Classification of pectus excavatum according to morphologic parameters measured in chest computed tomography

Abstract

Introduction: The present classification systems of pectus excavatum are based on subjective morphology. We herein describe a classification system of pectus excavatum based on objective data and validate the clinical relevance of the new classification.

Methods: Patients who underwent surgical repair of pectus excavatum without a history of chest surgery were included in this study. Classification was performed by hierarchical clustering of morphologic parameters obtained from chest computed tomography, including the pectus index, asymmetry index, flatness index, sternal torsion angle, and angle of Louis. Correlations among parameters were analyzed, and the clinical relevance of the suggested classification system was verified.

Results: In total, 230 patients who underwent operations from January 2001 to August 2013 were included in the study. The patients were classified into two major groups: the typical group (Group I, n = 197) and atypical group (Group II, n = 33). Group I was divided into three minor groups: the symmetric group (Ia, n = 82), asymmetric proper group (Ib, n = 108), and asymmetric with flat sternum group (Ic, n = 7). Group II was classified into four minor groups: the asymmetric with doubly distorted sternum group (IIa, n = 8), asymmetric with severe sternal torsion group (IIb, n = 8), reverse asymmetric group (IIc, n = 16), and extremely depressed group (IId, n = 1). The asymmetry of the pectus excavatum was associated with the sternal torsion angle (p < 0.0001, R = 0.625) and inversely associated with the angle of Louis (p = 0.013, R = –0.163). Scoliosis was more common in Group IIa than Ia (p = 0.0008, odds ratio [OR] =10.5, 95% confidence interval [CI] = 1.7–63.6), and multiple bar insertion was necessary more frequently in Groups Ic (p = 0.001, OR = 34.9, 95% CI = 3.9–310.1) and IIb (p = 0.007, OR = 22.7, 95% CI = 2.3–223.6) compared with Group Ia.

Conclusions: This new classification produced two major groups with typical and atypical morphology and seven minor subgroups. The classification was clinically relevant in that specific subgroups were associated with scoliosis and with the need for multiple bar insertion.