Baseline markers of treatment response to vagus nerve stimulation in difficult-to-treat depression: a retrospective data analysis

This retrospective analysis of 53 patients failed to replicate the previously reported correlation between baseline corrected QT interval and VNS treatment response in difficult-to-treat depression, but instead identified lower absolute neutrophil count, lower BMI, younger age, and higher LDL levels as potential baseline predictors of improvement.

The Gist

The Heartbeat, The Blood, and The Brain: A Simple Guide to This Study

Imagine your brain is a house that has been stuck in a dark, stormy room for years. This is Depression. For some people, trying different keys (medications, therapy, shock therapy) doesn't open the door. This is called Difficult-to-Treat Depression (DTD).

When the keys don't work, doctors sometimes try a different tool: Vagus Nerve Stimulation (VNS). Think of the vagus nerve as a super-highway cable running from your brain down to your heart and gut. VNS is like installing a tiny, permanent pacemaker on that cable. It sends gentle electrical signals to "reset" the brain's mood circuits.

However, there's a catch: It's expensive, requires surgery, and it doesn't work for everyone. About half the people who get the implant don't see much improvement. Doctors are desperate for a "Crystal Ball"—a way to look at a patient before the surgery and say, "Yes, this will work for you," or "No, let's try something else."

This study was a team of doctors trying to find that Crystal Ball. They looked at two main things: Heartbeats and Blood.

---

1. The Heartbeat Test (The QTc Interval)

The Theory:
A previous study suggested that the length of a specific part of a heartbeat (called the QTc interval) could predict success. Imagine the heartbeat is a drumbeat. If the "rest" between beats is unusually long, it might mean the brain is "listening" better to the VNS signal. The old study said: Longer rest = Better mood improvement.

What This Study Found:
The researchers gathered data from 53 patients across three different hospitals. They measured everyone's heartbeat before the surgery.

• The Result: They could not confirm the old finding. When they looked closely and accounted for things like age, gender, and body weight, the length of the heartbeat had no connection to whether the mood improved.
• The Analogy: It's like trying to predict if a car will run well just by listening to the engine idle. In the old study, a specific idle sound seemed to predict success. In this larger study, that sound turned out to be just background noise.

2. The Blood Test (The "Fingerprint" of the Body)

Since the heartbeat didn't work as a predictor, the researchers looked at the patients' blood samples (from 22 patients) to see if there were other clues. They looked for signs of inflammation (like a fire in the body) and metabolic health (how the body processes fuel).

The Surprising Clues Found:
They found four "fingerprint" markers that seemed to hint at who might get better:

1. White Blood Cells (Neutrophils):

   • The Finding: Patients with lower levels of a specific white blood cell (neutrophils) tended to get better.
   • The Analogy: Think of neutrophils as the body's "first responders" to a fire. If your body is constantly sending out too many firefighters (high inflammation), the brain might be too "noisy" to hear the VNS signal. A quieter, calmer body (lower neutrophils) might be more receptive to the treatment.

2. Body Weight (BMI):

   • The Finding: Patients with a lower Body Mass Index (BMI) tended to improve more.
   • The Analogy: Imagine trying to tune a radio. If the radio is covered in thick mud (high body weight/metabolic issues), the signal is fuzzy. A cleaner radio (lower BMI) picks up the VNS signal more clearly.

3. Cholesterol (LDL):

   • The Finding: Surprisingly, patients with higher levels of "bad" cholesterol (LDL) actually improved more.
   • The Analogy: This is counter-intuitive! Usually, high cholesterol is bad. But here, it might be like having a full tank of fuel. Perhaps the body's metabolic engine was revving high, and the VNS helped redirect that energy into healing the mood.

4. Age:

   • The Finding: Younger patients tended to see faster improvements in the first six months.
   • The Analogy: A young sapling bends and grows quickly when the wind (VNS) blows. An old, gnarled tree might take much longer to shift its branches.

The Big Picture: What Does This Mean?

The Good News:
We are getting closer to "Precision Psychiatry." Instead of guessing, we might one day be able to say, "Based on your blood work and body type, you have a high chance of responding to VNS."

The Bad News:

• The Heartbeat Myth: The "heartbeat length" trick from the old study doesn't seem to hold up when you look at a bigger group of people.
• The Blood Clues are "Exploratory": The blood findings (neutrophils, BMI, etc.) are interesting hints, but they are based on a small group of people. It's like finding a few footprints in the sand; you suspect a person walked there, but you need to see the whole trail to be sure.
• Not a Crystal Ball Yet: We cannot use these results to decide who gets surgery right now. The study was too small and didn't have a control group (a group that didn't get the treatment) to be 100% certain.

The Takeaway

This study is like a detective clearing up a false lead (the heartbeat) and finding some new, promising clues (the blood markers) at a crime scene. The detectives (doctors) are now saying: "We need to investigate these blood clues more thoroughly with a bigger team before we can solve the case."

Until then, VNS remains a powerful tool for difficult depression, but we still need to be careful about who we choose to implant it in.

Technical Summary

1. Problem Statement

Vagus Nerve Stimulation (VNS) is an FDA-approved invasive neuromodulatory treatment for Difficult-to-Treat Depression (DTD). However, clinical response rates are suboptimal, with approximately 50% of patients failing to achieve sufficient remission. Currently, there is a lack of reliable baseline biomarkers to predict which patients will respond to VNS, leading to a "trial-and-error" approach in patient selection.

A previous study (Longpré-Poirier et al., 2020) suggested that a longer baseline corrected QT (QTc) interval correlates with better treatment response. However, that study had a small sample size ($N=19$) and did not control for critical covariates such as sex, age, or medication effects on QTc. This study aims to:

1. Replicate the finding regarding baseline QTc intervals as a predictor of VNS response in a larger, independent cohort.
2. Explore other physiological and blood-based markers (inflammatory and metabolic) that might predict treatment outcomes.

2. Methodology

• Study Design: Retrospective data analysis pooling data from three VNS centers (Vienna, Münster, Frankfurt) participating in the RESTORE-LIFE study.
• Participants:
  • Total Cohort: 53 patients with DTD (unipolar or bipolar) meeting DSM-5 criteria.
  • Analysis Cohort: 48 patients with complete baseline ECG and BMI data.
  • Sub-cohort for Blood Analysis: 22 patients (Vienna site only) with available pre-surgical blood samples.
• Intervention: Adjunctive VNS implantation with standard titration protocols. Antidepressant treatments (pharmacotherapy, psychotherapy, ECT/TMS) were continued as per clinical judgment.
• Outcome Measures:
  • Primary: Change in Montgomery-Åsberg Depression Rating Scale (MADRS) scores at 6 and 12 months.
  • Response Definition: Remission (MADRS $\le$ 9) or $\ge$50% symptom reduction.
• Data Collection:
  • ECG: 12-lead ECGs recorded pre-implantation. QTc calculated using Bazett's formula. Medication risk scores for QTc prolongation were calculated.
  • Blood Markers: Complete blood counts (ANC, lymphocytes, monocytes, etc.), inflammatory ratios (NLR, PLR, SII), and metabolic panels (Glucose, TG, HDL, LDL, Cholesterol).
  • Anthropometrics: BMI, age, sex.
• Statistical Analysis:
  • QTc Analysis: Pearson correlation and multiple linear regression adjusting for age, sex, BMI, baseline MADRS, and medication risk.
  • Exploratory Analysis: Pearson correlations followed by stepwise forward multiple linear regression to identify independent predictors of MADRS reduction.

3. Key Contributions

• Replication Attempt: This is the first study to attempt replication of the QTc-VNS response correlation in a larger, multi-center sample with rigorous covariate adjustment.
• Novel Biomarker Exploration: It introduces a broad exploratory analysis of hematological (immune) and metabolic markers as potential predictors for VNS efficacy in DTD.
• Methodological Rigor: Unlike the previous study, this analysis explicitly controlled for sex, age, BMI, and psychotropic medication effects on QTc intervals.

4. Key Results

• QTc Interval Findings:
  • Unadjusted: A significant negative correlation was found between baseline QTc and 6-month MADRS reduction ($r = -0.34, p = .026$), but not at 12 months.
  • Adjusted: After controlling for sex, age, BMI, and medication risk, QTc length did not predict MADRS reduction at either 6 or 12 months ($p > .05$). This contradicts the previous study's findings.
• Exploratory Blood-Based Findings (N=22):
  • 6-Month Response: A stepwise regression model identified four significant predictors of MADRS reduction:
    1. Lower Absolute Neutrophil Count (ANC) ($\beta = -0.005, p = .020$)
    2. Lower BMI ($\beta = -1.140, p = .007$)
    3. Younger Age ($\beta = -0.499, p = .039$)
    4. Higher LDL Levels ($\beta = 0.169, p = .031$)
  • 12-Month Response: Only Lower ANC remained a significant independent predictor ($p = .020$).
  • Other Markers: While univariate analysis showed associations with NLR and SII at 12 months, these were not retained in the final multivariate models, suggesting their effect was captured by ANC.
• Clinical Response Rates:
  • 6-month response rate: 32.7% (16/45).
  • 12-month response rate: 37.0% (17/42).

5. Significance and Discussion

• QTc Interval: The study suggests that the previously reported link between QTc and VNS response may be confounded by sex and medication. The authors argue that without proper adjustment, QTc is not a reliable standalone predictor.
• Immune and Metabolic Markers:
  • ANC: Lower baseline neutrophil counts were associated with better outcomes. This aligns with literature linking elevated inflammation (high ANC) to depression severity and treatment resistance.
  • LDL & BMI: Counter-intuitively, higher LDL and lower BMI predicted better response. The authors note that while high BMI is generally a risk factor for depression, VNS has known metabolic effects (including weight management applications), and lipid profiles may reflect specific metabolic subtypes responsive to vagal modulation.
  • Age: Younger age predicted better 6-month response, contrasting with some recent RESTORE-LIFE data suggesting older age might be favorable, highlighting the complexity of age as a variable.
• Clinical Implications: The findings suggest that a "precision psychiatry" approach for VNS might benefit from screening immune (ANC) and metabolic (LDL, BMI) profiles rather than relying solely on cardiac metrics like QTc.
• Limitations:
  • Sample Size: The blood analysis was limited to 22 patients, increasing the risk of Type II errors and limiting the generalizability of the exploratory findings.
  • Retrospective Nature: Lack of a control group and lack of longitudinal ECG data post-implantation.
  • Confounding Variables: Physical activity and smoking status were not recorded, which could influence immune cell counts.

Conclusion: The study failed to replicate the QTc interval as a robust predictor of VNS response when covariates are controlled. However, it provides preliminary evidence that baseline immune (ANC) and metabolic (LDL, BMI) markers may offer valuable insights for patient selection, warranting verification in prospective, controlled trials.