Bulletin Volume 4, Issue 1, Winter 1997

Stereotactic Proton Irradiation of Pituitary Adenomas

Pituitary Home

Alan Thornton Jay Loeffler
by Allan F. Thornton, M.D. and Jay S. Loeffler, M.D.

NEPTCC Newsletter MGH Neuroendocrine Center Bulletin Volume 4, Issue 1, Winter 1997


With the opening of the Northeast Proton Therapy Center (NPTC) in the fall of 1998, a quantum increase in the available resources of proton therapy will occur, allowing many pituitary patients to realize the benefits of this important modality. The Massachusetts General Hospital has long been a pioneer in the development of stereotactic precision irradiation of pituitary neoplasia. Since 1963, the Departments of Neurosurgery, Endocrinology, and Radiation Oncology have used proton beam therapy (Bragg Peak Particle therapy) available through the Harvard Cyclotron Laboratory to irradiate precisely a variety of skull base tumors. Although this effort in its early years was a limited program, many of the seminal discoveries and elemental techniques of the field of radiosurgery (treatment of small volumes of tissue with high-dose, precision irradiation) were developed within the MGH-HCL proton program and later inspired the development of gamma-knife and focused stereotactic linear accelerator therapy. The treatment of benign pituitary neoplasia remains one of the most important applications of this therapy. This method allows physicians to deliver curative doses of irradiation in a sufficiently focused manner to preclude damage to adjacent tissues, while delivering sufficiently high doses to obtain lasting tumor mass and hormonal control.

The management of pituitary tumors has undergone major changes over the past 20 years, necessitating re-evaluation of the roles of both conventional radiotherapy and radiosurgical applications. Both the availability of MRI imaging offering resolution of greater than 2mm in an area previously difficult for CT to image, as well as transsphenoidal ressection as a safer method of surgery, have radically changed the management of these tumors. Development of the agents bromocriptine and cabergoline, and, the somatostatin analogue, Octreotide, has led medical management of functioning adenomas. Finally, the development of improved radioimmunoassay techniques now allows both early diagnosis and sensitive detection of recurrence of pituitary adenomas. As a consequence, these tumors are detected at an earlier stage, and alternatives to surgical resection and wide-field irradiation are now possible.

In general, radiation therapy, whether by conventional fractionation over a six week period, or by radiosurgery, has the advantage of non-invasively treating and potentially curing unresectable pituitary disease, either in the post-operative setting or for patients who are not surgical candidates. However, there are several disadvantages of current conventional irradiation techniques which may be improved by stereotactic proton radiotherapy. The first lies in the relatively slow (6 months to 3 years) decrease in hormone excess symptoms after irradiation. Second, although complications after fractionated irradiation are rare, given present conformal irradiation techniques and energies available, many patients develop some degree of hypopituitarism several years following the irradiation requiring replacement hormonal therapy. Third, current techniques often irradiate the visual apparatus unnecessarily, increasing the potential risk to the optic pathways either from the initial treatment (rare, occurring in less than 1% of cases when doses of less than 4600 cGy at 200 cGy per fraction are observed), or from reirradiation, should the tumor recur years later. And finally, the risk of second malignancies induced from large-field irradiation in not negligible, estimated at just under 3% in recent studies from Canada.

The use of conventional irradiation for the treatment of neoplasia of the pituitary region has routinely involved the use of relatively simple orientations of treatment portals intending to treat the MRI-defined pituitary volume in addition to a relatively generous margin of normal brain. These treatment plans customarily involve 2 or 3 axial-plane static fields, but may incorporate the use of dynamic, arcing treatment designs intending to focus on a confined volume of brain, thereby allowing a lower "safe" dose of radiation to be delivered to the normal brain. More complicated planning has been performed, usually incorporating multiple beam angles (5-6), all within an axial plane. However, inherent risks of irradiation employing these plans include temporal lobe damage and failure to include the marginal target zones, particularly important for larger lesions, including those with cavernous sinus extension. The recent advances in imaging of tumors of the pituitary region incorporating MRI now offer the potential of more precise dose confinement, thus decreasing the recognized risk of temporal lobe damage, while increasing confidence of adequate irradiation of tumor margins. Proton radiotherapy realizes this precise dose confinement through the marriage of MR-based 3-dimensional treatment planning with an irradiation modality capable of homogeneous (within ± 5%) treatment of small, irregularly shaped treatment volumes.

The advantages of proton radiotherapy are entirely provided by the physical properties of the beam. The finite range of penetration of protons is affected by both the initial beam energy and the electron density of the absorbing material. The rapid increase in the rate of energy loss near the end of the range of a particle (proton) results in a well-defined volume of increased dose, known as the Bragg peak. By appropriate distribution of proton energies, the Bragg peaks may be grouped so as to provide a uniform dose across the target. This absence of exit dose offers important advantages to patients with sellar pathology. Proton irradiation delivers substantially higher doses to the target tissues, while respecting accepted dose constraints on critical normal tissues (chiasm, optic nerve, brain stem).

Accurate treatment with particle irradiation requires accurate dosimetry, reflecting the correct prediction of proton absorption within the scattering material. Such dosimetry relies on complex algorithms to provide information on the likely patterns of scattering and absorption of incident protons and involves computer modeling incorporating beam’s eye view perspectives (BEV) of the relative positions of the tumor, target, and critical structure volumes. However, this accurate prediction of dose deposition mandates excellent, and consistent, patient immobilization and correlation of imaging studies. Patients are immobilized in the supine treatment position using thermoplastic cranial immobilization. Patients are imaged with CT and MRI, using minimum slice spacing and contrast, in the treatment position using the above masks. Implanted metallic fiducials are used within the cranium to further enhance the stereotaxic precision of the beam planning and delivery. Following image acquisition, treatment image correlation is performed using both CT and MRI information within an integrated 3-D RTTP system running on a microcomputer.

Current Protocols

Stereotactic irradiation at the MGH may be delivered either with fractionated therapy over 6-7 weeks, or in a single, radiosurgical method (e.g., 24 Gy). Currently, we reserve radiosurgery for lesions that are intrasellar and greater than 7mm to the optic apparatus. This distance is necessary to avoid excessive dose to the optic chiasm and represents the proton beam edge. Delivered with full incorporation of 3-dimensional treatment planning, stereotactic frame cranial immobilization, and cranial fiducial localization, patients treated with single-fraction treatment (radiosurgery) are rotated about the proton beam using a STAR patient immobilization system developed for the the Harvard Cyclotron in conjunction with the Department of Neurosurgery.

Patients with larger pituitary tumors are eligible for fractionated irradiation protocols. Currently, the Departments of Endocrinology, Neurosurgery, and Radiation Oncology are embarking on a randomized, dose-escalation protocol comparing standard (50.4 Gy) irradiation to escalated (59.4 Gy) doses delivered to secretory pituitary adenomas. This proton therapy effort has been piloted and demonstrates a significant increase in dose to the pituitary adenoma, while maintaining no increased risk to the optic apparatus. Because long-term, retrospective series of secretory pituitary adenomas treated with convention irradiation to 40-50 Gy have demonstrated hormonal control not exceeding 45% after fifteen years, this protocol will represent the first effort to improve on cure rates with fractionated irradiation. As proton therapy resources increase with the opening of the NPTC, non-functioning pituitary adenomas will be treated in a similar manner to slightly lower doses. For these patients, the advantage of proton therapy lies in avoidance of the visual system, affording the potential for retreatment with irradiation in the future with less visual risk.

Finally, patients with recurrent tumors, previously irradiated may be eligible for re-irradiation with proton therapy, provided the geometry of the recurrent tumor allows adequate sparing of the visual system. Two previous series suggest a significant salvage rate with reirradiation to conventional doses. However, temporal lobe and visual system damage remain concerns in this group of patients - risks that may be minimized with proton stereotaxy.

Inquiries regarding pituitary irradiation with proton therapy may be made to Dr.Swearingen (617) 726-3910, Dr. Klibanski (617) 726- 3874, or to Drs. Loeffler and Thornton (617) 726-8150

Acromegaly: Complications and Therapeutic Update

Pituitary Home

Laurence Katznelson
by Laurence Katznelson, M.D.

NEPTCC Newsletter MGH Neuroendocrine Center Bulletin Volume 4, Issue 1, Winter 1997

Acromegaly is characterized by enlargement of the hands and feet, facial changes including frontal bossing, enlarged mandible and increased dental spacing, arthralgias, fatigue, diaphoresis, sleep apnea, hypertension, diabetes mellitus, and hypertrophic cardiomyopathy. Because it is a rare disorder and development of these clinical features is insidious, patients typically have acromegaly for many years before the diagnosis is made. Approximately 90% of all somatotroph tumors, which cause this disorder, are macroadenomas (greater than 1 cm) at diagnosis. Therefore, these tumors frequently cause local anatomic compression, resulting in visual field deficits, headaches, hypopituitarism and cranial nerve palsies.

The pulsatile release of growth hormone (GH) by normal pituitary somatotroph cells is regulated by growth hormone releasing hormone (GHRH), which stimulates GH secretion, and somatostatin, which decreases secretion. At the liver, GH stimulates secretion of somatomedin C, also known as insulin-like growth factor I (IGF-I). IGF-I mediates many of the peripheral somatic effects of GH and feeds back at the level of the hypothalamus and pituitary resulting in a reduction in GH secretion. Therefore, GH and IGF-I levels are held in tight balance.

The diagnosis of acromegaly is based on three key findings: 1) clinical evidence, 2) demonstration of an elevated IGF-I level, and 3) inability to suppress serum GH to less than 2 ng/ml following an oral glucose challenge (OGTT).

Why do we treat? Short term benefits of therapy include improvement of symptoms such as headaches, which are often debilitating. In addition, there are long-term complications of acromegaly that are of concern. There is a 2 to 5 fold increase in the mortality rate in acromegalic patients and this is largely due to cardiovascular and cerebrovascular disease. In a recent long-term follow-up of 79 subjects, therapy (regardless of modality) of acromegaly with resultant reduction of GH to greater than 5 ng/ml was associated with a decrease in the risk of mortality to that expected for the population. Therefore, given this provocative although limited data, successful management of acromegaly may negate the mortality risk.

There are multiple medical complications associated with acromegaly. In part because of hypertension, there is cardiac involvement that includes left ventricular hypertrophy and congestive heart failure. Sleep apnea syndrome (both central and obstructive) is detected in up to 80% of subjects and may result in considerable morbidity. Acromegalics may also develop significant arthropathy that may lead to pain and necessitate joint replacement. Left ventricular mass, sleep apnea syndrome, and arthralgias may improve with therapy.

Patients with acromegaly may also be at enhanced risk for cancer, and colon cancer is the most prevalent. This risk is particularly increased in men over 40 years with a positive family history of colon cancer and multiple skin tags. Other malignancies, including breast cancer, have been described. Although it seems likely, it is unknown whether successful treatment of acromegaly will reduce the risk of neoplasia.

The primary mode of therapy for acromegaly is surgery to reverse the mass effect and attempt biochemical cure. Surgical cure is dependent on surgical skill and experience as well as the size of the tumor. Cure, defined as normalization of IGF-1 levels and normalization of the GH response to an OGTT, is demonstrated in up to 88% of patients with microadenomas ( greater than 1cm). In contrast, up to 50-65% of acromegalic patients with macroadenomas are cured following transsphenoidal surgery. Residual disease following transsphenoidal surgery is therefore common, indicating the need for adjuvant therapy. Radiation therapy is a potential adjuvant therapy for patients with residual disease, however, there is a delayed effect in that 1/2 to 2/3 of subjects attain GH levels greater than 5 ng/ml by 10 years. Hypopituitarism is a significant complication of radiation therapy. Therefore, in most patients, medical management may be necessary in surgically non-cured patients in lieu of or in combination with radiation.

Medical management is a highly useful adjuvant therapy for patients with residual disease. Dopamine agonists, including bromocriptine (parlodel) may normalize GH and IGF-1 levels, but in only 8% of patients. Therefore, it may be reasonable to attempt a course of bromocriptine as adjuvant medical therapy, but it may have limited value. In addition, large doses are often required and this therapy may be associated with significant side effects.

The most efficacious form of medical therapy available includes somatostatin analogs, such as octreotide. Many studies have demonstrated the efficacy of octreotide in the management of acromegaly. The initial octreotide dose is usually 50 mg b.i.d., and doses may be increased to 250 or 500 mg t.i.d. depending on the response of circulating GH and IGF-1 levels. However, most studies show 300-900 mg per day is an effective dose. Octreotide administration results in a decrease in GH and IGF-1 levels in a majority of patients with normalization of IGF-1 levels in up to 60% of patients, indicating biochemical remission. Most patients note a marked improvement in their symptoms of acromegaly very soon after starting octreotide therapy, including headaches, joint pains and diaphoresis. The most significant adverse effect of somatostatin analogs is the development of gallstones, so ultrasounds should be obtained initially. However, the development of symptomatic gallstones are very rare and the need for serial ultrasounds is controversial. Other side effects include gastrointestinal disturbances with nausea, abdominal pain and diarrhea which often occur after initiation of therapy but usually resolve within 1 to 2 weeks.

An exciting new approach to the management of acromegaly is the development of longer acting somatostatin analogs that may be administered intramuscularly at 2 to 4 week intervals. These analogs are currently under active investigation. Efficacy of these analogs appears similar to that of shorter acting preparations, and, in theory, long-acting analogs may have greater efficacy because of continuous versus intermittent GH suppression. The additional benefit of requiring injections at monthly intervals versus multiple times during the day makes these analogs preferable.

The MGH Neuroendo-crine Unit is currently initiating studies involving administration of these long-acting analogs to patients with acromegaly. Physicians interested in this study should contact Dr. Katznelson at 617-726-3874.


  1. Ho KY, Weissberger AJ, Marbach P, Lazarus MB. Therapeutic efficacy of the somatostatin analog SMS 201-995 (Octreotide) in acromegaly. Ann Int. Med. 1990; 112:173-181.
  2. Serri O, Somma M, Comtois R, Rasio E, Beauregard H, Jilwan N, Hardy J. Acromegaly: biochemical assessment of cure after long term follow-up of transsphenoidal selective adenomectomy. J Clin Endocrinol Metab. 1985; 61: 1185-1189.
  3. Bates A.S., Van’t Hoff W., Jones J.M. Does treatment of acromegaly affect life expectancy? Metab. 1995;44: 1-5.

Pituitary Journal Review: Discussion of Recent Articles of Interest Related to Pituitary Disease

Pituitary Home

Beverly MK Biller
by Beverly M.K. Biller, M.D.

NEPTCC Newsletter MGH Neuroendocrine Center Bulletin Volume 4, Issue 1, Winter 1997

Article 1 Review
"Prolactinomas Resistant to Standard Dopamine Agonists Respond to Chronic Cabergoline Treatment" - A Colao, A DiSarno, F Sarnacchiaro et al. 1997 J Clin Endocrinol and Metab 82:876-83

Anorexia nervosa is a devastating disease that affects approximately 1% of college-aged women. Although it is a psychiatric illness, the medical sequelae of prolonged starvation in women afflicted with the disease are myriad, serious and the focus of this article. A 5.6% mortality rate per decade – 12 times the rate for healthy young women — has been established, and is at least in part due to an increased risk of suicide in women with anorexia nervosa (1). The cause of death in other women with anorexia nervosa is often not clear, even after autopsy, and may be related to medical issues, particularly cardiac. Medical complications of anorexia nervosa are common and include bone loss, cardiac dysfunction, electrolyte disorders, and bone marrow suppression.

Bone loss is nearly universal in women with anorexia nervosa, due to the effects of severe undernutrition on endocrine regulators of skeletal homeostasis, and results in an increased fracture rate. We reported in the Archives of Internal Medicine that less than 15% of 214 young women with anorexia nervosa – average age 25 years — had normal bone density at all skeletal sites tested (2). Thirty-four percent of these women had osteoporosis, defined as having a bone density more than 2.5 standard deviations below the normal healthy mean for young women (T score greater than -2.5) (Figure 2). Whether this reduction in bone density translates into an increased risk of fractures is an important question. Although we could not investigate this question directly in a cross-sectional study, of note, 30% of the women studied reported histories of fractures (2). In over one-third of these cases, multiple fractures were reported, and in 42% of cases, the fractures were atraumatic, i.e. resulted from minimal trauma that does not usually cause fractures (2). These rates are much higher than expected for a young healthy population and confirm published data by Rigotti et al, who reported a non-spinal fracture rate seven times higher than for healthy young women without anorexia nervosa (3). Despite the presence of amenorrhea, estrogen therapy is not effective at reversing bone loss in women who have already achieved peak bone mass (4, 5). We are currently studying whether estrogen use during adolescence will counteract the deleterious effects of amenorrhea on peak bone mass accrual. We previously demonstrated that recombinant IGF-I administration increases bone formation and bone density in adult women with anorexia nervosa (5), and studies of other potential therapies are ongoing (see listing of studies elsewhere in this issue for referral information.)

With the recent United States approval of cabergoline for hyperprolactinemia, there is increasing interest in this new, long-acting dopamine agonist. A recent JCEM article provides interesting information about the effectiveness of this medication in patients who have prolactinomas which failed to respond to other dopamine agonists.

This study, conducted in Italy, evaluated the response to cabergoline in 27 patients who had previously been shown to be resistant to bromocriptine. Resistance was defined as absent or poor response of prolactin (PRL) and/or lack of tumor shrinkage despite at least 3 months of 15 mg bromocriptine daily. The majority of the patients were also resistant to quinagolide, another dopamine agonist available in Europe. Nineteen of the subjects had macroprolactinomas and 8 had microprolactinomas; 9 were men, 18 were women, and ages ranged from 15 to 64 years. The majority of patients (7/9 men and 17/18 women) had gonadal dysfunction.

Cabergoline was administered at a starting dose of 0.25 mg once weekly for the first week, twice weekly for the second week, and 0.5 mg twice weekly thereafter. Progressive upward adjustment of the dose was made on the basis of serum PRL levels, with a maximum dose in this study of 3 mg/wk, administered as 0.5 mg six days/week.

A significant finding of this study was that the majority of patients (15 of 19 macroadenomas and all 8 microadenomas) attained a normal PRL level during the 22 months of therapy, despite the fact that none of them had done so on a relatively high dose of bromocriptine. In three of the remaining patients, PRL levels declined substantially, with only one patient being withdrawn from the study at 3 months because of complete absence of effect.

Another important finding was that tumor shrinkage (which was defined conservatively, with at least 25% reduction required by MRI scan) occurred in 9/19 macroprolactinomas and 4/8 microprolactinomas. Gonadal dysfunction improved in two-thirds of patients, headaches resolved in the majority of patients and galactorrhea resolved in all women experiencing this symptom. No subject discontinued the medication due to intolerance, and it was well tolerated by the 16 patients who had experienced side effects on other dopamine agonists.

One criticism of the study, in a letter to the Editor (JCEM 1997, 82:2756), was that the claimed effectiveness of cabergoline for cases of bromocriptine resistance might have been overestimated, with the higher success rates actually due to greater tolerability, resulting in higher compliance. The authors countered that, while this may be true, the net result remained greater effectiveness of cabergoline.

This study is important because it suggests that the majority of patients previously unable to be treated with dopamine agonists can be successfully managed with cabergoline. While the number of subjects was small, the demonstration of PRL normalization in all microprolactinoma patients warrants a trial of this dopamine agonist in such patients not responsive to bromocriptine. This study also suggests that cabergoline may be particularly beneficial to patients with macroprolactinomas, as it will reduce the number of such patients who require transsphenoidal surgery due to failure of medical treatment.

Article 1 Review
"Pituitary Irradiation is Ineffective in Normalizing Plasma Insulin-Like Growth Factor-1 in Patients with Acromegaly" A Barkan, I Halasz K Dornfeld et al. 1997 J Clin Endocrinol Metab 82: 3187-91

Radiation therapy has been employed in patients with residual acromegaly following transsphenoidal surgery. However, most of the literature about its effectiveness antedated the use of IGF-1 normalization as a key criterion for cure, and therefore reported success based on lowering growth hormone (GH) to below 5 mcg/L. It is now recognized that this level is substantially higher than in normals, and does not represent acceptable control of acromegaly. A recent JCEM article addresses the effectiveness of radiation therapy for treatment of this disorder using IGF-1 measurements.

In this retrospective study, charts were reviewed from 140 acromegalics treated in Michigan over a 21 year period. Of these, data from 38 patients who underwent radiation therapy and had IGF-1 levels obtainable from the records were evaluated. The main finding of the study was that only 2 patients (5%) achieved age-and sex-adjusted normal IGF-1 levels while off medical therapy. An interesting observation was that the majority of these patients had GH levels below 5 mcg/L, again indicating that this criterion does not indicate adequately biochemical control.

There are several problems with this study. First, the number of patients analyzed was fairly small. Another issue was that because of the retrospective design spanning a 21 year period, and the fact that many patients had obtained blood tests at local labs, IGF-1 measurements were made by an enormous variety of methods at many different laboratories with no consistency in normal ranges. To address this problem, the authors report plasma IGF-1 values as a percentage of the upper limit of normal for each lab conducting the test. The most important problem with the study is that over half of these patients (20/38) had been followed for fewer than 5 years, and older data using GH levels suggest a continued effect of radiation even ten years after its administration.

In the accompanying JCEM editorial, van der Lely, de Herder and Lamberts suggest reserving radiation therapy for those patients with large, infiltrating pituitary tumors which cannot be cured surgically nor controlled medically with somatostatin analogue therapy. A critical question which remains is whether the newer stereotactic radiosurgical techniques such as gamma knife or proton beam (see article by Drs. Allan F. Thornton and Jay S. Loeffler in this issue) will be more successful at normalizing IGF-1 levels, using a careful prospective analysis.