|Home | About | Journals | Submit | Contact Us | Français|
In June 1992 at the age of 4 years, a Hispanic boy underwent placement of a right ventriculoperitoneal (VP) shunt in Mexico followed by placement of a left VP shunt in July 1992. He underwent a pterional craniotomy on July 16, 1992, with biopsy and aspiration, and a craniopharyngioma was diagnosed. In August 1992, the VP shunts were revised. In October 1992 in the United States, the patient underwent a transcallosal craniotomy for excision of the tumor. Postoperatively, he received 5400 cGy of radiation therapy. In 1997 the tumor recurred and at age 8, he underwent a right orbitozygomatic craniotomy with near–total excision of the tumor. In June 1999 (age 10) he underwent a right pterional orbitozygomatic craniotomy with excision of the tumor followed by Gamma knife therapy. At 13 years he was brought for treatment after experiencing increasing headaches, double vision, and decreasing vision for 4 months.
His current medications are hydrocortisone (10 mg a.m., 5 mg p.m.), levothyroxine (0.125 mg, 1/day); 1–deamino–8–D–arginine vasopressin (DDAVP; 0.1 ml intranasal spray twice daily). On April 18, 2000, magnetic resonance (MR) imaging showed a cystic recurrence of the tumor (1.7 × 1.1 cm). On October 16, 2000, the lesion measured 1.9 × 1.4 cm (Fig. 1); on March 14, 2001, it measured 2.6 × 2.1 × 1.7 cm; and on June 6, 2001, it measured 2.6 × 2.7 × 3.2 cm. The patient could not see hand motion with his right eye. Visual acuity was 20/200 in only a small nasal field in his left eye. He had complete temporal hemianopsia. He was almost completely and permanently blind. He has hypopituitarism, and the tumor threatens the hypothalamus.
This case represents a particularly difficult problem with regard to craniopharyngioma. Unfortunately, problems like this are not rare, especially in patients treated in childhood. Like many children with large craniopharyngiomas, this youngster presented with obstructive hydrocephalus and was initially treated with a shunt. One assumes that the tumor was largely cystic because the cyst was aspirated after a pterional craniotomy. He then underwent shunt revisions and subsequently a transcallosal approach for removal of the tumor followed by a standard course of 5400 cGy of conventional radiation therapy, a rather large dose for a child this age. At age 8 he underwent a right–sided orbitozygomatic craniotomy, which was repeated when he was 10 years old and followed by Gamma knife radiosurgery.
At presentation at the age of 14, the patient had increasing headache, diplopia, and progressive visual loss. Imaging studies showed a recurrent suprasellar tumor on the left, and the tumor appeared to be primarily cystic. Vision was severely threatened and he was almost blind.
When making a recommendation for treatment, the surgeon must consider the quality of life of this almost–blind 13–year–old with profound hypopituitarism. From the data provided, it appears that simple aspiration of the cystic recurrence will not restore vision. Typically, these cystic tumors recur rapidly unless injected with a radioisotope (P32) or bleomycin. Both injection strategies would be reasonable options in this situation, especially if vision is not anticipated to recover. These represent the most conservative alternatives and should help deal with the patient's ongoing headaches.
Another attempt at radical removal of this tumor could also be considered. The tumor is eccentrically placed on the left and could be approached through a left subfrontal craniotomy approach, with or without removal of the roof of the orbit. If the patient's vision is poor enough, sacrificing the optic nerve on this side to obtain better exposure is an option. An interhemispheric approach could also be considered but might be difficult and hazardous because of the patient's previous surgery. An anterior temporal craniotomy, as recommended by Lindsay Symon, can also be considered. In some patients the temporal pole can be mobilized and retracted directly posteriorly, thus gaining access to the tumor from the side and below the optic chiasm. In some cases, particularly in children where it is difficult to mobilize the relatively large temporal lobe, an anterior temporal lobectomy can provide the same access.
Finally, a trans–sphenoidal extended skull base approach could be considered. The tumor, which sits directly above the sella, could be approached inferiorly, and extracapsular removal of the lesion could be attempted. Because of this patient's prior surgery and the presumed adhesion of the dorsal aspect of the tumor capsule to the optic chiasm, hypothalamus, and vessels of the circle of Willis, this approach would be quite hazardous.
Like many patients with craniopharyngiomas, this patient could be managed with a number of different strategies. None, however, would likely restore this unfortunate young man to an excellent quality of life. Whichever method is chosen, the goal should be to relieve his current headaches, to preserve what little of his vision remains, and to provide some measure of long–term tumor control.
Considering all the options and the patient's specific clinical situation, my preference would probably be stereotactic aspiration of the cyst with installation of a beta–emitting radioisotope such as P32. The advantage of P32 over bleomycin is that only a single injection would be necessary. Repeated injections are ordinarily recommended with bleomycin and are best accomplished through a catheter inserted within the cyst under direct visualization.
This case represents a residual solid tumor that caused repeated cystic recurrences. These lesions are notoriously refractory to radiation therapy. I suspect that the cystic component has been addressed on previous occasions, but not the solid component. The position of this tumor lends itself well to an extended trans–sphenoidal approach that would provide access to the solid tumor in and above the sella with subsequent resection of the cystic component. If this strategy should fail, the cystic component could by treated with either P32 or bleomycin as a last resort. In such cases, we have had excellent results with the trans–sphenoidal route, which offers excellent access to the solid component that cannot be visualized adequately through transcranial approaches.
The patient was not a candidate for further conventional radiation therapy. Gamma knife therapy was inadvisable because of the proximity of the optic nerves. No effective parenteral or oral chemotherapy was available. Alternative treatments included stereotactic procedures such as yttrium–90 or P32. Surgical options included pterional, orbitozygomatic, subtemporal, transcallosal, suboccipital, translaminar, terminalis, transcortical, bifrontal interhemispheric, and trans–sphenoidal approaches. In the past, trans–sphenoidal surgery has only been recommended for tumors that are primarily intrasellar and cystic, for those associated with an enlarged sella, or for palliation of recurrent or persistent craniopharyngiomas.
This patient did not have an enlarged sella or a primarily intrasellar tumor. However, we opted for a trans–sphenoidal transtubercular approach. The transtubercular approach is best suited for lesions located between the optic nerves and carotid arteries, as well as anterior to the pituitary stalk. Because this patient already had hypopituitarism, including diabetes insipidus, pituitary function did not have to be preserved. The surgery was performed using contrasted MR image guidance. A sublabial incision was selected for its slightly improved angle anterior and superior to the sella compared to a transnasal approach. This incision also maintains a midline orientation and slightly wider approach compared to a unilateral transnasal approach. The sphenoid sinus was quite small with presellar–type pneumatization.
The bone was removed from the planum sphenoidale, tuberculum sella, and anterior sella with a 2–mm diamond burr and 2– and 3–mm 45–degree Kerrison rongeurs. The dura was opened into the floor of the frontal fossa, across the intracavernous sinus, and over the anterior sella. Gelfoam® (Upjohn, Kalamazoo, MI) soaked in thrombin and bipolar cauterization were used to control bleeding from the intracavernous sinus. The cystic mass was immediately apparent on opening the dura. The optic chiasm was identified, and the capsule was dissected from the chiasm. The cyst was drained, and the capsule was dissected carefully from the inferior surface of the chiasm. There was no recognizable pituitary stalk, which may have been removed from the third ventricle and hypothalamic area. Sharp dissection was required to remove the tumor and diaphragm from under the left optic nerve. The tumor was involved with the diaphragm and extended into the sella. The intrasellar tissue was fibrotic and excised along with the capsule and diaphragm. Gross total removal was achieved.
The defect was closed with an intracranial fat graft. The fat was supported anteriorly over the planum sphenoidale defect by allowing it to fold over a piece of septal cartilage placed transversely epidurally in the frontal region. A second piece of fat was placed within the sella, and a piece of cartilage was placed epidurally between the frontal and sellar grafts. The defect was covered with Gelfoam® soaked in thrombin and Tisseel® (Baxter Healthcare Corp., Deerfield, IL). The sphenoid sinus was packed with fat, and the Gelfoam® was placed over the fat at the level of the anterior wall of the sphenoid sinus. Postoperatively, a lumbar drain was placed and allowed to drain at 10 ml/hr for 5 days. There was no postoperative cerebrospinal fluid leak. Hydrocephalus recurred, requiring a VP shunt, presumably due to failure of the bilateral VP shunts. Postoperatively, the patient recovered well. Vision in the right eye did not recover, but vision in his left eye improved to 20/70. Postoperative MR imaging (Fig. 2) showed no residual tumor. The patient continues to require the endocrinologic replacement therapy that was needed before surgery but otherwise is neurologically intact.