Stereotactic radiosurgery for pituitary metastases: a systematic review and single-arm meta-analysis of local control and survival outcomes

Abstract

Background

Pituitary metastases (PM) are rare intracranial secondary tumors, most often arising from breast and lung primaries. They are clinically significant due to their association with advanced systemic malignancy and potential to cause visual, neurological, and endocrinological morbidity. Conventional therapies such as surgery and fractionated radiotherapy have limited efficacy or high morbidity. Stereotactic radiosurgery (SRS) has emerged as a minimally invasive alternative, but available evidence is limited to small, heterogeneous series.

Objective

To evaluate the effectiveness and safety of SRS in patients with PM by synthesizing data on tumor control, survival outcomes, progression-free survival (PFS), and treatment-related morbidity.

Methods

A systematic review and meta-analysis were performed according to PRISMA guidelines. PubMed, Embase, Scopus, Web of Science, and Cochrane Library were searched through June 2025. Eligible studies included patients with radiologically or histologically confirmed PM treated with SRS. Outcomes of interest were local tumor control, survival, PFS, radiological response, and complications, including hypopituitarism, optic neuropathy, and radiation necrosis. Data were pooled using a random-effects single-arm meta-analysis model.

Results

Six retrospective case series comprising 142 patients were included. Pooled local control following SRS was 89% (95% CI 75–96%). Six-month PFS was 97% (95% CI 83–100%). One- and two-year survival rates were 53% (95% CI 32–73%) and 40% (95% CI 15–72%), respectively. Tumor regression occurred in 34% and stable disease in 33%, while only 7% demonstrated progression. Treatment-related morbidity was low: new-onset hypopituitarism (1%), optic neuropathy (2%), and radiation necrosis (1%).

Conclusion

SRS provides effective and durable local control of pituitary metastases with a favorable safety profile and low rates of treatment-related toxicity. Despite the poor systemic prognosis associated with PM, SRS offers meaningful intracranial disease control and symptom relief, particularly in well-selected patients with stable systemic disease. Larger prospective, standardized studies are needed to further clarify long-term endocrine and neurological outcomes.

Introduction

Pituitary metastases (PM) are an uncommon manifestation of systemic malignancy, accounting for less than 1% of all intracranial secondary tumors [1]. Despite their rarity, PM are clinically significant due to their association with advanced systemic disease and their potential to cause substantial neurological and endocrinological morbidity, including visual compromise, cranial nerve palsies, diabetes insipidus, and anterior hypopituitarism [2–4]. Improvements in oncologic survival and increased use of high-resolution neuroimaging have contributed to more frequent recognition of this entity.

Historically, management strategies for PM have included surgical decompression, fractionated external beam radiotherapy, and systemic therapy, each associated with notable limitations such as surgical morbidity, delayed local control, or limited intracranial efficacy [5]. Over the past three decades, stereotactic radiosurgery (SRS) has emerged as a minimally invasive treatment modality capable of delivering highly conformal, ablative radiation doses while minimizing exposure to surrounding critical structures, particularly the optic apparatus [6]. Multiple platforms, including Gamma Knife, CyberKnife, and LINAC-based systems, have been employed for this indication, mirroring their established role in the management of other intracranial metastases [7].

However, evidence supporting SRS for PM remains fragmented and largely limited to small retrospective case series with heterogeneous patient populations, treatment parameters, and outcome reporting. The absence of prospective trials and standardized endpoints has hindered the drawing of definitive conclusions regarding local control durability, survival benefits, endocrine toxicity, and optimal patient selection.

To address these gaps, we performed a systematic review and single-arm meta-analysis to synthesize available evidence on SRS for pituitary metastases. The primary objectives were to evaluate pooled local tumor control and survival outcomes, while secondary objectives included assessment of progression-free survival and treatment-related morbidity. Given the rarity of PM, this study aims to provide a consolidated quantitative framework to contextualize the role of SRS within contemporary multidisciplinary oncologic care.

Methods

Protocol

In compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA), we performed our systematic review and meta-analysis [8] We also followed the Cochrane Handbook for Systematic Reviews of Interventions Methods [9].

Sources and search strategy

Up to June 2025, we searched five databases—PubMed, Embase, Scopus, Web of Science, and Cochrane Library—using a comprehensive search strategy. This search strategy consists of a combination of multiple sets of terms and synonyms, as each set addresses specified articles through the searched databases, while the combination addresses articles that align coherently with our hypothesis. Targeting pituitary metastasis, stereotactic radiosurgery, and related outcomes, we used these strings: (“pituitary metastases” OR “pituitary metastatic tumor” OR “sellar metastasis” OR “pituitary secondary tumors”), (“stereotactic radiosurgery” OR “Gamma Knife” OR “CyberKnife” OR “LINAC radiosurgery” OR “focused radiation therapy”), and (tumor control OR survival OR endocrinopathy OR complications), respectively.

Selection criteria

We included English studies in which patients were diagnosed with pituitary metastases confirmed radiologically and/or histopathologically and treated with SRS (Gamma Knife, CyberKnife, or LINAC-based). In the context of study design, we included randomized controlled trials (RCTs), cohort studies, case-control studies, and large case series that contain more than five patients. We included only studies published in the last two decades. Additionally, we included studies reporting at least one of the following outcomes: tumor control rates, overall survival (OS), progression-free survival (PFS), endocrinological function outcomes, or adverse events. In contrast, we excluded studies that adhere to the following criteria: studies with fewer than 5 patients, studies on pituitary adenomas or craniopharyngiomas without metastatic involvement, studies using interventions other than SRS, review articles, expert opinions, case reports, animal or in vitro studies, or studies reported in non-English languages.

Selection process

We uploaded articles that had resulted from database searches to the Rayyan website [10]. Using the same website, we removed duplicates and screened articles for inclusion in the studies. The screening process was conducted in two phases. Firstly, two authors, independently, screened titles and abstracts, minimizing the number of articles to be assessed in the second phase: full-text screening. Any disagreements between them were solved by discussion or the opinion of a third expert author. Additionally, lists of references and citations of the included studies and previous systematic reviews and meta-analyses were screened for any eligible study.

Data extraction

Two authors extracted the data using a pre-developed Google Sheet, revised by other two authors. Extracted data include four categories: study characteristics, patient demographics, treatment characteristics, and outcomes. As primary outcomes, we focused on local control rate, progression-free survival, radiological response, new-onset hypopituitarism, and survival outcomes, including 1-year and 2-year survival rates. We also assessed radiation necrosis, optic neuropathy, and hypopituitarism as secondary outcomes.

Quality assessment

To assess the quality of our included studies, we use the Risk of Bias in Non-randomized Studies of Interventions-1 (ROBINS-I) tool [11]. This tool evaluates the quality through seven domains, testing risks of bias, and every domain can be judged as low, moderate, critical, or serious risk of bias. These seven domains include bias due to confounding, bias due to selection of participants, bias in classification of interventions, bias due to deviation from intended interventions, bias due to missing data, bias in measurement of outcomes, and bias in selection of the reported result.

Statistical analysis

Since our included studies were of a single-arm design, we conducted a single-arm meta-analysis using the meta package in R (version 4.4.0, 2025 release). For stabilization of the variances and development of the normality of disruption, we applied the logit transformation. For all outcomes, the random effect model was used and embedded using the DerSimonian–Laird estimator; a p-value < 0.05 was considered statistically significant; results were reported as pooled proportions with a 95% confidence interval and visualized as forest plots. In order to assess the heterogeneity, we used Cochran’s Q-test and I² statistics, with I² values > 50% indicating substantial heterogeneity. When heterogeneity was detected, a leave-one-out test was used, in which one study was omitted at a time and the pooled proportion was recalculated. Publication bias was assessed using the funnel plot and Egger’s test [12].

Results

Study Selection and Characteristics

A total of 240 records were identified through four electronic databases. After removing 80 duplicates, 189 records were screened by title and abstract. Of these, 161 were excluded because they did not meet the inclusion criteria. Twenty-eight full-text articles were assessed for eligibility, with 22 excluded. Ultimately, six studies comprising 142 patients with pituitary metastases treated with stereotactic radiosurgery (SRS) were included in the systematic review and meta-analysis.

Definitions of oncological outcomes varied across included studies, with heterogeneity in radiographic thresholds and inclusion of death or distant intracranial progression in progression-free survival. These differences were accounted for during qualitative synthesis and interpretation of pooled outcomes. The definitions of oncological outcomes are summarized in Table 1.

Table 1.

Definitions of Local Tumor Control, Tumor Regression, and Progression-Free Survival Across Included Studies

Study (Ref)	Year	Outcome Reported	Local Tumor Control – Definition	Tumor Regression – Definition	Progression-Free Survival (PFS) – Definition
Chon et al. [13]	2017	Local control, radiographic response	Absence of radiographic tumor progression on serial MRI following SRS	Decrease in tumor size on follow-up MRI compared with baseline	Not explicitly defined or reported
Kano et al. [14]	2009	Local tumor control	No increase in tumor size at treated site on follow-up imaging	Tumor shrinkage or stable disease on post-treatment MRI	Time from SRS to radiographic tumor progression or death
Lin et al. [15]	2023	Local control, PFS	Lack of tumor enlargement or need for salvage treatment on serial neuroimaging	Any measurable reduction in tumor size on follow-up MRI	Time from initial GKRS to radiographic progression, new intracranial metastasis, or death
Mori et al. [16]	2006	Tumor control	Absence of local tumor growth at the treated pituitary or cavernous sinus lesion	Radiographic decrease in tumor size on follow-up imaging	Not formally defined or reported
Wei et al. [17]	2022	Local control	No radiographic progression of the treated lesion on follow-up MRI	Partial or complete radiographic response (tumor size reduction)	Time from GKRS to radiographic progression or death
Abou-Al-Shaar et al. [18]	2025	Local control, PFS	Stability or reduction in tumor size without need for salvage therapy	Reduction in tumor volume on follow-up imaging	Time from SRS to local progression, distant intracranial progression, or death

The study selection process is summarised in the PRISMA flow diagram (Fig. 1), and the characteristics of the included studies are presented in Table 2 [13–18].

Fig. 1.

PRISMA chart outlined flow and criteria for the included studies

Table 2.

Baseline characteristics of the included studies

Author name (reference number)	Study Characteristics					Patient Demographics				Treatment Characteristics
	Year of pubication	Country/region	Study design (RCT, cohort, case-control, case series)	Sample size	Follow-up duration	Age Mean(SD)	Gender distribution Mean(SD)	Primary malignancy source (breast, lung, others)	Duration between primary cancer diagnosis and PM detection	SRS modality (Gamma Knife, CyberKnife, LINAC-based)	Number of SRS fractions (single vs. hypofractionated)	SRS dose (Gy) and prescription isodose line	Tumor volume (cc)	Adjunct therapies (surgery, chemotherapy, systemic therapy)
Chon [13]	2017	South Korea	Retrospective case Series	7 Patients	Median 14 months (range 6–24 months)	Mean = 55.9, SD = 10.1	4 Males (57.1%), 3 Females (42.9%).	Lung (5), Breast (2)	Metachronous in 5 patients, synchronous in 2 patients; no specific duration reported.	CyberKnife system version 9.0	Hypofractionated: 5 daily fractions	31 Gy; isodose 80% (range 73–85%)	Mean 1.6 cc, Range: 0.5–3.2 cc	No concurrent therapy during SRS
Kano [14]	2009	USA	Retrospective case Series	18 Patients	Median 6.3 months (ranges 0.5–19.4)	57.33 years, SD: 12.34	11 Males (61.1%), 7 Females(38.9%)	Lung (6), Breast (4), Renal (4), Melanoma (2), Colon (1), Gastric (1).	12.2 Months	Gamma Knife	Single fraction	13 Gy (mean marginal dose), Max does: 27 Gy	Mean 3.5 cc (range 0.2–18.0).	No concurrent therapy during SRS
Lin [15]	2023	Taiwan	Retrospective review	23 patients	Median and mean clinical follow-up durations were 12.3 and 20.0 months, respectively, while median and mean neuroimaging follow-up durations were 10.6 and 18.7 months, respectively.	Mean: 54.3 years (Median: 52.2, Range: 30.3–83.6). SD not reported	11 males and 12 females	metastases involving the pituitary (Group 1, 11 patients): lung cancer (n = 5), breast cancer (n = 3), and renal cell carcinoma (n = 2). patients with metastasis involving only the cavernous sinus (Group 2, 12 patients): (n = 8, parotid adenoid cystic carcinoma, and head and neck squamous cell carcinoma), lung cancer (n = 1), and breast cancer (n = 1).	not reported.	Gamma Knife Radiosurgery (GKRS)	single fraction: Leksell Gamma Unit Model C (Elekta Instrument, Inc)	Median marginal dose: 15.0 Gy (11.5–18.0 Gy); Median maximum dose: 25.5 Gy (20.0–31.2 Gy); Isodose: median 58% (50–70%)	Median volume: 4.5 cc (range: 0.13–41.92 cc)	presenting with tumor progression: 4 patients underwent repeated GKRS or chemotherapy as salvage therapy. New brain metastases: 6 patients treated using whole brain radiotherapy or GKRS.
Mori [16]	2006	Japan (Nagoya)	Retrospective case Series	Total patients with brain metastases treated with GKS: 623 Patients with pituitary/cavernous sinus metastases: 13 (2.1% of total)	follow up for 9 out of 13 patients: range (2 to 12 moths), median: 4 months.	34 to 72 years.	7 males, 6 females	Lung cancer: 5 patients, Breast cancer: 2 patients, Parotid cancer: 2 patients. Others: Renal cell carcinoma: 1, Thyroid cancer: 1, Nasal cancer: 1, Gastrointestinal stromal tumor (GIST): 1	(Case 2) mentions a 1-year history of lung adenocarcinoma, but no consistent data across the cohort is provided.	Gamma Knife surgery(GKS)	Single fraction for all patients	A margin dose of 12 to 12.3 Gy (median 12 Gy) to pituitary metastases (Type 1). margin dose of 14.4 to 20 Gy (median 16 Gy) to cavernous lesions (Types 2 and 3). (median 50%) isodose curve was used.	median tumor volume was 3.4 ml (range 1.5–11.3 ml)	Chemotherapy: 3/13 patients after GKS therapy
Wei [17]	2022	USA	Retrospective case Series	18 Patients	6.5 months (range 0.5–58 months)	Median = 65.5 (44–89)	8 Males Females 10	(6) non- small cell lung cancer, (2) small cell lung can- cer, (4) breast cancer, (2) renal cell carcinoma, (2) prostate cancer, (1) colon cancer, and (1) eccrine carcinoma.	No specific duration reported	Gamma Knife	Single fraction	Median GKRS margin dose 13 Gy (range 8–16 Gy) Maximum point dose optic apparatus was 10.2 Gy (range 1.3–13.8 Gy)	median tumor volume was 0.78 cc (range 0.04–6.42 cc)	No concurrent therapy during SRS
Abou-Al-Shaar [18]	2025	USA	Retrospective Case Series	63 Patients	Median 25.4 months	Median = 62 (IQR: 57–71)	30 Males (47.62%) Females (52.38%)	Lung 25 (39.7%) Breast 16 (25.4%) Renal 5 (7.9%) Prostate 6 (9.5%) Colon 3 (4.8%) Other* 8 (12.7%)	1.7 years (IQR: 0-5.6)	Gamma Knife	Single fraction	Median margin dose for pituitary metastasis [IQR] (Gy)13 [12–16] Median maximum dose for pituitary metastasis [IQR] (Gy)25 [23.7–31.1]Median maximum dose received for optic apparatus [IQR] (Gy)10.2 [8.2–12]	Median 1.5 [IQR] [0.7–3.6]	Surgery 11 (17.5%) Radiation therapy 2 (3.2%) None 50 (79.4%)

Summary of included studies

All six included studies were retrospective case series published between 2006 and 2025. The sample sizes ranged from 7 to 63 patients, with median follow-up durations from 4 to 25.4 months. Patients’ mean or median age ranged from 54 to 65.5 years, and the sex distribution was relatively balanced across studies.

The patients have primary malignancies, which were most commonly lung and breast cancer, and others. SRS was delivered via Gamma Knife in five studies and CyberKnife in one. Most studies used single-fraction SRS, while one used hypofractionated treatment (five fractions). The marginal SRS doses ranged from 12 Gy to 31 Gy, and tumour volumes ranged from 0.04 cc to 41.92 cc. Adjunctive treatments such as surgery, chemotherapy, or whole-brain radiotherapy were reported in some studies, but the majority of patients received SRS alone [13–18]. Further details are summarized in Table 2.

Risk of bias assessment

The risk of bias was assessed using the ROBINS-I tool for the studies. All six included studies demonstrated a moderate overall risk of bias. Most studies have shown a fair quality across various domains, including bias due to confounding, participant selection, and outcome measurement. However, the risk of bias related to classifying interventions and deviations from intended interventions was consistently rated as low [11] (Fig. 2).

Fig. 2.

Risk of Bias Assessment of the Included Studies

Primary outcomes

Local tumor control

Five studies (n = 129 patients) reported local control rates following stereotactic radiosurgery (SRS) for pituitary metastases. Individual study estimates ranged from 65% to 100%, while the calculated pooled estimate of the local control rate was 89% (95% CI 75–96%), as shown in Fig. 3. Moderate heterogeneity was observed (I² = 63%, P = 0.029). Fig. 4 indicates that a Leave-One-out sensitivity analysis was employed, in which omitting (Kano 2009) resulted in the least heterogeneity (I² = 0%) [13, 14, 16–18], suggesting that this study has contributed to the between-study variance. The funnel plot in Fig. 5suggests no publication bias ( P-value of Egger Z test = 0.4211).

Fig. 3.

Forest plot showing pooled local control rates after stereostatic radiosurgery (SRS) for pituitary metastases

Fig. 4.

Leave-one-out sensitivity analysis for local control rates, demonstrating reduction of heterogeneity when omitting kano et al. (2009)

Fig. 5.

Funnel plot for local control rates showing no evidence of publication bias (P-value of Egger´s Z test = 0.4211)

New-Onset Hypopituitarism

Five studies (n = 129 patients) reported the new-onset hypopituitarism. The pooled incidence was 1% (95% CI: 0.00 to 0.048). All studies were homogeneous with no observed heterogeneity (I² = 0.0%, P = 1.000) (Fig. 6). A possible publication bias was established as shown in the funnel plot, Fig. 7, ( P-value of Egger Z test = 0.0377) [13–18].

Fig. 6.

Forest plot showing the pooled incidence of new-onset hypopituitarism after SRS for pituitary metastasis

Fig. 7.

Funnel plot for new-onset hypopituitarism indicating possible publication bias ( P-value of Egger Z test = 0.0377)

Survival outcomes

One-year survival

A pooled analysis of five studies (n = 129 patients) revealed a one-year survival rate of 53% (95% CI 32–73%), with substantial heterogeneity (I² = 75.9%, P = 0.0023), shown in Fig. 8. Through the Leave-One-out analysis in Fig. 9, no study has solved the heterogeneity completely. However, omitting the study of (Kano 2009), reduced heterogeity to (I² = 48.6%) [13, 14, 16–18]. Fig. 10 shows the funnel plot, in which no publication bias was observed ( P-value of Egger Z test = 0.4468).

Fig. 8.

Forest plot showing the pooled one-year survival rate following SRS for pituitary metastases

Fig. 9.

: Leave-one-out sensitivity analysis for one-year survival, showing persistence of heterogeneity despite omission of single studies

Fig. 10.

Funnel plot for one-year survival showing no publication bias (P-value of Egger Z test = 0.4468)

Two-year survival

Two studies (n = 30 patients) reported a pooled two-year survival rate of 40% (95% CI: 15–72%), with moderate heterogeneity (I² = 62.0%, P = 0.1047) [13, 16]. Figure 11.

Fig. 11.

Forest plot showing the pooled two-year survival rate after SRS for pituitary metastases

Six-month progression-free survival

Two studies (n = 36 patients) reported a pooled six-month progression-free survival rate of 97% (95% CI: 83–100%). No heterogeneity was observed (I² = 0%, P = 0.99) [16, 18]. Figure 12.

Fig. 12.

Forest plot showing the pooled six-month progression-free survival rate following SRS

Radiological response

Generally, the radiological response showed that the regression features or stable tumor features take the upper hand in comparison to the tumor progression.

Tumor regression

Three studies (n = 54 patients) reported a pooled regression rate of 34% (95% CI 9–73%) [14, 16, 18]. Substantial heterogeneity was observed (I² = 79.4%, p = 0.0078), Fig. 13. Leaving (Wei), through the Leave-One-Out analysis test, has resolved the heterogeneity (I² = 0%), Fig.14.

Fig. 13.

Forest plot showing the pooled tumor regression rate following SRS for pituitary metastases

Fig. 14.

Leave-one-out sensitivity analysis for tumor regression, demonstrating resolustion of heterogeneity when omitting Wei et al.

Stable tumor

Four studies (n = 72 patients) reported a pooled stable disease rate of 33% (95% CI 8–73%), with high heterogeneity (I² = 83.8%, P = 0.0004), Fig. 15. Sensitivity Leave-One-Out analysis revealed that no single study had a significant impact on the overall results, Fig. 16.

Fig. 15.

Forest plot showing the pooled rate of stable disease after SRS for pituitary

Fig. 16.

Leave-one-out sensitivity analysis for stable disease showing no single study significantly altered the overall estimate

Tumor progression

Five studies (n = 135 patients) reported a pooled progression rate of 7% (95% CI 3–16%), with low heterogeneity (I² = 26.4%, P = 0.245), Fig. 17. The Funnel plot in Fig. 18 shows no publication bias ( P-value of Egger Z test = 0.3131).

Fig. 17.

Forest plot showing the pooled tumor progression rate after SRS for pituitary metastases

Fig. 18.

Funnel plot for tumor progression showing no evidance of publication bias ( P-value of Egger Z test = 0.3131)

Secondary outcomes

Adverse events

Hypopituitarism

Six studies (n = 142 patients) assessed hypopituitarism following SRS. The pooled incidence was 1% (95% CI: 0.00 to 0.048), with no observed heterogeneity (I² = 0.0%, p = 1.000) (Fig. 19). Five out of six studies reported no events of hypopituitarism, and only one study (Abou-Al-Shaar 2025) reported five cases (8%) [13–18]. Depending on the Egger Z test and funnel plot in Fig. 20, publication bias was observed, ( P-value of Egger Z test = 0.0129).

Fig. 19.

Forest plot showing the pooled incidence of hypopituitarism after SRS

Fig. 20.

Funnel plot for hypopituitarism showing evidence of publication bias ( P-value of Egger Z test = 0.0129). :

Optic neuropathy

Six studies (n = 142 patients) reported a pooled incidence of optic neuropathy of 2% (95% CI 1–6%), with no heterogeneity (I² = 0%, P = 0.97) (Fig. 21). A total of three events were documented across all studies [13–15, 17], as each of these three studies ( Abou-Al-Shaar, Lin, Kano) reported one case of optic neuropathy. The Egger Z test of the funnel plot asymmetry showed no publication bias ( p-value = 0.7709) (Fig. 22).

Fig. 21.

Forest plot showing the pooled incidence of optic neuropathy following SRS for pituitary metastases

Fig. 22.

Funnel plot for optic neuropathy showing no publication bias (p-value = 0.7709)

Radiation necrosis

The pooled incidence of radiation necrosis was 1% (95% CI: 0–5%), with no heterogeneity (I² = 0%, P = 1). Only one event was reported across the six included studies (n = 142 patients), (Fig. 23). The P-value of The Egger Z test was higher than 0.05, (p-value = 0.2126), establishing a no publication bias (Fig. 24).

Fig. 23.

Forest plot showing the pooled incidence of radation necrosis after SRS

Fig. 24.

Funnel plot for radation necrosis showing no evidence of publication bias (p-value = 0.2126)

Discussion

Local control rate and tumor response rate

Across the six included series, stereotactic radiosurgery (SRS) achieved consistently high local control of pituitary metastases, with most reports clustering around 80–95% at 6–12 months and durable control in patients living long enough for extended follow-up. In a hypofractionated CyberKnife pilot (n = 7), local control was reported in all patients, with complete remission in 4 and partial remission in 3 on MRI; an outcome obtained with an optimized plan to protect the optic apparatus [13]. In another early single-center experience (n = 18), partial response was observed in 26.7% of patients, stable disease in 53.3%, and delayed in-field progression in 20%, yielding an overall tumor control rate of 80% at last follow-up; progression-free survival after SRS was 100% at 6 months and 66.7% at 12 months, indicating that most failures occurred beyond the first half-year [14].

More contemporary cohorts report similar or higher control rates. A 23-patient series that analyzed pituitary versus cavernous-sinus involvement documented local tumor control rates of 82.3% at 1 year, 82.3% at 2 years, and 65.9% at 3 years on Kaplan–Meier analysis, without a difference between pituitary and cavernous groups [15]. A single-institution study of 18 patients showed treated-lesion control in 17/18 (94.4%), with 15 stable, 1 complete regression, and 1 significant volume reduction (> 20%) at follow-up; one patient required a boost for progression [17]. Finally, the largest multicenter cohort (n = 63) reported local control of 93.1% at 6 and 12 months and 85.4% at 18 months, underscoring durable in-field control with appropriately selected SRS dosing [18].

Radiological tumor response following stereotactic radiosurgery was reported in several studies, with generally favorable outcomes. In the series by Mori et al. , six of nine evaluable tumors demonstrated regression greater than 50% in size, while three remained stable, resulting in no cases of local failure at a median follow-up of four months [16]. In the 2009 Kano et al. cohort, a partial radiological response was observed in 26.7% (4/15) of patients, stable disease in 53.3% (8/15), and delayed in-field tumor progression in 20% (3/15), occurring at approximately 12–15 months after treatment [14]. Lin et al. described volumetric regression in 75% of patients with available follow-up imaging, with stable disease in 5% and tumor growth in 20% [15]. Abou-Al-Shaar et al. similarly reported durable control, with most patients showing stability or reduction in tumor size during follow-up, though exact volumetric breakdown was not consistently available [18]. In the large multi-institutional analysis by Wei et al., detailed volumetric data were not uniformly reported, but the authors emphasized high local tumor control with frequent radiographic stability across institutions [17].

Collectively, these findings align with our meta-analytic estimate of high local control and suggest that dose selection and proximity to the optic apparatus are the dominant practical modulators of radiological response within the included series (e.g. hypofractionation or careful isodose shaping near the chiasm/optics in the CyberKnife and Gamma Knife plans) [13–18]. Clinically, this supports SRS as a reliable local therapy for pituitary metastases, with expectations of > 80–90% control at one year and meaningful regression in a majority.

Endocrinological and neurological safety

Across studies, stereotactic radiosurgery (SRS) for pituitary metastases was consistently associated with a relatively low rate of new endocrinopathies and neurological toxicities. In Mori et al., three tumors remained stable while six of nine evaluable lesions showed ≥ 50% volumetric regression [16]. Most patients demonstrated stability or improvement in visual and cranial nerve function. One patient experienced intratumoral hemorrhage within a Type 1 tumor at one month post-treatment; however, visual function recovered by Month 3, and no cases of radiation-induced optic neuropathy were observed [16]. Kano et al. found that 42.9% of patients with diabetes insipidus improved following SRS, while those with panhypopituitarism required continued hormone replacement. Among patients with neurological symptoms, 50% showed improvement, including recovery of optic neuropathy (1 case), cranial nerve III palsy (2 of 3 cases), and trigeminal neuropathy (1 of 2 cases). However, 16.7% developed new or progressive deficits, including abducens paresis and one case of optic neuropathy related to tumor progression and radiation effects [14].

Wei et al. observed that most patients with preexisting endocrinopathies remained stable after GKRS, with only two developing new hypothyroidism controlled by replacement therapy, while visual outcomes were favorable, with improvements in acuity or diplopia in several patients and no reported deterioration [17]. Lin et al. further confirmed the safety of SRS, noting no new endocrinopathy and reporting that most patients experienced stability or improvement in visual and cranial nerve function, even for cavernous sinus lesions close to the optic chiasm [15]. Chon et al., using hypofractionated CyberKnife SRS, likewise reported no new cranial neuropathies or endocrinological deficits, supporting the potential protective role of fractionation in high-risk lesions [13].

Finally, the large multi-institutional series by Abou-Al-Shaar et al. corroborated these findings, with only a minority of patients developing new anterior pituitary hormone deficiencies (5/63) or transient arginine vasopressin deficiency (2/63), while the majority remained endocrinologically stable [18]. Additionally, 17 of 18 patients with pre-existing visual deficits showed favorable outcomes, including 6 with improvement and 11 with stable vision, while only 1 patient experienced worsening. Cranial nerve recovery was also reported.

Overall, the studies indicate that SRS achieves good outcomes as long as current dose limits to nearby structures are followed. SRS offers effective tumor control with low rates of new endocrinopathy or cranial nerve injury, and in some cases may even reverse preexisting neurological deficits.

Survival outcomes

Survival outcomes across the included series highlight generally poor prognosis of pituitary metastases, reflecting advanced systemic disease rather than local treatment failure. Mori et al. primarily focused on radiological outcomes and reported imaging follow-up in nine patients for a median of four months (range 2–12 months), without providing detailed survival rates, underscoring the short survival window in this cohort [16]. Kano et al. (2009) reported a median survival of 5.2 months, with overall survival rates of 66.7%, 33.3%, and 13.3% at 1, 3, and 5 years, respectively [14].

More favorable outcomes were observed in later studies. Chon et al. (2017) found a median survival of 14 months after hypofractionated CyberKnife treatment [13]. Lin et al. (2023) reported one-, two-, and three-year survival rates of 72.9%, 51.8%, and 45.3%, respectively, with a median survival after GKRS of 30.8 [15]. In the multi-institutional cohort by Wei et al. (2022), the median survival was 6.5 months (range 0.5–58 months) [17]. The largest series by Abou-Al-Shaar et al. (2025) demonstrated the longest survival, with a median of 25.4 months, substantially exceeding that of earlier single-center studies [18].

Together, these findings demonstrate that while stereotactic radiosurgery ensures consistent local control, systemic progression drives overall survival, with outcomes varying by study design, follow-up duration, and patient selection.

Clinical outcomes

The collective findings across these six series underline several key clinical implications of stereotactic radiosurgery (SRS) for pituitary metastases. First, SRS offers an effective local treatment modality for lesions otherwise difficult to access surgically due to proximity to the optic chiasm or cavernous sinus. In both single- and multi-institutional reports, local tumor control consistently exceeded 90% at one year, with frequent volumetric regression and rare local failures, suggesting that SRS can serve as a reliable means of focal disease management [13–18]. Importantly, this effect was achieved with a favorable safety profile, with new endocrinopathies and visual deterioration being uncommon and typically medically manageable [14–18].

Second, because systemic disease progression remains the predominant cause of death in this population, the role of SRS must be viewed in the context of multimodality care. Several series noted that patients often died from progression of their primary cancer rather than from intracranial failure, despite durable local control [13–16]. This highlights the importance of integrating SRS with systemic therapies such as chemotherapy, targeted agents, or immunotherapy, rather than considering it a curative intervention in isolation.

Finally, these data suggest a practical treatment algorithm: SRS is best suited for patients with limited intracranial disease, small-to-moderate lesion volume, and controlled or stable systemic cancer [14–18]. In such patients, SRS can achieve durable symptom relief (particularly for cranial nerve deficits or diabetes insipidus) while minimizing treatment morbidity. Conversely, in patients with rapidly progressive systemic disease, the benefits of SRS may be limited to short-term palliation. This reinforces the need for careful patient selection and individualized treatment planning to balance local efficacy against systemic prognosis.

Limitations

The limitations of the available literature must be acknowledged when interpreting outcomes of stereotactic radiosurgery (SRS) for pituitary metastases. Across the six included studies, most cohorts were retrospective and single-institution, introducing inherent biases related to patient selection, reporting accuracy, and heterogeneity in treatment protocols [13–18]. Mori et al. and Kano et al. both highlighted small sample sizes (n = 13 and n = 18, respectively) as restricting statistical power and generalizability, while also limiting subgroup analyses [16, 14]. Lin et al. similarly acknowledged the difficulty of drawing firm conclusions due to heterogeneity in tumor location (pituitary vs. cavernous sinus) and the retrospective study design [15]. Wei et al. noted that despite pooling data from multiple institutions, variability in radiation dose, margin selection, and patient baseline systemic disease status remained important confounders [17]. The most contemporary international series by Abou-Al-Shaar et al. attempted to overcome this by including a larger multi-institutional dataset, yet still recognized limitations of retrospective data collection and the absence of uniform follow-up imaging or endocrine testing [18]. Furthermore, across all studies, survival outcomes were confounded by systemic cancer progression, making it difficult to isolate the effect of SRS on long-term prognosis. Finally, detailed neuro-ophthalmological and endocrine follow-up was inconsistently reported, raising the possibility of underestimating late toxicities. Collectively, these limitations underscore the need for larger, prospective, and standardized studies to better define the role of SRS in this clinical context.

Future directions

Future research on stereotactic radiosurgery (SRS) for pituitary metastases should address several gaps identified across the current evidence base. First, all six included studies were retrospective, limiting the strength of causal inferences; prospective, multi-institutional registries or clinical trials are needed to establish standardized dosing protocols, follow-up schedules, and consistent reporting of endocrine and neuro-ophthalmic outcomes [13–18]. Second, heterogeneity in patient populations including differences in primary tumor histology, systemic treatment status, and lesion location (pituitary versus cavernous sinus) complicates interpretation of results. Future studies should look at outcomes in groups based on these variables to see which patients gain the most benefit.

Third, the short survival windows observed in most cohorts suggest that SRS should be studied as part of integrated multimodal care, alongside advances in systemic therapies such as targeted agents and immunotherapy [17, 18]. Investigating how local control with SRS interacts with systemic disease control will help refine patient selection and optimize sequencing of treatment. Finally, longer-term prospective data are essential to clarify the risk of late endocrine and visual toxicities. Addressing these questions will be critical in defining the precise role of SRS in modern neuro-oncology practice.

Conclusion

Stereotactic radiosurgery (SRS) has emerged as an effective and durable treatment modality for achieving local control of pituitary metastases, offering a favorable safety profile and low rates of radiation-induced toxicity. The results of this meta-analysis demonstrate that SRS provides meaningful intracranial disease stabilization and symptomatic relief, even in the setting of advanced systemic malignancy. Importantly, SRS allows for precise tumor targeting while preserving adjacent critical neurovascular and endocrine structures, minimizing the risks traditionally associated with open surgical resection or fractionated radiotherapy.

Although the overall prognosis of patients with pituitary metastases remains largely determined by the aggressiveness of the primary malignancy and systemic disease burden, local intracranial control achieved through SRS contributes significantly to maintaining neurological function, visual preservation, and quality of life. These findings underscore the value of SRS as both a definitive and palliative strategy in carefully selected patients with limited or controlled extracranial disease.

Nevertheless, the current body of evidence is largely retrospective and heterogeneous, with variable reporting of endocrinological and neuro-ophthalmological outcomes. Future prospective, multicenter studies employing standardized outcome measures and longer follow-up intervals are warranted to more accurately delineate the long-term impact of SRS on endocrine function, tumor progression, and overall survival. Such studies will be essential in refining patient selection criteria and optimizing multidisciplinary management algorithms for this rare but clinically challenging condition.

Author contributions

Ahmed Hafez Mousa – Conceptualization, study design, data extraction supervision, statistical analysis, manuscript drafting, and critical revision for important intellectual content.- Gergis Altalab – Data acquisition, quality assessment, and statistical validation.- Anas Mansour – Methodology design, data verification, and table generation.- Ahmed Harb – Formal analysis, figure preparation, and manuscript editing.- Salma Ismail Daffa – Literature search, data curation, and drafting of the results section.- Kinan Mahmoud Radwan Alitaki – Data extraction, screening of full texts, and reference management.- Omar Hamadi – Data verification, study selection, and critical review of the discussion section.- Babacar Cisse – Senior supervision, conceptual guidance, and final manuscript approval.All authors read and approved the final manuscript.

Funding

The authors declare that no funding was received for this study.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable – as the study did not include new human or animal data requiring ethical approval. This study is a systematic review and meta-analysis based entirely on previously published data and does not involve any new human participants, identifiable patient data, or experimental interventions. Therefore, institutional review board (IRB) approval and informed consent were not required.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.