Radiological and clinical outcomes of pituitary apoplexy: comparison of conservative management versus early surgical intervention

View More View Less
  • 1 Department of Neurosurgery, University of Virginia Health System, Charlottesville, Virginia;
  • | 2 Department of Neurosurgery, MD Anderson Cancer Center, Houston, Texas; and
  • | 3 Surgical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland
Full access

OBJECTIVE

Early surgical intervention for patients with pituitary apoplexy (PA) is thought to improve visual outcomes and decrease mortality. However, some patients may have good clinical outcomes without surgery. The authors sought to compare the radiological and clinical outcomes of patients with PA who were managed conservatively versus those who underwent early surgery.

METHODS

Patients with symptomatic PA were identified. Radiological, endocrinological, and ophthalmological data were reviewed. Patients with progressive visual deterioration or ophthalmoplegia were candidates for early surgery (within 7 days). Patients without visual symptoms or whose symptoms improved on high-dose steroids were treated conservatively. Log-rank and univariate analysis compared clinical and radiological outcomes between those receiving early surgery and those who underwent intended conservative management.

RESULTS

Sixty-four patients with PA were identified: 47 (73.4%) underwent intended conservative management, while 17 (26.6%) had early surgery. Patients receiving early surgery had increased rates of impaired visual acuity (VA; 64.7% vs 27.7%, p = 0.009); visual field (VF) deficits (64.7% vs 19.2%, p = 0.002); and cranial neuropathies (58.8% vs 29.8%, p < 0.05) at presentation. Tumor volumes were greater in the early surgical cohort (15.1 ± 14.8 cm3 vs 4.5 ± 10.3 cm3, p < 0.001). The median clinical and radiological follow-up visits were longer in the early surgical cohort (70.0 and 64.4 months vs 26.0 and 24.7 months, respectively; p < 0.001). Among those with VA/VF deficits, visual outcomes were similar between both groups (p > 0.9). The median time to VA improvement (2.0 vs 3.0 months, p = 0.9; HR 0.9, 95% CI 0.3–3.5) and the median time to VF improvement (2.0 vs 1.5 months; HR 0.8, 95% CI 0.3–2.6, p = 0.8) were similar across both cohorts. Cranial neuropathy improvement was more common in conservatively managed patients (HR 4.8, 95% CI 1.5–15.4, p < 0.01). Conservative management failed in 7 patients (14.9%) and required surgery. PA volumes spontaneously regressed in 95.0% of patients (38/40) with successful conservative management, with a 6-month regression rate of 66.2%. Twenty-seven patients (19 in the conservative and 8 in the early surgical cohorts) responded to a prospectively administered Visual Function Questionnaire-25 (VFQ-25). VFQ-25 scores were similar across both cohorts (conservative 95.5 ± 3.8, surgery 93.2 ± 5.1, p = 0.3). Younger age, female sex, and patients with VF deficits or chiasmal compression were more likely to experience unsuccessful conservative management. Surgical outcomes were similar for patients receiving early versus delayed surgery.

CONCLUSIONS

These data suggest that a majority of patients with PA can be successfully managed without surgical intervention assuming close neurosurgical, radiological, and ophthalmological follow-up is available.

ABBREVIATIONS

CN = cranial nerve; DI = diabetes insipidus; GCS = Glasgow Coma Scale; PA = pituitary apoplexy; VA = visual acuity; VF = visual field; VFQ-25 = Visual Function Questionnaire-25.

OBJECTIVE

Early surgical intervention for patients with pituitary apoplexy (PA) is thought to improve visual outcomes and decrease mortality. However, some patients may have good clinical outcomes without surgery. The authors sought to compare the radiological and clinical outcomes of patients with PA who were managed conservatively versus those who underwent early surgery.

METHODS

Patients with symptomatic PA were identified. Radiological, endocrinological, and ophthalmological data were reviewed. Patients with progressive visual deterioration or ophthalmoplegia were candidates for early surgery (within 7 days). Patients without visual symptoms or whose symptoms improved on high-dose steroids were treated conservatively. Log-rank and univariate analysis compared clinical and radiological outcomes between those receiving early surgery and those who underwent intended conservative management.

RESULTS

Sixty-four patients with PA were identified: 47 (73.4%) underwent intended conservative management, while 17 (26.6%) had early surgery. Patients receiving early surgery had increased rates of impaired visual acuity (VA; 64.7% vs 27.7%, p = 0.009); visual field (VF) deficits (64.7% vs 19.2%, p = 0.002); and cranial neuropathies (58.8% vs 29.8%, p < 0.05) at presentation. Tumor volumes were greater in the early surgical cohort (15.1 ± 14.8 cm3 vs 4.5 ± 10.3 cm3, p < 0.001). The median clinical and radiological follow-up visits were longer in the early surgical cohort (70.0 and 64.4 months vs 26.0 and 24.7 months, respectively; p < 0.001). Among those with VA/VF deficits, visual outcomes were similar between both groups (p > 0.9). The median time to VA improvement (2.0 vs 3.0 months, p = 0.9; HR 0.9, 95% CI 0.3–3.5) and the median time to VF improvement (2.0 vs 1.5 months; HR 0.8, 95% CI 0.3–2.6, p = 0.8) were similar across both cohorts. Cranial neuropathy improvement was more common in conservatively managed patients (HR 4.8, 95% CI 1.5–15.4, p < 0.01). Conservative management failed in 7 patients (14.9%) and required surgery. PA volumes spontaneously regressed in 95.0% of patients (38/40) with successful conservative management, with a 6-month regression rate of 66.2%. Twenty-seven patients (19 in the conservative and 8 in the early surgical cohorts) responded to a prospectively administered Visual Function Questionnaire-25 (VFQ-25). VFQ-25 scores were similar across both cohorts (conservative 95.5 ± 3.8, surgery 93.2 ± 5.1, p = 0.3). Younger age, female sex, and patients with VF deficits or chiasmal compression were more likely to experience unsuccessful conservative management. Surgical outcomes were similar for patients receiving early versus delayed surgery.

CONCLUSIONS

These data suggest that a majority of patients with PA can be successfully managed without surgical intervention assuming close neurosurgical, radiological, and ophthalmological follow-up is available.

ABBREVIATIONS

CN = cranial nerve; DI = diabetes insipidus; GCS = Glasgow Coma Scale; PA = pituitary apoplexy; VA = visual acuity; VF = visual field; VFQ-25 = Visual Function Questionnaire-25.

In Brief

A subset of patients with pituitary apoplexy (PA) may benefit from conservative management in lieu of surgery. In this study, the authors report the clinical, radiological, and ophthalmological outcomes of patients with PA who underwent intended conservative management versus early transsphenoidal surgery. The results suggest that excellent clinical and ophthalmological outcomes are possible for patients with PA and that the majority of patients with PA can be managed in a conservative manner.

Pituitary apoplexy (PA) refers to a clinical syndrome caused by the rapid enlargement of a pituitary adenoma, usually in the setting of infarction or hemorrhage within the tumor.1 PA classically presents with sudden onset of headaches, nausea, emesis, meningismus, altered mental status, decreased vision, or ophthalmoplegia.1–5 Traditionally, PA has been thought to represent a neurosurgical emergency, requiring urgent decompression of the sella and nearby neurovascular structures to improve visual outcomes and decrease mortality.6,7 This assertion, however, has been increasingly challenged by several authors who have performed retrospective analyses on individuals with PA who were managed successfully in a conservative fashion.3,5,8–12 In these series, comparable rates of visual and endocrinological outcomes were achieved in the conservatively and surgically treated patients, although in several of the studies there was a selection bias favoring conservative management for those with better baseline visual and ocular function.3,11 Furthermore, several of the studies were performed in Europe and included patients who developed PA in the 1980s, prior to the routine use of MRI.3,8,9,11 Thus, the generalizability of these studies to North American patients who are currently receiving a diagnosis of PA is uncertain. Furthermore, reports advocating early surgery in the contemporary literature remain common.13,14

We therefore sought to define the clinical, endocrinological, and radiological outcomes of patients undergoing management of PA at a single institution. We attempted to further clarify which patients with PA may benefit from a trial of conservative therapy versus early surgical intervention.

Methods

Study Design

A retrospective, IRB-approved case series was performed at the University of Virginia. Patients diagnosed with PA were identified from a prospectively maintained database and from billing records between January 1, 2007, and February 9, 2019. Patients were excluded if they did not have at least one radiological and clinical follow-up visit. In this series, patients were eligible for analysis if they had 1) radiological imaging consistent with hemorrhagic or infarctive PA, and 2) headaches, visual impairment, or diplopia. All neuroimaging was reviewed by a neuroradiologist and neurosurgeon. Clinical data were censored as of October 26, 2019.

Patient charts were reviewed for demographic, clinical, and endocrinological information upon first presentation and over subsequent follow-up visits. Patients typically underwent a comprehensive ophthalmological evaluation and the presence of visual acuity (VA) loss, visual field (VF) deficits, and cranial nerve (CN) palsies was noted. Improvement in VA, VF deficits, and cranial neuropathies was noted based on subsequent ophthalmological evaluations and on patient-reported symptoms at follow-up visits. An endocrinopathy was noted if a patient had one or more anterior hormone pituitary deficiencies or diabetes insipidus (DI).

All patients who met criteria for this study were then offered to take the National Eye Institute Visual Function Questionnaire-25 (VFQ-25) via telephone at the time of the study. Raw scores were converted into a composite score per survey instructions. The VFQ-25 is a metric of overall visual disability, with higher scores indicating better overall visual function.15

Surgical Intervention

Patients were defined to have undergone early surgery if they required surgery within 1 week of PA diagnosis. In general, we attempt to manage patients presenting with apoplexy at our institution in a conservative fashion. Thus, when a patient presents with apoplexy, they begin to receive high-dose steroids. In general, only those who have worsening VA, cranial neuropathies, or progressive VF deterioration are offered surgery within a week of presentation. In this series, all patients underwent an endoscopic endonasal transsphenoidal approach.

Conservative Management

It is our practice that patients without ophthalmological deficits are not offered upfront surgery and are managed conservatively. Patients who present with visual symptoms or cranial neuropathies, but whose deficits improve on steroids, are also managed conservatively. These patients typically undergo a short-interval MRI (6–8 weeks post-apoplexy) with close neurosurgical, endocrinological, and ophthalmological follow-up. Surgery is offered in a delayed fashion if patients develop progressive VA/VF loss or worsening cranial neuropathy, or if the first follow-up MRI does not show any radiological improvement. For this study, we defined patients as undergoing conservative management if they did not require surgery within 1 week of PA. We noted that conservative measures were unsuccessful in some patients. Here, we defined that a patient required delayed surgery if they failed initial conservative management and required surgery within 3 months of PA.

Tumor Volumes

PA volumes (equal to the adenoma volume plus the hemorrhagic/infarctive components) were calculated on initial imaging and at each follow-up MRI. PA volumes were approximated using the formula abc/2 as previously described.16 PA regression was defined as a 20% reduction of the PA volume, whereas PA progression was defined as a 20% or greater increase in the PA volume.

Statistical Analysis

Descriptive statistics were performed using the mean/median for continuous variables and frequency and percentages for categorical data. For univariate analysis, the Student t-test was used to compare continuous variables, whereas categorical data were compared using chi-square analysis. All tests were 2-sided. Kaplan-Meier analysis was performed to determine actuarial data and to compare time-dependent variables across two populations. A p value < 0.05 was considered significant.

Results

Clinical and Radiological Characteristics at Apoplexy

Sixty-four patients with PA met inclusion criteria for this study. The baseline demographic information for patients is summarized in Table 1. The majority of patients (47, 73.4%) underwent an initial trial of conservative management, whereas 17 patients (26.6%) underwent early surgery. There was no statistical difference between patients in the conservative cohort and those in the early surgical cohort with respect to age, sex, or the presence of pituitary insufficiency at presentation. Similarly, there was no difference in risk factors for patients treated with initial surgery versus conservative management.

TABLE 1.

Clinical characteristics of patients presenting with PA

VariableAll PatientsConservativeEarly Surgeryp Value*
No. of patients644717
Mean age at apoplexy ± SD, yrs53.2 ± 16.653.4 ± 17.952.5 ± 12.70.8
Females, n (%)29 (45.3)21 (44.7)8 (47.1)0.9
Symptoms at diagnosis, n (%)
 Headache50 (78.1)40 (85.1)10 (58.8)0.04
 Visual changes24 (37.5)13 (27.7)11 (64.7)<0.01
 Diplopia26 (40.6)18 (38.3)8 (47.1)0.6
 Nausea25 (39.1)20 (42.6)5 (29.4)0.4
 Photophobia3 (4.7)3 (6.4)0 (0.0)0.6
 Confusion4 (6.3)3 (6.4)1 (5.9)0.9
Clinical findings at diagnosis, n (%)
 Decreased VA15 (23.4)6 (12.8)9 (52.9)<0.01
  Unilateral7 (10.9)1 (2.1)6 (35.3)<0.01
  Bilateral8 (12.5)5 (10.6)3 (17.7)0.4
 VF deficit20 (31.3)9 (19.2)11 (64.7)<0.01
  Bilateral13 (20.3)8 (17.0)5 (29.4)0.3
  Quadrantanopsia1 (1.6)1 (2.1)0 (0.0)0.9
  Hemianopsia17 (26.6)8 (17.0)9 (52.9)<0.01
  Complete field loss2 (3.1)0 (0.0)2 (11.8)0.07
 CN palsy24 (37.5)14 (29.8)10 (58.8)<0.05
  Bilateral0 (0.0)0 (0.0)0 (0.0)0.9
  Complete palsy4 (6.3)1 (2.1)3 (17.7)0.05
  Partial palsy20 (31.3)13 (27.7)7 (41.2)0.4
  CN III20 (31.3)11 (23.4)9 (52.9)0.03
  CN IV1 (1.6)0 (0.0)1 (5.9)0.3
  CN V3 (4.7)0 (0.0)3 (17.6)0.02
  CN VI8 (12.5)4 (8.5)4 (23.5)0.2
Endocrine findings at diagnosis, median (IQR)
 Prolactin11.4 (2.25–35.5)10.4 (2.3–30.5)13.8 (2.1–53.8)0.1
 Cortisol6.9 (3.4–14.3)7.7 (4.3–14.8)4.9 (1.0–11.9)0.8
 ACTH18.9 (7.0–28.0)18.9 (13.0–28.0)15.5 (5.5–43.3)0.9
 TSH0.8 (0.4–1.9)1.1 (0.6–2.2)0.5 (0.2–0.7)0.2
 Free T40.9 (0.7–1.1)0.9 (0.9–1.1)0.7 (0.7–1.0)0.1
 GH0.7 (0.3–1.7)0.7 (0.2–1.7)0.5 (0.3–1.6)0.7
 IGF-192.0 (66.3–219.3)91.5 (64.8–209.8)97.5 (72.3–228.0)0.8
 LH1.9 (0.7–3.9)2.4 (0.9–4.5)1.0 (0.2–1.7)0.1
Endocrinopathy at diagnosis, n (%)40 (62.5)27 (55.3)13 (76.5)0.2
 Adrenal sufficiency14 (21.9)11 (23.4)3 (17.7)0.7
 Central hypothyroidism12 (18.8)7 (14.9)5 (29.4)0.3
 GH/IGF-1 deficiency6 (9.4)4 (8.5)2 (11.8)0.7
 Gonadotropin deficiency16 (25.0)12 (25.5)4 (23.5)0.9
 DI2 (3.1)2 (4.3)0 (0.0)0.9
Functional adenoma, n (%)
 CD0 (0.0)0 (0.0)0 (0.0)>0.99
 Acromegaly0 (0.0)0 (0.0)0 (0.0)>0.99
 Prolactinoma4 (6.3)2 (4.3)2 (11.8)0.3
Risk factors for apoplexy, n (%)
 DM15 (23.4)12 (25.5)3 (17.7)0.7
 HTN28 (43.8)23 (48.9)5 (29.4)0.3
 Recent surgery11 (17.2)10 (21.3)1 (5.9)0.3
 Dopamine agonist usage4 (6.3)3 (6.4)1 (5.9)0.9
Median clinical follow-up (IQR), mos35.5 (14–64.8)26.0 (12–53.0)70.0 (40.5–98.5)<0.001
Median radiological follow-up (IQR), mos29.3 (13.8–63.3)24.7 (8.4–50.0)64.4 (31.2–99.9)<0.001

ACTH = adrenocorticotropic hormone; CD = Cushing’s disease; DM = diabetes mellitus; GH = growth hormone; HTN = hypertension; IGF-1 = insulin-like growth factor–1; LH = luteinizing hormone; TSH = thyroid-stimulating hormone.

Comparison of conservative versus early surgical cohorts.

In the early surgical cohort, 64.7% (11/17) of patients endorsed visual changes at presentation compared with 27.7% (13/47) of those managed conservatively (p = 0.009). Similarly, diminished VA was more prevalent in the early surgical cohort (9/17, 52.9%) compared with the conservative management cohort (6/47, 12.8%; p = 0.002). Bilateral VA loss was present in each group equally (p = 0.4). There were increased rates of VF deficits in the surgical cohort (11/17 [64.7%] vs 9/47 [19.2%], p = 0.002). Most of the VF deficits noted in this series were hemianopsias (early surgery 9/17 [52.9%] vs conservative 8/47 [17.0%], p = 0.009). Complete field loss was only seen in the surgical cohort (11.8% [2/17] vs 0.0%), but given the small sample size, statistical significance was not achieved (p = 0.07).

Likewise, there were increased rates of CN deficits noted in the early surgical cohort (10/17, 58.8%) compared with the conservative cohort (14/47, 29.8%; p < 0.05). CN III palsies accounted for the majority of CN deficits observed for all patients (20, 31.3%). As a result, those in the surgical cohort tended to have increased rates of CN III palsies (9/17 [52.9%] vs 11/47 [23.4%], p = 0.03). Complete CN palsies were present in only 4 patients, 3 of whom underwent surgery. Trigeminal nerve dysfunction was apparent exclusively in the surgical cohort (17.6% [3/17] vs 0.0%, p = 0.016). Altered mental status (i.e., Glasgow Coma Scale [GCS] score = 14) was present in 4 individuals (6.3%) in this series. There was no statistical difference between the conservative and early surgical cohorts with regard to patients presenting with confusion (6.4% vs 5.9%, p = 0.9).

Overall, 24 patients (51.1%) treated conservatively had no ophthalmological findings compared with no patients (0.0%) who underwent early surgery (p < 0.0001).

Table 2 summarizes the salient radiological findings of the apoplectic tumors at initial diagnosis. Hemorrhagic apoplexy was the predominant subtype of PA and was represented more commonly in those undergoing conservative management of PA (43/47 [91.5%] vs 12/17 [70.6%], p < 0.05). Not surprisingly, chiasmal compression and cavernous sinus invasion were more common in the surgical cohorts (Table 2). Similarly, patients undergoing early surgery had larger tumors (maximum PA diameter 2.8 ± 0.9 cm vs 2.1 ± 0.7 cm, p = 0.002; PA volume 15.1 ± 14.8 cm3 vs 4.5 ± 10.3 cm3, p < 0.001).

TABLE 2.

Radiological findings of patients presenting with PA

Radiological FindingAll PatientsConservativeEarly Surgeryp Value*
Hemorrhagic apoplexy, n (%)55 (85.9)43 (91.5)12 (70.6)<0.05
Chiasmal compression, n (%)35 (54.7)22 (46.8)13 (76.5)<0.05
Cavernous sinus invasion, n (%)25 (39.1)14 (29.8)11 (64.7)0.02
Thickened sphenoidal mucosa, n (%)9 (14.1)5 (10.6)4 (23.5)0.20
Mean max diameter of apoplectic tumor ± SD, cm2.3 ± 0.82.1 ± 0.72.8 ± 0.9<0.01
Mean apoplectic tumor volume ± SD, cm36.0 ± 7.04.5 ± 10.315.1 ± 14.8<0.001

Comparison between conservative and early surgical cohorts.

The clinical and radiological follow-up visits for patients undergoing conservative management (26.0 and 24.7 months, respectively) in this study were shorter than in those who underwent early surgical intervention (70.0 and 64.4 months, respectively; p < 0.001). The ophthalmological follow-up was likewise shorter in the conservative cohort (conservative 19.3 months, early surgery 48.5 months, p = 0.017).

Clinical and Radiological Outcomes

Table 3 represents an intention-to-treat analysis comparing the clinical and radiological outcomes of patients undergoing conservative versus surgical management of PA. In examining patients who originally presented with VA or VF loss, improvement occurred at statistically similar rates across both cohorts. Importantly, there was no statistically significant rate of VA/VF worsening across either group. Those undergoing surgery had slightly shorter median times to VA/VF improvement (VA 2.0 vs 3.0 months, p = 0.9; VF 1.5 vs 2.0 months, p = 0.8), but these were not statistically significant (Fig. 1A).

TABLE 3.

Clinical outcomes between patients intended to be treated conservatively and those undergoing early surgery

VariableConservative Management*Early SurgeryHR (95% CI)p Value
VA improvement5 (83.3%)7 (77.8%)>0.9
Median time to VA improvement, mos3.02.00.9 (0.3–3.5)0.9
VA worsening1 (2.1%)0 (0.0%)>0.9
VF improvement§8 (88.9%)10 (90.9%)>0.9
Median time to VF improvement, mos2.01.50.8 (0.3–2.6)0.8
VF worsening1 (2.1%)0 (0.0%)>0.9
Cranial neuropathy improvement14 (100.0%)6 (60.0%)0.02
Median time to cranial neuropathy improvement, mos1.535.54.8 (1.5–15.4)<0.01
Cranial neuropathy worsening1 (2.1%)1 (5.9%)0.5
New endocrinopathy28 (59.6%)9 (52.9%)0.8
Median time to new endocrinopathy, mos6.0NR1.6 (0.7–3.7)0.3
PA progression0 (0.0%)NA
PA recurrence1 (2.1%)1 (5.9%)0.5
Mean VFQ-25 score ± SD**95.5 ± 3.8 93.2 ± 5.10.3

NA = not applicable, because the apoplectic tumors were removed (and thus could not progress) before the first follow-up MRI; NR = not reached.

Values are expressed as number of patients (%) unless otherwise indicated.

Intention-to-treat analysis.

Conservative versus early surgery.

Of 6 and 9 patients in the conservative and early surgical groups, respectively.

Of 9 and 11 patients in the conservative and early surgical groups, respectively.

Of 14 and 10 patients in the conservative and early surgical groups, respectively.

Of 19 and 8 patients in the conservative and early surgical groups, respectively.

FIG. 1.
FIG. 1.

A: The rates of VF improvement were similar between those patients treated with surgery and those who underwent intended conservative management (HR 0.8, 95% CI 0.3–2.6, p = 0.8). B: There was a decreased time to CN improvement in patients treated with intended conservative management compared to early surgery (HR 4.8, 95% CI 1.5–15.4, p < 0.01). C: The actuarial rates of apoplectic tumor regression in those undergoing intended conservative management were 27.3%, 66.2%, and 77.5%, at 2, 6, and 12 months, respectively (patients who underwent surgery were censored at the date of tumor progression necessitating surgery).

Interestingly, the rate of cranial neuropathy improvement was higher in the conservative cohort (14/14 [100%] vs 6/10 [60%], p = 0.02). Likewise, the median time to cranial neuropathy improvement was faster in the conservative cohort (1.5 vs 35.5 months; HR 4.8, 95% CI 1.5–15.4, p < 0.01; Fig. 1B). Of the 14 patients who had CN deficits improve in the conservative cohort, 10 (71.4%) resolved completely. By comparison, of the 6 patients who had improved cranial neuropathies in the surgical group, 5 (83.3%) completely resolved. By the last follow-up, of those with CN deficits at initial presentation, 28.6% (4/14) of patients in the conservative cohort and 20.0% (2/10) in the early surgical cohort had persistent diplopia (p > 0.9).

Of the 64 patients analyzed in this study, 27 (42.2%) were reachable and willing to complete the VFQ-25. The response rates were 40.4% (19/47) and 47.1% (8/17) for patients in the conservative and early surgical cohorts, respectively. Overall, the mean composite scores indicated that patients managed in both groups had low levels of visual disability (conservative 95.5 ± 3.8 vs surgery 93.2 ± 5.1). There was no significant statistical difference between the two groups with respect to VFQ-25 scores (p = 0.3) on intention-to-treat analysis (Table 3).

Spontaneous PA regression rates were favorable in the conservative cohort. The actuarial spontaneous PA regression rates were 27.3%, 66.2%, and 77.5% at 2, 6, and 12 months, respectively, following apoplexy in the conservative group (Fig. 1C).

Of 47 patients who underwent initial conservative management, 7 patients (14.9%) ultimately required surgery. The median time to surgery after apoplexy for these 7 individuals was 50 days (range 17–81 days). Of the 7 patients who required surgery, 3 (42.9%) had no improvement in their radiological findings on their initial follow-up MRI and were indicated for surgery. One patient presented with a second apoplectic event, and 3 (42.9%) had worsening vision and were thus indicated for surgery. All 3 patients whose vision deteriorated had VA or VF deficits at initial presentation.

All 24 patients who presented without visual or CN deficits were managed conservatively. Three of these patients (12.5%) required surgery within 3 months of apoplexy. No patient with normal vision developed worsening visual function. Of the 3 patients who required surgery, 1 presented with a second apoplectic episode and the other 2 had persistent chiasmal compression on their first follow-up MRI.

Good ophthalmological outcomes were still appreciated when examining only those who were successfully managed in a conservative fashion. All patients (4/4, 100%) with VA loss showed clinical improvement, 100% of patients (5/5) with VF deficits noted improvement, and 100% of patients (14/14) with CN deficits had cranial neuropathy improvement. The actuarial PA regression rates (excluding those who underwent surgery) were 30.0%, 70.0%, and 80.0% at 2, 6, and 12 months, respectively, following PA. Of the 40 patients who did not undergo surgery within 3 months, 28 (70.0%) had some radiological evidence of persistent adenoma on MRI. None of those tumors were causing chiasmal compression, however. Seven of the 40 patients eventually underwent scheduled resection of their pituitary adenoma (median time to surgery 4.1 months, IQR 3.4–14.1 months).

No significant difference was noted when comparing visual outcomes for all patients who underwent surgical management versus those who were treated successfully in a conservative fashion. Indeed, the rates of VA improvement (successful conservative treatment 100.0% [4/4], surgery 72.7% [8/11], p = 0.52) and VF improvement (successful conservative treatment 100.0% [5/5], surgery 86.7% [12/15], p > 0.9) were statistically similar across both groups. No patient in whom intended conservative management failed presented with a CN deficit; thus, the CN improvement rate is the same as was listed in Table 3. Of the 19 respondents to the VFQ-25 survey who were intended to be treated in a conservative fashion, 5 ultimately required surgery within 3 months of PA. When these individuals were analyzed with the 8 other surgically treated patients, VFQ-25 composite scores remained similar across both groups (successful conservative treatment 95.7 ± 3.3 [n = 14], surgery 93.8 ± 5.0 [n = 13], p = 0.3).

Factors Associated With Failed Conservative Management

We examined factors associated with failed conservative management (Table 4). On univariate analysis, younger age (p = 0.03), female sex (p = 0.03), presence of VF deficits (p = 0.02), and radiological evidence of chiasmal compression (p = 0.04) were associated with failed conservative management. Interestingly, decreased VA, the presence of CN palsies, tumor size, and cavernous sinus invasion were not associated with the need for delayed surgery.

TABLE 4.

Univariate analysis factors associated with failed conservative management

FactorSuccessful Conservative (%)Delayed Surgery (%)*p Value
No. of patients407
Mean age ± SD, yrs55.8 ± 17.839.9 ± 10.40.03
Female15 (37.5)6 (85.7)0.03
Decreased VA4 (10.0)2 (28.6)0.2
VF deficit5 (12.5)4 (57.1)0.02
CN palsy14 (35.0)0 (0.0)0.09
Mean tumor diameter ± SD, cm2.0 ± 0.72.3 ± 0.80.4
Mean tumor volume ± SD, cm36.1 ± 7.45.7 ± 4.30.9
Chiasmal compression16 (40.0)6 (85.7)0.04
Cavernous sinus invasion13 (32.5)1 (14.3)0.7
Hemorrhagic apoplexy37 (92.5)6 (85.7)0.5
Thickened sphenoidal mucosa5 (12.5)0 (0.0)>0.9
Endocrinopathy at presentation24 (60.0)3 (42.9)0.4

Defining delayed surgery as within 3 months of apoplexy.

Surgical Details

In total, 24 patients (37.5%) required surgery for PA in this study. All patients underwent an endoscopic endonasal transsphenoidal approach to their tumor. Gross-total resection of the tumor was achieved in 19 patients (79.2%). Intraoperative CSF leaks were evident in 11 of the surgical cases (45.8%), whereas postoperative leaks occurred in 2 patients (8.3%; Table 5). There was no difference in intraoperative or postoperative outcomes for patients undergoing early versus delayed surgery (Table 5). Notably, the rate of gross-total resection and the need for repeat resection or postoperative radiosurgery were similar in both cohorts. There was a trend toward higher postoperative DI rates in the delayed surgical cohort, but this was not statistically significant (p = 0.2).

TABLE 5.

Surgical outcomes for PA

VariableAll (%)Early Surgery (%)Delayed Surgery (%)p Value*
No. of patients24177
Early surgery17 (70.8)NA
Endoscopic approach24 (100.0)17 (100.0)7 (100.0)>0.9
Gross-total resection19 (79.2)13 (76.5)6 (85.7)>0.9
Intraop CSF leak11 (45.8)8 (47.1)3 (42.9)>0.9
Postop CSF leak2 (8.3)2 (11.8)0 (0.0)>0.9
Postop DI3 (12.5)1 (5.9)2 (28.6)0.2
Adjuvant SRS4 (16.7)2 (11.8)2 (28.6)0.6
Repeat resection2 (8.3)1 (5.9)1 (14.3)0.5

SRS = stereotactic radiosurgery.

Early versus delayed surgery.

Discussion

Individuals with PA usually present to neurosurgical attention after developing the acute onset of headaches, nausea, emesis, visual loss, or ocular paresis.1,2 Although rare, with an incidence of 0.17 episodes per 100,000 person-years, it is estimated that 2%–12% of individuals with pituitary adenomas may develop apoplexy.1,2,17 Given that the majority of patients present with life-threatening adrenal insufficiency and that more than half of affected patients may present with visual loss or ophthalmoplegia, it is critical that the optimal medical and surgical management of these patients be clearly defined.

Recently, guidelines have been published that detail the optimal diagnosis and early medical management strategies for individuals with PA.4 All individuals with suspected PA should obtain expedited neuroimaging by MRI and should have a comprehensive metabolic and endocrinological evaluation performed. Individuals who are hemodynamically unstable or those with altered mental status (GCS score < 15), reduced VA, or VF deficits are to be started on empirical steroids for the treatment of presumed adrenal insufficiency and to reduce swelling in the parasellar region.4

While guidelines are generally agreed upon for the initial diagnosis and medical management of individuals with PA, there remains a significant degree of controversy in the literature regarding the use and timing of surgery in these patients.4,8,18 In the largest published series of PA, Singh et al. described the outcomes of 87 patients with PA at a single institution.12 In their series, 26 patients (29.9%) underwent a trial of conservative management, with approximately one-third of these patients ultimately requiring surgery. The estimation of failure of conservative management in this series may be somewhat overstated, as 25% of patients underwent delayed surgery 4 or more months after the initial apoplectic event. While patients treated with surgery and conservative management had good ophthalmological and radiological outcomes in their series, almost 50% of patients in the surgical arm had no visual deficits at presentation. Therefore, it is plausible that a considerable proportion of the patients treated with surgery in their series may have been able to be successfully managed without surgical intervention.

The reported rates of conservative management of PA have ranged between 20.7% and 74.4% in other series.3,9–12,19,20 The time to surgery among surgical cohorts described in the literature has likewise varied considerably. Indeed, the time frame in which patients were said to have undergone surgical management of PA has ranged from 7 days to 4 months.3,11 Failures of conservative management (i.e., those requiring “delayed surgery”) have been described as patients requiring surgery up to 1 year after PA.10 To make informed, evidence-based decisions on the management of patients, we performed an intention-to-treat analysis in which patients who were intended to be managed conservatively were compared to those who required early surgery. Here, we standardized the definition of early surgery (within 1 week of PA) and the definition of delayed surgery (within 3 months).

Overall, our study represents the largest series of conservatively managed cases of PA published to date and the second largest series of apoplexy reported in the recent literature. In this report, the majority of patients (73.4%) underwent a trial of conservative management with good clinical and radiological outcomes. Even some patients who presented with ophthalmological deficits on initial presentation had clinical improvement without surgery. In these cases, patients received high-dose steroids and had objective and/or subjective improvement in their ophthalmological examination during their initial hospitalization. In our intention-to-treat analysis, patients who underwent earlier surgery had higher rates of VA/VF and CN deficits, which is consistent with other reports in the literature.9,12,20 Nevertheless, when examining only patients with ophthalmological deficits, the rates of VA/VF improvement in this study were comparable to those who underwent surgery. This notion has been described previously with VA improvement rates consistently reported between 86% and 100% for patients treated without surgery.9–12,20 Likewise, the rates of VF recovery have been reported to be between 75% and 100% for patients treated conservatively.9,12 Our findings, like those reported previously, likely reflect a selection bias in which patients with more extensive VA/VF deficits are more likely to be selected for surgical intervention.

Interestingly, in our study, surgery was associated with decreased rates of cranial neuropathy recovery and slower times to cranial neuropathy improvement. Although there was a small trend toward patients having complete cranial neuropathies in the surgical cohort, this was not statistically significant. While a selection bias does likely exist here, these data do suggest that a subset of patients (even with ophthalmological deficits) can be successfully treated without surgery.

Apoplexy regression following PA has been previously reported, but no study to date has quantified the actuarial rate of PA regression.12 The 6-month PA regression rate reported in this study was approximately 66% (censoring patients who required surgery). Seo et al. recently reported that apoplectic adenomas that had noticeable tumor regression volumes 6 months following PA were associated with good outcomes.20

Overall, of those who underwent an intended trial of medical management in this series, approximately 15% required delayed surgery (within 3 months of PA). Younger patients with VF deficits and who had evidence of chiasmal compression on MRI were more likely to experience unsuccessful conservative management in this study. Thus, for patients who undergo conservative management, close neurosurgical, radiological, and neuroophthalmological follow-up is essential.

In our series, 7 of 40 patients (17.5%) who were deemed to be managed successfully in a conservative fashion eventually underwent elective transsphenoidal surgery. In these cases, surgery was not performed in an urgent fashion and was deemed indicated irrespective of the apoplectic event. As gross-total resection rates of apoplectic tumors are high in our series, we understand the argument that early surgery may provide a surgical cure for many patients. Nevertheless, as many tumors regress following PA, we feel it is a reasonable strategy to offer patients a chance to avoid surgery completely.

Our study suggests that the clinical and ophthalmological outcomes are good in most patients with PA based both on last clinical follow-up and on VFQ-25 survey results. Surgery continues to play an important role in select patients with severe and progressive ophthalmological deficits. Nevertheless, our data suggest that a trial of conservative management is reasonable in most patients and that the majority of patients with PA do not require urgent surgery. This appears to especially be the case when patients lack any chiasmal compression or visual symptoms at presentation. No patient with preserved VA/VFs at initial presentation had deterioration of their vision within 3 months of surgery. While some of these patients may require delayed surgery, most do not.

Like any retrospective analysis, our study has several limitations. First, there is undoubtedly a selection bias in favoring surgery in more severe cases of PA. Thus, our results should be interpreted with caution. The main findings here are not that surgery and medical management are equivalent for all patients. Rather, our data suggest that in carefully selected cases, close follow-up is a reasonable upfront approach. This appears to especially be the case when no visual signs or symptoms are present. Furthermore, there is a notable difference in the duration of radiological and clinical follow-up for patients undergoing surgery. We attempted to mitigate this by comparing actuarial data across both populations rather than comparing patients at their last follow-up. The shorter follow-up in the conservative cohort may underestimate rates of tumor recurrence/progression. Certainly, these data continue to be hypothesis-generating and future, multicenter registries or prospective trials may be needed to further clarify the role of medical management for patients with PA.

Conclusions

This study suggests that conservative management is a viable strategy in the majority of patients with PA, particularly those who present with no visual deficits or less severe deficits (i.e., incomplete bitemporal hemianopsia or partial cranial neuropathy) that improve with steroids, as long as close neurosurgical and endocrine follow-up is available. Conversely, surgery may be considered in patients with severe (i.e., dense bitemporal VF deficit, blindness) and/or progressive ophthalmological deficits.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Jane, Shepard. Acquisition of data: all authors. Analysis and interpretation of data: all authors. Drafting the article: Jane, Shepard, Snyder. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Statistical analysis: Shepard, Snyder. Administrative/technical/material support: Jane. Study supervision: Jane, Shepard.

References

  • 1

    Randeva HS, Schoebel J, Byrne J, et al. . Classical pituitary apoplexy: clinical features, management and outcome. Clin Endocrinol (Oxf). 1999;51(2):181188.

    • Search Google Scholar
    • Export Citation
  • 2

    Briet C, Salenave S, Bonneville JF, et al. . Pituitary apoplexy. Endocr Rev. 2015;36(6):622645.

  • 3

    Bujawansa S, Thondam SK, Steele C, et al. . Presentation, management and outcomes in acute pituitary apoplexy: a large single-centre experience from the United Kingdom. Clin Endocrinol (Oxf). 2014;80(3):419424.

    • Search Google Scholar
    • Export Citation
  • 4

    Rajasekaran S, Vanderpump M, Baldeweg S, et al. . UK guidelines for the management of pituitary apoplexy. Clin Endocrinol (Oxf). 2011;74(1):920.

    • Search Google Scholar
    • Export Citation
  • 5

    Singh R, Zhou Z, Tisnado J, et al. . A novel magnetic resonance imaging segmentation technique for determining diffuse intrinsic pontine glioma tumor volume. J Neurosurg Pediatr. 2016;18(5):565572.

    • Search Google Scholar
    • Export Citation
  • 6

    Bills DC, Meyer FB, Laws ER Jr, et al. . A retrospective analysis of pituitary apoplexy. Neurosurgery. 1993;33(4):602609.

  • 7

    da Motta LA, de Mello PA, de Lacerda CM, et al. . Pituitary apoplexy. Clinical course, endocrine evaluations and treatment analysis. J Neurosurg Sci. 1999;43(1):2536.

    • Search Google Scholar
    • Export Citation
  • 8

    Ayuk J, McGregor EJ, Mitchell RD, Gittoes NJL. Acute management of pituitary apoplexy—surgery or conservative management?. Clin Endocrinol (Oxf). 2004;61(6):747752.

    • Search Google Scholar
    • Export Citation
  • 9

    Gruber A, Clayton J, Kumar S, et al. . Pituitary apoplexy: retrospective review of 30 patients—is surgical intervention always necessary?. Br J Neurosurg. 2006;20(6):379385.

    • Search Google Scholar
    • Export Citation
  • 10

    Leyer C, Castinetti F, Morange I, et al. . A conservative management is preferable in milder forms of pituitary tumor apoplexy. J Endocrinol Invest. 2011;34(7):502509.

    • Search Google Scholar
    • Export Citation
  • 11

    Sibal L, Ball SG, Connolly V, et al. . Pituitary apoplexy: a review of clinical presentation, management and outcome in 45 cases. Pituitary. 2004;7(3):157163.

    • Search Google Scholar
    • Export Citation
  • 12

    Singh TD, Valizadeh N, Meyer FB, et al. . Management and outcomes of pituitary apoplexy. J Neurosurg. 2015;122(6):14501457.

  • 13

    Tu M, Lu Q, Zhu P, Zheng W. Surgical versus non-surgical treatment for pituitary apoplexy: a systematic review and meta-analysis. J Neurol Sci. 2016;370:258262.

    • Search Google Scholar
    • Export Citation
  • 14

    Turgut M, Özsunar Y, Başak S, et al. . Pituitary apoplexy: an overview of 186 cases published during the last century. Acta Neurochir (Wien). 2010;152(5):749761.

    • Search Google Scholar
    • Export Citation
  • 15

    Mangione CM, Lee PP, Gutierrez PR, et al. . Development of the 25-item National Eye Institute Visual Function Questionnaire. Arch Ophthalmol. 2001;119(7):10501058.

    • Search Google Scholar
    • Export Citation
  • 16

    Lundin P, Pedersen F. Volume of pituitary macroadenomas: assessment by MRI. J Comput Assist Tomogr. 1992;16(4):519528.

  • 17

    Raappana A, Koivukangas J, Ebeling T, Pirilä T. Incidence of pituitary adenomas in Northern Finland in 1992-2007. J Clin Endocrinol Metab. 2010;95(9):42684275.

    • Search Google Scholar
    • Export Citation
  • 18

    Bi WL, Dunn IF, Laws ER Jr. Pituitary apoplexy. Endocrine. 2015;48(1):6975.

  • 19

    Almeida JP, Sanchez MM, Karekezi C, et al. . Pituitary apoplexy: results of surgical and conservative management clinical series and review of the literature. World Neurosurg. 2019;130:e988e999.

    • Search Google Scholar
    • Export Citation
  • 20

    Seo Y, Kim YH, Dho YS, et al. . The outcomes of pituitary apoplexy with conservative treatment: experiences at a single institution. World Neurosurg. 2018;115:e703e710.

    • Search Google Scholar
    • Export Citation

Contributor Notes

Correspondence John A. Jane Jr.: University of Virginia Health System, Charlottesville, VA. johnjanejr@virginia.edu.

INCLUDE WHEN CITING Published online April 30, 2021; DOI: 10.3171/2020.9.JNS202899.

Disclosures The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

  • View in gallery

    A: The rates of VF improvement were similar between those patients treated with surgery and those who underwent intended conservative management (HR 0.8, 95% CI 0.3–2.6, p = 0.8). B: There was a decreased time to CN improvement in patients treated with intended conservative management compared to early surgery (HR 4.8, 95% CI 1.5–15.4, p < 0.01). C: The actuarial rates of apoplectic tumor regression in those undergoing intended conservative management were 27.3%, 66.2%, and 77.5%, at 2, 6, and 12 months, respectively (patients who underwent surgery were censored at the date of tumor progression necessitating surgery).

  • 1

    Randeva HS, Schoebel J, Byrne J, et al. . Classical pituitary apoplexy: clinical features, management and outcome. Clin Endocrinol (Oxf). 1999;51(2):181188.

    • Search Google Scholar
    • Export Citation
  • 2

    Briet C, Salenave S, Bonneville JF, et al. . Pituitary apoplexy. Endocr Rev. 2015;36(6):622645.

  • 3

    Bujawansa S, Thondam SK, Steele C, et al. . Presentation, management and outcomes in acute pituitary apoplexy: a large single-centre experience from the United Kingdom. Clin Endocrinol (Oxf). 2014;80(3):419424.

    • Search Google Scholar
    • Export Citation
  • 4

    Rajasekaran S, Vanderpump M, Baldeweg S, et al. . UK guidelines for the management of pituitary apoplexy. Clin Endocrinol (Oxf). 2011;74(1):920.

    • Search Google Scholar
    • Export Citation
  • 5

    Singh R, Zhou Z, Tisnado J, et al. . A novel magnetic resonance imaging segmentation technique for determining diffuse intrinsic pontine glioma tumor volume. J Neurosurg Pediatr. 2016;18(5):565572.

    • Search Google Scholar
    • Export Citation
  • 6

    Bills DC, Meyer FB, Laws ER Jr, et al. . A retrospective analysis of pituitary apoplexy. Neurosurgery. 1993;33(4):602609.

  • 7

    da Motta LA, de Mello PA, de Lacerda CM, et al. . Pituitary apoplexy. Clinical course, endocrine evaluations and treatment analysis. J Neurosurg Sci. 1999;43(1):2536.

    • Search Google Scholar
    • Export Citation
  • 8

    Ayuk J, McGregor EJ, Mitchell RD, Gittoes NJL. Acute management of pituitary apoplexy—surgery or conservative management?. Clin Endocrinol (Oxf). 2004;61(6):747752.

    • Search Google Scholar
    • Export Citation
  • 9

    Gruber A, Clayton J, Kumar S, et al. . Pituitary apoplexy: retrospective review of 30 patients—is surgical intervention always necessary?. Br J Neurosurg. 2006;20(6):379385.

    • Search Google Scholar
    • Export Citation
  • 10

    Leyer C, Castinetti F, Morange I, et al. . A conservative management is preferable in milder forms of pituitary tumor apoplexy. J Endocrinol Invest. 2011;34(7):502509.

    • Search Google Scholar
    • Export Citation
  • 11

    Sibal L, Ball SG, Connolly V, et al. . Pituitary apoplexy: a review of clinical presentation, management and outcome in 45 cases. Pituitary. 2004;7(3):157163.

    • Search Google Scholar
    • Export Citation
  • 12

    Singh TD, Valizadeh N, Meyer FB, et al. . Management and outcomes of pituitary apoplexy. J Neurosurg. 2015;122(6):14501457.

  • 13

    Tu M, Lu Q, Zhu P, Zheng W. Surgical versus non-surgical treatment for pituitary apoplexy: a systematic review and meta-analysis. J Neurol Sci. 2016;370:258262.

    • Search Google Scholar
    • Export Citation
  • 14

    Turgut M, Özsunar Y, Başak S, et al. . Pituitary apoplexy: an overview of 186 cases published during the last century. Acta Neurochir (Wien). 2010;152(5):749761.

    • Search Google Scholar
    • Export Citation
  • 15

    Mangione CM, Lee PP, Gutierrez PR, et al. . Development of the 25-item National Eye Institute Visual Function Questionnaire. Arch Ophthalmol. 2001;119(7):10501058.

    • Search Google Scholar
    • Export Citation
  • 16

    Lundin P, Pedersen F. Volume of pituitary macroadenomas: assessment by MRI. J Comput Assist Tomogr. 1992;16(4):519528.

  • 17

    Raappana A, Koivukangas J, Ebeling T, Pirilä T. Incidence of pituitary adenomas in Northern Finland in 1992-2007. J Clin Endocrinol Metab. 2010;95(9):42684275.

    • Search Google Scholar
    • Export Citation
  • 18

    Bi WL, Dunn IF, Laws ER Jr. Pituitary apoplexy. Endocrine. 2015;48(1):6975.

  • 19

    Almeida JP, Sanchez MM, Karekezi C, et al. . Pituitary apoplexy: results of surgical and conservative management clinical series and review of the literature. World Neurosurg. 2019;130:e988e999.

    • Search Google Scholar
    • Export Citation
  • 20

    Seo Y, Kim YH, Dho YS, et al. . The outcomes of pituitary apoplexy with conservative treatment: experiences at a single institution. World Neurosurg. 2018;115:e703e710.

    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 719 719 114
Full Text Views 110 110 28
PDF Downloads 98 98 44
EPUB Downloads 0 0 0