- Email: [email protected]
Pathological correlates of pituitary adenomas presenting with apoplexy
Pathological Correlates of Pituitary Adenomas Presenting With Apoplexy BEI-i-E K, KLEINSCHMIDT-DEMASTERS, MD AND KEVIN O, LILLEHEI, MD Pituitary adenomas presenting de novo with apoplectic symptoms are uncommon. Several series on pituitary adenomas with apoplexy have been reported but have emphasized clinical aspects, incidence, or radiographic appearance by computed tomography (CT) scan. We reviewed our 13 years' experience, focusing on pathological appearance. We identified 15 patients with pituitary adenomas whose first presentation to o u r institution was with apoplexy. By combining histological and intraoperative findings with more sensitive magnetic resonance imaging (MRI) scans, we were able to attribute 5 of 15 cases to bland infarction, five cases to hemorrhagic infarction, four cases to pure clot, and one to a remote hemorrhagic infarction leaving extensive residual fibrosis and hemosiderin. Despite extensive tumor necrosis, reticulin staining still allowed identification of specimens as adenomas. Immunohistochemical staining was informative in 13 of 15 cases and endocrine-inactive, weak gonado-
troph and null cell adenomas predominated. Cases with the greatest interval between symptom onset and surgery showed peripheral rim enhancement by MRI. Pathologically, this corresponded to granulation tissue, T cell lymphocytic infiltration, and atrophic pituicytes at the edge of the infarction or hemorrhage. Occasionally this reaction overshadowed the necrotic adenoma and contributed to diagnostic confusion. Our finding of lymphocytic inflammation in pituitary adenomas that have undergone an apoplectic event may suggest at least a modest immune-mediated response to damaged anterior pituitary tissue. HuM PATHOL29:1255-1265. Copyright © 1998 by W.B. Saunders Company Key words: pituitary adenoma, magnetic resonance imaging, pituitary apoplexy, pathology, immunohistochemistry. Abbreviations: MRI, magnetic resonance imaging.
Pituitary apoplexy is defined as sudden onset of a clinical syndrome consisting of headache, visua.1 deficit, ophthalmoplegia, or altered mental status caused by h e m o r r h a g e or infarction of the pituitary gland. 13 Apoplexy most commonly occurs in the setting of a pituitary a d e n o m a and only rarely in association with a n o n a d e n o m a t o u s pituitary gland,>~.Rathke cleft cyst,<5 or craniopharyngioma. 6 Often a pitmtary mass has already been identified and apoplectic symptoms may follow dynamic endocrine testing (reviewed by Rolih and Ober2), bromocriptine therapy, 5-7 or radiation therapy. 5,8,9 T h e first presentation of a pituitary ade n o m a as an apoplectic event is less common. 7-9 Only 1.6% of 300 pituitary adenomas had apoplexy as their initial manifestation of the neoplasm in a series by Weisberg, 8 and 2.7% of 560 adenomas in the series by Wakai et al. 9 Estimates of the incidence of apoplexy in pituitary adenomas are c o n f o u n d e d by the inclusion in some series of pituitary tumors with pathological or radiological evidence of h e m o r r h a g e but without sudden onset of a clinical syndrome. 5-1° H e m o r r h a g e within pituitary adenomas is more readily recognized on radiographic studies than infarction and occurs in up to 10% 11 to 15% 12 of adenomas, albeit often as a clinically silent event.5,z,10,n
Several large series on pituitary apoplexy associated with pituitary adenomas have been published and are summarized in this pub!icat[on. 5q9 Most series, however, have consisted of 15 or fewer patients, emphasized clinical rather than pathological findings, and antedated the magnetic resonance imaging (MRI) era. Since 1985, we have e n c o u n t e r e d 15 cases of pituitary a d e n o m a patients presenting as apoplexy, either from our institution or in consultation, and u n d e r t o o k a systematic review of our experience. We had n o t e d in several cases that the pathological features were underrepresentative of the extent of the apoplectic process in comparison with radiographic or intraoperative findings that were subsequently presented at patient conferences. We hypothesized that by combining pathological and intraoperative findings with more sensitive MRI scan information available on most of our patients, we could better characterize whether apoplexy had occurred on the basis, of bland infarction, hemorrhagic infarction, or clot. We also were interested in whether histological changes in damaged a(lenoma matched the timing of the apoplectic event ,clinically. Although a detailed radiological study of the MRI appearance of pituitary adenomas with apoplexy was not our focus per se, we were interested in identifying the histological correlates of any major MRI features seen in our cases. We had observed the most interesting pathological features in cases with a longer delay between clinical presentation and operation. Several biopsy specimens had shown p r o m i n e n t lymphocytic infiltration and granulation tissue, in one case overshadowing the neci-otic a d e n o m a and contributing to initial diagnostic confusion. We sought to characterize the lymphocytic infiltrates immunohistochemically and compare the degree of lymphocytic inflammation with that f o u n d in a known immune-mediated pituitary disorder, lymphocytic hyphophysitis2 ° We hypothesized that destruction of a d e n o m a tissue with apoplexy might lead to anti-
From the Departments of Pathology and Neurology, and the Department of Surgery (Neurosurgery), University of Colorado Health Sciences Center. 4200 East Ninth Ave, Denver. CO 80262. Accepted for publication May 13. 1998. Presented in Abstract Format at the 87th Annual Meeting of the United States and Canadian Academy of Pathology; Boston. MA, March 1998. Address correspondence and reprint requests to B.K. Kleinschmidt-DeMasters. MD. Department of Pathology (B-2161, University of Colorado Health Sciences Center, 4200 East Ninth Ave, Denver. CO 80262. Copyright © 1998 by W.B. Saunders Company 0046-8177/98/2911=001358.00/0
1255
HUMAN PATHOLOGY
Volume 29, No. 11 (November 1998)
TABLE 1. Pathological and MRI Features of Pituitary Adenomas Presenting With Apoplexy Patient and Operation Age/ Date Sex Patient1
Clinical Presentation
Interval of Symptom Onset to Operation
MR/ Appearance
Intraoperative Findings
Bland infarction;
39/M Acute onset of headaches, nausea and vomiting, cornplete left CN Ill, IV, VI paralysis
25 days
Patient2
47/M Sudden-onset severe headaches and severe decrease in vision in left eye and left 3rd nerve palsy; 6 months intermittent blurring in left eye
Patient3
71/M Sudden collapse-taken to ER, somnolent and unarousable; had acute GI bleed night before with hypotension and acidosis; history of pituitary adenoma partially removed 1991, followed by radiotherapy 68/M Sudden onset of headache above left eye and blurred vision
Sellar/suprasellar Hemorrhagic, Hemorrhagic infarcmass with mild necrotic tumor tion; Necrotic right cavernous tissue but some sinus compression; nuclear detail; ¼ minimal periphhemorrhagic; no eral rim enhanceorganization or ment with gadolymphocytic linium (outside response MRI scans no longer available to assess T1 weighting) 1 day Se!lar/suprasellar Acutely infarcted Hemorrhagic infarcmass isointense on tumor with righttion; Necrotic T1 with foci of sided acute hemortumor with thromhyperintensity, rhage intermixed boses and differing hypointense on T2; with tumor and ages of tissue peripheral rim older, organized damage, ranging enhancement with hemorrhage from one to several gadqlinium days old; ¼ hemorrhagic; scant organization; brisk focal lymphocytic response 2 weeks Sellar/suprasellar Firm tumor with Clot; Majority of mass is0intense on clotted, coalesced tumor non-neT1, minimally blood 1-2 weeks crotic, ¼0 hemorhyperintense on old in superior rhagic with organiT2; mildly homogeaspect of tumor zation; minimal neously enhancing lymphocytic with gadolinium response 6 weeks Sellar/suprasellar ½ Of tumor old clot Clot; Non-necrotic mass isointense on superiorly, ½ nonadenoma without T1, hypointense on necrotic adenoma hemorrhage, T2; ~ mass coninferiorly hemosiderin, sisting of clot fibrosis, organizashowing fluid/ tion, or lymphofluid level in supecytic response rior aspect; peripheral rim enhancement with gadolinium 4 days Sellar/suprasellar % Hemorrhagic Hemorrhagic infarction mass hyperintense necrotic tumor & clot; 80% on T1, hypointense with ~ old organecrotic adenoma with areas without on T2; clot in supenizing hemorrhage rior aspect hypersuperiorly nuclear detail, intense on T1 and hemorrhagic T2; no peripheral necrosis with early rim enhancement organization; brisk with gadolinium focal lymphocytic response
Patient4
Patient5
72/F Sudden onsetof severe decreased acuity in left eye
Patient6
64/M Sudden onset of headaches and righ t 3rd nerve palsy, bitemporal visual field deficit and decreased acuity
Sellar/suprasellar Homogeneous mass hyperintense necrotic tissue; no on T1, mildly hemorrhage or hyperintense T2; viable tumor idenunenhanced study tiffed only
Pathological Features
1 day
12S6
Completely necrotic tissue without nuclear detail or hemorrhage; scant organization or lym~ phocytic response
Immunohistochemistry Results No immunostaining; too necrotic to analyze
Weak gonadotroph cell adenoma ' (20% cells + for ASU, FSH, LH)
Null cell adenoma
Null cell adenoma
Weak gonadotroph cell adenoma (10% cells + FSH, LH, ASU)
Weak gonadotroph cell adenoma (25% cells + LH, FSH, ASU)
PITUITARYADENOMAS WITH APOPLEXY (Kleinschmidt-DeMasters & Lillehei)
TABLE | . Patient and Operation Age/ Date Sex Patient 7
Pathological and MRI Features of Pituitary Adenomas Presenting With Apoplexy (Continued)
Clinical Presentation
Interval of Symptom Onset to Operation
'54/M Sudden Onset 0fleft retro-orbital headache, nausea, vomiting, stiff neck, obtundation
6 days
Patient 8
36/F Sudden onset of severe headache, nausea, vomiting, and blurred vision
10 days
Patient 9
46/M Sudden onset of severe headache, nausea, vomiting, followed by progressive visual loss
Patient 10 81/M Sudden onS& of right 3rd nerve palsy
Patient 11 64/M Sudden visual loss and near blindness
Patient 12 41/M Sudden onset of headache, nausea, vomiting, no visnal defects
MR1 Appearance
Intraoperative Findings
Pathological Features
Sellar/suprasel'lar Necrotic tumor mass, isointense on admixed with T1 with focal small amount of hyperintensity, hemorrhage hyperintense on T2; no contrast enhancement
Hemorrhagic infarction; necrotic
Necrotic tumor Sellar/suprasellar mass, hyperintense admixed with on T1, hyperinquestionable "old hemorrhage" tense on Y2, peripheral rim enhancement with gadolinium
Bland infarction;
adenoma with focal preservation of nuclear detail; focal oiganization and scant lymphocytes at ~dge! ¼0 hemorrhagic
Immunohistochemistry Results Prola'cti~oma (scattered PRL + viable cells)
Corticotroph cell nearly completely adenoma (scatnecrotic adenoma tered with coagulhtive ACTH + viable necrosis, scant cells) islands of viable adenoma cells, • promilgent organiZation and brisk lymphocytosis at edge Viable adenoma with Clot;Non-necrotic Weak gonadotroph large central liqueadenoma, ½ hemcell adenoma fled hematoma orrhagic with (!0%-25% with "subdural-like prominent organicells + ASU, LH) neomembrane" at zation and lymphoedge cytosis at edge
Sellar/suprasellar mass hypointense on Ti, hyperintense on T 2 with fluid/fluid level; peripheral rim enhancement with gadolinium 1 day Sellar/suprasellar Necrotic, "gritty Bland infarction; 75 % Null cell hdenoma mass with right purple tumor" with necrotic adenoma cavernous and partially firm tissue with some preservation of nuclear sphen0id sinus superiorly detail, no hemorextension; rhage, organizairregular peripheral rim enhancetion or lymphocytic response ment with gadolinium (T1 weighting unspecified) Corticotroph cell 1 day Sellar/suprasellar Hemorrhagic tumor Hemorrhagic infarction; 90% necrotac adenoma (ACTH mass on CT; no with cystic centeI, only) adenoma, ½ hempreoperative MR][ fibrous tissue at edge orrhage£ and coagulative necrosis with organization and brisk lymphocytic response at edge Remote hemorrhagic Insufficient adenoma 5 weeks Sellar/suprasellar Necrotic "cheesy" mass irregularly material, no gross tissue; too necrotic infarction; 90% hyperintense on blood or old blood granulation tissue, to evaluate hemosiderin, and T1, T2 not specilymphocytic fied; peripheral rim enhancement response; ¼0 necrotic adenoma with gadolinum with severe coagulative necrosis, clumps of atrophic, perivascular pituicytes
1 week
1257
HUMAN PATHOLOGY
Volume 29, No, 11 (November 1998)
TABLE 1. Pathological and MRI Features of Pituitary Adenomas Presenting With Apoplexy (Continued) Patient and Operation Age/ Date Sex
Clinical Presentation
Interval of Symptom Onset to Operation
Patient 13 67/M Sudden onset visual loss in right eye and ophthalmoplegia during nasal packing for epistaxis, progressive over 36 hours
6 days
Patient 14 53/M Sudden-onset severe headache, left 3rd nerve palsy
1 day
Patient 15 42/F Sudden onset of headache with partial right 3rd nerve palsy, progressive over final 48 hours
5 days
MRI Appearance
Intraoperative Findings
Sellar/suprasellar mass, peripheral rim enhancement with gadolinium
Necrotic rumor, no hemorrhage idenrifled
Pathological Features
Immunohistochemistry Results
Bland infarction; 75% Null cell adenoma necrotic adenoma with coagulative necrosis, extensive organization an d focal lymphocytic response at edge, clumps of atrophic, perivascular pituicytes Clot; ½ necrotic Null cell adenoma adenoma with some preservation of nuclear detail, ½ hemorrhage without organization
Sellar/suprasellar Mostly cystic mass with left cavadenoma conernous sinus commining xanthopression, hypoinchromic fluid and tense on T1, liquefied hemahyperintense on T2 toma with fibrotic with fluid/fluid capsule level; peripheral rim enhancement with gadolinium Bland infarction; 80% Sellar/suprasellar Necrotic "cheesy" necrotic adenoma mass with right material, no gross cavernous sinus blood or old blood with coagulative compression, necrosis, some preservation of irregularly hyperintense on T1, nuclear detail, irregularly hyperorganization and focal lymphocytic intense on Ti; peripheral rim response at edge, enhancement with clumps of atrogadolinium phic, perivascular piuficyteS
Null cell adenoma (rare ceils positive for ASU and FSH)
Abbreviations: M. male; E female; CN, cranial nerve; ER. emergency room; ACTH, adrenocorticotrophic hormone; LH. luteinizing hormone; FSH, follicle-stimulating hormone; GI, gastrointestinal; ASU, alpha-subunit of glycoprotein hormones: PRL. prolactin.
genic exposure, stimulation of the i m m u n e system, and inflammation.
METHODS Fifteen patients with pituitary a d e n o m a s presenting with apoplexy between 1985 and 1997 were identified from o u r Departments of N e u r o s u r g e r y and Pathology files and via review of the University of C o l o r a d o Health Science C e n t e r medical r e c o r d database. Only patients with a clinical history of s u d d e n onset of clinical symptoms, usually consisting of h e a d a c h e or visual disturbances, who were f o u n d radiographically and mtraoperatively to have pituitary adenomas, were included in the study. Most patients were operate d o n by the same n e u r o s u r g e o n (K.O.L.). and details of the clinical and intraoperative history were o b t a i n e d from personal observation, review of intraoperative dictations, clinic records, and medical chart review. MRI scans were obtained for review w h e n e v e r possible, a l t h o u g h in several of the Older cases in the series, the actual films had b e e n destroyed. Also, in several of the outside consultation cases, actual MRI scans could n o t be located, and i n f o r m a t i o n was taken f r o m any radiologieal reports that Could be obtained. Specific attention was paid to correlation between findings at the time of surgical exploration and MRI appearance in an attempt to identify which p o r t i o n of an a d e n o m a h a d h e m o r r h a g e and w h e t h e r this was included in the sample sent for pathological e x a m i n a t i o n or was lost to intraoperative suctioning.
All tissue specimens were submitted in toto, fixed in 10% buffered formalin, as well as Bouin's fixative in some cases for optimization of i m m u n o h i s t o c h e m i s t r y staining, and processed routinely for light microscopy. Five-micron sections were stained with hematoxylin and eosin. Gomori's reticulin, and periodic acid-Schiff-orange G stains. I m m u n o h i s t o c b e m i cal staining for anterior pituitary h o r m o n e s , CD45 (leukocytic c o m m o n a n n g e n ) , CD3 (T cell marker), and C D i 0 (B cell marker) was p e r f o r m e d , using the peroxidase, anti-peroxidase c o m p l e x t e c h n i q u e with light hematoxylin counterstain. Antisera and their sources that were used i n c l u d e d prolacfin, growth h o r m o n e , and adrenocorticorrophic h o r m o n e (all three polyclenal, Signet, D e d h a m , MA); follicle,stimulating h o r m o n e and luteinizing h o r m o n e (both polyclonal, Dako, Santa Barbara, CA): thyroid-stimulating h o r m o n e and alpha= subunit (both m o n o c l o n a l . AMAC, Westbrook, ME); CD45 and C D i 0 (both m o n o c l o n a l ) and CD3 (po!yclonal) (Dako); and pancytokeratin (AE1/AE3. m o n o c l o n a l , Signet).
RESULTS T h e age of the patients, their clinical presentations, and the intervals f r o m s y m p t o m onset to time o f surgery are detailed in Table 1. T h e apoplectic event was the initial presentation o f the pituitary a d e n o m a in 14 o f 15 patients. O n e of the elderly patients, patient 3, first presented to our medical system somnolent and unarous-
1258
PITUITARYADENOMAS WITH APOPLEXY (Kleinschmidt-DeMasters & Liilehei)
FIGURE I. MRI scans showed striking fluid-fluid levels (arrows) within the adenoma in apoplexy patients who had clot as their underlying pathogenic mechanism. T~-weighted, sagittal (A) and T2-weighted, horizontal (B) views. Patient 5 is illustrated.
FIGURE 2, MRI scans, after gadolinium enhancement, showed peripheral rim enhancement (arrows) in several cases with longer intervals between symptom onset and surgery. This corresponded to prominent granulation tissue response to necrotic areas of adenoma. Sagittal (A) and coronal (B) views. Patient 9 is illustrated.
1259
FIGURE 3. Severely necrotic adenomas showed maintenance of the papillary or sheetiike architectural pattern characteristic of pituitary adenomas in at least some areas. Patient 1 is illustrated. (Hematoxylin and eosin; original magnification ×400.)
FIGURE 4. Despite necrosis, reticulin staining still illustrated the disruption of acinar pattern that is diagnostic for a pituitary adenoma. Patient 1 is illustrated. (Gomori's reticulin stain; original magnification x200.)
FIGURE 5. Immunostaining for anterior pituitary hormones was still informative because clumps of viable adenoma cells were identified in most adenomas in this study (13 of 15 patients). Patient 2 is illustrated. (Immunostaining for alpha-subunit, peroxidase/antiperoxidase technique with light-hematoxylin counterstain; original magnification ×400.)
FIGURE 6. The biopsy material from patient 12, who had the longest interval between symptom onset and surgery (5 weeks), showed extensive granulation tissue that overshadowed the smalL amount of residual necrotic adenoma (arrows) and led to diagnostic confusion. (Hematoxylin and eosin; original magnification x200.)
FIGURE 7. Adding to the diagnostic confusion was the presence of small dark cells in a perivascular location, which prompted a referral diagnosis of possible lymphoma. Patient 12 is illustrated. (Hematoxylin and eosin; original magnification ×200.)
FIGURE 8. Pancytokeratin immunostain of these cells with dark "naked" nuclei and nuclear molding identifies them as atrophic pituicytes, not lymphoma. (Immunostaining for pancytokeratin, peroxidase/antiperoxidase technique with lighthematoxylin counterstain; original magnification ×600.)
FIGURE 9. Most apoplectic adenomas showed lymphocytic infiltration of varying amounts, which consisted predominantly of CD3-positive, T cells (A) rather than CD20-positive, B cells (B). Photographs taken from approximately the same region in patient 9. (Immunostaining for CD3 and CD20, peroxidase/antiperoxidase technique with Iight-hematoxylin counterstain;
HUMAN PATHOLOGY
Volume 29, No, 11 (November 1998)
able and was found to have a suprasellar mass. He was taken to surgery within 24 hours and, during the operative approach, it was recognized that he had had previous transsphenoidal surgery. Postoperatively, a history was obtained that he had had a pituitary adenoma partially removed 5 years previously at another institution, followed by radiation therapy. Radiographically, all patients had sellar and suprasellar macroadenomas. Four of 15 cases showed cavernous sinus compression or invasion; in addition, patient 10 had sphenoid sinus extension. Eleven scans were informative for T1 weighting. The apoplectic adenoma in four patients showed hyperintensity on T1, in four patients isointensity on T1; and in two patients hypointensity on Tv Three of the cases showed striking fluid/fluid levels, all corresponding to the intraoperative finding of clot in the same area (Fig 1). The intervals between symptom onset and surgery were variable in these three cases with fluid/fluid levels, ranging from I day to I week to 6 weeks. Peripheral rim enhancement with gadolinium was seen in 10 cases (Fig 2). Again, the intervals between symptom onset and surgery in these cases with rim enhancement varied from 1 day to 6 weeks. Radiographic appearance on Tz-weighted scans was predominately that of hyperintensity (Table 1). Combining the MRI appearance with intraoperative findings and pathological features allowed characterization of the apoplectic process. Five cases were classified as bland (ie, with either minimal or no hemorrhagic component) infarctions, four as hemorrhagic infarctions, one as mixed hemorrhagic infarction and clot, and three as pure clot. In one case, the pathological specimen consisted of extensive granulation tissue and hemosiderin deposition with scant amounts of residual, severely necrotic adenoma, consistent with remote hemorrhagic infarction (patient 12). No correlation between type of apoplectic process and patient age was observed. In several cases, the pathological specimen alone severely underrepresented the apoplectic event. In two cases in which pure clot constituted the apoplectic process, the pathological specimen consisted almost completely of non-necrotic adenoma, but no clot. These cases would have been difficult to characterize as adenomas with apoplexy without correlation of the pathological features with the MRI appearance and intraoperative findings (patients 4 and 5). In cases in which necrotic adenoma constituted the majority of the specimen, the characteristic papillary or sheetlike architectural pattern of a pituitary adenoma could still be discerned in some areas (Fig 3). The timing of the pathological features in terms of loss of nuclear detail and extent of organization within the necrotic adenoma generally correlated well with the stated clinical interval between symptom onset and operation. Nonetheless, exceptions were observed. Patient 3 showed radiographic, intraoperative, and pathological features of older, as well as more recent, necrosis and thromboses. The clinical interval between collapse and surgery was 1 day. However, he had had a hypoten-
sive episode 48 hours previously and had received radiation therapy 5 years before his apoplectic presentation, both of which may have contributed to the apoplexy. Similarly, patient 11 showed pathological evidence of organization and severe dissolution of cellular detail within his necrotic adenoma, but symptom interval between onset and operation was also only 1 day. Discordance was less pronounced in patient 9, whose symptom interval was 1 week, but his MRI, intraoperative, and pathological features were all more consistent with longer interval between the apoplectic event and surgery. Specifically, the intraoperative finding of a "subdural-like n e o m e m b r a n e " at the edge of the hematoma, coupled with the pathological findings of prominent organization and lymphocytosis in the specimen, was more consistent with several weeks elapsed time between tissue damage and surgery. Special stains remained a useful adjunct for diagnosis despite the extensive tissue damage. Reticulin staining still clearly showed loss of acinar pattern within the tissue, verifying the necrotic or hemorrhagic tissue as adenoma, rather than normal anterior pituitary gland (Fig 4). Immunohistochemistry could be evaluated in all but two cases in which 100% of the adenoma tissue obtained showed severe, advanced necrosis with complete loss of nuclear detail (patients 1 and 12). The patchy or less advanced state of necrosis allowed evaluation ofimmunostains in the remaining 13 of 15 patients (Fig 5). Endocrine-inactive adenomas of null cell and weak gonadotroph cell types predominated and constituted 10 of 13 cases that could be evaluated. The remaining three cases included two corticotroph cell adenomas and one prolactinoma (Table 1). The most confusing pathological feature occurred in the biopsy material from patient 12, where the scant residual necrotic adenoma was completely overshadowed by the granulation tissue and fibrotic response (Fig 6). Admixed were clumps of atrophic pituicytes that showed "naked nuclei," nuclear molding, clustering around blood vessels, and entrapment within the dense lymphocytic-rich granulation tissue (Fig 7). Immunostaining with pancytokeratin showed residual patchy staining in these clusters, clarifying their identity (Fig 8). Clumps of atrophic pituicytes were also seen in patients 13 and 15. Most of our cases, especially those with the longest interval between clinical onset of symptoms and operation, showed influx of lymphocytes either within necrotic adenoma or at the edge within the granulation tissue (Fig 9). Although both CD3- and CD20-positive lymphocytes were present, CD3-positive T lymphocytes predominated. This feature contributed to a confusing histological appearance only in biopsy specimens where they were coupled with pituicytes showing severe atrophy. We compared the degree of lymphocytic infiltrate with that of a known inflammatory, autoimmune disorder, lymphocytic hypophysitis, we had previously published. 2° Nowhere did the lymphocytic intensity equal lymphocytic hypophysitis, although in both disorders T lymphocytes predominated. No case with delayed inter-
1262
PITUITARYADENOMAS WITH APOPLEXY(Kleinschmidt-DeMasters & Lillehei)
val between symptoms and operation showed either dystrophic calcification or squamous metaplasia. DISCUSSION This study of patients with pituitary adenomas who presented with apoplexy provides pathologists and clinicians with several insights to the disorder. First, not all apoplectic pituitary adenomas have a significant hemorrhagic component. Second, histological features of the process can be confusing, especially in cases with delayed interval to surgery. Third, special stains such as reticulin and immunohistochemistry stains are necessary for making the diagnosis, remained a useful adjunct to diagnosis despite the severe tissue damage, and should be performed. Finally, our identification of T-cell inflammation in later stages of apoplectic adenomas may suggest at least a modest immune response to TABLE 2.
Large Series on Clinical Sympomatic Pituitary Adenoma and Apoplexy Patients (English Language Literature)
No. o f Patients
Authors
damaged tissue. The size of our series is comparable to others in the literature, most of which have focused on clinical features, incidence, or outcome of the disease, rather than pathological aspects, as reviewed in Table 2. 519 Our investigation showed that a nearly equal number of apoplectic adenomas had resulted from bland infarction, hemorrhagic infarction, or pure clot. This detailed pathology correlation study confirms the observations made by earlier investigators (reviewed in detail by Cardoso and Peterson 1) that the underlying pathophysiological process in some cases of pituitary apoplexy seems to be pure hemorrhage, rather than hemorrhagic infarction. Postulated mechanisms for pure hemorrhage in apoplectic adenomas have centered on abnormalities in the blood vessels of the tumors.1 Our findings are similar to those of McFadzean et al. 13 Mixed pathological features in the nine cases of
Focus of Study
McFadzean, et aP 2
15
Clinical features, visual deficits, CT features
A h m e d et al is
13
Presenting features a n d follow-up
Muller-Jensen et al 6
58
Clinical features, incidence
Fraioli et al n
13
H e m o r r h a g e in pituitary adenomas
Tsitsopoulos et al TM
13
Incidence; clinical features, effect o f surgery on visual outcome
Wakai et al 9
38 (Plus 13 additional with " m i n o r " apoplexy)
Incidence; risk factors for apoplexy
Weisburg s
14
Onesti et al 5
15
Ebersold et al a5
13
CT detection; association with radiation therapy Clinical presentation, surgical m a n a g e m e n t , o u t c o m e of visual deficits, MRI correlation Clinical, pathologic, CT features
Rovit & Fein 16
9
Clinical presentation
Bills et a117
Clinical presentation a n d visual outcome
Vidal et aP s
37 (including 13 from series by Ebersold et al) 12
Bonecki et al 7
23
Clinical features, incidence, predisposing factors for apoplexy
Clinical a n d endocrinology presentation, clinical o u t c o m e
1263
N u m b e r of Patients With MRI Scans
Mention of Pathological Features
ImmunohistoChemistry
5 With ischemic necrosis, 5 with h e m o r r h a g e , 4 m i x e d ische m i a / h e m o r r h a g e , 2 with clot, 2 neither ischemia or hemorrhage 13 With h e m o r r h a g e , 7 with hemorrhage/infarction, 6 with fibrosis/hemosiderin " E x t e n d e d cystic necrotic areas, most frequently with a h e m o r rhagic c o m p o n e n t " Seven with blood collection, two with h e m o r r h a g i c necrosis, four with b o t h Seven with recent h e m o r r h a g e ; four, with old h e m o r r h a g e , two, infarcted with hemorrhage 14 With h e m a t o m a or old bloody fluid, 11 with xanthoc h r o m i c fluid or e m p t y sella, 8 with solid tumor, one with solid t u m o r a n d s u b a r a c h n o i d hemorrhage " H e m o r r h a g e , necrosis, a n d cystic changes in t u m o r " 15 With h e m o r r h a g e , 13 with h e m o r r h a g e / n e c r o s i s , 2 with hemosiderin
No
4 o f 15
No
0 of 13
No
0 of 58
No
0 of 13
No
0 of 13
No
0 o f 38
No
0 o f 14
Yes
10 of 15
No
0 of I3
No
0 of 9
Yes
3 of 37
Yes
0 ofli
12 With h e m o r r h a g e , bloody fluid or clot, 1 necrotic, 4 also with hemosiderin/fibrosis Variable h e m o r r h a g e , liquefaction, softening "Many specimens consisted mainly of necrotic tissue" 8 With h e m o r r h a g e / n e c r o s i s , two with no a d e n o m a identified 14 With h e m o r r h a g e , 8 with ischemia, 1 necrotic
No
Not specified
HUMAN PATHOLOGY
Volume 29, No, 11 (November 1998)
McFadzean et a113 included ischemic necrosis (five cases), h e m o r r h a g e (five cases), i s c h e m i a / h e m o r r h a g e (four cases), b l o o d clot (two cases), and neither ischemia or h e m o r r h a g e (two cases). T h e latter two cases may have b e e n examples in which the pathological specimen u n d e r r e p r e s e n t e d the apoplectic event, possibly because of blood clot being lost to intraoperative suctioning procedures. Such was the case in patients 4 and 5 of our series. In contrast, our study differs f r o m those by A h m e d et al, ~4 Fraioli et al, 11 and Onesti et al, 5 who h a d suggested that h e m o r r h a g e was a feature in all of their apoplectic cases. By detailed correlation between MRI scans, intraoperative records, and histological findings, we concluded f r o m this study that at least some apoplectic a d e n o m a s result on the basis of bland infarction. Few previous studies have tried to correlate clinical timing of the apoplectic event with histological dating. Rovit and Fein 17 had observed that pathologic changes "correlated reasonably well with the rapidity and severity o f . . . clinical course." Although we would concur, o u r attempts to m a t c h the dating between histological features and the clinical interval between symptom onset and surgery did show discordance in some cases. Perhaps, clinical symptoms may develop or worsen with the onset of edema, which may follow tissue necrosis by several days. From a practical pathological standpoint, our study showed the utility of reticulin stains in pituitary adenomas, even if they displayed severe necrosis. T h e relative preservation of vasculature, despite tissue necrosis, still allowed ascertainment of a disrupted acinar pattern in the necrotic tissues and c o n f i r m e d the underlying process as an adenoma, not n o r m a l anterior pituitary gland. Because of the patchy nature of the necrosis, i m m u n o h i s t o c h e m i s t r y was still informative in 13 of 15 cases, and should be p e r f o r m e d . Only a few of the previous large series had included i m m u n o h i s t o c h e m i cal categorization of pituitary a d e n o m a s with apoplexy (see Table 2). Like Bills et al, is and Mfiller-Jensen a n d Lfldecke, 6 we f o u n d most of apoplectic a d e n o m a s in our series to be endocrine-inactive null cell a d e n o m a s and weak g o n a d o t r o p h cell adenomas. Similar to the study of Onesti et al, 5 occasional corticotroph and prolactin cell a d e n o m a s were identified. Unlike Bonicki et al, 7 we f o u n d no p r e d o m i n a n c e of large growth h o r m o n e - s e c r e t i n g adenomas, Nelson's syndrome, or r e c u r r e n t tumors in o u r patient population. Although four of o u r cases showed cavernous or sphenoid sinus compression or invasion, we disagree with Bonicki et al 7 that apoplexy in pituitary a d e n o m a s is a m a r k e r suggesting invasiveness of the tumor. Eleven of our apoplexy cases showed neither invasion or compression of cavernous sinus by MRI. No correlation between immunohistochemical staining pattern of the a d e n o m a s and type of apoplectic event (ie, bland infarction v h e m o r r h a g i c infarction v clot) was n o t e d in this study. T h e main focus of this study was on pathological features of apoplectic pituitary adenomas, and this was not a detailed MRI study p e r se, but we were interested in correlating any MRI findings in o u r patients with the
matching histological features. Similar to the c o m p u t e d t o m o g r a p h y study of A h m e d et al, 14 we f o u n d fluid/ fluid levels by MRI in three apoplectic a d e n o m a s with acute (one case) or subacute (two cases) clot. Peripheral rim e n h a n c e m e n t ("ring e n h a n c e m e n t " ) could be f o u n d in response to either clot or infarction (Table 1) and c o r r e s p o n d e d to the histological finding of granulation tissue and lymphocytosis. Like A h m e d et al, 14 we n o t e d that peripheral rim e n h a n c e m e n t could be f o u n d in processes of varying acute or chronic duration. In our institutional experience on m o r e than 350 pituitary adenomas, we have n o t e n c o u n t e r e d peripheral rim e n h a n c e m e n t in adenomas, except in the setting of h e m o r r h a g e or infarction. Finally, we included in o u r series patients with longer intervals between symptom onset and surgery. Several of these cases h a d b e e n referred to us because of their confusing histological features. Atrophic pituicytes coupled with lymphocytosis h a d mimicked the a p p e a r a n c e of l y m p h o m a or carcinoma. Immunostaining with pancytokeratin distinguished the apoplectic process f r o m a lymphoma. Although immunoreactivity with pancytokeratin can be seen in carcinomas, our c u r r e n t study illustrates for pathologists the histological features of apoplectic a d e n o m a s and should eliminate an incorrect diagnosis o f carcinoma. Atrophic pituicytes have similarly b e e n described at the edge o f large Rathke cleft cysts, as n o t e d in o u r previous study, 4 where they also may contribute to diagnostic confusion. Although calcification 21 and squamous metaplasia 22 have been r e p o r t e d as late sequelae o f apoplexy in pituitary adenomas, we did not identify these features in o u r series, perhaps because of their rarity. T h e finding of lymphocytosis with a p r e d o m i n a n c e of CD3-positive T lymphocytes may suggest that at least a m o d e s t i m m u n e response develops to d a m a g e d pituitary a d e n o m a cells.
Acknowledgment. The authors thank Drs Steven Cullen and Leonard Zemel for their invaluable assistance in providing clinical information about their patients. We are grateful to Drs Steven Cullen, Walter E. Madsen, Paul M. Scheele, and Philip T. Stoffel, who submitted cases. The photographic support of Bob McCullough, the secretarial help of Ginger Woodward, and the histological expertise of David Davis and Vickie McHenry are greatly appreciated. REFERENCES 1. Cardoso ER, Peterson EW: Pituitary apoplexy: A review. Neurosurgery 14:363-373, 1984 2. Rolih CA, Ober KP: Pituitary apoplexy.Endocrinol Metab Clin North Am 22:291-302, 1993 3. Reid RL, Quigley ME, Yen SSC: Pituitary apoplexy. Arch Neuro142:712-719, 1985 4. Kleinschmidt-DeMasters BtL Lillehei KO, Stears JC: The pathologic, surgical, and MR spectrum of Rathke cleft cysts. Surg Neuro144:19-27, 1995 5. Onesti ST, Wisniewski T, Post KD: Clinical versus subclinical pituitary apoplexy: Presentation, surgical management, and outcome in 21 patients. Neurosurgery 26:980-986, 1990 6. M/iller-Jensen A, Lfidecke D: Clinical aspects of spontaneous necrosis of pituitary tumors (pituitary apoplexy). J Neurol 224:267271, 1981 7. Bonicki W, Kasperlik-ZaluskaA, KoszewskiW, et al: Pituitary apoplexy: Endocrine, surgical and oncological emergency. Incidence,
1264
PITUITARYADENOMAS WITH APOPLEXY (Kleinschmidt-DeMasters & Lillehei) clinical course and treatment with reference to 799 cases of pituitary adenomas. Acta Neurochir 120:118-122, 1993 8. Weisberg LA: Pituitary apoplexy: Association of degenerative change in pituitary adenoma with radiotherapy and detection by cerebral computed tomography. AmJ Med 63:109-115, 1977 9. Wakai S, Fukushima T, Teramoto A, et al: Pituitary apoplexy: Its incidence and clinical significance.J Neurosurg 55:187-193, 1981 10. Ostrov SG, Quencer RM, Hoffman JC, et al: Hemorrhage within pituitary adenomas: How often associated with pituitary apoplexy" syndrome? AJR 153:153-160, 1989 11. Fraioli B, Esposito V, Palma L, et al: Hemorrhagic pituitary adenomas: Clinicopathological features and surgical treatment. Neurosurgery 27:741-748, 1990 12. Wakai S, Yamakawa K, Manaka S, et al: Spontaneous intracranial hemorrhage caused by brain tumor: Its incidence and clinical significance. Neurosurgery 10:437-444, 1982 13. McFadzean RM, Doyle D, Rampling R, et ah Pituitary apoplexy and its effect on vision. Neurosurgery 29:669-675, 1991 14. Ahmed M, Rifai A, A1-Jurf M, et al: Classical pituitary apoplexy presentation and a follow-up of 13 patients. Hormone Res 31:125-132, 1989
15. Tsitsopoulos P, Andrew J, Harrison MJG: Pituitary apoplexy and haemorrhage into adenomas. Postgrad MedJ 62:623-626, 1986 16. Ebersold MJ, Laws ER, Scheithauer BW, et al: Pituitary apoplexy treated by transsphenoidal surgery: A clinocopathological and immunocytochemical study.J Neurosurg 58:315-320, 1983 17. Rovit RL, Fein JM: Pituitary apoplexy: A review and reappraisal.J Neurosurg 37:280-288, 1972 18. Bills DC, Meyer FB, Laws ER, et al: A restrospective analysis of pituitary apoplexy. Neurosurgery 33:602-609, 1993 19. Vidal E, Cevallos R, Vidal J, et al: Twelve cases of pituitary apoplexy. Arch Intern Med 152:1893-1899, 1992 20. Reusch JE, Kleinschmidt-DeMasters BK, Lillehei KO, et al: Preoperative diagnosis of lymphocytic hypophysitis (adenohypophysitis) unresponsive to short course dexamethasone: Case report. Neurosurgery 30:268-272, 1992 21. Kepes JJ, Sayler J, Hiszczynskyj R: Squamous metaplasia following necrosis of the adenohypophysis and of a chromophobe adenoma of the pituitary. Virchows Arch A Pathol Anat Histol 395:69-76, 1982 22. Horiuchi T, Tanaka Y, Kobayashi S, et al: Total capsular calcification in a prolactinoma: Case report. Neurol Med Chirurg 36:729-732, 1996
1265
troph and null cell adenomas predominated. Cases with the greatest interval between symptom onset and surgery showed peripheral rim enhancement by MRI. Pathologically, this corresponded to granulation tissue, T cell lymphocytic infiltration, and atrophic pituicytes at the edge of the infarction or hemorrhage. Occasionally this reaction overshadowed the necrotic adenoma and contributed to diagnostic confusion. Our finding of lymphocytic inflammation in pituitary adenomas that have undergone an apoplectic event may suggest at least a modest immune-mediated response to damaged anterior pituitary tissue. HuM PATHOL29:1255-1265. Copyright © 1998 by W.B. Saunders Company Key words: pituitary adenoma, magnetic resonance imaging, pituitary apoplexy, pathology, immunohistochemistry. Abbreviations: MRI, magnetic resonance imaging.
Pituitary apoplexy is defined as sudden onset of a clinical syndrome consisting of headache, visua.1 deficit, ophthalmoplegia, or altered mental status caused by h e m o r r h a g e or infarction of the pituitary gland. 13 Apoplexy most commonly occurs in the setting of a pituitary a d e n o m a and only rarely in association with a n o n a d e n o m a t o u s pituitary gland,>~.Rathke cleft cyst,<5 or craniopharyngioma. 6 Often a pitmtary mass has already been identified and apoplectic symptoms may follow dynamic endocrine testing (reviewed by Rolih and Ober2), bromocriptine therapy, 5-7 or radiation therapy. 5,8,9 T h e first presentation of a pituitary ade n o m a as an apoplectic event is less common. 7-9 Only 1.6% of 300 pituitary adenomas had apoplexy as their initial manifestation of the neoplasm in a series by Weisberg, 8 and 2.7% of 560 adenomas in the series by Wakai et al. 9 Estimates of the incidence of apoplexy in pituitary adenomas are c o n f o u n d e d by the inclusion in some series of pituitary tumors with pathological or radiological evidence of h e m o r r h a g e but without sudden onset of a clinical syndrome. 5-1° H e m o r r h a g e within pituitary adenomas is more readily recognized on radiographic studies than infarction and occurs in up to 10% 11 to 15% 12 of adenomas, albeit often as a clinically silent event.5,z,10,n
Several large series on pituitary apoplexy associated with pituitary adenomas have been published and are summarized in this pub!icat[on. 5q9 Most series, however, have consisted of 15 or fewer patients, emphasized clinical rather than pathological findings, and antedated the magnetic resonance imaging (MRI) era. Since 1985, we have e n c o u n t e r e d 15 cases of pituitary a d e n o m a patients presenting as apoplexy, either from our institution or in consultation, and u n d e r t o o k a systematic review of our experience. We had n o t e d in several cases that the pathological features were underrepresentative of the extent of the apoplectic process in comparison with radiographic or intraoperative findings that were subsequently presented at patient conferences. We hypothesized that by combining pathological and intraoperative findings with more sensitive MRI scan information available on most of our patients, we could better characterize whether apoplexy had occurred on the basis, of bland infarction, hemorrhagic infarction, or clot. We also were interested in whether histological changes in damaged a(lenoma matched the timing of the apoplectic event ,clinically. Although a detailed radiological study of the MRI appearance of pituitary adenomas with apoplexy was not our focus per se, we were interested in identifying the histological correlates of any major MRI features seen in our cases. We had observed the most interesting pathological features in cases with a longer delay between clinical presentation and operation. Several biopsy specimens had shown p r o m i n e n t lymphocytic infiltration and granulation tissue, in one case overshadowing the neci-otic a d e n o m a and contributing to initial diagnostic confusion. We sought to characterize the lymphocytic infiltrates immunohistochemically and compare the degree of lymphocytic inflammation with that f o u n d in a known immune-mediated pituitary disorder, lymphocytic hyphophysitis2 ° We hypothesized that destruction of a d e n o m a tissue with apoplexy might lead to anti-
From the Departments of Pathology and Neurology, and the Department of Surgery (Neurosurgery), University of Colorado Health Sciences Center. 4200 East Ninth Ave, Denver. CO 80262. Accepted for publication May 13. 1998. Presented in Abstract Format at the 87th Annual Meeting of the United States and Canadian Academy of Pathology; Boston. MA, March 1998. Address correspondence and reprint requests to B.K. Kleinschmidt-DeMasters. MD. Department of Pathology (B-2161, University of Colorado Health Sciences Center, 4200 East Ninth Ave, Denver. CO 80262. Copyright © 1998 by W.B. Saunders Company 0046-8177/98/2911=001358.00/0
1255
HUMAN PATHOLOGY
Volume 29, No. 11 (November 1998)
TABLE 1. Pathological and MRI Features of Pituitary Adenomas Presenting With Apoplexy Patient and Operation Age/ Date Sex Patient1
Clinical Presentation
Interval of Symptom Onset to Operation
MR/ Appearance
Intraoperative Findings
Bland infarction;
39/M Acute onset of headaches, nausea and vomiting, cornplete left CN Ill, IV, VI paralysis
25 days
Patient2
47/M Sudden-onset severe headaches and severe decrease in vision in left eye and left 3rd nerve palsy; 6 months intermittent blurring in left eye
Patient3
71/M Sudden collapse-taken to ER, somnolent and unarousable; had acute GI bleed night before with hypotension and acidosis; history of pituitary adenoma partially removed 1991, followed by radiotherapy 68/M Sudden onset of headache above left eye and blurred vision
Sellar/suprasellar Hemorrhagic, Hemorrhagic infarcmass with mild necrotic tumor tion; Necrotic right cavernous tissue but some sinus compression; nuclear detail; ¼ minimal periphhemorrhagic; no eral rim enhanceorganization or ment with gadolymphocytic linium (outside response MRI scans no longer available to assess T1 weighting) 1 day Se!lar/suprasellar Acutely infarcted Hemorrhagic infarcmass isointense on tumor with righttion; Necrotic T1 with foci of sided acute hemortumor with thromhyperintensity, rhage intermixed boses and differing hypointense on T2; with tumor and ages of tissue peripheral rim older, organized damage, ranging enhancement with hemorrhage from one to several gadqlinium days old; ¼ hemorrhagic; scant organization; brisk focal lymphocytic response 2 weeks Sellar/suprasellar Firm tumor with Clot; Majority of mass is0intense on clotted, coalesced tumor non-neT1, minimally blood 1-2 weeks crotic, ¼0 hemorhyperintense on old in superior rhagic with organiT2; mildly homogeaspect of tumor zation; minimal neously enhancing lymphocytic with gadolinium response 6 weeks Sellar/suprasellar ½ Of tumor old clot Clot; Non-necrotic mass isointense on superiorly, ½ nonadenoma without T1, hypointense on necrotic adenoma hemorrhage, T2; ~ mass coninferiorly hemosiderin, sisting of clot fibrosis, organizashowing fluid/ tion, or lymphofluid level in supecytic response rior aspect; peripheral rim enhancement with gadolinium 4 days Sellar/suprasellar % Hemorrhagic Hemorrhagic infarction mass hyperintense necrotic tumor & clot; 80% on T1, hypointense with ~ old organecrotic adenoma with areas without on T2; clot in supenizing hemorrhage rior aspect hypersuperiorly nuclear detail, intense on T1 and hemorrhagic T2; no peripheral necrosis with early rim enhancement organization; brisk with gadolinium focal lymphocytic response
Patient4
Patient5
72/F Sudden onsetof severe decreased acuity in left eye
Patient6
64/M Sudden onset of headaches and righ t 3rd nerve palsy, bitemporal visual field deficit and decreased acuity
Sellar/suprasellar Homogeneous mass hyperintense necrotic tissue; no on T1, mildly hemorrhage or hyperintense T2; viable tumor idenunenhanced study tiffed only
Pathological Features
1 day
12S6
Completely necrotic tissue without nuclear detail or hemorrhage; scant organization or lym~ phocytic response
Immunohistochemistry Results No immunostaining; too necrotic to analyze
Weak gonadotroph cell adenoma ' (20% cells + for ASU, FSH, LH)
Null cell adenoma
Null cell adenoma
Weak gonadotroph cell adenoma (10% cells + FSH, LH, ASU)
Weak gonadotroph cell adenoma (25% cells + LH, FSH, ASU)
PITUITARYADENOMAS WITH APOPLEXY (Kleinschmidt-DeMasters & Lillehei)
TABLE | . Patient and Operation Age/ Date Sex Patient 7
Pathological and MRI Features of Pituitary Adenomas Presenting With Apoplexy (Continued)
Clinical Presentation
Interval of Symptom Onset to Operation
'54/M Sudden Onset 0fleft retro-orbital headache, nausea, vomiting, stiff neck, obtundation
6 days
Patient 8
36/F Sudden onset of severe headache, nausea, vomiting, and blurred vision
10 days
Patient 9
46/M Sudden onset of severe headache, nausea, vomiting, followed by progressive visual loss
Patient 10 81/M Sudden onS& of right 3rd nerve palsy
Patient 11 64/M Sudden visual loss and near blindness
Patient 12 41/M Sudden onset of headache, nausea, vomiting, no visnal defects
MR1 Appearance
Intraoperative Findings
Pathological Features
Sellar/suprasel'lar Necrotic tumor mass, isointense on admixed with T1 with focal small amount of hyperintensity, hemorrhage hyperintense on T2; no contrast enhancement
Hemorrhagic infarction; necrotic
Necrotic tumor Sellar/suprasellar mass, hyperintense admixed with on T1, hyperinquestionable "old hemorrhage" tense on Y2, peripheral rim enhancement with gadolinium
Bland infarction;
adenoma with focal preservation of nuclear detail; focal oiganization and scant lymphocytes at ~dge! ¼0 hemorrhagic
Immunohistochemistry Results Prola'cti~oma (scattered PRL + viable cells)
Corticotroph cell nearly completely adenoma (scatnecrotic adenoma tered with coagulhtive ACTH + viable necrosis, scant cells) islands of viable adenoma cells, • promilgent organiZation and brisk lymphocytosis at edge Viable adenoma with Clot;Non-necrotic Weak gonadotroph large central liqueadenoma, ½ hemcell adenoma fled hematoma orrhagic with (!0%-25% with "subdural-like prominent organicells + ASU, LH) neomembrane" at zation and lymphoedge cytosis at edge
Sellar/suprasellar mass hypointense on Ti, hyperintense on T 2 with fluid/fluid level; peripheral rim enhancement with gadolinium 1 day Sellar/suprasellar Necrotic, "gritty Bland infarction; 75 % Null cell hdenoma mass with right purple tumor" with necrotic adenoma cavernous and partially firm tissue with some preservation of nuclear sphen0id sinus superiorly detail, no hemorextension; rhage, organizairregular peripheral rim enhancetion or lymphocytic response ment with gadolinium (T1 weighting unspecified) Corticotroph cell 1 day Sellar/suprasellar Hemorrhagic tumor Hemorrhagic infarction; 90% necrotac adenoma (ACTH mass on CT; no with cystic centeI, only) adenoma, ½ hempreoperative MR][ fibrous tissue at edge orrhage£ and coagulative necrosis with organization and brisk lymphocytic response at edge Remote hemorrhagic Insufficient adenoma 5 weeks Sellar/suprasellar Necrotic "cheesy" mass irregularly material, no gross tissue; too necrotic infarction; 90% hyperintense on blood or old blood granulation tissue, to evaluate hemosiderin, and T1, T2 not specilymphocytic fied; peripheral rim enhancement response; ¼0 necrotic adenoma with gadolinum with severe coagulative necrosis, clumps of atrophic, perivascular pituicytes
1 week
1257
HUMAN PATHOLOGY
Volume 29, No, 11 (November 1998)
TABLE 1. Pathological and MRI Features of Pituitary Adenomas Presenting With Apoplexy (Continued) Patient and Operation Age/ Date Sex
Clinical Presentation
Interval of Symptom Onset to Operation
Patient 13 67/M Sudden onset visual loss in right eye and ophthalmoplegia during nasal packing for epistaxis, progressive over 36 hours
6 days
Patient 14 53/M Sudden-onset severe headache, left 3rd nerve palsy
1 day
Patient 15 42/F Sudden onset of headache with partial right 3rd nerve palsy, progressive over final 48 hours
5 days
MRI Appearance
Intraoperative Findings
Sellar/suprasellar mass, peripheral rim enhancement with gadolinium
Necrotic rumor, no hemorrhage idenrifled
Pathological Features
Immunohistochemistry Results
Bland infarction; 75% Null cell adenoma necrotic adenoma with coagulative necrosis, extensive organization an d focal lymphocytic response at edge, clumps of atrophic, perivascular pituicytes Clot; ½ necrotic Null cell adenoma adenoma with some preservation of nuclear detail, ½ hemorrhage without organization
Sellar/suprasellar Mostly cystic mass with left cavadenoma conernous sinus commining xanthopression, hypoinchromic fluid and tense on T1, liquefied hemahyperintense on T2 toma with fibrotic with fluid/fluid capsule level; peripheral rim enhancement with gadolinium Bland infarction; 80% Sellar/suprasellar Necrotic "cheesy" necrotic adenoma mass with right material, no gross cavernous sinus blood or old blood with coagulative compression, necrosis, some preservation of irregularly hyperintense on T1, nuclear detail, irregularly hyperorganization and focal lymphocytic intense on Ti; peripheral rim response at edge, enhancement with clumps of atrogadolinium phic, perivascular piuficyteS
Null cell adenoma (rare ceils positive for ASU and FSH)
Abbreviations: M. male; E female; CN, cranial nerve; ER. emergency room; ACTH, adrenocorticotrophic hormone; LH. luteinizing hormone; FSH, follicle-stimulating hormone; GI, gastrointestinal; ASU, alpha-subunit of glycoprotein hormones: PRL. prolactin.
genic exposure, stimulation of the i m m u n e system, and inflammation.
METHODS Fifteen patients with pituitary a d e n o m a s presenting with apoplexy between 1985 and 1997 were identified from o u r Departments of N e u r o s u r g e r y and Pathology files and via review of the University of C o l o r a d o Health Science C e n t e r medical r e c o r d database. Only patients with a clinical history of s u d d e n onset of clinical symptoms, usually consisting of h e a d a c h e or visual disturbances, who were f o u n d radiographically and mtraoperatively to have pituitary adenomas, were included in the study. Most patients were operate d o n by the same n e u r o s u r g e o n (K.O.L.). and details of the clinical and intraoperative history were o b t a i n e d from personal observation, review of intraoperative dictations, clinic records, and medical chart review. MRI scans were obtained for review w h e n e v e r possible, a l t h o u g h in several of the Older cases in the series, the actual films had b e e n destroyed. Also, in several of the outside consultation cases, actual MRI scans could n o t be located, and i n f o r m a t i o n was taken f r o m any radiologieal reports that Could be obtained. Specific attention was paid to correlation between findings at the time of surgical exploration and MRI appearance in an attempt to identify which p o r t i o n of an a d e n o m a h a d h e m o r r h a g e and w h e t h e r this was included in the sample sent for pathological e x a m i n a t i o n or was lost to intraoperative suctioning.
All tissue specimens were submitted in toto, fixed in 10% buffered formalin, as well as Bouin's fixative in some cases for optimization of i m m u n o h i s t o c h e m i s t r y staining, and processed routinely for light microscopy. Five-micron sections were stained with hematoxylin and eosin. Gomori's reticulin, and periodic acid-Schiff-orange G stains. I m m u n o h i s t o c b e m i cal staining for anterior pituitary h o r m o n e s , CD45 (leukocytic c o m m o n a n n g e n ) , CD3 (T cell marker), and C D i 0 (B cell marker) was p e r f o r m e d , using the peroxidase, anti-peroxidase c o m p l e x t e c h n i q u e with light hematoxylin counterstain. Antisera and their sources that were used i n c l u d e d prolacfin, growth h o r m o n e , and adrenocorticorrophic h o r m o n e (all three polyclenal, Signet, D e d h a m , MA); follicle,stimulating h o r m o n e and luteinizing h o r m o n e (both polyclonal, Dako, Santa Barbara, CA): thyroid-stimulating h o r m o n e and alpha= subunit (both m o n o c l o n a l . AMAC, Westbrook, ME); CD45 and C D i 0 (both m o n o c l o n a l ) and CD3 (po!yclonal) (Dako); and pancytokeratin (AE1/AE3. m o n o c l o n a l , Signet).
RESULTS T h e age of the patients, their clinical presentations, and the intervals f r o m s y m p t o m onset to time o f surgery are detailed in Table 1. T h e apoplectic event was the initial presentation o f the pituitary a d e n o m a in 14 o f 15 patients. O n e of the elderly patients, patient 3, first presented to our medical system somnolent and unarous-
1258
PITUITARYADENOMAS WITH APOPLEXY (Kleinschmidt-DeMasters & Liilehei)
FIGURE I. MRI scans showed striking fluid-fluid levels (arrows) within the adenoma in apoplexy patients who had clot as their underlying pathogenic mechanism. T~-weighted, sagittal (A) and T2-weighted, horizontal (B) views. Patient 5 is illustrated.
FIGURE 2, MRI scans, after gadolinium enhancement, showed peripheral rim enhancement (arrows) in several cases with longer intervals between symptom onset and surgery. This corresponded to prominent granulation tissue response to necrotic areas of adenoma. Sagittal (A) and coronal (B) views. Patient 9 is illustrated.
1259
FIGURE 3. Severely necrotic adenomas showed maintenance of the papillary or sheetiike architectural pattern characteristic of pituitary adenomas in at least some areas. Patient 1 is illustrated. (Hematoxylin and eosin; original magnification ×400.)
FIGURE 4. Despite necrosis, reticulin staining still illustrated the disruption of acinar pattern that is diagnostic for a pituitary adenoma. Patient 1 is illustrated. (Gomori's reticulin stain; original magnification x200.)
FIGURE 5. Immunostaining for anterior pituitary hormones was still informative because clumps of viable adenoma cells were identified in most adenomas in this study (13 of 15 patients). Patient 2 is illustrated. (Immunostaining for alpha-subunit, peroxidase/antiperoxidase technique with light-hematoxylin counterstain; original magnification ×400.)
FIGURE 6. The biopsy material from patient 12, who had the longest interval between symptom onset and surgery (5 weeks), showed extensive granulation tissue that overshadowed the smalL amount of residual necrotic adenoma (arrows) and led to diagnostic confusion. (Hematoxylin and eosin; original magnification x200.)
FIGURE 7. Adding to the diagnostic confusion was the presence of small dark cells in a perivascular location, which prompted a referral diagnosis of possible lymphoma. Patient 12 is illustrated. (Hematoxylin and eosin; original magnification ×200.)
FIGURE 8. Pancytokeratin immunostain of these cells with dark "naked" nuclei and nuclear molding identifies them as atrophic pituicytes, not lymphoma. (Immunostaining for pancytokeratin, peroxidase/antiperoxidase technique with lighthematoxylin counterstain; original magnification ×600.)
FIGURE 9. Most apoplectic adenomas showed lymphocytic infiltration of varying amounts, which consisted predominantly of CD3-positive, T cells (A) rather than CD20-positive, B cells (B). Photographs taken from approximately the same region in patient 9. (Immunostaining for CD3 and CD20, peroxidase/antiperoxidase technique with Iight-hematoxylin counterstain;
HUMAN PATHOLOGY
Volume 29, No, 11 (November 1998)
able and was found to have a suprasellar mass. He was taken to surgery within 24 hours and, during the operative approach, it was recognized that he had had previous transsphenoidal surgery. Postoperatively, a history was obtained that he had had a pituitary adenoma partially removed 5 years previously at another institution, followed by radiation therapy. Radiographically, all patients had sellar and suprasellar macroadenomas. Four of 15 cases showed cavernous sinus compression or invasion; in addition, patient 10 had sphenoid sinus extension. Eleven scans were informative for T1 weighting. The apoplectic adenoma in four patients showed hyperintensity on T1, in four patients isointensity on T1; and in two patients hypointensity on Tv Three of the cases showed striking fluid/fluid levels, all corresponding to the intraoperative finding of clot in the same area (Fig 1). The intervals between symptom onset and surgery were variable in these three cases with fluid/fluid levels, ranging from I day to I week to 6 weeks. Peripheral rim enhancement with gadolinium was seen in 10 cases (Fig 2). Again, the intervals between symptom onset and surgery in these cases with rim enhancement varied from 1 day to 6 weeks. Radiographic appearance on Tz-weighted scans was predominately that of hyperintensity (Table 1). Combining the MRI appearance with intraoperative findings and pathological features allowed characterization of the apoplectic process. Five cases were classified as bland (ie, with either minimal or no hemorrhagic component) infarctions, four as hemorrhagic infarctions, one as mixed hemorrhagic infarction and clot, and three as pure clot. In one case, the pathological specimen consisted of extensive granulation tissue and hemosiderin deposition with scant amounts of residual, severely necrotic adenoma, consistent with remote hemorrhagic infarction (patient 12). No correlation between type of apoplectic process and patient age was observed. In several cases, the pathological specimen alone severely underrepresented the apoplectic event. In two cases in which pure clot constituted the apoplectic process, the pathological specimen consisted almost completely of non-necrotic adenoma, but no clot. These cases would have been difficult to characterize as adenomas with apoplexy without correlation of the pathological features with the MRI appearance and intraoperative findings (patients 4 and 5). In cases in which necrotic adenoma constituted the majority of the specimen, the characteristic papillary or sheetlike architectural pattern of a pituitary adenoma could still be discerned in some areas (Fig 3). The timing of the pathological features in terms of loss of nuclear detail and extent of organization within the necrotic adenoma generally correlated well with the stated clinical interval between symptom onset and operation. Nonetheless, exceptions were observed. Patient 3 showed radiographic, intraoperative, and pathological features of older, as well as more recent, necrosis and thromboses. The clinical interval between collapse and surgery was 1 day. However, he had had a hypoten-
sive episode 48 hours previously and had received radiation therapy 5 years before his apoplectic presentation, both of which may have contributed to the apoplexy. Similarly, patient 11 showed pathological evidence of organization and severe dissolution of cellular detail within his necrotic adenoma, but symptom interval between onset and operation was also only 1 day. Discordance was less pronounced in patient 9, whose symptom interval was 1 week, but his MRI, intraoperative, and pathological features were all more consistent with longer interval between the apoplectic event and surgery. Specifically, the intraoperative finding of a "subdural-like n e o m e m b r a n e " at the edge of the hematoma, coupled with the pathological findings of prominent organization and lymphocytosis in the specimen, was more consistent with several weeks elapsed time between tissue damage and surgery. Special stains remained a useful adjunct for diagnosis despite the extensive tissue damage. Reticulin staining still clearly showed loss of acinar pattern within the tissue, verifying the necrotic or hemorrhagic tissue as adenoma, rather than normal anterior pituitary gland (Fig 4). Immunohistochemistry could be evaluated in all but two cases in which 100% of the adenoma tissue obtained showed severe, advanced necrosis with complete loss of nuclear detail (patients 1 and 12). The patchy or less advanced state of necrosis allowed evaluation ofimmunostains in the remaining 13 of 15 patients (Fig 5). Endocrine-inactive adenomas of null cell and weak gonadotroph cell types predominated and constituted 10 of 13 cases that could be evaluated. The remaining three cases included two corticotroph cell adenomas and one prolactinoma (Table 1). The most confusing pathological feature occurred in the biopsy material from patient 12, where the scant residual necrotic adenoma was completely overshadowed by the granulation tissue and fibrotic response (Fig 6). Admixed were clumps of atrophic pituicytes that showed "naked nuclei," nuclear molding, clustering around blood vessels, and entrapment within the dense lymphocytic-rich granulation tissue (Fig 7). Immunostaining with pancytokeratin showed residual patchy staining in these clusters, clarifying their identity (Fig 8). Clumps of atrophic pituicytes were also seen in patients 13 and 15. Most of our cases, especially those with the longest interval between clinical onset of symptoms and operation, showed influx of lymphocytes either within necrotic adenoma or at the edge within the granulation tissue (Fig 9). Although both CD3- and CD20-positive lymphocytes were present, CD3-positive T lymphocytes predominated. This feature contributed to a confusing histological appearance only in biopsy specimens where they were coupled with pituicytes showing severe atrophy. We compared the degree of lymphocytic infiltrate with that of a known inflammatory, autoimmune disorder, lymphocytic hypophysitis, we had previously published. 2° Nowhere did the lymphocytic intensity equal lymphocytic hypophysitis, although in both disorders T lymphocytes predominated. No case with delayed inter-
1262
PITUITARYADENOMAS WITH APOPLEXY(Kleinschmidt-DeMasters & Lillehei)
val between symptoms and operation showed either dystrophic calcification or squamous metaplasia. DISCUSSION This study of patients with pituitary adenomas who presented with apoplexy provides pathologists and clinicians with several insights to the disorder. First, not all apoplectic pituitary adenomas have a significant hemorrhagic component. Second, histological features of the process can be confusing, especially in cases with delayed interval to surgery. Third, special stains such as reticulin and immunohistochemistry stains are necessary for making the diagnosis, remained a useful adjunct to diagnosis despite the severe tissue damage, and should be performed. Finally, our identification of T-cell inflammation in later stages of apoplectic adenomas may suggest at least a modest immune response to TABLE 2.
Large Series on Clinical Sympomatic Pituitary Adenoma and Apoplexy Patients (English Language Literature)
No. o f Patients
Authors
damaged tissue. The size of our series is comparable to others in the literature, most of which have focused on clinical features, incidence, or outcome of the disease, rather than pathological aspects, as reviewed in Table 2. 519 Our investigation showed that a nearly equal number of apoplectic adenomas had resulted from bland infarction, hemorrhagic infarction, or pure clot. This detailed pathology correlation study confirms the observations made by earlier investigators (reviewed in detail by Cardoso and Peterson 1) that the underlying pathophysiological process in some cases of pituitary apoplexy seems to be pure hemorrhage, rather than hemorrhagic infarction. Postulated mechanisms for pure hemorrhage in apoplectic adenomas have centered on abnormalities in the blood vessels of the tumors.1 Our findings are similar to those of McFadzean et al. 13 Mixed pathological features in the nine cases of
Focus of Study
McFadzean, et aP 2
15
Clinical features, visual deficits, CT features
A h m e d et al is
13
Presenting features a n d follow-up
Muller-Jensen et al 6
58
Clinical features, incidence
Fraioli et al n
13
H e m o r r h a g e in pituitary adenomas
Tsitsopoulos et al TM
13
Incidence; clinical features, effect o f surgery on visual outcome
Wakai et al 9
38 (Plus 13 additional with " m i n o r " apoplexy)
Incidence; risk factors for apoplexy
Weisburg s
14
Onesti et al 5
15
Ebersold et al a5
13
CT detection; association with radiation therapy Clinical presentation, surgical m a n a g e m e n t , o u t c o m e of visual deficits, MRI correlation Clinical, pathologic, CT features
Rovit & Fein 16
9
Clinical presentation
Bills et a117
Clinical presentation a n d visual outcome
Vidal et aP s
37 (including 13 from series by Ebersold et al) 12
Bonecki et al 7
23
Clinical features, incidence, predisposing factors for apoplexy
Clinical a n d endocrinology presentation, clinical o u t c o m e
1263
N u m b e r of Patients With MRI Scans
Mention of Pathological Features
ImmunohistoChemistry
5 With ischemic necrosis, 5 with h e m o r r h a g e , 4 m i x e d ische m i a / h e m o r r h a g e , 2 with clot, 2 neither ischemia or hemorrhage 13 With h e m o r r h a g e , 7 with hemorrhage/infarction, 6 with fibrosis/hemosiderin " E x t e n d e d cystic necrotic areas, most frequently with a h e m o r rhagic c o m p o n e n t " Seven with blood collection, two with h e m o r r h a g i c necrosis, four with b o t h Seven with recent h e m o r r h a g e ; four, with old h e m o r r h a g e , two, infarcted with hemorrhage 14 With h e m a t o m a or old bloody fluid, 11 with xanthoc h r o m i c fluid or e m p t y sella, 8 with solid tumor, one with solid t u m o r a n d s u b a r a c h n o i d hemorrhage " H e m o r r h a g e , necrosis, a n d cystic changes in t u m o r " 15 With h e m o r r h a g e , 13 with h e m o r r h a g e / n e c r o s i s , 2 with hemosiderin
No
4 o f 15
No
0 of 13
No
0 of 58
No
0 of 13
No
0 of 13
No
0 o f 38
No
0 o f 14
Yes
10 of 15
No
0 of I3
No
0 of 9
Yes
3 of 37
Yes
0 ofli
12 With h e m o r r h a g e , bloody fluid or clot, 1 necrotic, 4 also with hemosiderin/fibrosis Variable h e m o r r h a g e , liquefaction, softening "Many specimens consisted mainly of necrotic tissue" 8 With h e m o r r h a g e / n e c r o s i s , two with no a d e n o m a identified 14 With h e m o r r h a g e , 8 with ischemia, 1 necrotic
No
Not specified
HUMAN PATHOLOGY
Volume 29, No, 11 (November 1998)
McFadzean et a113 included ischemic necrosis (five cases), h e m o r r h a g e (five cases), i s c h e m i a / h e m o r r h a g e (four cases), b l o o d clot (two cases), and neither ischemia or h e m o r r h a g e (two cases). T h e latter two cases may have b e e n examples in which the pathological specimen u n d e r r e p r e s e n t e d the apoplectic event, possibly because of blood clot being lost to intraoperative suctioning procedures. Such was the case in patients 4 and 5 of our series. In contrast, our study differs f r o m those by A h m e d et al, ~4 Fraioli et al, 11 and Onesti et al, 5 who h a d suggested that h e m o r r h a g e was a feature in all of their apoplectic cases. By detailed correlation between MRI scans, intraoperative records, and histological findings, we concluded f r o m this study that at least some apoplectic a d e n o m a s result on the basis of bland infarction. Few previous studies have tried to correlate clinical timing of the apoplectic event with histological dating. Rovit and Fein 17 had observed that pathologic changes "correlated reasonably well with the rapidity and severity o f . . . clinical course." Although we would concur, o u r attempts to m a t c h the dating between histological features and the clinical interval between symptom onset and surgery did show discordance in some cases. Perhaps, clinical symptoms may develop or worsen with the onset of edema, which may follow tissue necrosis by several days. From a practical pathological standpoint, our study showed the utility of reticulin stains in pituitary adenomas, even if they displayed severe necrosis. T h e relative preservation of vasculature, despite tissue necrosis, still allowed ascertainment of a disrupted acinar pattern in the necrotic tissues and c o n f i r m e d the underlying process as an adenoma, not n o r m a l anterior pituitary gland. Because of the patchy nature of the necrosis, i m m u n o h i s t o c h e m i s t r y was still informative in 13 of 15 cases, and should be p e r f o r m e d . Only a few of the previous large series had included i m m u n o h i s t o c h e m i cal categorization of pituitary a d e n o m a s with apoplexy (see Table 2). Like Bills et al, is and Mfiller-Jensen a n d Lfldecke, 6 we f o u n d most of apoplectic a d e n o m a s in our series to be endocrine-inactive null cell a d e n o m a s and weak g o n a d o t r o p h cell adenomas. Similar to the study of Onesti et al, 5 occasional corticotroph and prolactin cell a d e n o m a s were identified. Unlike Bonicki et al, 7 we f o u n d no p r e d o m i n a n c e of large growth h o r m o n e - s e c r e t i n g adenomas, Nelson's syndrome, or r e c u r r e n t tumors in o u r patient population. Although four of o u r cases showed cavernous or sphenoid sinus compression or invasion, we disagree with Bonicki et al 7 that apoplexy in pituitary a d e n o m a s is a m a r k e r suggesting invasiveness of the tumor. Eleven of our apoplexy cases showed neither invasion or compression of cavernous sinus by MRI. No correlation between immunohistochemical staining pattern of the a d e n o m a s and type of apoplectic event (ie, bland infarction v h e m o r r h a g i c infarction v clot) was n o t e d in this study. T h e main focus of this study was on pathological features of apoplectic pituitary adenomas, and this was not a detailed MRI study p e r se, but we were interested in correlating any MRI findings in o u r patients with the
matching histological features. Similar to the c o m p u t e d t o m o g r a p h y study of A h m e d et al, 14 we f o u n d fluid/ fluid levels by MRI in three apoplectic a d e n o m a s with acute (one case) or subacute (two cases) clot. Peripheral rim e n h a n c e m e n t ("ring e n h a n c e m e n t " ) could be f o u n d in response to either clot or infarction (Table 1) and c o r r e s p o n d e d to the histological finding of granulation tissue and lymphocytosis. Like A h m e d et al, 14 we n o t e d that peripheral rim e n h a n c e m e n t could be f o u n d in processes of varying acute or chronic duration. In our institutional experience on m o r e than 350 pituitary adenomas, we have n o t e n c o u n t e r e d peripheral rim e n h a n c e m e n t in adenomas, except in the setting of h e m o r r h a g e or infarction. Finally, we included in o u r series patients with longer intervals between symptom onset and surgery. Several of these cases h a d b e e n referred to us because of their confusing histological features. Atrophic pituicytes coupled with lymphocytosis h a d mimicked the a p p e a r a n c e of l y m p h o m a or carcinoma. Immunostaining with pancytokeratin distinguished the apoplectic process f r o m a lymphoma. Although immunoreactivity with pancytokeratin can be seen in carcinomas, our c u r r e n t study illustrates for pathologists the histological features of apoplectic a d e n o m a s and should eliminate an incorrect diagnosis o f carcinoma. Atrophic pituicytes have similarly b e e n described at the edge o f large Rathke cleft cysts, as n o t e d in o u r previous study, 4 where they also may contribute to diagnostic confusion. Although calcification 21 and squamous metaplasia 22 have been r e p o r t e d as late sequelae o f apoplexy in pituitary adenomas, we did not identify these features in o u r series, perhaps because of their rarity. T h e finding of lymphocytosis with a p r e d o m i n a n c e of CD3-positive T lymphocytes may suggest that at least a m o d e s t i m m u n e response develops to d a m a g e d pituitary a d e n o m a cells.
Acknowledgment. The authors thank Drs Steven Cullen and Leonard Zemel for their invaluable assistance in providing clinical information about their patients. We are grateful to Drs Steven Cullen, Walter E. Madsen, Paul M. Scheele, and Philip T. Stoffel, who submitted cases. The photographic support of Bob McCullough, the secretarial help of Ginger Woodward, and the histological expertise of David Davis and Vickie McHenry are greatly appreciated. REFERENCES 1. Cardoso ER, Peterson EW: Pituitary apoplexy: A review. Neurosurgery 14:363-373, 1984 2. Rolih CA, Ober KP: Pituitary apoplexy.Endocrinol Metab Clin North Am 22:291-302, 1993 3. Reid RL, Quigley ME, Yen SSC: Pituitary apoplexy. Arch Neuro142:712-719, 1985 4. Kleinschmidt-DeMasters BtL Lillehei KO, Stears JC: The pathologic, surgical, and MR spectrum of Rathke cleft cysts. Surg Neuro144:19-27, 1995 5. Onesti ST, Wisniewski T, Post KD: Clinical versus subclinical pituitary apoplexy: Presentation, surgical management, and outcome in 21 patients. Neurosurgery 26:980-986, 1990 6. M/iller-Jensen A, Lfidecke D: Clinical aspects of spontaneous necrosis of pituitary tumors (pituitary apoplexy). J Neurol 224:267271, 1981 7. Bonicki W, Kasperlik-ZaluskaA, KoszewskiW, et al: Pituitary apoplexy: Endocrine, surgical and oncological emergency. Incidence,
1264
PITUITARYADENOMAS WITH APOPLEXY (Kleinschmidt-DeMasters & Lillehei) clinical course and treatment with reference to 799 cases of pituitary adenomas. Acta Neurochir 120:118-122, 1993 8. Weisberg LA: Pituitary apoplexy: Association of degenerative change in pituitary adenoma with radiotherapy and detection by cerebral computed tomography. AmJ Med 63:109-115, 1977 9. Wakai S, Fukushima T, Teramoto A, et al: Pituitary apoplexy: Its incidence and clinical significance.J Neurosurg 55:187-193, 1981 10. Ostrov SG, Quencer RM, Hoffman JC, et al: Hemorrhage within pituitary adenomas: How often associated with pituitary apoplexy" syndrome? AJR 153:153-160, 1989 11. Fraioli B, Esposito V, Palma L, et al: Hemorrhagic pituitary adenomas: Clinicopathological features and surgical treatment. Neurosurgery 27:741-748, 1990 12. Wakai S, Yamakawa K, Manaka S, et al: Spontaneous intracranial hemorrhage caused by brain tumor: Its incidence and clinical significance. Neurosurgery 10:437-444, 1982 13. McFadzean RM, Doyle D, Rampling R, et ah Pituitary apoplexy and its effect on vision. Neurosurgery 29:669-675, 1991 14. Ahmed M, Rifai A, A1-Jurf M, et al: Classical pituitary apoplexy presentation and a follow-up of 13 patients. Hormone Res 31:125-132, 1989
15. Tsitsopoulos P, Andrew J, Harrison MJG: Pituitary apoplexy and haemorrhage into adenomas. Postgrad MedJ 62:623-626, 1986 16. Ebersold MJ, Laws ER, Scheithauer BW, et al: Pituitary apoplexy treated by transsphenoidal surgery: A clinocopathological and immunocytochemical study.J Neurosurg 58:315-320, 1983 17. Rovit RL, Fein JM: Pituitary apoplexy: A review and reappraisal.J Neurosurg 37:280-288, 1972 18. Bills DC, Meyer FB, Laws ER, et al: A restrospective analysis of pituitary apoplexy. Neurosurgery 33:602-609, 1993 19. Vidal E, Cevallos R, Vidal J, et al: Twelve cases of pituitary apoplexy. Arch Intern Med 152:1893-1899, 1992 20. Reusch JE, Kleinschmidt-DeMasters BK, Lillehei KO, et al: Preoperative diagnosis of lymphocytic hypophysitis (adenohypophysitis) unresponsive to short course dexamethasone: Case report. Neurosurgery 30:268-272, 1992 21. Kepes JJ, Sayler J, Hiszczynskyj R: Squamous metaplasia following necrosis of the adenohypophysis and of a chromophobe adenoma of the pituitary. Virchows Arch A Pathol Anat Histol 395:69-76, 1982 22. Horiuchi T, Tanaka Y, Kobayashi S, et al: Total capsular calcification in a prolactinoma: Case report. Neurol Med Chirurg 36:729-732, 1996
1265