Magnetic resonance imaging as a predictor of submucous cleft palate severity and guide for surgical intervention.

BACKGROUND

Diagnosis of submucous cleft palate (SMCP) is frequently delayed, adversely affecting speech outcomes. Previous studies show that MRI reliably identifies structural abnormalities in velopharyngeal musculature. This information has potential to assist with diagnosis and treatment decisions.

AIMS: The objectives of this study were to (1) develop a clinician-friendly MRI grading scale of SMCP anatomy, (2) identify correlations between radiographic cleft severity and clinical severity using Pittsburgh Weighted Speech Scores (PWSS), and (3) determine if MRI is a predictor of surgical efficacy in improving PWSS.

DESIGN

Thirty patients presenting to our Cleft Palate-Craniofacial Clinic for evaluation of velopharyngeal insufficiency (VPI) and suspected SMCP were reviewed. VPI severity was clinically graded using PWSS. All patients underwent MRI to grade palatal abnormalities, using a novel MRI grading scale. PWSS and cleft severity on MRI were compared. A subgroup of patients (n = 19) underwent palatoplasty. Preoperative and postoperative PWSS were compared. Degree of PWSS improvement was then correlated with the preoperative MRI grade.

RESULTS

Twenty-nine out of 30 MRIs demonstrated abnormal palate anatomy. Of the 30 patients evaluated, 5 clinically improved with speech therapy alone. In this subgroup, MRI severity did not correlate with PWSS (P = 0.06-0.6). Nineteen patients underwent palatoplasty. Of these, 14 demonstrated improved postoperative PWSS. There were no significant correlations between severity of cleft on imaging and preoperative PWSS or score improvement (P = 0.056-0.65).

CONCLUSION

While MRI accurately identifies structural abnormalities of the soft palate, these abnormalities do not reliably correspond to clinical severity. Clinical examination including speech scores and dynamic speech testing, rather than static MRI, should guide treatment decisions and surgical indications.