Ophthalmic surveillance before and after treatment of non-functioning pituitary adenomas with optic pathway involvement: a systematic review protocol

Abstract

Introduction

Non-functioning pituitary adenomas (NFPAs) account for approximately 30–40% of pituitary tumours. They frequently compress the optic chiasm, leading to visual field defects, reduced visual acuity and colour vision impairment. Ophthalmic assessment is essential for informing treatment choices and monitoring recovery or disease progression. However, monitoring assessments and frameworks to guide test selection, timing and follow-up vary widely across centres and lack international consensus.

To systematically examine and synthesise ophthalmic surveillance in adults with radiologically confirmed NFPAs involving the optic pathway, before and after intervention.

Methods

We will search Ovid MEDLINE, Ovid Embase, CENTRAL, Scopus, Web of Science and grey literature sources from inception. Eligible studies will include adults (≥18 years) with radiologically confirmed NFPAs and documented optic pathway involvement. Both randomised and non-randomised studies that report ophthalmic assessments will be included. Case series with fewer than 10 participants will be excluded.

Two reviewers will independently and in duplicate screen titles/abstracts and full texts, extract data and assess risk of bias using design-specific tools (including Risk-of-Bias Tool for Randomised Trials, Risk of Bias in Non-randomised Studies of Interventions, Joanna Briggs Institute checklists and the Appraisal tool for Cross-Sectional Studies tool).

Ethics and dissemination

No ethical approval is required as this review synthesises previously published data. Findings will be submitted for peer-reviewed publication and presented at national and international conferences. The protocol was prospectively registered with PROSPERO (CRD420261337096).

PROSPERO registration number

CRD420261337096.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• A comprehensive multi-database search strategy (including citation searching and grey literature sources) will maximise identification of relevant studies.

• Two independent reviewers will screen titles/abstracts and full texts, extract data and assess risk of bias, enhancing methodological rigour.

• Given the expected clinical and methodological heterogeneity, this may preclude a quantitative meta-analysis.

• The literature is dominated by retrospective and observational studies with inconsistent reporting of ophthalmic outcomes, which may limit the strength and generalisability of findings.

• No language restrictions will be applied—however, translation capacity for non-English records may be limited and will be documented transparently in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram.

Introduction

Pituitary tumours account for approximately 10–20% of all intracranial tumours^{1 2} with non-functioning pituitary adenomas (NFPAs) comprising approximately 30–40% of all pituitary tumours.³ NFPAs frequently compress the optic chiasm, resulting in reduced visual acuity, impaired colour vision and bitemporal visual field defects.^{4 5} Visual field defects are present in up to 58% of NFPA macroadenomas at presentation,⁶ making ophthalmic assessment central to clinical management and a key determinant of surgical timing. Depending on the size, growth and symptoms of the tumour, management options include observation, surgery or radiotherapy.⁷

Ophthalmic evaluation is critical for determining intervention thresholds and monitoring postoperative recovery or progression.⁶ Functional assessments include visual acuity, automated perimetry, colour vision, pupillary examination and orthoptics; structural assessments include optical coherence tomography (OCT) and funduscopy.⁸

Despite their clinical importance, ophthalmic surveillance practices vary markedly across centres, with no consensus guidelines on modalities, timing or frequency.

To our knowledge, no systematic review has specifically synthesised ophthalmic surveillance modalities and intervals in adults with NFPAs and optic pathway involvement; existing reviews⁹ focus on broader post-treatment follow-up rather than ophthalmic practice specifically.

Objectives

This review aims to:

1. Identify ophthalmic assessment modalities used in adults with NFPAs.

2. Evaluate the timing, frequency and clinical context of assessments pre- and post-intervention.

3. Highlight variations and gaps in current ophthalmic surveillance practices to inform the development of future guidance on neuro-ophthalmic surveillance in NFPAs.

Methods and analysis

This protocol will be reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P) 2015, and the review will be reported following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020.^{10 11} This systematic review protocol adopts the Population-Concept-Context framework—recommended for scoping and evidence-mapping reviews and defined as follows: population—adults (≥18 years) with radiologically confirmed NFPA and optic pathway involvement; concept—ophthalmic assessment modalities, monitoring intervals and surveillance pathways used in clinical practice; context—observation, pre-treatment assessment, post-surgical follow-up, post-radiotherapy follow-up and surveillance for recurrent or residual disease.

For the secondary prognostic question, a PECO framework (Population, Exposure, Comparator, Outcome) is adopted, rather than the conventional PICO framework, as this review does not evaluate a therapeutic intervention. The exposure of interest is ophthalmic assessment and surveillance practice, encompassing diagnostic and monitoring behaviours rather than therapeutic interventions.

Eligibility criteria

Population

Studies involving adult human participants (aged ≥18 years) with radiologically confirmed NFPA will be included. Radiological confirmation requires imaging (eg, MRI or CT) demonstrating a sellar or suprasellar lesion without clinical or biochemical evidence of hormone hypersecretion. The specific hormonal axes tested will be extracted as a data field to enable assessment of diagnostic consistency across included studies—no minimum hormonal workup is required for inclusion.

Eligible studies must demonstrate radiological and/or clinical evidence of optic pathway involvement (affecting, compressing or distorting the optic chiasm or visual pathways). Histologically confirmed subtypes including null cell adenomas and gonadotroph adenomas are eligible—histological confirmation is not required where imaging and biochemistry are consistent with NFPA.

Studies that concentrate on other sellar or parasellar pathologies, such as meningiomas, craniopharyngiomas, Rathke’s cleft cysts, optic pathway gliomas, metastatic lesions, vascular lesions or infectious or inflammatory diseases that affect the optic pathways, such as hypophysitis or sarcoidosis, will be excluded. Paediatric populations (<18 years) will be excluded and studies involving functioning (hormone-secreting) pituitary adenomas or pituitary apoplexy will be excluded. Where apoplexy arises during the follow-up period of an otherwise eligible study, the relevant surveillance episodes will be excluded from analysis.

Where studies include both functioning and non-functioning pituitary adenomas without separate reporting of NFPA subgroup data, authors will be contacted to request disaggregated data. If this is unavailable, the study will be excluded from primary synthesis and listed in a supplementary table of potentially relevant studies.

Exposure

Any ophthalmic assessment or monitoring approach in patients with NFPAs is the exposure of interest. This includes post-intervention evaluations after procedures such as transsphenoidal surgery, craniotomy or radiotherapy, as well as pre-intervention (baseline) evaluations before surgical or radiotherapeutic treatment. Ophthalmic assessments performed during observation (conservative management) for NFPAs not meeting surgical thresholds are eligible and will be categorised under the pre-specified context of observation surveillance.

Eligible assessments include:

• Functional assessments include best-corrected visual acuity (visual acuity will be presented as both corrected and uncorrected, findings will be extracted separately and will be clearly distinguished in the synthesis and tables.) (best corrected visual acuity (BCVA), utilising standardised charts), automated static perimetry (alongside, confrontation visual fields will be recorded and distinguished separately at extraction. Perimetry device, test programme, reliability indices (fixation losses, false positives, false negatives) and the visual field defect definition used will be extracted.) (eg, Humphrey 24–2 or 30–2 SITA standard, Octopus G perimetry programme or kinetic Goldmann perimetry), colour vision testing, pupillary examination (including assessment for a relative afferent pupillary defect) and orthoptic evaluation (involving ocular motility, ocular alignment, binocular single vision, presence of diplopia and cranial nerve examination).

• Structural assessments include funduscopy and retinal fundus photography (where photographic documentation is performed—funduscopy without photography will not be categorised as retinal imaging), and OCT for analysis of the peripapillary retinal nerve fibre layer (pRNFL), macular ganglion cell layer (GCL), ganglion cell complex and ganglion cell-inner plexiform layer (GCIPL).^{8 12}

Comparator

No formal comparator is required for the primary mapping question. For the secondary prognostic question, studies with or without a comparator group are both eligible—comparator details will be extracted where present. Eligible studies will include descriptive studies, single-arm cohorts and reports of ophthalmic monitoring procedures without a control group.

Outcomes

This review addresses two distinct but complementary questions:

1. Descriptive mapping—Which ophthalmic assessment modalities and surveillance schedules are reported in the literature for adults with NFPAs and optic pathway involvement?

2. Exploration—Are surveillance modalities or scheduling associated with earlier detection of visual recovery or deterioration?

Each question will be synthesised separately; the mapping question will be synthesised narratively following the Synthesis Without Meta-analysis (SWiM) framework.¹³ The prognostic question will be synthesised narratively with effect estimates extracted (where reported); quantitative pooling will be considered only where ≥3 studies report comparable populations, modalities, timing and outcome definitions using a compatible effect measure.

Primary outcomes

Primary outcomes will be categorised according to the type of ophthalmic assessment: functional assessment and structural assessments. Functional assessments will include visual acuity, automated perimetry, colour vision testing, pupillary examination, orthoptic evaluation, timing of evaluations (eg, pre-operative baseline and post-operative intervals) and longitudinal monitoring strategies. Structural assessments will include OCT scans (providing quantitative retinal layer thickness and structural measurements), retinal imaging and funduscopy (providing qualitative visualisation of the retina).

These results will be used to characterise the variety of ophthalmic tests used to evaluate the structural integrity and function of the visual pathway in this patient group.

Secondary outcomes

• The relationship between ophthalmic monitoring strategies—including the choice of assessment modality, timing and frequency—and reported visual outcomes, such as detection of recovery or deterioration following intervention. The following effect measures will be extracted where reported: mean differences or the standardised mean differences (for continuous outcomes), risk ratios, HRs or odds ratios (for binary or time-to-event outcomes), Pearson or Spearman correlations (for association analyses) and the sensitivity, specificity or AUC (where diagnostic or monitoring performance is assessed).

• The application of structural biomarkers, such as OCT parameters.¹⁴

• Identification of practice gaps or heterogeneity.

• Any reported rationale for assessment strategies.

Studies that report particular clinical efficacy outcomes will not be disqualified; however, if such information is available, it will be extracted.

Study design

Studies reporting ophthalmic assessment or monitoring in the target population will be eligible, including observational and interventional designs. Eligible observational designs will include prospective and retrospective cohort studies, cross-sectional studies and surveys of clinical practice. Eligible interventional designs will include randomised controlled trials, non-randomised trials and before-and-after studies (provided they report relevant ophthalmic assessment or monitoring data).

For the purposes of this review, ophthalmic surveillance will be defined as the planned assessment of visual function or structural integrity over time. A study will be considered to report ophthalmic surveillance if it meets at least one of the following criteria:

1. An explicit ophthalmic follow-up schedule.

2. Two or more planned ophthalmic assessment time points.

3. A stated ophthalmic monitoring pathway or clinical protocol.

4. A defined trigger for repeat ophthalmic testing or change in management; or

5. A survey reporting routine surveillance practice.

Studies reporting a single isolated measurement without an explicitly stated surveillance strategy will not be excluded solely on this basis, as such studies may provide relevant information for the descriptive mapping component of the review. However, these studies will be identified at data extraction and analysed separately from studies describing longitudinal or protocolised surveillance.

Case reports and case series with fewer than 10 participants will be excluded because of their potential for substantial selection bias and limited contribution to the assessment of surveillance practices. Case series including 10–19 participants will be included but flagged as small series during data extraction and considered accordingly in the narrative synthesis.

Conference abstracts without sufficient methodological or outcome information, editorials, commentaries, opinion pieces, narrative or systematic reviews and non-human studies will be excluded. Studies that do not provide primary data on ophthalmic assessment or monitoring will also be excluded.

Where a preprint is identified that has since been published, the peer-reviewed publication will be used for primary synthesis. Unreviewed preprints will be flagged in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram and sensitivity-analysed separately.

Report characteristics

No restrictions on publication date or language will be applied. If adequate methodological information is supplied, peer-reviewed publications and grey literature (such as theses and preprints) will be taken into consideration. Conference abstracts will not be accepted unless complete methodological details are available (eg, via linked full-text or author contact).

Where potentially eligible records are identified in languages other than English, translation will be sought through collaboration with multilingual team members or institutional translational services where available.

Search strategy

MEDLINE (Ovid), Embase (Embase.com), CENTRAL (Cochrane Library), ProQuest Central, Scopus, Web of Science (Clarivate) and grey literature sources will be searched from inception. Grey literature sources will include Google Scholar (first 200 results by relevance), ProQuest Dissertations & Theses, ClinicalTrials.gov, the WHO International Clinical Trials Registry Platform and relevant conference proceedings. Reference lists of included studies and relevant reviews will also be hand-searched.

The MEDLINE strategy was developed with input from an information specialist (DB), with formal PRESS peer review to be completed prior to full-text screening and documented in the final review.¹⁵ Search sensitivity will be assessed against a sample of known relevant studies identified through expert consultation, with search terms revised if eligible studies are not consistently retrieved.

The search will combine five conceptual domains: (1) NFPA/pituitary, including *Pituitary Neoplasms* [medical subject headings (MeSH)], non-functioning/nonfunctioning adenoma, null cell adenoma and gonadotroph adenoma; (2) optic pathway, including *Optic Chiasm* [MeSH] and chiasmal compression; (3) visual function, including *Visual Fields* [MeSH], perimetry, visual acuity, hemianopia, colour vision and relative afferent pupillary defect (RAPD); (4) structural imaging, including *Optical Coherence Tomography* [MeSH], OCT, RNFL, GCL, GCIPL and optic atrophy; and (5) surveillance, including follow-up, monitoring, surveillance, interval and recovery. Full electronic search strategies, including platforms, dates, syntax and retrieval counts, will be provided in the online supplemental appendix.

Grey literature and trial registry searches will be reported narratively, with search dates, queries and screening numbers documented. Conference proceedings from the Pituitary Society, European Congress of Endocrinology and European Society of Ophthalmology will be searched for the preceding 5 years and trial registries will be searched for potentially relevant unpublished prospective cohorts and practice surveys.

Database searches were initially conducted in March 2023 and updated on 13 February 2026. A further update will be undertaken if more than 6 months elapse between the most recent search and commencement of full-text screening. If more than 12 months elapse between the updated search and completion of synthesis—an additional update will be performed.

Approximately 50 title/abstract records were informally reviewed by two reviewers during the initial phase. Full-text screening, data extraction, risk-of-bias assessment and synthesis had not commenced at the time of protocol revision. Any subsequent protocol deviations will be documented, dated and justified in the final review and reflected in the PROSPERO record.

Study selection

Two reviewers (SMN and PR) will independently and in duplicate screen titles/abstracts and full texts using Rayyan. Disagreements will be resolved by discussion or arbitration by a third reviewer (VN). Reasons for exclusions at full-text stage will be documented and reported in the PRISMA Flow Diagram.

Data extraction

Two reviewers will independently extract tumour size, volume, suprasellar extension, chiasmal relationship; biochemical NFPA confirmation method; presence and severity of visual dysfunction at baseline; BCVA and reporting scale (LogMAR/Snellen/ETDRS); perimetry device, programme, strategy, reliability criteria and visual field defect definition; colour vision test method; RAPD assessment findings; orthoptic findings (motility, alignment, binocular vision, diplopia, cranial nerve function); OCT device, segmentation software, pRNFL sectors assessed, GCIPL sectors assessed, scan quality criteria and normative database; intervention type and date; exact timing of each ophthalmic assessment relative to intervention; stated rationale for surveillance interval; definitions of visual recovery and deterioration; handling of missing or ungradable tests; study funding and conflicts of interest using an Excel spreadsheet.

Tumour size will be extracted as reported (maximum diameter or volumetric estimate), Knosp grade will be extracted separately as an index of cavernous sinus invasion, not as a measure of tumour size. All forms will be recorded and reported in tabular summaries.

Consensus or a third reviewer will resolve discrepancies. To account for clinical heterogeneity, included studies will be categorised during extraction into the following pre-specified subgroups where data permits: Dimension 1—structural/radiological (tumour-chiasm relationship); (a) radiological contact (tumour abutting chiasm without compression); (b) radiological compression or distortion of the chiasm. Dimension 2—functional (visual deficit); (c) no attributable visual deficit; (d) documented visual deficit attributable to the lesion (visual field loss, reduced BCVA or RAPD). Context categories; (e) pre-intervention surveillance; (f) post-intervention surveillance; (g) surveillance for residual disease or recurrent disease.

Where key methodological data is unclear or unavailable from the published report, the corresponding author will be contacted via email with a maximum of two follow-up attempts over 4 weeks. The outcome of all author contact attempts will be documented in the online supplemental appendix of the final review. Furthermore, where multiple publications appear to derive from the same cohort—the most complete report will be used for primary synthesis—all linked publications will be cross-referenced and listed. Authors will be contacted where overlap is uncertain.

Risk of bias assessment

Risk of bias will be assessed independently and in duplicate by two reviewers (SMN and PR) using the following design-specific tools: Revised Cochrane Risk-of-Bias Tool for Randomised Trials for any randomised controlled trials¹⁶; Risk Of Bias In Non-randomised Studies of Interventions for non-randomised interventional studies¹⁷; Joanna Briggs Institute (JBI) cohort checklist for cohort studies¹⁸; Appraisal tool for Cross-Sectional Studies tool for analytical cross-sectional studies¹⁹; JBI prevalence checklist for practice surveys²⁰; and the JBI case-series checklist for eligible case series.¹⁸ No pooled quality score will be calculated across study designs, and any discrepancies will be resolved by discussion or by a third reviewer.

Data synthesis

Narrative synthesis will follow the SWiM framework,¹³ evidence will be grouped and reported across the following pre-specified categories: (1) observation prior to intervention; (2) pre-operative baseline assessments; (3) early post-operative assessments (≤3 months); (4) late post-operative assessments (>3 months); (5) post-radiotherapy surveillance; (6) and surveillance for residual or recurrent disease.

Within each category, findings will be stratified by assessment modality (functional vs structural), and by the degree of optic pathway involvement (see subgroup categories under population). The direction and consistency of findings across studies, and the influence of risk of bias on conclusions will be discussed. Findings will be characterised as consistent where a clear directional majority of studies reporting a given modality or interval reach the same conclusion, in accordance with SWiM guidance.¹³

For primary descriptive mapping questions, formal grading of recommendations assessment, development, and evaluation (GRADE) assessments will not be applied, as GRADE is designed for intervention evidence and does not translate to surveillance-mapping. For the secondary prognostic association question, certainty of evidence will be assessed using GRADE where quantitative prognostic data predominate, or confidence in the evidence from reviews of qualitative research (GRADE-CERQual) where qualitative or mixed-methods predominates. The framework will be selected prospectively and justified in the final review.

Where feasible, the following pre-specified sub-group analyses will be conducted:

1. Degree of optic pathway involvement (contact vs compression vs functional deficit).

2. Visual deficit at baseline (present vs absent).

3. Management approach (observation vs radiotherapy vs surgery).

4. Primary vs recurrent disease.

5. OCT technology generation (time-domain vs spectral-domain vs swept-source OCT).

6. OCT parameter assessed (pRNFL vs macular GCL/GCIPL).

Studies spanning multiple involvement categories will be categorised under all applicable subgroups and flagged as mixed—sensitivity analyses will assess whether excluding mixed-population studies alters overall conclusions.

Data statement

The piloted data extraction form and any code used for tabular summaries of the narrative synthesis will be made available on request. The form will be piloted independently by two reviewers on a random sample of five included studies before commencement of full extraction—discrepancies will be discussed and the form revised accordingly. No primary dataset will be generated, and individual patient data sharing is not applicable.

Patient and public involvement

Patients were not involved in designing or conduct of this protocol. Patient representatives from The Pituitary Foundation will be invited to review dissemination materials prior to publication. This review directly addresses a patient-identified research need—optimising ophthalmic follow-up after pituitary surgery. This was among the top research priorities identified through the PitCOP international Delphi consensus process—incorporating the perspectives of patients, caregivers and healthcare professionals—across multiple countries.²¹

Findings will be disseminated via patient advocacy groups (eg, The Pituitary Foundation), with educational materials (eg, leaflets) and guidelines developed to support clinical practice.