients with bilateral adrenalectomy. Medications that block adrenal production of cortisol were administered by the patient’s local physicians, with ketoconazole the most common choice, followed by metyrapone, and then aminoglutethimide. Attempts to wean cortisol-suppressive medications were made after patients had sustained normal UFC levels for at least 3 months. If discontinuation of medication resulted in high UFCs, medical treatment was resumed. Medications were permanently discontinued once complete remission, defined as successive normal UFCs on no medical treatment, was documented.
Patients were screened for pituitary dysfunction at least annually by checking levels of free T4, prolactin (women), and testosterone (men), as well as documenting the menstrual status of women who were premenopausal at PSR. Early morning cortisol more than or equal to 18 µg/dl or cortisol more than equal to 18 µg/dl after Cortrosyn injection (Amphastar Pharmaceuticals, Inc., Rancho Cucamonga, CA) was used to define normal glucocorticoid production.
The years encompassed by the follow-up predated the Food and Drug Administration approval of GH replacement in adults, therefore, systematic evaluation for GH deficiency was not performed in all patients but was available in 23. Glucocorticoid replacement therapy was administered to patients with subsequent adrenal insufficiency. MRI scans of the pituitary were obtained regularly (at least annually for 2 yr and then at least every 3 yr) to follow for evidence of radiographic recurrence and screen for secondary tumors, or when prompted by clinical or biochemical findings.
Endpoints–CD
Complete response (CR) was defined as sustained (3 months) normalization of UFC after the completion of a washout period during which medical therapy was withdrawn. Local tumor control, defined as no tumor enlargement on imaging, was also required for the definition of CR.
Endpoints–NS
CR was defined as normalization of plasma ACTH and local tumor control.
Endpoints–all patients
Before PSR, 22 patients (58%) already required replacement of one or more pituitary hormones, whereas the remainder of patients had no deficit. New partial pituitary dysfunction was defined as the need for initiating replacement of any additional pituitary axis hormones, including thyroid hormone, GH, sex hormones, or glucocorticoids, after PSR. In patients who were menstruating at PSR, the onset of amenorrhea without elevated prolactin level or clear evidence of menopause based on clinical symptoms and elevated FSH was counted as a new pituitary deficit. New complete pituitary dysfunction was defined as the requirement of replacement of all of these hormones among patients who had two or less deficiencies before PSR. Patients were considered deficient at the point when replacement medication was initiated.
Statistical analysis
Actuarial rates of CR to single dose PSR, and new pituitary deficiencies, were calculated using Kaplan-Meier estimates. Variables evaluated for statistical association with biochemical outcomes (CR and hypopituitarism) included age, sex, cavernous sinus invasion, visible MRI residual, treatment volume, pretreatment UFC or ACTH level, medical therapy at PSR, number of prior surgeries, history of prior radiation, and radiation dose. The group comparisons were performed using log-rank tests.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
CD
A CR after PSR was achieved in 17 patients (52%) with CD. Among patients with CR, median time to CR was 14 months (range 5–49). Actuarial rates of CR at 1–6 yr were 21, 45, 49, 49, 55, and 55%, respectively (Fig. 2). During follow-up, 50% of the patients were in CR at 25 months after PSR. No complete responder developed recurrent CD at last follow-up. There were 12 (36%) additional patients with CD who had achieved stable, normal UFC levels but have not been weaned off medical therapy. None of the patients with normal UFC had evidence for tumor growth on follow-up imaging. Among the four patients (12%) with persistently elevated UFC levels, two developed radiographic evidence of local tumor progression and underwent subsequent TSS during the follow-up period.
|
NS
A CR to PSR was achieved in all five patients (100%) with NS (Fig. 3). Median time to CR was 22 months (range 15–27). There has been no evidence for biochemical or radiographic progression after a median 9 yr (range 8–11) of follow-up in these patients.
|
No visual complications or clinical evidence of brain injury was observed. No patients had clinical or radiographic evidence suggestive of a cerebrovascular event during the follow-up period. No secondary tumors were noted on MRI scans. MRI scans in two of the four previously irradiated patients demonstrated new temporal lobe enhancement that was not associated with any symptoms. Before PSR, both patients had been previously treated with fractionated radiotherapy for involvement of the cavernous sinus directly adjacent to the area of enhancement. There were no changes noted among patients with no history of prior radiation.
Five patients had panhypopituitarism before PSR and, therefore, were not evaluated for new pituitary deficits. Of the 33 patients at risk, 17 (52%) developed new pituitary deficits during follow-up and were administered replacement hormones (Fig. 4): panhypopituitarism in two (6%), thyroid and GH in one (3%), adrenal and GH in one (3%), GH and estrogen in one (3%), thyroid hormone and estrogen in one (3%), thyroid only in six (18%), GH only in four (12%), and estrogen only in one (3%). The median time to pituitary hormone deficiency was 27 months (range 9–60) as seen in Fig. 5. On univariate analysis, the incidence of hypopituitarism was not associated with any of the tested variables.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Most conventional fractionated radiotherapy series have reported CR and tumor control in 50–57% of patients (9, 10, 11, 12, 13, 14); one study demonstrated a higher rate (15). Fractionated stereotactic radiotherapy, which delivers the same 6-wk daily treatment but uses improved patient positioning for tighter treatment margins, has achieved CR and tumor control in 54% of patients (16). Linear accelerator-based SRS and GKRS series showed CR and tumor control in 43–56% of patients (17, 18, 19, 20, 21, 22, 23, 24) treated with a single large radiation dose. Median time to CR has been relatively similar, consistently between 6 months and 2 yr. Our findings indicate that PSR has similar efficacy and timing to other radiation techniques. This similarity demonstrates a fundamental concept in the field of radiation oncology: the ability to deliver an adequate radiation dose to a target volume can be achieved by different techniques. The challenge lies in reaching this goal while optimally reducing the radiation dose to the normal surrounding tissue.
Based on the proton vs. photon dose characteristics, PSR has several potential advantages over other radiation techniques. Protons deposit a radiation dose over a finite distance (called the Bragg peak) with essentially no exit dose beyond this region. In contrast, photons (GKRS and linear accelerator-based SRS) deposit a maximum dose at a specific depth, then continue to deliver an attenuated but significant dose to the remainder of the tissue traversed before the photon exits the patient. Because of this difference, the irradiated volume conforms more closely to the target with protons than photons (6). This allows the delivery of a desired treatment dose to the target volume while optimally reducing radiation of surrounding normal tissues. This selectivity has several potential benefits in the treatment of pituitary tumors.
First, by reducing the optic apparatus dose, PSR may provide a radiosurgical treatment option when the proximity of the target volume to the chiasm is dose limiting for other SRS techniques. Second, the dose to the adjacent medial temporal lobes and vascular structures of the cavernous sinus is also reduced with PSR. Studies have suggested that irradiation of these tissues with conventional fractionated radiation increases the risk of long-term neurocognitive and cerebrovascular sequelae (25, 26, 27, 28). Limiting the radiation dose to these tissues by using PSR may lower the incidence of these complications. Finally, the most notable difference using PSR is the reduction in the integral dose. Integral dose represents the summation of radiation dose received by all tissues in the patient and is generally considered to relate to the risk of radiation-induced neoplasia. Reports have demonstrated an increase in the incidence of secondary brain tumors in areas previously irradiated using conventional fractionated radiotherapy for pituitary tumors (29, 30, 31, 32, 33). The most common tumors described after pituitary irradiation include malignant gliomas and meningiomas. The significant reduction in volume of normal tissue irradiated using PSR may reduce the risk for secondary tumors in this patient population. However, the median follow-up of 62 months in this series is not sufficient to confirm this hypothesis, given that the latency ranges from 5–34 yr in the literature (30); longer follow-up is needed.
Fractionated stereotactic radiotherapy and other forms of SRS also significantly reduce the exposure of normal structures to radiation when compared with conventional fractionated radiation. Although PSR achieves the least exposure of these tissues to radiation, and, thus, in theory should result in the lowest incidence of long-term sequelae, the difference between techniques may be small and could require large numbers of patients followed for long periods to demonstrate any clinically significant difference.
New pituitary deficiencies developed in 52% of patients in this series. This is somewhat higher than previously published radiosurgical series (17, 18, 19, 20, 21, 22, 23, 24). Differences in the percentage of patients developing hypopituitarism after radiation could result from a number of factors. One contributing factor is that 30 of 38 patients (79%) in this series had two or more surgeries, and 58% already required replacement of at least one pituitary hormone before PSR. Patients who have undergone more extensive surgical treatment may have less functional reserve and could be more likely to develop future deficits with or without radiation. In addition, at our institution the entire sellar contents, including the adjacent portions of the cavernous sinuses, are targeted in all patients treated with PSR for CD. This approach has been adopted based on the theory and literature suggesting that surgical recurrence may result from microscopic seeding in the medial wall of the cavernous sinuses or dural invasion. By design, we intentionally make no attempt to use the improved targeting possible with protons to reduce the dose to the pituitary gland. Instead, the selectivity of proton technique is used to deliver a high dose to this larger target area to ensure adequate treatment while reducing the dose to the optic apparatus, adjacent neurovascular structures, and temporal lobe. Another potential factor is that the median time to pituitary deficit was 27 months (range 9–60) in the current series. The shorter follow-up reported in most of the aforementioned series would be unlikely to have detected some of these events; the one study with longer follow-up noted deficits in 67% of patients, some occurring more than 10 yr after treatment (20).
Two patients developed new temporal lobe enhancement without clinical symptoms. Both had received prior conventional radiotherapy. In the recent publication describing the largest GKRS series to date, Jagannathan et al. (23) reported that two of three patients with prior fractionated radiotherapy developed new cranial neuropathies after GKRS. Both neuropathies occurred within 15-month treatment. In contrast, the two patients who developed new temporal lobe enhancement in the current series have been followed for 92 and 108 months after PSR, respectively, with no clinical evidence for cranial neuropathy or other deficit thus far. This highlights the fact that administering a second course of radiation may carry more risk. In addition, it indicates that although protons allow for substantially reduced irradiation to surrounding tissues, this dose is not zero and must be considered carefully.
One limitation of this study was the length of follow-up (median 62 months). Although this represents the longest follow-up for any cohort treated with radiosurgery for ACTH-secreting adenomas in the modern era, additional follow-up is required to evaluate the long-term sequelae of pituitary irradiation using PSR and all other radiotherapy techniques. One of the basic principles of radiation oncology is to reduce the volume of normal tissue exposed to ionizing radiation while ensuring adequate tumor dosing. In this respect, PSR currently provides the best technique for achieving this goal. As the number of proton facilities in this country continues to increase, further data on PSR will become important in guiding clinical practice.
Conclusions
These results demonstrate that PSR is effective for patients with persistent corticotroph adenomas, with 58% of patients attaining biochemical control off all medication after a median follow-up of 62 months. Our findings indicate that PSR achieves biochemical control with low morbidity; longer follow-up is warranted to assess for late radiation-related sequelae.
|
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online November 20, 2007
Abbreviations: CD, Cushing’s disease; CGE, Cobalt Gray Equivalent; CR, complete response; CT, computed tomography; GKRS, Gamma Knife radiosurgery; MRI, magnetic resonance imaging; NS, Nelson’s syndrome; PSR, proton stereotactic radiotherapy; PSRS, proton stereotactic radiosurgery; SRS, stereotactic radiosurgery; TSS, transsphenoidal resection of the tumor; UFC, urinary free cortisol.
Received June 4, 2007.
Accepted November 13, 2007.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. M. K. Biller, A. B. Grossman, P. M. Stewart, S. Melmed, X. Bertagna, J. Bertherat, M. Buchfelder, A. Colao, A. R. Hermus, L. J. Hofland, et al. Treatment of Adrenocorticotropin-Dependent Cushing's Syndrome: A Consensus Statement J. Clin. Endocrinol. Metab., July 1, 2008; 93(7): 2454 - 2462. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
