Should the webspaces be a review area for Mortons neuroma on computed tomographies performed for metatarsalgia?

Christine Azzopardi, E. Smith, James Kho, Steven James, Rajesh BotchuPolish 2021 · DOI

Abstract

Sensitivity and specificity of computed tomography (CT) for Morton’s neuroma-bursal complex (MNBC) are poor. CT should not be used as an isolated modality to make a definite diagnosis regarding the presence or absence of a MNBC.

Introduction

Metatarsalgia, or pain in the ball of the foot, is a frequent clinical complaint that often necessitates diagnostic imaging. One of the most common underlying causes is Morton’s neuroma-bursal complex (MNBC), a condition characterized by nerve fibrosis and tissue inflammation in the interdigital webspaces.

While ultrasound is currently considered the "gold standard" due to its high resolution and dynamic testing capabilities, patients are frequently referred for CT scans to rule out other pathologies like stress fractures or arthritis. This study investigates whether radiologists can reliably identify MNBC on CT scans originally ordered for other indications, potentially streamlining the diagnostic process.

Methods

Researchers conducted a retrospective study at a tertiary orthopedic hospital, reviewing 43 examinations from 41 patients who had undergone both a CT scan and an ultrasound for metatarsalgia between 2007 and 2019.

To evaluate the reliability of CT, two fellowship-trained musculoskeletal radiologists performed a blinded review of the scans, looking for soft tissue density in the webspaces that obliterated normal fat. Ultrasound reports were utilized as the reference standard (ground truth). Statistical analysis included the calculation of sensitivity, specificity, accuracy, and inter-observer agreement (Cohen’s kappa).

Results

The quantitative performance of CT for detecting MNBC was found to be poor across all metrics. The mean sensitivity was 45.5%, meaning the modality missed more than half of the actual neuromas. The mean specificity was 62.5%, and the overall accuracy was only 52.3%—barely better than a coin flip.

Figure 2: Axial CT scan of the foot. The arrow indicates a soft tissue density in the webspace, which is often difficult to distinguish from normal anatomy on CT.

Furthermore, inter-observer agreement was only 62.8%, categorized as "fair" to poor. This indicates significant disagreement even among experts on what constitutes a neuroma on a CT scan.

Discussion

The primary hurdle for CT in diagnosing MNBC is its poor soft tissue contrast discrimination. Unlike ultrasound or MRI, CT struggles to differentiate a small fibrotic mass from the surrounding connective tissue. Additionally, CT is a static "snapshot," losing the diagnostic advantage of dynamic imaging. In ultrasound, a sonographer can perform the Mulder maneuver (squeezing the foot) to see if a mass displaces, a critical dimension missing in CT.

Figure 3: Ultrasound image of the same region. The arrow points to a well-defined hypoechoic mass, the hallmark of Morton’s neuroma, which is more clearly visualized than on CT.

The study also highlighted a high risk of false positives. Even when radiologists identified a potential neuroma on CT, it was frequently unconfirmed by ultrasound. This suggests that the webspaces on a CT scan are a "noisy" area that can easily mislead clinicians.

Conclusion

CT is remarkably unreliable for detecting Morton’s neuroma-bursal complex. It should never be used as a standalone tool to diagnose or rule out this condition. If MNBC is clinically suspected, clinicians should proceed directly to ultrasound or MRI, regardless of CT findings. Future research may explore whether newer technologies, such as Dual-Energy CT (DECT), offer improved soft-tissue characterization for this specific application.

Summary updated

Abstract

Sensitivity and specificity of CT for MNBC are poor. CT should not be used as an isolated modality to make a definite diagnosis regarding the presence or absence of a MNBC.
BibTeX
@inproceedings{azzopardi2021should,
  author = {Christine Azzopardi and E. Smith and James Kho and Steven James and Rajesh Botchu},
  title = {{Should the webspaces be a review area for Mortons neuroma on computed tomographies performed for metatarsalgia?}},
  booktitle = {Polish},
  year = {2021},
  doi = {10.5114/pjr.2021.106692},
}
Figure