Levator scapulae muscle variability: pathology or anatomical variant? A narrative review with pragmatic...

The levator scapulae muscle (LSM) is traditionally described as an anatomically stereotyped structure; however, anatomical and imaging studies have reported notable variability, with potential relevance in the neurovascularly complex cervicoscapular region. This narrative review aimed to (1) map reported LSM variants and (2) introduce a pragmatic, author-proposed classification capturing cranial and caudal attachment patterns, architectural configurations (including accessory bands), and neurovascular relationships, with consideration of imaging and clinical context. A structured literature search of PubMed and Google Scholar (1867-January 2025) was performed to identify human cadaveric, MRI/CT/ultrasound, and clinical or case-based reports providing primary anatomical data on LSM variability. Records were screened for relevance, and the extracted data were synthesized qualitatively. The synthesis was supplemented with anonymized original CT and MRI material for illustrative purposes. Reported variants clustered into: (I) cranial extensions or alternative origins (e.g., mastoid/occipital region), (II) caudal or additional insertions or slips (most commonly involving serratus anterior, serratus posterior superior, and ribs I-II), (III) multi-slip or multi-belly architectures, and (IV) less frequently documented neurovascular variants, predominantly associated with accessory structures. The reported prevalence of LSM variability differed between cadaveric and imaging-based studies, with additional slips identified in 13.1% of cases in a large anatomical series (n = 300), and variants reported in 43% (16/37) of individuals in an MRI cohort. The proposed classification provides a conceptual framework for organizing heterogeneous anatomical reports and may facilitate structured description and interpretation of LSM variants in anatomical and imaging contexts. However, given the heterogeneity and predominantly case-based nature of the available evidence, these findings should be interpreted with caution and considered hypothesis-generating.