Significance of arterial spin labeling in relation to sertraline treatment response in adolescents with major depressive disorder

Abstract

BACKGROUND

Major depressive disorder (MDD) in adolescents is a serious global mental health issue, often accompanied by significant functional impairment and limited treatment response. A deeper understanding of neurobiological predictors of antidepressant efficacy may therefore enable more targeted and effective interventions. Altered cerebral blood flow (CBF) has been implicated in the pathophysiology of depression. However, the relationship between regional perfusion changes and subsequent treatment outcomes in adolescents remains underexplored.

AIM

To investigate changes in CBF after eight weeks of sertraline treatment in adolescents with MDD and to determine whether baseline CBF is associated with subsequent clinical response.

METHODS

Thirty-three adolescents with MDD underwent magnetic resonance imaging scans as well as clinical evaluations at baseline and after eight weeks of sertraline treatment. Specifically, CBF was quantified using arterial spin labeling, while severity of depressive symptoms was assessed with the 17-item Hamilton Depression Rating Scale (HAMD-17). Additionally, voxel-wise analyses compared pre-treatment and post-treatment CBF, with correlation analyses subsequently conducted to examine associations between regional CBF and changes in depressive symptom severity (ΔHAMD scores).

RESULTS

Sertraline treatment led to a significant reduction in HAMD-17 scores, with 51.5% of participants classified as treatment responders. Voxel-wise analysis further revealed a significant decrease in CBF in the right insular cortex after treatment (t = 4.751, P < 0.001). Notably, higher baseline CBF in the right insula was positively correlated with greater reductions in depressive symptoms (R² = 0.183, P = 0.013), whereas post-treatment CBF showed no such association.

CONCLUSION

Baseline perfusion in the right insular cortex showed an association with subsequent antidepressant response in adolescents with MDD. These exploratory findings do not establish causality or predictive utility; confirmation in randomized, placebo or psychotherapy-controlled trials is required before informing clinical decision making, although they may contribute to understanding neurovascular features related to clinical recovery in youth depression.

Key Words: Adolescent depression; Functional magnetic resonance imaging; Major depressive disorder; Cerebral blood flow; Arterial spin labeling

Core Tip: Adolescent major depressive disorder is a global concern with limited treatment efficacy. This study used arterial spin labeling magnetic resonance imaging to evaluate cerebral blood flow changes during sertraline treatment and to examine whether baseline perfusion was associated with subsequent response. Thirty-three adolescents with adolescent major depressive disorder were assessed before and after eight weeks of treatment. A significant reduction in depressive symptoms was observed over eight weeks, and responders showed higher baseline cerebral blood flow in the right insula. These exploratory findings suggest that insular perfusion may index treatment-relevant states, but do not establish causality or predictive utility; confirmation in randomized, placebo or psychotherapy-controlled trials is required before informing clinical decision-making.

INTRODUCTION

Adolescent major depressive disorder (MDD) has emerged as a significant global public health concern, exerting profound effects on individual psychological development, family dynamics, and broader societal functioning[1,2]. However, addressing adolescent depression is particularly challenging due to three main factors. Firstly, its clinical manifestations during adolescence are often subtle and insidious[3], with emotional disturbances frequently presenting as irritability[4], academic decline[5] or unexplained somatic complaints[6]. These atypical features contribute to delayed diagnosis and low utilization of psychiatric services. Secondly, even when timely and appropriate clinical interventions are provided, treatment outcomes are often suboptimal[7-9], with a substantial proportion of adolescents either experiencing only partial symptom remission or relapsing shortly after initial improvement[10]. Finally, the etiology of adolescent depression remains incompletely understood, especially in reflecting the complex interplay of genetic vulnerabilities[11], neurodevelopmental difficulties[12], environmental stressors[13] and dysfunctions in neural circuits involved in emotion regulation[14].

Among existing therapeutic strategies, pharmacological treatment remains a cornerstone in managing moderate to severe depressive episodes in adolescents[15]. In this context, sertraline, a widely prescribed selective serotonin reuptake inhibitor, has demonstrated a favorable safety profile as well as consistent efficacy in alleviating depressive symptoms and improving psychosocial functioning in this population[16]. Nonetheless, due to considerable interindividual variability, reported response rates in adolescents with MDD approximately 69% based on Children’s Depression Rating Scale-Revised criteria[17], thus highlighting the urgent need for objective neurobiological indicators that may help anticipate treatment response prior to initiation. Indeed, in the absence of reliable validated biomarkers, clinicians often have to adopt a trial-and-error approach, which prolongs patient suffering and places a significant burden on healthcare systems. Moreover, the neurobiological mechanisms differentiating responders from non-responders remain poorly understood, particularly in the context of the dynamic and still-developing adolescent brain.

Arterial spin labeling (ASL), a non-invasive magnetic resonance imaging technique, offers promising opportunities for elucidating the neural substrates of depressive disorders[18]. This approach utilizes magnetically-labeled arterial blood water as an endogenous tracer to quantify regional cerebral blood flow (rCBF), thereby serving as a surrogate marker of resting-state brain activity without the need for exogenous contrast agents. Given the close coupling between rCBF and neuronal activity, ASL has been increasingly applied in the investigation of psychiatric conditions[19,20]. Notably, in adolescents with MDD, emerging evidence indicates aberrant perfusion patterns, particularly within the fronto-limbic circuitry and the default mode network[21], regions critically involved in emotion regulation, cognitive control and self-referential processing. These functional abnormalities represent potential candidates for biomarkers of both disease diagnosis and treatment prognosis. Additionally, compared with conventional blood-oxygen-level-dependent functional magnetic resonance imaging, ASL has demonstrated good test-retest reliability across healthy and clinical populations, including psychiatric cohorts[22,23].

ASL building on these insights, the present study was designed to explore the longitudinal associations of ASL-derived cerebral perfusion characteristics with subsequent therapeutic response to sertraline in adolescents diagnosed with MDD. By examining baseline rCBF profiles and comparing them between treatment responders and non-responders, we observed neuroimaging signals associated with subsequent response. Ultimately, this research aimed to facilitate the development of individualized pharmacological interventions, thereby enhancing therapeutic efficacy, reducing unnecessary exposure to ineffective treatments and promoting a more precise and biologically informed approach to managing adolescent depression.

MATERIALS AND METHODS

Ethics declaration

Written informed consent was provided by the legal guardians of all participants prior to enrollment. The study protocol was reviewed and approved by the Medical Ethics Committee of the First Affiliated Hospital of Chongqing Medical University (Approval No. 2020-779). All procedures were conducted in accordance with the ethical standards set forth in the Declaration of Helsinki.

Participants

A total of 40 adolescents, aged 12 years to 17 years and clinically diagnosed with MDD, were enrolled in this study. Participants were consecutively recruited from the Department of Psychiatry at the First Affiliated Hospital of Chongqing Medical University between January 2021 and December 2022. The diagnosis of MDD was independently confirmed by two experienced clinicians the mini-international neuropsychiatric interview for children and adolescents based on the criteria outlined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition. In addition to the categorical diagnosis, the severity of depressive symptoms was also assessed using the 17-item Hamilton Depression Rating Scale (HAMD-17).

The following inclusion criteria were applied: (1) A new diagnosis of MDD, either treatment-naïve or no exposure to antidepressant medication or electroconvulsive therapy within the previous month; (2) Han ethnicity; (3) Aged between 12 years and 17 years; (4) Right-handedness; (5) At least primary school level education; and (6) A baseline HAMD-17 score greater than 17. Additionally, the exclusion criteria included: (1) Comorbid psychiatric disorders; (2) A history of significant neurological disease or structural brain abnormalities; (3) Severe physical illness; (4) Current or past substance use disorders; (5) Inability or unwillingness to comply with study procedures; and (6) Contraindications to magnetic resonance imaging. All participants underwent comprehensive baseline assessments and were reassessed following an eight-week course of treatment.

Intervention procedure and treatment evaluation

Sertraline was used as the primary pharmacological intervention for managing depressive symptoms. The initial dosage was 50 mg/day, with titration up to 100-200 mg/day based on clinical assessment of symptom severity and individual tolerability. The treatment period was standardized to eight weeks for all participants. Treatment response was defined as a ≥ 50% reduction in HAMD-17 scores from baseline, indicating an improvement in the severity of depressive symptoms. Conversely, participants not meeting this threshold were classified as non-responders. This criterion is consistent with widely accepted standards in psychopharmacological research and clinical trials[24,25].

Data collection and evaluation measures

At baseline, all 40 adolescents diagnosed with MDD underwent assessment using the HAMD-17. Of these, 33 participants completed a follow-up evaluation after eight weeks of treatment. Sociodemographic information: Basic demographic data, including age, sex and educational attainment, were collected using a structured questionnaire. Depression severity assessment: The HAMD-17 was employed to quantify the severity of depressive symptoms. In this case, symptom severity was categorized according to established thresholds: 0-7 (no depression), 8-17 (mild), 18-24 (moderate), and > 24 (severe).

Magnetic resonance imaging acquisition protocol

Neuroimaging data were acquired using a 3.0 Tesla Signa MRI scanner (General Electric Medical Systems, Waukesha, WI, United States). The imaging protocol comprised high-resolution T1-weighted anatomical scans, ASL perfusion imaging and T2-fluid-attenuated inversion recovery sequences. In this case, participants with abnormal findings on T2-fluid-attenuated inversion recovery imaging were excluded from subsequent analyses to avoid potential confounding effects of underlying neuropathology and were referred to the department of neurology for further evaluation. Participants were scanned in the supine position and instructed to remain awake, relaxed and with eyes closed, while avoiding active cognitive engagement. Earplugs were also provided for acoustic protection, and head motion was minimized using foam padding. All imaging sessions were performed using a standard eight-channel head coil, with identical acquisition parameters employed for both baseline and follow-up scans. Structural imaging: T1-weighted anatomical images were acquired using the following parameters: Repetition time = 8.4 milliseconds, echo time = 3.3 milliseconds, number of excitations = 3, flip angle = 12°, field of view = 240 mm × 240 mm and matrix = 256 × 256, with 156 axial slices at a slice thickness of 1 mm. Perfusion imaging used 3D pseudo-continuous ASL (PCASL) with labeling duration (LD) 1800 milliseconds and post-labeling delay 1525 milliseconds. Dual background suppression (BS) pulses were applied. A separate M0 proton-density-weighted image was acquired for absolute cerebral blood flow (CBF) calibration. Parameter choices follow consensus recommendations for clinical ASL (PCASL, LD ≈ 1.8 seconds, BS, separate M0)[26]. Other key parameters are as follows: Repetition time = 4639 milliseconds, echo time = 9.8 milliseconds, number of excitations = 3, field of view = 240 mm × 240 mm, slice thickness = 4.0 mm, and a 3D spiral k-space readout covering 40 slices with 512 sampling points and 8 arms. Participant wakefulness during scanning was confirmed immediately after image acquisition.

Quality control and exclusion

Scans were excluded if mean framewise displacement or any inter-pair displacement exceeded 2.5 mm, if BS failure or coregistration failure occurred, or if gray-matter CBF fell outside 10-120 mL/100 g/minute. Scans failing BS integrity checks, extreme gray-matter CBF (< 10 mL/100 g/minute or > 120 mL/100 g/minute), or coregistration failure were excluded from inferential analyses.

Image processing

All raw digital imaging and communication in medicine images were initially converted to neuroimaging informatics technology initiative format using the dcm²nii tool (https://neuroelf.net/wiki/doku.php?id=dcm2nii). Subsequent preprocessing was conducted using SPM12, implemented in MATLAB R2013b. Data were processed with a standardized ASL pipeline (including motion correction, BS verification, coregistration to T1, and absolute CBF computation). A one-step registration procedure was then employed to align each subject’s images to a standardized anatomical space, followed by meticulous visual inspection to ensure proper image alignment. Furthermore, image normalization was performed using DPABI v4.3 (http://rfmri.org/dpabi) to maintain anatomical consistency across participants. Spatial smoothing was then applied in SPM12 using an 8-mm full-width at half maximum gaussian kernel before eventually standardizing signal intensity based on the global mean signal across the entire brain to enhance inter-subject comparability.

Statistical analysis

Paired t-tests were used to assess within-subject differences in HAMD-17 scores and CBF between baseline and post-treatment scans. In this case, voxel-wise statistical significance was set at P < 0.001, with correction for multiple comparisons performed at the cluster level using the family-wise type I error rate method (corrected P < 0.05). All statistical analyses were performed within the SPM12 framework. Statistically significant clusters were then visualized using XjView10, and the mean CBF values from these clusters were extracted using DPABI v4.3.

RESULTS

Clinical outcomes following eight weeks of sertraline treatment in adolescents with MDD

A total of 33 adolescents diagnosed with MDD completed both magnetic resonance imaging scans and clinical assessments at baseline and after an eight-week course of sertraline treatment. The demographic and clinical characteristics of these participants are summarized in Table 1. Initially, 40 adolescents were recruited, but four did not complete the post-treatment magnetic resonance imaging and clinical assessments, two discontinued sertraline prematurely and one was lost to follow-up (Figure 1).

Figure 1 Flow diagram illustrating participant selection and exclusion criteria for the study. MINI-KID: Mini International Neuropsychiatric Interview for children; HAMD-17: 17-item Hamilton Depression Rating Scale; MRI: Magnetic resonance imaging.

Table 1 Demographic and clinical characteristics of adolescents with major depressive disorder, mean ± SD.

Item	Pre-treatment (n = 33)	Post-treatment (n = 33)	t	P value
Gender (female/male)	24/9	24/9	-	-
Age (years)	15.18 ± 1.93	15.18 ± 1.93	-	-
Education (years)	9.15 ± 1.50	9.15 ± 1.50	-	-
HAMD-17	23.48 ± 3.83	12.79 ± 6.36	11.467	< 0.001

HAMD-17: 17-item Hamilton Depression Rating Scale.

The final cohort consisted of 24 females and 9 males, with a mean age of 15.18 ± 1.93 years and an average educational attainment of 9.15 ± 1.50 years. After the eight-week treatment period, a significant reduction in depressive symptom severity was observed. Notably, the mean HAMD-17 scores declined from 23.48 ± 3.83 at baseline to 12.79 ± 6.36 post-treatment (t = 11.467, P < 0.001). Treatment response was defined a priori as a ≥ 50% reduction in HAMD-17 scores from baseline, consistent with established clinical criteria for antidepressant efficacy. Based on this criterion, 17 participants (51.5%) were classified as responders, while 16 (48.5%) were categorized as non-responders.

Changes in CBF following treatment

Voxel-wise analysis of CBF revealed a significant post-treatment reduction in perfusion within the right insular cortex (Insula_R) following eight weeks of sertraline administration (Figure 2). Quantitative analysis subsequently confirmed this finding: The mean CBF in the right insula decreased from (66.71 ± 7.91) mL/100 g/minute at baseline to (62.72 ± 6.39) mL/100 g/minute after treatment, thus representing a statistically significant difference (t = 4.751, P < 0.001) (Table 2).

Figure 2 Voxel-wise comparison of cerebral blood flow in the right insular cortex between baseline and post-treatment with sertraline. Significant decreases were observed in the right insula (P < 0.001, family-wise type I error rate: Corrected).

Table 2 Regional cerebral blood flow in the right insular cortex before and after sertraline treatment, mean ± SD.

CBF value (mL/100 g/minute)	Pre-treatment (n = 33)	Post-treatment (n = 33)	t	P value
Insula_R	66.71 ± 7.91	62.72 ± 6.39	4.751	< 0.001

CBF: Cerebral blood flow.

To further clarify treatment effects, subgroup analyses were performed for responders and non-responders. Responders (n = 17) demonstrated a significant reduction in HAMD-17 scores (22.59 ± 3.32 to 8.06 ± 3.60, P < 0.001), whereas non-responders (n = 16) showed numerically smaller reduction that did not reach statistical significance (24.44 ± 4.21 to 17.81 ± 4.51, P = 0.077). Similarly, right insular CBF significantly decreased in responders (67.35 ± 8.03 to 61.42 ± 5.89, P < 0.001), but not in non-responders (66.02 ± 7.81 to 64.12 ± 6.62, P = 0.053). These results suggest that CBF reduction in the right insula is more pronounced among clinical responders (Table 3).

Table 3 Responders vs non-responders (17-item Hamilton Depression Rating Scale and right insular cerebral blood flow), mean ± SD.

Group	HAMD-17 baseline	HAMD-17 post treatment	P value	Insula_R CBF baseline	Insula_R CBF post-treatment	P value
Responders (n = 17)	22.59 ± 3.32	8.06 ± 3.60	< 0.001	67.35 ± 8.03	61.42 ± 5.89	< 0.001
Non-responders (n = 16)	24.44 ± 4.21	17.81 ± 4.51	0.077	66.02 ± 7.81	64.12 ± 6.62	0.053

HAMD-17: 17-item Hamilton Depression Rating Scale; Insula_R: Right insular; CBF: Cerebral blood flow.

Correlation between right insular cortex CBF and treatment response

Correlation analyses were also performed to assess the relationship between CBF in the Insula_R and therapeutic efficacy, as reflected by changes in depressive symptom severity (ΔHAMD scores). At baseline, higher CBF values within the Insula_R were significantly associated with greater reductions in HAMD-17 scores following eight weeks of sertraline treatment (R² = 0.183, P = 0.013), hence indicating that elevated baseline perfusion in this region was associated with a more favorable clinical response (Figure 3). In contrast, no significant correlation was found between post-treatment CBF in the right insula and symptom improvement (R² = 0.054, P = 0.193).

Figure 3 Relationship between baseline right insular cortex cerebral blood flow and changes in 17-item Hamilton Depression Rating Scale scores post-treatment. CBF: Cerebral blood flow; HAMD: Hamilton Depression Rating Scale.

DISCUSSION

This study examined the effects of an eight-week course of sertraline on CBF and clinical outcomes in adolescents diagnosed with MDD. Overall, a significant reduction in depressive symptoms was observed, with approximately half of the participants meeting the criteria for treatment response. Voxel-wise analysis further revealed a significant decrease in CBF in the right insular cortex following treatment. Moreover, baseline CBF in this region was positively correlated with symptom improvement, whereas post-treatment CBF showed no significant association with clinical outcomes. These findings suggest that pre-treatment neurovascular activity in the right insula was associated with subsequent antidepressant response in adolescent MDD, this pattern may offer insights into the neurobiological processes related to treatment efficacy.

The observed reduction in right insular CBF following sertraline treatment is consistent with previous findings[27,28] implicating this region in the pathophysiology of depression. The insula plays a critical role in emotion regulation[29,30], interoceptive awareness[31] and the integration of affective and cognitive processes[32-34], with these different domains commonly disrupted in MDD[32]. Decreased perfusion in this area post-treatment may reflect the normalization of aberrant neural activity associated with depressive symptomatology. Therefore, the current findings extend existing literature by demonstrating that baseline CBF in the insula, rather than post-treatment changes, is more strongly associated with therapeutic response in adolescents. While previous studies[35-37] have reported that resting-state functional alterations in the insula may predict antidepressant response, these works primarily focused on functional connectivity rather than CBF metrics. Hence, the present study provides a novel hemodynamic perspective by identifying a perfusion-based correlate of treatment efficacy.

Right anterior insula shows lateralized specialization for interoceptive awareness and sympathetic autonomic control and serves as a hub of the salience network that mediates switching between default-mode and executive systems. Accordingly, baseline right anterior insula perfusion can plausibly act as a predictor of antidepressant response via salience-driven regulation, consistent with ASL/functional studies, including sertraline cohorts and emerging adolescent evidence, linking insula-anchored network activity to treatment outcomes.

Additionally, the observed positive correlation between baseline right insular CBF and treatment response underscores the potential of neuroimaging biomarkers to inform personalized treatment strategies. Adolescents exhibiting relatively preserved or elevated insular perfusion may possess enhanced neuroplastic capacity, making them more receptive to pharmacological modulation of mood-related neuronal circuits. This interpretation aligns with emerging neurodevelopmental frameworks which suggest that early identification of biological resilience factors can improve treatment stratification and optimize clinical outcomes[38]. While previous research has predominantly focused on structural markers such as cortical thickness[39] or functional connectivity measures[40] as predictors of treatment response, the current findings highlight the prognostic relevance of perfusion metrics. These measures, which reflect both vascular integrity and metabolic activity of key brain regions, may offer complementary insights into the neurobiological mechanisms underlying therapeutic efficacy.

In this study we implemented a 3D PCASL protocol-LD 1800 milliseconds, PLD 1525 milliseconds, dual BS, separate M0, and partial-volume correction-consistent with clinical/clinical-research recommendations. Nevertheless, arterial transit time lengthens with age and varies across individuals; consequently, a PLD of 1525 milliseconds may be suboptimal for older adolescents, potentially underestimating gray-matter CBF and altering regional contrasts. Looking ahead, multi-PLD (or time-encoded) acquisitions that jointly estimate arterial transit time and CBF should be prioritized.

Our open-label, single-arm design precludes separation of putative medication effects from time-dependent and expectancy-related influences, including spontaneous remission, regression to the mean, clinical contact, and non-specific factors. These issues are especially pertinent in pediatric depression, where placebo response can be substantial. Recent analyses highlight that early ASL changes may also track placebo response, underscoring the design sensitivity of neuroimaging claims[41]. Accordingly, the present findings should be interpreted as longitudinal associations rather than causal effects or predictive biomarkers suitable for clinical decision-making.

Despite the promising findings, several limitations should be acknowledged. Firstly, the sample size, although comparable to that of other neuroimaging studies in adolescent depression, was relatively modest, potentially limiting statistical power and the generalizability of results. Secondly, this study focused exclusively on a single antidepressant (sertraline) and a fixed treatment duration (eight weeks), thus, making it unclear whether similar perfusion patterns would emerge with other pharmacological agents or extended treatment timelines. Thirdly, the absence of a healthy control group limited the ability to determine whether baseline CBF levels were pathologically elevated or fell within normative developmental parameters. Finally, while ASL offers a non-invasive approach to quantifying CBF, its spatial resolution and susceptibility to motion artifacts, especially in adolescent populations, warrant cautious interpretation of the imaging results.

CONCLUSION

This study demonstrated that higher baseline CBF in the right insular cortex, as measured by ASL, was significantly correlated with greater clinical improvement following an eight-week course of sertraline in adolescents diagnosed with MDD. These results suggest that right insular perfusion was associated with subsequent antidepressant treatment response in this open-label, single-arm cohort and does not establish predictive utility. Accordingly, while ASL-derived perfusion metrics may inform hypothesis generation, their incorporation into early clinical decision-making is premature and requires confirmation in randomized, placebo or psychotherapy-controlled trials; if validated, they could help support more personalized treatment strategies for adolescents.

ACKNOWLEDGEMENTS

We extend our sincere gratitude to all participating adolescents and their families for their commitment to the study. We also acknowledge the use of DPABI and SPM12 software for processing neuroimaging data and recognize the contributions of the research assistants involved in data management and quality control.