Department of Neurosurgery, University of CaliforniaSan Francisco, San Francisco, California 94143
Address correspondence and requests for reprints to: Charles B. Wilson, Department of Neurosurgery, University of CaliforniaSan Francisco, 505 Parnassus Ave, Box 0112, San Francisco, California 94143.
![]() |
Introduction |
---|
![]() |
Transsphenoidal Surgery |
---|
We subscribe to the principle of specialized care, and if it makes sense for interested internists and pediatricians to acquire special knowledge and experience in endocrinology, it seems reasonable to apply the same rationale to specialization in pituitary surgery. Particularly in neurosurgery practice makes perfect, and concentrating, rather than diffusing, surgical referrals by encouraging one of several neurosurgeons in larger communities to become the local expert in pituitary surgery can provide improved outcomes for your patients. Care of high quality is rarely cost-ineffective in children, and particularly in pediatric patients with Cushings disease, in which inexpert pituitary surgery can be very expensive, financially and in quality of life, in both the short-term and the long-term. Specialists in endocrinology have a critical role in the diagnosis, preoperative preparation, and postoperative management of all pediatric patients with pituitary adenomas. The short- and long-term follow-up of pediatric patients treated by surgery is best placed in the hands of an endocrinologist because critical decisions regarding recurrence of endocrine-active adenomas and the subtleties of anterior pituitary insufficiency require continuous monitoring by an expert for optimal outcomes.
![]() |
Pathology |
---|
The distribution of tumor cell types in the pediatric adenomas treated
by CBW at UCSF is listed in Table 1. For
several reasons, the distribution of cell types does not reflect the
true overall proportion of cell types of pediatric pituitary adenomas.
Because this is a series of tumors removed surgically, it does not
recognize asymptomatic adenomas, or PRL-secreting adenomas that are
managed satisfactorily by medical means and, therefore, not referred
for surgery. Furthermore, because of an institutional interest in
Cushings disease, the proportion of ACTH-secreting adenomas in the
series may be misleadingly high.
|
![]() |
Clinical Presentation |
---|
|
|
![]() |
Corticotrophic Adenomas (Cushings Disease) |
---|
All patients should undergo magnetic resonance imaging (MRI) with the administration of gadolinium. Our previous review shows this imaging modality had a 72% sensitivity in localizing the adenoma (1). However, their small size and the likelihood of a false positive MRI in 15% of the normal population argues for preoperative venous sinus sampling in selected cases. Bilateral venous sampling is technically difficult in children, and the risk of morbidity from the procedure or anesthesia is not entirely insignificant. Therefore, we recommend such exploration occur at centers with wide experience because interpretation of such a study can also be difficult. If a patient with no anomalous venous drainage patterns exhibits a lateralizing ACTH gradient of 2:1 or greater, then removal of the appropriate half of the anterior pituitary gland will be curative in 80% of cases. In the presence of a negative surgical exploration, a guide to the probable location of the adenoma is invaluable. Under the right circumstances, a hemihypophysectomy is appropriate and, in most cases, will be successful. Clinical observations support the view that one third, or even slightly less, of a normal anterior pituitary gland can provide normal anterior pituitary function.
Treatment of Cushings disease is transsphenoidal adenomectomy. The surgical exploration must be meticulous and must include the posterior lobe. The corticotrophic adenoma has distinctive gross characteristics. It is dirty-white and soft, bordering on semiliquid, and is often freckled by petechial hemorrhage. The adenoma often has a diameter of 2 mm or less. It can reach the surface of the anterior lobe, can be exophytic into the subarachnoid space, can invade or even lie entirely within the cavernous sinus or its wall (3), and, rarely, it can originate in the posterior lobe. Exploration of the pituitary gland begins on the side with the radiographic abnormality or the side of the ACTH gradient on venous sampling. If no tumor is found, a second exploration of the "abnormal" side is performed and then careful exploration of the opposite "normal" side is performed. Several patients have been operated on after a hemihypophysectomy was performed elsewhere based on an assumed infallibility of inferior petrosal venous sampling: at a second operation, the microadenoma was removed from the unexplored "normal" remaining anterior lobe. Hemihypophysectomy is indicated after a negative exploration (i.e. in a child who has a clearly lateralized venous gradient with a normal venous anatomy and with informed parental consent), at which time the stalk must not be injured. Radiation therapy is suboptimal in the pediatric population because of the risks of injuring pituitary function and because its effects in correcting hypercortisolemia are delayed and uncertain.
Cushings disease is a rare disease that, if left untreated, is fatal. With no other tumor is the surgeons experience more critical in determining outcome, and for this reason inexperienced pituitary surgeons must gain experience with less complex tumor types, initially referring pediatric patients with Cushings disease to an experienced colleague.
![]() |
Prolactin-Secreting (Lactotrophic) Adenomas |
---|
Based on the basal prolactin level and a high-resolution MRI of the sella, one can make a prediction for the likelihood of surgical cure. If cure is not possible because of extrasellar spread, particularly cavernous sinus invasion, surgery is not advised unless the patient fails to respond to medical management. From our experience, extension into the cavernous sinus is overestimated based on the MRI alone because macroadenomas can push the wall of the cavernous sinus laterally, giving the appearance of invasion. When there is no chance of surgical cure, the objective of surgery is to reduce the adenomas mass, in order to reduce the production of PRL to a level that can be further reduced into a desirable range by a tolerated dose of a dopamine agonist or to relieve symptoms of compression and reduce the bulk of the adenoma before irradiation. Even after radiation therapy, prolactinomas continue to secrete PRL for several years and, therefore irradiation is a poor and very delayed means of treating the hyperprolactinemic syndrome.
In general, the level of PRL in the blood corresponds to the tumor size, such that hyperprolactinemia from a large adenoma with a mildly elevated PRL reflects stalk compression. The postoperative PRL value, obtained 12 days after surgery, provides an accurate prediction of outcome: an unmeasurable (<2 ug/L) value predicts a cure with more than 90% probability, and higher values within the normal range indicate incomplete removal of the adenoma. Surgery of prolactinomas, even in the pediatric population, has good outcome with long-term surgical cure of 82% for all prolactinomas (median follow-up, 5 yr) (2), with very low morbidity and no mortality.
Dopamine agonists, such as Parlodel, inhibit tumor-cell replication in the majority of prolactinomas and can result in normalization of PRL levels in these patients. Despite a dramatic reduction in tumor size in the majority of patients treated with dopamine agonists, elimination of the tumor has not been reported, and, subsequently, these patients will require life-long pharmacological treatment. For patients with residual or recurrent adenoma after surgery, often a small dose of Parlodel, taken at bedtime, is sufficient to maintain or restore normal PRL levels and prevent tumor regrowth. Irradiation is reserved for those few patients with surgically unresectable disease who are intolerant of medications. The risk of developing or aggravating pituitary insufficiency remains a major drawback to parasellar radiotherapy.
![]() |
GH (Somatotrophic) Adenomas |
---|
![]() |
Endocrine-Inactive Adenomas |
---|
Nearly all symptomatic endocrine-inactive macroadenomas have extended beyond the confines of an expanded sella turcica, the most common pattern of extrasellar growth being directly upward into the suprasellar space. Neither suprasellar extension nor focal perforation of the sellar floor, alone or combined, precludes complete removal. With some exceptions, the only patterns of growth that preclude total removal are lateral extension into the cavernous sinus or middle fossa. Invasion of the dura separating the sella from the cavernous sinus, even without extension into the sinus itself, renders the tumor surgically unresectable unless the invasion is focal, in which case the involved dura can sometimes be removed. A large tumor can displace the intact dural wall of the cavernous sinus far laterally without perforating the dura and mimic extension into the cavernous sinus on MRI.
The surgical objective in treating large endocrine-inactive adenomas is decompression of the optic nerves and chiasm. Unless the tumor invades the cavernous sinus, complete removal of the tumor with preservation of the compressed anterior lobe is possible in the majority of cases, and todays objective should be curative surgery with preservation of anterior pituitary lobe function. In the majority of large adenomas, a clean surgical plane separates the surface of the tumor from the compressed normal structures. By finding and developing this plane early in the process of tumor removal, a surprisingly large proportion of macroadenomas can be resected completely. Our belief is that more than half of the endocrine-inactive adenomas can be cured by surgery alone, and for this reason the surgeon should go into the operation with complete removal, rather than decompression, as the goal. Loss of preoperative anterior pituitary function has been an infrequent complication; more often, function that was lost preoperatively later returned spontaneously.
![]() |
Incidental Adenomas |
---|
Indications for removal of incidental intrasellar tumors and cysts are
individualized according to the patients age, sex, development, and
the size of the lesion. Table 3 provides
relative indications for surgical intervention. An asymptomatic adenoma
that has doubled in size within a year is far more troubling than an
adenoma that has shown slight but unquestioned growth over the course
of 5 yr. We have included a pure cyst that is larger than 1cm because
its potential for further expansion at some unpredictable growth rate
is different than the small cysts seen commonly in the pars intermedia.
In other words, we are convinced that in this age group, a cyst that
has reached a diameter of 1cm is "committed."
|
![]() |
Conclusion |
---|
We express particular gratitude to the pediatric endocrinologists at UCSF from whom we have learned so much.
Received September 17, 1999.
Revised October 5, 1999.
Accepted October 5, 1999.
![]() |
References |
---|