Neurocognitive outcome of HS-WBRT vs WBRT in patients with brain metastases: A systematic review and meta-analysis

Abstract

Background

Whole-brain radiation therapy (WBRT) is used prophylactically and therapeutically in patients with brain metastases, effectively controlling intracerebral tumors and reducing neurological mortality. However, WBRT poses a significant risk of cognitive decline. Hippocampus-sparing WBRT (HS-WBRT) offers a potential solution by preserving memory and other cognitive functions. This study evaluates neurocognitive outcomes of HS-WBRT compared to WBRT in patients with brain metastases.

Methods

A systematic search was conducted in MEDLINE, Google Scholar, Embase, and CENTRAL for cohort studies and clinical trials reporting neurocognitive outcomes of HS-WBRT vs WBRT, up to March 2024. Non-English studies and those lacking neurocognitive outcomes were excluded. Eligible studies underwent data extraction and analysis focused on neurocognitive function testing.

Results

Of 9 eligible studies, 7 were included in the quantitative analysis. HS-WBRT significantly reduced cognitive decline compared to WBRT, with improvements in Hopkins Verbal Learning Test (HVLT) scores for total recall (SMD = 0.42; P = .02) and delayed recall (SMD = 0.25; P = .02). Cognitive impairment measured by the Montreal Cognitive Assessment (MoCA) was also significantly lower in the HS-WBRT group (SMD = 1.21; P < .00001).

Conclusion

HS-WBRT demonstrates a clear advantage over WBRT in preserving neurocognitive function in patients with brain metastases, as reflected in HVLT and MoCA scores. Future studies should further explore adverse effects and survival outcomes to guide clinical practice.

Key Points.

• HS-WBRT preserves memory and cognition in contrast to WBRT alone.

• HVLT-R showed a significant decline for HS-WBRT overall at the 6-month follow-up.

• MoCA revealed a significantly lower cognitive impairment in HS-WBRT than in WBRT

Importance of the Study.

Brain metastasis is a leading cause of death in patients with systemic malignancies. For survivors, these may have a long-lasting detrimental impact on their overall health. Cognitive function has also been observed to be significantly impaired as an aftermath of brain metastasis in such diseases. While WBRT has been used widely in the past, it is now being replaced by HS-WBRT which preserves areas of the brain vital for memory and learning from receiving radiation. However, the outcomes of this new treatment route need to be assessed thoroughly. In our study, we have analyzed the cumulative neurocognitive scores reported across various studies to delve into the possible effects of HS-WBRT on brain function. We have used validated scales including the Hopkins Verbal Learning Test-Revised (HVLT-R) and Montreal Cognitive Assessment (MoCA) to determine how HS-WBRT can be efficacious in preventing cognitive decline and sustaining mental functioning.

Intracranial or brain metastases constitute a major source of morbidity and mortality in patients with systemic malignancies.¹ The incidence rates of intracranial metastases are comparable to the incidence estimates of primary brain neoplasms combined, ranging from 20%–40% in patients with solid tumors.^2,3 Brain metastases have a profound impact on the clinical prognosis of patients, with many patients presenting with neurologic impairment. These patients also exhibit cognitive decline and impaired performance status and quality of life.⁴ Technological progress in the field has led to the development of more effective treatment strategies, improving survival outcomes.³ Current therapeutic modalities of brain metastases include radiation, medical treatment, and surgical intervention.⁵ Unlike surgery and radiation therapy, systemic drugs have limited efficacy owing to the lack of permeability across the blood-brain barrier.⁶

Historically, brain metastases patients have been treated with whole-brain radiation therapy (WBRT), which combined with surgery yielded survival benefits and improved disease control.⁶ Early investigators also demonstrated minimal toxicity and morbidity in brain metastases patients undergoing radiation therapy.^5,7 While conventional WBRT has long been a cornerstone in the management of patients with multiple brain metastases, increasing yield of evidence related to neurotoxicity and radiation-induced brain injury has contributed to reluctance towards the use of this intervention in brain metastases patients.^8,9 WBRT significantly reduces hippocampal neurogenesis, which is responsible for cognitive deterioration in brain metastases patients. Additionally, WBRT induces an inflammatory response in the hippocampus, which remains persistent for months. Radiation therapy further decreases the perivascular clusters of neural precursors in the hippocampus and impairs the physiological functioning of the mature hippocampal neurons.¹⁰ These effects of radiation on neural stem cells cumulatively translate into a decline in neurocognitive functioning.¹¹

In order to avoid neurocognitive function decline in brain metastases patients, focused approaches such as hippocampus-sparing whole-brain radiation therapy (HS-WBRT) can be used.^11,12 Hippocampal contour relevant to the successful administration of HS-WBRT focuses on the hippocampal dentate gyrus, which is critical to memory formation functions. The main aim of HS-WBRT is to minimize the irradiation dose delivered to the hippocampus while providing adequate irradiation to the whole-brain.^12,13 In the United States, the utilization of HS-WBRT has seen a notable rise during 2016–2022, increasing from 33% to 73%.¹ Moreover, the percentage of radiation oncologists using HS-WBRT, at least occasionally (<50% of cases), increased from 56% in 2018 to 89% in 2023.^14,15 Compared to the US, the reported prevalence of HS-WBRT use in Europe is 4.6%, which can be attributed to the paucity of evidence on treatment patterns.¹⁴ The Federation of Asian Organizations for Radiation Oncology (FARO) survey conducted in 2023 included radiation oncologists from various countries. The survey demonstrated that 50% of the brain metastases cases utilized HS-WBRT.¹⁶

Compared to conventional WBRT, the use of HS-WBRT in brain metastases patients is associated with significantly reduced neurocognitive failure risk and lower deterioration of learning memory, executive function, and verbal memory.^14,17 While initial observational studies and clinical trials have demonstrated promising results with HS-WBRT in brain metastases patients, there are potential disparities in the outcome measures, irradiation dose, and population sizes, making it difficult to draw definitive conclusions.

In this systematic review and meta-analysis, the authors aimed to integrate and analyze the current body of evidence to compare the neurocognitive outcomes of HS-WBRT and WBRT in patients diagnosed with brain metastases.

Materials and Methods

This study was conducted while strictly adhering to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines.¹⁸ The PRISMA checklist is reported in the Supplementary Table S1. This systematic review and meta-analysis is registered on PROSPERO under the ID CRD42024535789.

Data Sources and Search Strategy

Four electronic databases including Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (PubMed), Google Scholar, and Embase were searched for observational cohort studies and clinical trials that have compared WBRT and HS-WBRT in brain metastases patients for assessing neurocognitive outcomes, from inception to November 2024. The literature search comprised the following Medical Subject Headings (MeSH terms): “whole-brain radiotherapy,” “whole-brain radiation therapy,” “hippocampus,” “cognition,” and “brain metastases.” The complete search strategy used in the literature is provided in Supplementary Table S2.

Study Selection

Two independent authors (AS, UN) screened the retrieved articles to match the above-mentioned PICO criteria. The process of study selection was divided into 2 phases. During the first phase, the articles were assessed for their eligibility based on the abstracts whereas the second phase of the study involved full-read texts to finalize the cohort studies and clinical trials. Studies conforming to the PICO eligibility criteria (population: brain metastases patients aged ≥ 18 years; intervention: HS-WBRT; comparison: WBRT; outcome: neurocognitive outcomes) were included in this study. Studies that did not meet the eligibility criteria or published in non-English language(s) along with case reports, reviews, letters, and meta-analyses were excluded. Any disagreements between the 2 authors were resolved after consulting a third author (AJ).

Data Extraction

Three independent authors (AJ, AS, and UN) extracted data about baseline characteristics from each of the included studies and recorded the data into a Microsoft Excel spreadsheet. The baseline characteristics included study design, location, duration, sample size, age of participants, tumor type, primary tumor site, hippocampal avoidance area, and radiation dose. Extracted data on the neurocognitive outcomes of radiation therapy included cognitive test results, psychomotor speed, verbal fluency, processing speed, and executive function.

Quality Assessment

The GRADE Quality Assessment Tool¹⁹ was used by 2 authors (SG and UN) to assess the quality of the included studies. This was followed by a third author (AJ) who resolved any inconsistency in the assessment of the risk of bias.

Statistical Analysis

All statistical analysis was performed on Review Manager (Version 5.4.1, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). The outcomes were pooled using a random effects model. The random effects model assumes that different studies estimated different intervention effects, partly explaining the heterogeneity between studies. Standardized mean differences were used as the primary effect measure. We used the DerSimonian and Laird variance estimator for tau. The Higgins (I²) statistic was used to evaluate heterogeneity; a value of 25%–50% was considered low, 50%–75% as moderate, and > 75% as high heterogeneity. The tolerated level of heterogeneity, meriting little further discussion, is set at less than or equal to 40%, a benchmark decided upon by reviewing the Cochrane Handbook. In all cases, a P-value of .05 or less was considered significant. Publication bias was assessed via visual inspection of Begg’s funnel plots.

Results

Characteristics of Studies

A total of 7 studies were included in the analysis, comprising a mix of randomized controlled trials and retrospective cohort studies,^14,17–27 with a cumulative sample size of 904 patients. The PRISMA flow diagram is demonstrated in Figure 1. These studies were conducted across various geographic locations, including China, the USA, Canada, Spain, Taiwan, and Egypt. The mean age of participants ranged from 41.5 to 63.7 years, with an overall average of approximately 58 years. The study populations were predominantly female, although gender distribution varied across studies. Lung cancer was the most common primary tumor site, followed by breast cancer and other malignancies, including small cell lung cancer, ovarian, and colorectal cancers. The majority of studies defined the hippocampal avoidance region as the hippocampus plus a 5-mm radial expansion, except for one study that used a 3-mm margin.^20,23 Radiation doses were largely standardized at 30 Gy in 10 fractions, with one study employing a 25 Gy protocol. Memantine was used as an adjunct therapy in 2 studies to mitigate cognitive decline. Baseline neurocognitive assessments included the Hopkins Verbal Learning Test (HVLT) for total recall and delayed recall, the Montreal Cognitive Assessment (MoCA), and various tests for verbal fluency, processing speed, and executive function. Follow-up durations across studies ranged from 4 to 60 months, enabling longitudinal comparisons of neurocognitive outcomes. The baseline patient and study characteristics are compiled in Table 1. The neurocognitive outcomes across WBRT and HS-WBRT groups are reported in Table 2.

Figure 1.

PRIMA flow chart

Table 1.

Study and Baseline Patient Characteristics

Author, Year	Study design	Study location	Study duration	Sample size		Gender (male/female)		Mean age	Tumor type	Primary tumor site (HS-WBRT/WBRT)	Other treatment(s)	Hippocampal avoidance area	Radiation dose
				HS-WBRT	WBRT	HS-WBRT	WBRT
Wang, 202120	Retrospective cohort	China	36 mo	n = 27	n = 20	21/6	12/8	62.09 ± 8.86 y	Brain metastases	Lung adenocarcinoma = 15/13 Lung squamous cell carcinoma = 12/7	None	Hippocampus plus 5 mm radial expansion	30 Gy in 10 fractions
Rodriguez de Dios, 202121	RCT	Spain	60 mo	n = 75	n = 75	52/23	53/22	63.65 ± 8.42 y	Brain metastases	Limited SCLC = 53/54 Extensive SCLC = 22/21	None	Hippocampus plus 5 mm radial expansion	25 Gy in 10 fractions
Yang, 202117	RCT	Taiwan	45 mo	n = 33	n = 32	14/19	13/19	58.35 y	Brain metastases	Lungs = 32/29 Breast = 0/2 Other sites = 1/1	None	Hippocampus plus 5 mm radial expansion	30 Gy in 10 fractions
Vinai Gondi, 202322	RCT	USA Canada	32 mo	n = 261	n = 257	N/A	N/A	58.5 ± 20.49 y	Malignant tumors	Lungs*	Memantine	Hippocampus plus 5 mm radial expansion	30 Gy in 10 fractions
Jiangrong Du, 202423	RCT	China	45 mo	n = 15	n = 21	N/A	N/A	N/A	Malignant tumors	Lungs = 11/17 Breast = 1/4 Ovary = 1/0 Colon = 1/0 Esophagus = 1/0	None	Hippocampus plus 3 mm radial expansion	30 Gy in 10 fractions
Weijie Shang, 202224	RCT	China	24 mo	n = 20	n = 20	N/A	N/A	58.00 ± 13.01 y	Malignant tumors	Lungs = 10/11 Breast = 7/5 Other sites = 3/4	None	Hippocampus plus 5 mm radial expansion	30 Gy in 10 fractions
Sherif Elzawawy, 202425	RCT	Egypt	13 mo	n = 21	n = 20	6/15	5/15	41.5 ± 13.4 y	Malignant tumors/Brain metastases	Breast = 15/15 Lungs = 3/4 Other sites = 3/1	None	Hippocampus plus 5 mm radial expansion	30 Gy in 10 fractions
Elaine A. C. Albers, 202326	RCT	The Netherlands	60 mo	n = 82	n = 79	38/44	42/53	62.75 ± 14.7 y	Brain metastases	SCLC*	Platinumeto-poside	N/A	25 Gy in 10 fractions
Paul D. Brown, 202027	RCT	USA Canada	32 mo	n = 261	n = 257	111/150	108/149	61.35 ± 11.55 y	Brain metastases	Bone = 1/1 Breast = 51/45 Colon = 4/6 Esophagus = 6/7 GE junction = 1/1 Kidney = 5/8 Lungs = 156/151 Ovary = 3/3 Skin = 15/7 Anal canal = 1/2 Pancreas = 1/1 Other sites = 17/25	Memantine	Hippocampus plus 5 mm radial expansion	30 Gy in 10 fractions

^*Number of participants not reported in the study.

Abbreviations: Gy = gray; HS-WBRT = hippocampus-sparing whole-brain radiation therapy; N/A = not available; RCT = randomized controlled trial; SCLC = small cell lung cancer; WBRT = whole-brain radiation therapy.

Table 2.

Neurocognitive Outcomes in HS-WBRT vs HBRT

Author, Year	HVLT-total recall	HVLT-delayed recall	HVLT-recognition index	MoCA	Verbal fluency COWA	Memory and verbal learning FCSRT* (WBRT vs HS-WBRT)	Processing speed TMT-A	Executive function TMT-B	Endpoint patient compliance/retention
Wang, 202120	N/A	N/A	N/A	Baseline–6 mo HS-WBRT = + 28.1 WBRT = + 27.028 Baseline–12 mo HS-WBRT = + 26.74 WBRT = + 21.78 P < .001	N/A	N/A	N/A	N/A	6 mo - 91% 12 mo—21%
Rodriguez de Dios, 202121	N/A	N/A	N/A	N/A	N/A	Baseline–3 mo FCSRT-TR = 20.6% vs 8.7% (P = .049) FCSRT-TFR = 14.7% vs 7.2% (P = .16) FCSRT-DFR = 23.5% vs 5.8% (P = .003) FCSRT-DTR = 25.0% vs 13.0% (P = .074) Baseline–6 mo FCSRT-TR = 38.9% vs 20.3% (P = .035) FCSRT-TFR = 31.5% vs 14.8% (P = .040) FCSRT-DFR = 33.3% vs 11.1% (P = .005) FCSRT-DTR = 38.9% vs 22.2% (P = .06) Baseline–12 mo FCSRT-TR = 15.0% vs 12.5% (P = .74) FCSRT-TFR = 12.5% vs 22.5% (P = .23) FCSRT-DFR = 17.5% vs 5.0% (P = .072) FCSRT-DTR = 25.0% vs 12.5% (P = .15) Baseline–24 mo FCSRT-TR = 47.6% vs 14.2% (P = .019) FCSRT-TFR = 23.8% vs 14.2% (P = .43) FCSRT-DFR = 33.3% vs 14.2% (P = .14) FCSRT-DTR = 42.9% vs 28.6% (P = .33)	N/A	N/A	N/A
Yang, 202117	Baseline–4 mo HS-WBRT = + 0.84 WBRT = + 1.67 P = .626 Baseline–6 mo HS-WBRT = + 1.65 WBRT = −0.95 P = .079	Baseline–4 mo HS-WBRT = −0.52 WBRT = + 0.25 P = .313 Baseline–6 mo HS-WBRT = + 0.35 WBRT = −0.65 P = .160	Baseline–4 mo HS-WBRT = −1.72 WBRT = −0.96 P = .392 Baseline–6 mo HS-WBRT = + 0.53 WBRT = −1.25 P = .019	N/A	Baseline–4 mo HS-WBRT = −0.48 WBRT = + 1.65 P = .205 Baseline–6 mo HS-WBRT = + 1.47 WBRT = + 1.00 P = .817	N/A	Baseline–4 mo HS-WBRT = + 15.67 WBRT = + 1.56 P = .177 Baseline–6 mo HS-WBRT = −0.53 WBRT = + 4.10 P = .528	Baseline–4 mo HS-WBRT = + 24.36 WBRT = + 9.71 P = .361 Baseline–6 mo HS-WBRT = −3.18 WBRT = + 9.15 P = .400	N/A
Vinai Gondi, 202322	N/A	Baseline–4 mo HS-WBRT = + 6.65 WBRT = + 5.91 P = .05 Baseline–6 mo HS-WBRT = + 7.87 WBRT = + 6.7 P = .011	N/A	N/A	WBRT vs HS-WBRT* Baseline–4 mo = 12.0% vs 10.5% (P = .73) Baseline–6 mo = 6.2% vs 11.8% (P = .23) Baseline–12 mo = 16.7% vs 8.5% (P = .21)	N/A	WBRT vs HS-WBRT* Baseline–4 mo = 24.8% vs 20.4% (P = .46) Baseline–6 mo = 28.0% vs 17.6% (P = .13) Baseline–12 mo = 31.7% vs 27.7% (P = .65)	WBRT vs HS-WBRT* Baseline–4 mo = 40.4% vs 23.3% (P = .012) Baseline–6 mo = 35.9% vs 23.9% (P = .12) Baseline–12 mo = 44.8% vs 37.0% (P = .42)	N/A
Jiangrong Du, 202423	N/A	N/A	N/A	Baseline–6 mo HS-WBRT = + 26.6 WBRT = + 24.48 Baseline–12 mo HS-WBRT = + 26.4 WBRT = + 24.95	N/A	N/A	N/A	N/A	N/A
Weijie Shang, 202224	N/A	N/A	N/A	Baseline–6 mo HS-WBRT = + 26.9 WBRT = + 25.35 P = .002 Baseline–12 mo HS-WBRT = + 26.11 WBRT = + 24.19 P = .001	N/A	N/A	N/A	N/A	N/A
Sherif Elzawawy, 202425	Baseline–4 mo HS-WBRT = + 27.64 WBRT = + 22.32 P = .001	Baseline–4 mo HS-WBRT = + 8.64 WBRT = + 7.32 P = .003	N/A	N/A	N/A	N/A	N/A	N/A	N/A
Elaine A. C. Albers, 202326	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A	N/A
Paul D. Brown, 202027	Baseline–4 mo HS-WBRT = −1.52 WBRT = −1.77 Baseline–6 mo HS-WBRT = −0.82 WBRT = −1.34 P = .049	Baseline–4 mo HS-WBRT = −1.62 WBRT = −1.86 Baseline–6 mo HS-WBRT = −0.95 WBRT = −1.38	N/A	N/A	N/A	N/A	N/A	N/A	4 mo—40% 6 mo—31% 1 y—16%

^*Decline/deterioration of scores from the baseline; Abbreviations: COWA = Controlled Oral Word Association test; DFR = delayed free recall; DTR = delayed total recall; FCSRT = Free and Cued Selective Reminding Test; HS-WBRT = hippocampus-sparing whole-brain radiation therapy; HVLT = Hopkins Verbal Learning Test; MoCA = Montreal Cognitive Assessment; N/A = not available; TMT = Trail Making Test; TR = total recall; TFR = total free recall; WBRT = whole-brain radiation therapy.

Outcome Analysis

1. HVLT-TR

Three studies assessed HVLT-TR score differences between WBRT and HS-WBRT groups (Figure 2). Treatment with HS-WBRT led to a statistically significant reduction in score decline overall (SMD = 0.42 [0.07,0.77]; I² = 65%; P = .02). This trend was not observed at 4 months (SMD = 0.37 [−0.25,1.00]; I² = 79%; P = .24) but was observed at 6 months (SMD = 0.49 [0.18,0.79]; I² = 0%; P = .002). Heterogeneity observed among studies was high moderate (I² = 65%. P = .02).

Figure 2.

Hopkins verbal learning test (HVLT) scores for total recall (HVLT-TR) at 4 and 6 mo

1. HVLT-DR

Four studies assessed HVLT-DR score differences between WBRT and HS-WBRT groups (Figure 3). Treatment with HS-WBRT led to a statistically significant reduction in score decline overall (SMD = 0.25 [0.02,0.49]; I² = 62%; P = .04). This trend was not observed at 4 months (SMD = 0.15 [−0.27,0.58]; I² = 78%; P = .48) but was observed at 6 months (SMD = 0.37 [0.17,0.57]; I² = 0%; P = .0004). Heterogeneity observed among studies was moderate (I² = 62%; P = .02).

Figure 3.

Hopkins verbal learning test (HVLT) scores for delayed recall (HVLT-DR) at 4 and 6 mo

1. MoCA

Three studies assessed MoCA score differences between WBRT and HS-WBRT groups (Figure 4). Cognitive impairment was lower overall in the HS-WBRT group compared to the WBRT group (SMD = 1.21 [0.35,2.07]; I² = 88%; P = .006). This trend was not significant at 6 months (SMD = 0.52 [0.15,0.89]; I² = 55%; P = .006), but was significant at 12 months (SMD = 2.06 [0.13,3.99]; I² = 94%; P = .04). Heterogeneity among studies was high (I² = 88%; P < .000001).

Figure 4.

Montreal Cognitive Assessment (MoCA) scores at 6 and 12 mo

Quality Assessment

Provided in Supplementary Table S3.

Discussion

WBRT therapy has been qualified for treating brain metastasis²⁸ and its combination with other treatment modalities is under investigation. The Congress of Neurological Surgeons has concluded that using surgery for treating multiple brain metastasis would be advantageous for patients with accessible symptomatic lesions, and controlled or treatable primary disease.²⁸ Since the hippocampus is sensitive to radiation damage, HS-WBRT has emerged as a plausible approach for reducing cognitive damage due to radiation.²⁹ It is documented to limit the hippocampal dose during WBRT and eventually suppress cognitive impairment by preventing intracranial lesions. These are assumed to improve the patient’s quality of life.^30,31 Nevertheless, while an improvement in median survival 1–2 months to 4–6 months or more has been recorded through radiotherapy for brain metastasis³² other effects of this route are under consideration. Although neurocognitive assessments including MoCA and MMSE have supported HS-WBRT as a feasible technique for preserving neurocognitive function,^17,33,34 acute and chronic adverse events are speculated to occur such that neurocognition could be detrimentally affected. These can be sequelae of brain metastasis or plausibly be a consequence of irradiation itself.³² Since studies discussing this aspect are limited and preliminary, we sought to investigate how much neurocognitive outcomes are affected for patients receiving HS-WBRT vs WBRT alone.

Patients are reported to encounter various challenges including memory, attention, and motor control.³⁵ In our comprehensive review, we noted changes through validated neurocognitive test batteries including the HVLT for total recall and delayed recall, the Montreal Cognitive Assessment (MoCA), the Mini-Mental State Examination (MMSE), and various tests for verbal fluency, processing speed, and executive function. Brown et al in their phase III trial investigated WBRT combined with memantine in contrast with WBRT alone and noted that the cognitive failure risk was significantly lower in the HS-WBRT plus memantine arm compared with the WBRT plus memantine arm (HR = 0.76; 95% CI = 0.60 to 0.98; P = .03).²⁷ The findings further the understanding that hippocampal radiosensitivity exists and that HA techniques such as intensity-modulated radiation therapy conserve hippocampal neurogenesis by reducing the radiation dose delivered to the mitotically active neural stem cells located in the hippocampal dentate gyrus. Additionally, the authors record a notable positive impact on neurocognition after following up for 4 months which provides evidence that a survival span of patients greater than equal to 4 months warrants the use of HS-WBRT that preserves brain function.²⁷ It has been hypothesized that memantine can reverse radiation-induced hippocampal excitatory synaptic remodeling. The study by Duman et al elaborates on memantine’s role and implies that there might be a potential synergistic effect of HS-WBRT and memantine on neuroprotection.³⁶ Gondi et al, in their study, divulged that for the HS-WBRT + memantine arm, the deterioration in HVLT-R DR from baseline to 4 months was 7.0% (95% CI = −4.7% to 18.7%), which was significantly lower in comparison with the historical control (P < .001).²² This was consistent with the trend suggested by the assessment of Brown et al noted above.

Moreover, Yang et al demonstrated a slight benefit with HVLT-R total recall and significant protection in HVLT-R recognition index and memory score after HS-WBRT. After employing the memory score (the sum of HVLT-R total recall and recognition index), the authors noted significant stability in the score for the HS-WBRT arm at 6-month follow-up (mean difference = 4.38, 95% CI: 0.72–8.03, P = .020) and till completion till 24 months.¹⁷ In contrast to Brown et al, the present study did not demonstrate any benefit of hippocampal avoidance in neurocognitive preservation by HVLT-R delayed recall at 4 months after treatment which could be assumed to be due to the lack of memantine in this study.¹⁷ Furthermore, a similar trend was reported by Elzawawy et al, 2024, where the HS-WBRT did not show significant results between the mean baseline and after 4 months of assessments for Total and delayed recalls tested by HVLT (P = .746 and .47).²⁵ In the phase 3 trial conducted by Chang et al, a greater than 5-point decrease from baseline to the 4-month follow-up was established as neurocognitive functional decline.³⁶ HVLT-R was utilized and an increase in the mean incidence of neurocognitive function decline was observed (49%, SRS + WBRT, vs 23%, SRS alone) in patients who received WBRT.³⁶

Shang et al utilized the MoCA and found comparable differences between HS-WBRT and WBRT groups respectively. MoCA scores decreased at 3, 6, and 12 months after the therapy, with higher scores in the HS-WBRT than the WBRT group (P < .05). The authors surmise that hippocampal sparing alleviates cognitive function and the differences were because of WBRT damaging neurocognitive function in these patients.²⁴

In addition, Jiangrong Du et al deduced no significant differences in MoCA scores between the HS-WBRT group and the conventional WBRT group before radiotherapy, and 24 hours and 2 months after radiotherapy (P > .05).²³ Within each group, pairwise comparisons showed a steady decline in MoCA scores after radiotherapy in contrast to before radiotherapy and these differences were statistically significant at 6 months and 12 months after radiotherapy (P < .05).²³

Moreover, Wang et al showed in their study revealed that there was a significant difference in neurocognitive function scores observed at 9 months by the MMSE scale (P < .05) and at 12 and 24 months by both MMSE and MoCA scales (P < .001). These differences could be linked to the small number of patients in the study.²⁰ Wang et al also observed that the MMSE has a lower sensitivity that may have affected the detection rates of neurocognitive failure.²⁰ Nonetheless, past research indicates that MMSE is linked with survival in brain metastasis patients.²⁰ Nonetheless, past research indicates that MMSE is linked with survival in brain metastasis patients.³⁷ In a retrospective study, when a group of non-small cell lung cancer patients were analyzed after having received WBRT combined with sequentially (SEB group) or simultaneously (SIB group) integrated boost for brain metastasis; the MMSE score of the SEB group at 3 months after radiation was higher than that of the SIB group (P < .05), showing the larger extent of neurocognitive impairment upon treatment with WBRT + SIB. The neurocognitive dysfunction occurred 3 months after WBRT³⁸ corroborating results in previously mentioned studies.

In our analysis, HVLT-TR scores at the 6-month follow-up showed a statistically significant reduction [SMD = 0.49 [0.18,0.79]; I² = 0%; P = .002] and HVLT-DR also exhibited a significant decline at the 6-month follow-up only [SMD = 0.37 [0.17,0.57]; I² = 0%; P = .0004]. In their systemic review, Rambaldi et al assessed HVLT-R in patients undergoing HS-WBRT. They recognized that HS-WBRT helped retain neurological function as neurocognitive scores were mostly constant between baseline and 4 months with only a significantly lower mean decline in HVLT-R DR at 4 months (7%) compared to historical control (30%).³⁹

In our review, we also evaluated MoCA and demonstrated that cognitive impairment was lower overall in the HS-WBRT group compared to the WBRT group [SMD = 1.21 [0.35,2.07]; I² = 88%; P = .006] and this was significant at 12 months [SMD = 2.06 [0.13,3.99]; I² = 94%; P = .04]. Comparable to these findings, Yan et al analyzed the efficacy and toxicity of WBRT-based combination therapies. After a robust review of several studies, the authors outlined meaningful differences in neurocognitive function between the HS-WBRT group and the historical control group (P < .001; P < .000) Studies included used the Hopkins Verbal Learning Test-Revised (HVLT-R) test, Montreal Cognitive Assessment (MoCA) among other test options to discern these neurocognitive outcomes.⁴⁰

Understanding that clinically significant outcomes are not always the same as statistically significant results is critical. Findings that improve medical treatment and a person’s capacity for social interaction, mental health, and physical function are considered clinically significant.⁴¹ The effect size is one metric used to evaluate the clinical significance of our findings. Effect sizes were classified as modest (d = 0.2), medium (d = 0.5), and big (d > 0.8) by Cohen et al. The authors highlighted that a little influence of 0.2 is less obvious but still significant, but a medium effect of 0.5 is discernible to a keen observer. Just as a small effect is noticeably smaller than a medium effect, a large effect of 0.8 is noticeably larger than a medium effect.⁴²

Accordingly, our findings suggest that for HVLT-Total recall, treatment with HS-WBRT showed a moderately clinically significant reduction in overall cognitive decline at six months. For HVLT-delayed recall, reduction with HS-WBRT showed a small overall clinical significance. No meaningful difference was observed at 4 months, and by 6 months, the clinical significance was small to moderate size. For MoCA, HS-WBRT resulted in lower overall cognitive impairment with a large clinical impact. At 6 months, the effect was moderate but still meaningful, and by 12 months, the cognitive benefit remained significant with a substantial effect size.

Because the MoCA is a thorough test that assesses several cognitive domains, it offers a more comprehensive evaluation of cognitive function, which may explain the strong clinical significance seen in MoCA scores when compared to HVLT assessments. By comparison, HVLT offers a more limited breadth of cognitive examination and concentrates exclusively on verbal memory. This contrast suggests that HS-WBRT may improve patients’ quality of life by providing more significant effects across numerous cognitive domains instead of only memory.

Limitations

In our study, we have examined in detail the degree of neurocognitive effects because of radiation treatment. We have explored in depth how sparing the hippocampal region can maintain brain function and the length of time after which the potential demise of function can be anticipated. Notwithstanding, there were several limitations. For instance, the rate of compliance with neurocognitive tests differed across the studies, and in some instances, even dropped below 50%; however, the compliance rate was seldom reported across all studies. Our studies were selected from publicly available databases and the articles included were strictly in the English language which could have excluded relevant studies available in other languages. Also, a few studies had small sample sizes and overall study characteristics were not congruous which may be associated with the moderate to high heterogeneity across our results. Some of the trials included in our review had not concealed treatment allocation to participants and providers that might have affected neurocognitive test scores, although some trialists have dismissed this as a potential confounding factor. Further, it has been established that the brain tumor status and health of the underlying brain tissue affect neurocognitive which could not be addressed in our study. Notwithstanding, while it may limit our results, the outcome measures of our study addressed differences in neurocognitive outcomes relative to each other based on the intervention that was being investigated. Given that the baseline tumor status did not have significant differences between the treatment groups, the baseline brain metastasis would have a trivial impact influencing our data. Another limitation of our review is that we were not able to examine the impact of HS-WBRT on critical treatment outcomes, such as progression-free survival and overall survival, which are essential to assessing the overall quality of life for patients. However, this was beyond the scope of our study as this research was centered towards assessing neurocognitive outcomes. Lastly, we analyzed the combination of 30Gy-in-10 and 25Gy-in-10 regimens due to the limited number of studies available for each. While this approach maximized statistical power, it may overlook potential differences between the 2 regimens.

Withal, our thorough findings that comprise high-quality studies have paved the way for determining the superiority and benefits of HS-WBRT in comparison to WBRT strictly in the realm of neurocognitive outcomes. However, it is also important to note that this study did not assess outcomes of survival and progression-free survival, which would be far greater determinants of the clinical utility of any treatment.

Conclusion

Prevention of intracranial lesions and curbing or averting cognitive impairment are the core objectives for patients with brain metastasis. Dynamic changes in neurocognitive function after HS-WBRT or WBRT can be decisive when specifying a patient’s treatment plan. Our systematic review and meta-analysis widen the pool of literature and give insight to further the current knowledge regarding the potential cognitive side effects. With respect to the available therapy options, we have rigorously assessed studies and assembled relevant outcomes that can guide future actions for this patient population and their caregivers.

Funding

None.

Conflict of interest statement. None declared.

CRediT authorship statement

Afia Salman (Writing—original draft, Writing—reviewing and editing, Validation, Data curation, Visualization, Conceptualization, Supervision). Unaiza Naeem (Writing—original draft, Writing—reviewing and editing, Data curation, Visualization). Shamas Ghazanfar (Writing—reviewing and editing, Data curation, Visualization, Analysis, Resources). Areesha Jawed (Data curation, Visualization, and Writing—reviewing and editing). Minaam Farooq (Validation, Supervision, Writing—reviewing and editing)

Data Availability

The supplementary data will be made available upon reasonable request.