Repeated transsphenoidal surgery to treat recurrent or residual pituitary adenoma

Transcription

1 J Neurosurg 102: , 2005 Repeated transsphenoidal surgery to treat recurrent or residual pituitary adenoma RONALD J. BENVENISTE, M.D., PH.D., WESLEY A. KING, M.D., JANE WALSH, R.N., N.P., JACOB S. LEE, M.D., BRADLEY N. DELMAN, M.D., AND KALMON D. POST, M.D. Departments of Neurosurgery and Radiology, Mt. Sinai School of Medicine, New York, New York Object. In this paper the authors describe the indications for and the results and complications of repeated transsphenoidal surgery (RTSS) to treat recurrent or residual pituitary adenoma. Methods. A retrospective review was conducted of 96 consecutive patients who underwent RTSS to treat recurrent or residual pituitary adenoma. Ninety-six patients underwent RTSS: 42 to treat a recurrent or residual pituitary mass and 54 to treat a recurrent or persistent hormone hypersecretion. There was no case of perioperative death and there was a 1% incidence of major complications. Postoperative endocrinological deficiencies were uncommon unless planned total hypophysectomy was performed to treat Cushing disease. Clinical remission occurred in 93% of patients undergoing RTSS to treat a tumor mass, and 15% of patients initially experienced remission only to face a relapse after a mean of 32 months. Endocrinological remission occurred in 57% of patients undergoing RTSS to treat hormone hypersecretion; most of these patients had Cushing disease. Thirty-five percent of patients with an initial endocrinological remission experienced a relapse of their symptoms after a mean of 31 months (thus, 37% of patients achieved sustained endocrinological remission). We failed to identify factors that accurately predicted initial symptom remission or delayed relapse following RTSS. Ten patients in our series eventually underwent a third transsphenoidal surgery without major complications. Conclusions. Repeated transsphenoidal surgery is a more effective treatment for recurrent or residual mass than it is for hormone hypersecretion and has acceptable rates of morbidity and mortality. If hypophysectomy is not performed, endocrinological deficiencies are unlikely following RTSS. KEY WORDS pituitary adenoma recurrence transsphenoidal surgery Cushing disease complication T RANSSPHENOIDAL resection of a pituitary adenoma is one of the most commonly performed surgeries for intracranial tumors. 7 Outcomes following an initial TSA are generally excellent, with high rates of clinical improvement and endocrinological remission, and minimal rates of morbidity and mortality. Unfortunately, a significant proportion of both nonfunctioning and hormone-secreting pituitary adenomas recur following a successful TSA; in addition, a residual tumor is often left behind following this procedure. Residual or recurrent tumors can be treated with observation, medical therapy, radiotherapy, stereotactic radiosurgery, or repeated surgery (TSA or craniotomy). Each of these treatment options has advantages and disadvantages; in some cases, more than one treatment option may be a reasonable choice. Abbreviations used in this paper: ACTH = adrenocorticotropic hormone; CSF = cerebrospinal fluid; DI = diabetes insipidus; GH = growth hormone; IPSS = inferior petrosal sinus sampling; MR = magnetic resonance; NFPA = nonfunctioning pituitary adenoma; RTSS = repeated transsphenoidal surgery; TSA = transsphenoidal adenectomy Like most other surgical procedures, an RTSS is generally more difficult to perform than the initial operation. Obliteration of surgical landmarks, formation of scar tissue, and the effects of preoperative radiotherapy or medical treatment may contribute to the difficulty encountered in an RTSS. Investigators in earlier studies suggested that complications occur relatively frequently following an RTSS and the rates of clinical remission are lower than those observed following a primary TSA. 11,13,31 The difficulties associated with an RTSS and the growing evidence of the safety and efficacy of SRS have made the latter modality the treatment of choice for many patients with residual or recurrent pituitary adenoma, such as those in whom the adenoma is predominantly located in the cavernous sinus. In other patients, such as those with deteriorating vision due to chiasmal compression, RTSS is the preferred option. In a third group of patients, more than one treatment option may be appropriate; for example, RTSS may be chosen by patients in whom the tumor abuts the optic chiasm, who wish to avoid the risks of radiotherapy, or by patients who wish to avoid the risk of hypopituitarism associated with stereotactic radiosurgery.

2 Repeated transsphenoidal surgery for pituitary adenomas The purpose of our study was to determine the complications and clinical outcomes associated with RTSS in a contemporary series. In recent years, technological advances such as MR imaging and IPSS have improved the preoperative assessment of patients with recurrent or residual pituitary adenomas. At the same time, with tremendous improvements in stereotactic radiosurgery, the patient population considered for RTSS has changed significantly. Patient Selection Clinical Material and Methods Approval was obtained from the Mount Sinai School of Medicine Institutional Review Board prior to commencing this retrospective study. We reviewed a database including more than 1695 TSAs for pituitary adenoma performed by the senior author (K.D.P.) between 1976 and Patients in whom the following four criteria were met were included in our study: 1) performance of an initial TSA for confirmed or suspected pituitary tumor; 2) performance of a repeated operation for a recurrent or residual tumor mass or a recurrent or persistent endocrine hypersecretion; 3) attainment of a confirmed histopathological diagnosis in at least one TSA; and 4) performance of at least one of the TSAs by the authors. Ninety-six patients meeting these criteria were identified. We reviewed all available charts, laboratory results, and imaging studies for these patients. Whenever possible, the imaging studies were retrospectively and independently reviewed by two fellowship-trained neuroradiologists at our institution (J.S.L. and B.N.D.) who were blinded to the patients clinical histories; any disagreements were resolved by consensus. In other cases, data were obtained from contemporary reports in the patients charts. Patients were contacted when necessary to request follow-up information or imaging studies. Surgical Technique Standard transsphenoidal surgery was performed in all patients. The sublabial approach was used exclusively during the earlier years covered by this study; more recently, the authors used the endonasal transseptal approach when feasible. 34 Lumbar drains were placed preoperatively in selected cases. The operating microscope was used during the approach in all cases. Before opening the sellar floor, intraoperative fluoroscopy was used to ensure that the approach was directed beneath the sella turcica. Intraoperative MR imaging or ultrasonography and endoscopy were not used, but frameless stereotactic guidance was used in a small number of cases. The sellar floor was reconstructed after tumor resection whenever possible; sellar fat packing was used only when the arachnoid was violated intraoperatively or when there was significant destruction of the sellar architecture and anticipated postoperative radiation therapy. Lumbar drains were left in postoperatively only when significant intraoperative CSF leakage occurred. Follow-Up Review All patients were monitored closely postoperatively for signs of DI, CSF leakage, neurological deterioration, and endocrine insufficiency. After discharge, most patients with nonfunctioning pituitary adenomas underwent repeated MR or computerized tomography imaging after approximately 3 months. Whenever possible, these patients also underwent serial imaging studies at intervals of 1 to several years following their surgeries. Most patients with hormone-secreting tumors received close biochemical follow up in cooperation with an endocrinologist. Remission of Cushing disease was defined as normal 24-hour urine cortisol or plasma ACTH levels, or both. Remission of acromegaly was defined as normal age-appropriate insulin-like growth factor I and/or GH levels. An oral glucose tolerance test was performed in some, but not all, patients who had acromegaly. Further Treatment Patients were referred for additional postoperative adjunctive treatment as needed. These treatments included medical therapy (for example, octreotide therapy for acromegaly or ketoconazole therapy for Cushing disease), stereotactic radiosurgery, radiotherapy, and bilateral adrenalectomy for refractory Cushing disease. Statistical Analysis Statistical program analysis (SPSS Student Version 12.0 for Windows; SPSS, Inc., Chicago, IL) was performed using commercially available software. The chi-square and Fisher exact tests were used for a comparison of categorical variables as appropriate, with one- or two-tailed tests selected depending on whether the direction of association could be predicted. The Student t-test and the Mann Whitney U- test were used for a comparison of continuous variables as appropriate. A probability value of 0.05 was selected as denoting statistical significance. Results Patient Characteristics at the Time of the Second TSA The characteristics of patients are summarized in Table 1. Ninety-six patients were included in our study. Of the 42 patients who underwent RTSS to treat a tumor mass, 12 (29%) underwent the procedure for a residual tumor, defined here as sufficient tumor on initial postoperative imaging following the first TSA to warrant a repeated operation. Thirty patients (71%) underwent RTSS for tumor recurrence, defined here as a regrowth of the tumor after an initial postoperative imaging study had shown adequate tumor decompression. Of the patients with residual tumor following the first surgery, one patient underwent radiotherapy between the first and second TSA. Forty-four (81%) of 54 patients who underwent RTSS for hormone hypersecretion had Cushing disease. Among the 54 patients who underwent RTSS for hormone hypersecretion, 29 (54%) had persistent hypersecretion following the first TSA, 24 (44%) had recurrent hypersecretion following remission after a first TSA, and one (1.9%) began to experience new Cushing disease after previous resection of a Rathke cleft cyst. Twenty-three patients had subjective or objective visual loss as the indication for a repeated operation: 14 (61%) had bitemporal hemianopia (four of these 14 patients had defects extending into the nasal or central visual fields as well and two also had unilateral cavernous sinus syndrome with cranial nerve palsies and Horner syndrome); three (13%) 1005

3 R. J. Benveniste, et al. TABLE 1 Characteristics of 96 patients who underwent RTSS for pituitary adenoma* Characteristic No. of patients (%) age (yrs) (9 87) sex male 32 (33) female 64 (67) type of tumor nonfunctioning 37 (39) ACTH-secreting 44 (46) GH-secreting 7 (7) PRL-secreting 4 (4) GH- & PRL-secreting 3 (3) FSH- & LH-secreting 1 (1) indication for reop visual loss 23 (24) asymptomatic chiasmal 9 (9) compression other symptoms due to mass 4 (4) effect residual/recurrent tumor, 6 (6) no mass effect hormone hypersecretion 54 (56) previous treatment radiotherapy 6 (6) medical 8 (8) craniotomy 2 (2) tumor size (53 patients) microadenoma ( 1 cm) 22 (42) macroadenoma ( 1 cm) 31 (58) extracellular extension of tumor (76 patients) suprasellar 35 (46) cavernous sinus 21 (28) sphenoid sinus 5 (7) retrosellar cisterns 2 (3) 3rd ventricle/hypothalamus 2 (3) interval btwn 1st & 2nd TSA wks (2 wks 19 yrs) * FSH = follicle-stimulating hormone; LH = luteinizing hormone; PRL = prolactin; SD = standard deviation. Value is expressed as the mean SD (range). Six NFPAs demonstrated focal immunoreactivity to hormones other than ACTH and five were silent ACTH-positive tumors; two NFPAs were in patients with multiple endocrine neoplasia I. Patients in whom tumor size could be measured. Patients in whom extension could be assessed. had bilateral superior or inferior temporal quadrantanopias; three (13%) had unilateral quadrantanopias; one (4.3%) had normal visual fields with diminished central visual acuity; one (4.3%) had normal findings; and one patient s examination findings (4.3%) were not available. All 54 patients who were undergoing RTSS for hormone hypersecretion had normal findings on their visual examinations. Preoperative endocrinological deficits were present in nine patients (9.4%): six patients had panhypopituitarism after prior surgeries and three had isolated hormone deficiencies. Neuroimaging findings (size and extrasellar extension of the tumor) are also presented in Table 1; the interpretation of MR imaging studies was often difficult because of postoperative changes. Surgical Techniques TABLE 2 Surgical techniques and intraoperative findings in 96 patients who underwent RTSS for pituitary adenoma* Surgical techniques and intraoperative findings are summarized in Table 2. The sublabial approach was used somewhat more often for repeated operations for hormone hypersecretion, although this difference did not reach statistical significance. Partial and total hypophysectomies were performed for tumors producing ACTH hypersecretion but not for other types of tumors. Intraoperative CSF leakage and the use of postoperative lumbar drains occurred in similar proportions of surgeries for mass effect and for hormone hypersecretion. Complications Following RTSS No. of Patients (%) Factor Overall Reop for Mass Reop for HH p Value approach sublabial 60 (64) 22 (52) 38 (73) 0.06 endonasal 34 (36) 20 (48) 14 (27) unknown surgical intent adenomectomy/ 61 (64) 42 (100) 19 (36) debulking partial hypo- 25 (26) 0 25 (47) physectomy hypophysectomy 9 (9.5) 0 9 (17) unknown arachnoid violated yes 41 (44) 16 (38) 25 (49) 0.40 no 52 (56) 26 (62) 26 (51) unknown lumbar drain left in postop yes 13 (14) 5 (12) 8 (15) 0.89 no 82 (86) 37 (88) 45 (85) unknown * HH = hormone hypersecretion. For comparison between repeated operation for mass and repeated operation for hormone hypersecretion, chi-square test. Complications occurring after RTSS are summarized in Table 3. No perioperative deaths occurred. One patient suffered a major complication. She experienced repeated episodes of epistaxis following repeated operation for Cushing disease and required packing of the nasal cavity followed by operative ligation of a bleeding branch of the internal maxillary artery. Other nonendocrinological complications were uncommon and included several cases of sinusitis and one case of wound infection at a fat graft harvest site; all these complications were successfully treated with short courses of oral antibiotic agents. One patient experienced postoperative CSF rhinorrhea, which was successfully treated with 4 days of lumbar drainage. Endocrinological complications, including permanent DI and panhypopituitarism, were relatively frequent following hypophysectomy for Cushing disease but were uncommon following other transsphenoidal procedures (p = 0.05 for DI, p for anterior pituitary deficiencies). Immediate and Long-Term Outcomes Following RTSS for a Tumor Mass The immediate outcomes following RTSS to treat tumor mass are summarized in Table 4. Three patients (7.1%) 1006

4 Repeated transsphenoidal surgery for pituitary adenomas TABLE 3 Complications following RTSS for pituitary adenoma* Complication No. of Patients (%) p Value total major complications 1 (1) major epistaxis 1 (1) total minor complications 28 (29) transient DI 18 (19) new permanent DI hypophysectomy 2 (25) 0.05 other surgery 3 (3) new ant pituitary deficiency hypophysectomy 5 (63) other surgery 3 (3) new panhypopituitarism hypophysectomy 5 (63) other surgery 2 (2) sinusitis 5 (5) fat graft harvest site infection 1 (1) transient cranial nerve palsy 1 (1) CSF leak 1 (1) * Eight patients underwent hyperphysectomy and 88 patients other types of surgery. For comparison of hypophysectomy and other types of surgeries. received inadequate tumor decompression, which was manifested by a failure of their vision to improve following surgery and immediate postoperative imaging studies demonstrating significant amounts of residual tumor. Two of these patients underwent craniotomy on postoperative Days 1 and 6, respectively; neither noted an improvement in vision. The third patient underwent a third RTSS on postoperative Day 4, also without any improvement in vision. The majority of patients (39 [93%]) received adequate decompression as a result of the RTSS; this was manifested by an improvement in their symptoms and/or neuroimaging evidence of decompression. Improvement or resolution of presenting symptoms occurred in 56% of patients with visual loss and 67% of patients with headaches. The two patients with preoperative cranial nerve palsies did not improve. The one patient presenting with sinus congestion due to a regrowth of tumor in the sphenoid sinus improved significantly. Of 39 patients who received adequate decompression, 33 (85%) had no relapse over a mean follow-up period of 32 months (range months), with relapse defined as either neuroimaging findings of regrowth of tumor or recurrence of symptoms. Six patients experienced a relapse at a mean of 31 months postoperatively. One patient began to exhibit an altered mental status and a new abducent nerve palsy 4 months postoperatively, with MR images demonstrating a massive regrowth of tumor. Two patients were found to have asymptomatic regrowth of tumor in the cavernous sinus at 5 and 48 months postoperatively. Two patients had asymptomatic regrowth of tumor in the sellar region at 18 and 38 months postoperatively; in one patient, the regrowth extended into the suprasellar cistern. One patient experienced new sinus congestion 73 months postoperatively; in that case MR imaging revealed regrowth of tumor in the sphenoid sinus. One patient with regrowth of tumor in the cavernous sinus underwent stereotactic radiosurgery and the other underwent radiotherapy. Three patients underwent a third TSA (see later section). The patient with asymptomatic regrowth of tumor confined to the sella turcica did not undergo further treatment. TABLE 4 Outcomes following RTSS for a tumor mass Outcome & Other Factors No. of Patients (%) immediate outcome inadequate decompression 3 (7) adequate decompression 39 (93) visual loss (23 patients) resolved 4 (17) improved 9 (39) unchanged 10 (43) headache (7 patients) resolved 4 (67) unchanged 2 (33) unknown 1 duration of stay (days)* (2 9) finding on postop imaging gross-total resection 10 (26) residual 28 (72) increased size 1 (3) no postop imaging 3 long-term outcome (39 patients) delayed relapse 6 (15) no relapse 33 (85) postop therapy (42 patients) repeated op 6 (14) stereotactic radiosurgery 1 (2) radiotherapy 8 (19) * Values are expressed as the mean SD (range). At a mean follow up of 31 months. At a mean follow up of 32 months. Four patients underwent TSA and two craniotomy. Factors Predicting Immediate and Long-Term Outcomes Following RTSS for a Tumor Mass Preoperative treatment with radiotherapy or bromocriptine therapy was not correlated with inadequate tumor decompression: one (12%) of eight patients who had previously undergone radiotherapy or bromocriptine therapy received inadequate decompression compared with two (5.9%) of 34 patients who had not undergone prior radiotherapy or bromocriptine therapy (p = 0.48, Fisher exact test). Preoperative tumor measurements could be calculated for only one of the three patients with inadequate decompression; therefore, we could not assess the effect of tumor size on immediate surgical outcome. Effects of various factors on the incidence of relapse are summarized in Table 5. No factors that we assessed showed statistically significant associations with relapse. Immediate and Long-Term Outcomes Following RTSS for Hormone Hypersecretion Outcomes following surgery for hormone hypersecretion are summarized in Table 6. Postoperative endocrinological data were not available for five patients. Of the remaining 49 patients, 28 (57%) initially experienced biochemical remissions. The effects of various factors on the likelihood of remission are summarized in Table 7. Neither the presence of tumor in pathology specimens obtained during surgery, the presence of identifiable tumor on preoperative MR images, nor the presence or absence of Cushing disease, compared with other types of pituitary adenoma, affected the likelihood of remission. For patients with Cushing disease, the preoperative IPSS and the performance of hypophysectomy rather than a more conservative procedure 1007

5 R. J. Benveniste, et al. TABLE 5 Factors predicting a delayed relapse following TSR for a tumor mass Factor Relapse No Relapse p Value no. of patients w/ postop radiotherapy (%) yes 1 (12) 7 (88) 0.64* no 5 (16) 26 (84) tumor vol (cm 3 ) interval btwn 1st & 2nd TSA (mos) patient age at 2nd TSA (yrs) no. of patients w/ cavernous sinus invasion on MRI (%) yes 4 (24) 13 (76) 0.28* no 2 (11) 17 (89) no. of patients w/ silent ACTH (%) yes 0 (0) 4 (100) 0.60* no 6 (17) 29 (83) * Fisher exact test. Values are expressed as means SDs. Mann Whitney U-test. Student t-test. did not affect the likelihood of remission. Thirty-five percent of patients who initially experienced biochemical remission suffered a relapse after a mean of 31 months postoperatively. The effects of several variables on the likelihood of relapse following an initial remission are summarized in Table 8. A nonsignificant trend toward a lower likelihood of relapse was found when patients with Cushing disease underwent hypophysectomy, compared with more conservative procedures. TABLE 6 Outcomes following RTSS for hormone hypersecretion Outcomes & Other Factors No. of Patients (%) immediate outcome remission 28 (57) no remission 21 (43) unknown 5 duration of stay (days)* (2 9) long-term outcome following initial remission relapse 8 (35) no relapse 15 (65) unknown 5 postop therapy (49 patients) reop TSA 6 (12) adrenalectomy 2 (4) stereotactic radiosurgery 6 (12) radiotherapy 6 (12) medical 7 (14) unknown 4 * Values are expressed as the mean SD (range). At a mean follow up of 31 months. At a mean follow up of 42 months. TABLE 7 Factors predicting remission following RTSS for hormone hypersecretion No. of Patients (%) Factor Remission No Remission p Value pathological findings at 2nd TSA present 18 (62) 11 (38) 0.38* absent 8 (44) 10 (56) Cushing disease present 24 (57) 18 (43) 1.0 absent 4 (57) 3 (43) visible tumor on preop MRI present 8 (53) 7 (47) 0.73* absent 7 (54) 6 (46) IPSS preop (Cushing disease only) IPSS 6 (50) 6 (50) 0.81* no IPSS 18 (60) 12 (40) procedure (Cushing disease only) hypophysectomy 5 (62) 3 (38) 0.53 other 19 (56) 15 (44) * Chi-square test. Fisher exact test. Third TSA for Treatment of Recurrent or Persistent Pituitary Adenoma Ten patients in our study underwent third TSAs. The indications for and results of this procedure are summarized in Table 9. No major complications occurred after the third TSA; panhypopituitarism occurred in two patients following hypophysectomies for treatment of Cushing disease, with permanent DI occurring in one of these patients. Two of the three patients who underwent third TSAs to treat a symptomatic tumor mass had visual loss. Improvement in preoperative visual deficits occurred in only one of these; TABLE 8 Factors predicting relapse following initial remission after RTSS for hormone hypersecretion No. of Patients (%) Factor Relapse No Relapse p Value* pathological findings at 2nd TSA present 5 (28) 13 (72) 0.74 absent 2 (25) 6 (75) Cushing disease present 6 (25) 18 (75) 0.55 absent 2 (50) 2 (50) visible tumor on preop MRI present 2 (25) 6 (75) 0.55 absent 1 (14) 6 (86) IPSS preop (Cushing disease only) IPSS 1 (17) 5 (83) 0.52 no IPSS 5 (28) 13 (72) procedure (Cushing disease only) hypophysectomy 0 (0) 5 (100) 0.20 other 6 (32) 13 (68) * Fisher exact test. 1008

6 Repeated transsphenoidal surgery for pituitary adenomas TABLE 9 Characteristics of patients who underwent a third TSA for residual or recurrent pituitary adenoma* Case Age (yrs), Path Mos Since Compli- Initial No. Sex Finding 2nd TSA Indications for Surgery Procedure cations Remission Follow Up 1 66, F PRLN 6 rec tumor, visual loss, adenomectomy none yes recurrence 2 mos postop, abducent nerve required 4th TSA w/ palsy, lethargy Gliadel wafer placement; died of disease 2 56, F NFPA 75 rec tumor, sinus symp adenomectomy none yes no relapse at 8 mos 3 44, F NFPA 4 days pers tumor, visual loss adenomectomy none no unchanged vision at 1 mo 4 14, M ACTH 78 Nelson syndrome post- sellar biopsy none yes no tumor growth at 30 mos adrenalectomy (following radiotherapy) 5 21, F ACTH 24 pers tumor, HS adenomectomy none yes further follow-up data unavailable 6 49, F ACTH 2 pers tumor, HS hypophysectomy perm DI, no ND PHP 7 36, F ACTH 15 pers tumor, HS partial hypophys- trans DI yes remission at 4 mos ectomy 8 45, F ACTH 7 pers tumor, HS sellar exploration none no required radiotherapy & ketoconazole treatment 9 45, F ACTH 43 rec tumor, HS hypophysectomy PHP yes remission at 2 mos 10 47, M NFPA 40 rec tumor, asymp adenomectomy none ND further follow-up data unavailable * Asymp = asymptomatic; HS = hypersecretion; ND = no data; path = pathological; perm = permanent; pers = persistent; PHP = panhypopituitarism; PRLN = prolactinoma; symp = symptoms; rec = recurrent; trans = transient. the patient eventually died of an extremely aggressive prolactinoma. Outcomes following third TSAs for persistent or recurrent Cushing disease were more encouraging. Among six patients with Cushing disease who underwent third TSAs, four patients (67%) had initial remissions. Further endocrinological follow-up review was not available for one of these patients, who underwent her third TSA at another institution; the other three patients had sustained remission at 2, 4, and 30 months of follow up. Discussion Complications Following RTSS for a Recurrent or Residual Pituitary Adenoma Nonoperative treatment options for recurrent or residual pituitary adenoma may include observation, medical therapy, radiotherapy, and stereotactic radiosurgery. The choice of which option(s) to choose for an individual patient must take into account the potential complications of each option. Radiotherapy carries a significant risk of hypopituitarism and, rarely, damage to the optic apparatus or medial temporal lobe structures or induction of secondary gliomas or meningiomas. 4,19 In appropriately selected patients, stereotactic radiosurgery is less likely to cause these complications. Nonendocrinological complications of this treatment modality are rare, if the radiation dose to the optic chiasm is less than approximately 9 Gy. In published studies of stereotactic radiosurgery for pituitary adenoma, nearly all patients had recurrent or persistent disease following a previous TSA. The incidence of visual deterioration in these series varied from 0% (most commonly) to 7%. 9,16,20,22,26,33 Other complications, such as transient cranial nerve palsies and secondary tumors, were also rare. In contrast, stereotactic radiosurgery for pituitary tumors carries a significant risk of hypopituitarism. Up to 41% of patients experience anterior pituitary hormone deficiencies following stereotactic radiosurgery in this setting, but new cases of DI have not been reported. 3,16,22 In our series, one serious surgical complication, delayed postoperative epistaxis requiring ligation of the internal maxillary artery, occurred among 96 patients undergoing RTSS (1% of patients). Severe, delayed epistaxis as a result of internal maxillary or sphenopalatine artery injury during endonasal RTSS has been previously reported. 2 Visual deterioration did not occur in any patient. Other minor nonendocrinological complications occurred in 8.3% of patients. Endocrinological complications such as DI or anterior pituitary hormone deficiencies were uncommon (3.4 and 3.6%, respectively) following second transsphenoidal procedures other than hypophysectomy. These complication rates are somewhat lower than rates previously described by Laws, et al., 13 for transsphenoidal surgery following unsuccessful prior treatment; however, their series included many patients undergoing repeated surgery for a CSF leak or for more aggressive tumors such as chordoma and craniopharyngioma, which could be associated with higher complication rates. Investigators in two small series of patients undergoing RTSS for hormone-secreting adenomas reported significantly higher complication rates. Ram, et al., 24 reported no major complications but an 18% incidence of postoperative CSF leak among 17 patients undergoing early RTSS for Cushing disease. In the same study, the incidence of new anterior pituitary deficiencies was 67% following hypophysectomy and 36% following other procedures, whereas the incidence of new permanent diabetes insipidus was 33 and 14%, respectively. Long and associates 14 reported major nonendocrinological complications in three (19%) of 16 patients undergoing RTSS for acromegaly; the incidence of new anterior pituitary deficiencies was 62.5%. In more recent years, no major complications were reported in two small series of patients undergoing RTSS for pituitary adenoma. 10,12 Our rates of nonendocrinological complications were comparable to those observed in published series 1009

7 R. J. Benveniste, et al. of primary transsphenoidal surgery for pituitary adenoma, although permanent DI occurred more frequently in our series (compared with 0.4% after primary transsphenoidal surgery). 1 Several groups report the use of various image guidance techniques to identify the midline so that the risk of surgical complications can be minimized during RTSS. 6,12,17,18 Although we report low complication rates without the use of image guidance other than lateral intraoperative fluoroscopy, more sophisticated intraoperative image-guidance techniques may help accurately identify the midline and should minimize the risk of vascular injury. We conclude that compared with stereotactic radiosurgery, RTSS for residual or recurrent pituitary adenoma is associated with a slightly higher incidence of nonendocrinological complications; however, the risk of endocrinological complications is considerably lower. Outcomes Following RTSS for Recurrent or Residual Tumor Mass In our series, most of the tumors requiring a repeated operation to treat a residual or recurrent mass were NFPAs. The rate of symptomatic recurrence of an NFPA following an initial TSA ranges between 6 and 21% in long-term studies. The recurrence of a hormone-secreting tumor usually does not result in a mass effect 30 because the patient presents with hormonal symptoms before the tumor has grown large enough to cause a mass effect. Several treatment options are available to treat a recurrent or residual NFPA. Choices include observation, radiotherapy, stereotactic radiosurgery, and repeated TSA or craniotomy. Observation including serial MR imaging and visual field examinations may be the most appropriate case management option for selected asymptomatic patients in whom there is no compression of the optic chiasm, such as elderly patients, in whom residual NFPAs grow relatively slowly. 15,29 Radiotherapy is effective in controlling NFPA growth in 75 to 100% of selected patients, but the incidence of visual or neurological complications is higher than that associated with stereotactic radiosurgery. 4,19 For this reason, radiotherapy is generally used only when the proximity of the tumor to the optic apparatus makes stereotactic radiosurgery unsafe. Stereotactic radiosurgery has become the treatment of choice for the recurrent or residual pituitary tumor mass in many clinical scenarios. It controls tumor growth in 86 to 100% of pituitary adenomas, according to reports of recent series. 9,16,20,22,26,33 In our series, second TSAs performed to treat a residual or recurrent mass were initially successful, as defined by an improvement in visual deficits or neuroimaging evidence of adequate decompression of the optic apparatus (93% of patients). Among the patients who underwent a second TSA that was initially successful, significant tumor regrowth was not evident in 85% at a mean of 32 months postoperatively. We failed to identify factors that predicted tumor regrowth, except for a nonsignificant trend toward an increased risk of regrowth with larger tumors. We acknowledge that the relatively small number of patients in our study limits its power to detect risk factors for regrowth. We conclude that RTSS is less effective than stereotactic radiosurgery at controlling tumor growth. Therefore, stereotactic radiosurgery may be the preferred option to treat a residual or recurrent adenoma that is not abutting the optic chiasm, if the increased risk of hypopituitarism is acceptable. Repeated transsphenoidal surgery is still the best option for treating residual or recurrent pituitary adenomas that abut or compress the optic chiasm, because of its ability to provide decompression of the optic apparatus and because the risk of radiation-induced optic neuropathy is high in this setting. Nevertheless, new radiosurgical techniques, such as staged radiotherapy, may eventually allow stereotactic radiosurgery to be used in this setting. 21 In our series, 75% of patients undergoing RTSS for a tumor mass had residual tumor, as demonstrated on MR images following the RTSS. Only three (12%) of these patients, however, had regrowth during the follow-up period. Therefore, we suggest that further therapy, such as stereotactic radiosurgery or radiotherapy, should be reserved for documented tumor regrowth. In this situation, frequent follow-up imaging studies are necessary so that if tumor regrowth does occur, the tumor can be treated before it abuts the optic chiasm and stereotactic radiosurgery becomes unsafe. Outcomes After RTSS for Recurrent or Residual Hormone Hypersecretion In our series, 81% of the patients who underwent RTSS for hormone hypersecretion had Cushing disease. The preponderance of patients with Cushing disease among patients undergoing RTSS may be due to the fact that better medical treatments exist for treating acromegaly and hyperprolactinemia. Options for treating persistent or recurrent Cushing disease include the suppression of cortisol secretion by using medications or bilateral adrenalectomy, but these treatments may result in Nelson syndrome. Definitive treatments for the pituitary tumor include radiotherapy, stereotactic radiosurgery, and repeated surgery. Radiotherapy results in endocrinological remission in 50 to 90% of unselected patients with Cushing disease, but requires up to 2 years before the therapeutic effect is apparent. In addition, radiotherapy carries a significant risk of hypopituitarism and neurological complications (see earlier discussion). 5,19,28 Stereotactic radiosurgery is increasingly recognized as a safe and effective option for recurrent or persistent Cushing disease 5 and probably provides similar outcomes in patients with other types of hormone-secreting pituitary adenomas. 23 Remission rates in this setting range from 63 to 100% if sufficient radiation doses are delivered, and a delayed relapse occurs in 11 to 20% of patients following remission. 8,25,32 In our series, initial remission following RTSS was achieved in 57% of patients with hormone-secreting tumors; 35% of these patients subsequently experienced a relapse during the follow-up period, leaving a sustained remission rate of 37%. Among the subset of patients with Cushing disease, initial remission was achieved in 57%, and 25% experienced a relapse during the follow-up period, leaving a sustained remission rate of 43%. Neither the initial nor the sustained remission rates were significantly different among all patients with hormone-secreting tumors and the subset of patients with Cushing disease (p = 1.0 for the comparison of initial remission rates and p = 0.55 for the comparison of relapse rates, Fisher exact test). Nevertheless, we acknowledge that the small number of patients with noncorticotroph tumors limits the power of our study to detect differences between outcomes for corticotroph and noncorticotroph tumors. Previous reports of series of 1010

8 Repeated transsphenoidal surgery for pituitary adenomas patients who underwent RTSS for hormone-secreting tumors provided initial remission rates of 62 to 71% for patients with Cushing disease, 24,27 and sustained remission rates of 31% for patients with acromegaly 14 and 53% for patients with Cushing disease. 24 The higher remission rates in these studies were accompanied by a much higher incidence of endocrinological and nonendocrinological complications (see earlier discussion). More recently, Kurosaki, et al., 10 reported the use of intraoperative GH measurements to guide resection of residual or recurrent GH-secreting adenomas, with a sustained remission rate of 59% attributed to surgery alone. The remission and relapse rates achieved with stereotactic radiosurgery compare favorably with those in our series and in those previously published reports of RTSS, indicating that in general, stereotactic radiosurgery may be the preferred option for patients with recurrent or persistent Cushing disease after an initial TSA. Stereotactic radiosurgery becomes an even more preferable treatment if surgical hypophysectomy is contemplated because we found that hypophysectomy was associated with a very high likelihood of panhypopituitarism and DI, and did not significantly increase the likelihood of endocrinological cure. We attempted to find factors that predicted remission and relapse following RTSS in patients with hormone-secreting tumors, to identify subsets of patients in whom RTSS would be more likely to succeed, sparing them the risk of hypopituitarism associated with stereotactic radiosurgery; however, we were unable to identify any such factors, probably due to the small size of our patient population. Conclusions Nonendocrinological complications observed after RTSS has been performed to treat recurrent or residual pituitary adenoma are rare, despite the technical difficulty of repeated operation. Endocrinological complications are also uncommon, except after planned total hypophysectomy. Repeated transsphenoidal surgery to treat recurrent or residual tumor mass is associated with a 93% rate of clinical remission and a 15% rate of tumor regrowth over 31 months of follow-up review. In contrast, RTSS to treat recurrent or persistent hormone hypersecretion results in only a 57% rate of initial endocrinological remission, and a 37% likelihood of sustained endocrinological remission at a mean of 31 months. Although RTSS may be the preferred option for treating residual or recurrent adenomas that cause a mass effect, stereotactic radiosurgery may be a better treatment option for persistent or recurrent hormone hypersecretion. References 1. Black PM, Zervas NT, Candia GL: Incidence and management of complications of transsphenoidal operation for pituitary adenomas. Neurosurgery 20: , Cockroft KM, Carew JF, Trost D, Fraser RAR: Delayed epistaxis resulting from external carotid artery injury requiring embolization: a rare complication of transsphenoidal surgery: case report. Neurosurgery 47: , Feigl GC, Bonelli CM, Berghold A, Mokry M: Effects of gamma knife radiosurgery of pituitary adenomas on pituitary function. J Neurosurg 97 (Suppl 5): , Flickinger JC, Kondziolka D: Fractionated radiotherapy for pituitary tumors, in Winn HR (ed): Youmans Neurological Surgery, ed 5. Philadelphia: Saunders, 2003, pp Hentschel SJ, McCutcheon IE: Stereotactic radiosurgery for Cushing s disease. Neurosurg Focus 16(4):E5, Jane JA Jr, Thapar K, Alden TD, Laws ER Jr: Fluoroscopic frameless stereotaxy for transsphenoidal surgery. Neurosurgery 48: , Jane JA Jr, Thapar K, Kaptain GJ, Maartens N, Laws ER Jr: Pituitary surgery: transsphenoidal approach. Neurosurgery 51: , Kobayashi T, Kida Y, Mori Y: Gamma knife radiosurgery in the treatment of Cushing disease: long-term results. J Neurosurg 97 (Suppl 5): , Kondziolka D, Nathoo N, Flickinger JC, Niranjan A, Maitz AH, Lunsford LD: Long-term results after radiosurgery for benign intracranial tumors. Neurosurgery 53: , Kurosaki M, Luedecke DK, Abe T: Effectiveness of secondary transnasal surgery in GH-secreting pituitary macroadenomas. Endocr J 50: , Landolt AM: Surgical management of recurrent pituitary tumors, in Schmidek HH (ed): Operative Neurosurgical Techniques, ed 4. Philadelphia: WB Saunders, 2000 pp Lasio G, Ferroli P, Felisati G, Broggi G: Image-guided endoscopic transnasal removal of recurrent pituitary adenomas. Neurosurgery 51: , Laws ER Jr, Fode NC, Redmond MJ: Transsphenoidal surgery following unsuccessful prior therapy. An assessment of benefits and risks in 158 patients. J Neurosurg 63: , Long H, Beauregard H, Somma M, Comtois R, Serri O, Hardy J: Surgical outcome after repeated transsphenoidal surgery in acromegaly. J Neurosurg 85: , Losa M, Franzin A, Mangili F, Terreni MR, Barzaghi R, Veglia F, et al: Proliferation index of nonfunctioning pituitary adenomas: correlations with clinical characteristics and long-term follow-up results. Neurosurgery 47: , Losa M, Valle M, Mortini P, Franzin A, da Passano CF, Cenzato M, et al: Gamma knife surgery for treatment of residual nonfunctioning pituitary adenomas after surgical debulking. J Neurosurg 100: , Ohhashi G, Kamio M, Abe T, Otori N, Haruna S: Endoscopic transnasal approach to the pituitary lesions using navigation system (InstaTrak System). Acta Neurochir 143: , Pergolizzi RS, Nabavi A, Schwartz RB, Hsu L, Wong TZ, Martin C, et al: Intra-operative MR guidance during trans-sphenoidal pituitary resection: preliminary results. J Magn Reson Imaging 13: , Petrovich Z, Jozsef G, Yu C, Apuzzo ML: Radiotherapy and stereotactic radiosurgery for pituitary tumors. Neurosurg Clin N Am 14: , Petrovich Z, Yu C, Giannotta SL, Zee CS, Apuzzo ML: Gamma knife radiosurgery for pituitary adenoma: early results. Neurosurgery 53:52 61, Pham CJ, Chang SD, Gibbs IC, Jones P, Heilbrun MP, Adler JR Jr: Preliminary visual field preservation after staged CyberKnife radiosurgery for perioptic lesions. Neurosurgery 54: , Pollock BE, Carpenter PC: Stereotactic radiosurgery as an alternative to fractionated radiotherapy for patients with recurrent or residual nonfunctioning pituitary adenomas. Neurosurgery 53: , Pollock BE, Nippoldt TB, Stafford SL, Foote RL, Abboud CF: Results of stereotactic radiosurgery in patients with hormone-producing pituitary adenomas: factors associated with endocrine normalization. J Neurosurg 97: , Ram Z, Nieman LK, Cutler GB Jr, Chrousos GP, Doppman JL, Oldfield EH: Early repeat surgery for persistent Cushing s disease. J Neurosurg 80:37 45, Sheehan JM, Vance ML, Sheehan JP, Ellegala DB, Laws ER Jr: Radiosurgery for Cushing s disease after failed transsphenoidal surgery. J Neurosurg 93: , Sheehan JP, Kondziolka D, Flickinger J, Lunsford LD: Radiosur- 1011

9 R. J. Benveniste, et al. gery for residual or recurrent nonfunctioning pituitary adenoma. J Neurosurg 97 (Suppl 5): , Shimon I, Ram Z, Cohen ZR, Hadani M: Transsphenoidal surgery for Cushing s disease: endocrinological follow-up monitoring of 82 patients. Neurosurgery 51:57 62, Swearingen B, Zervas NT: Surgical management of Cushing s disease, in Schmidek HH (ed): Operative Neurosurgical Techniques, ed 4. Philadelphia: Saunders, pp Tanaka Y, Hongo K, Tada T, Sakai K, Kakizawa Y, Kobayashi S: Growth pattern and rate in residual nonfunctioning pituitary adenomas: correlations among tumor volume doubling time, patient age, and MIB-1 index. J Neurosurg 98: , Thapar K, Laws ER Jr: Pituitary tumors, functioning and nonfunctioning, in Winn HR (ed): Youmans Neurological Surgery, ed 5. Philadelphia: Saunders, 2003, pp Thapar K, Laws ER Jr: Transsphenoidal surgery for recurrent pituitary tumors, in Kaye AH, Black PMcLB (eds): Operative Neurosurgery. London: Churchill Livingstone, pp Wong GK, Leung CH, Chiu KW, Ma R, Cockram CS, Lam MJ, et al: LINAC radiosurgery in recurrent Cushing s disease after transsphenoidal surgery: a series of 5 cases. Min Invasive Neurosurg 46: , Wowra B, Stummer W: Efficacy of gamma knife radiosurgery for nonfunctioning pituitary adenomas: a quantitative follow up with magnetic resonance imaging-based volumetric analysis. J Neurosurg 97 (Suppl 5): , Zada G, Kelly DF, Cohan P, Wang C, Swerdloff R: Endonasal transsphenoidal approach for pituitary adenomas and other sellar lesions: an assessment of efficacy, safety, and patient impressions. J Neurosurg 98: , 2003 Manuscript received August 4, Accepted in final form January 25, Address reprint requests to: Kalmon Post, M.D., Department of Neurosurgery, Annenberg 8th floor, Mount Sinai School of Medicine, 1 Gustave Levy Place, New York, New York kalmon.post@mssm.edu. 1012