Correlation of Elastography and Ultrasound in Patients Undergoing Vaginal Birth

Objectives: During vaginal birth (VB), the pelvic floor (PF) can sustain many types of injuries. How the pre-delivery elasticity and how these injuries change the intrinsic elasticity of the PF is poorly understood. Shear wave elastography (SWE) offers insights into tissue elasticity, 3D endovaginal ultrasound (EVUS) shows the structural integrity of the PF. We hypothesized that women who have PF injury during VB had different tissue characteristics from uninjured women that predisposed them. The secondary aim was to see if the tissue characteristics between the two groups recovered similarly after VB.

Methods: A preliminary observational longitudinal cohort study of 36-42-week pregnant primigravid women underwent three visits: pre-delivery (v1), immediate postpartum (v2), and 6-months post-partum (v3). Patients who completed three visits were included. Using SWE, we measured elasticity at four different planes: 1) sagittal view of the urethra: rhabdomyosphincter, sub-urethra, and trigone; 2) perineum in axial view: external anal sphincter (EAS), and perineum (Fig.1), 3) levator plate (LP) in mid-sagittal view: perineum and the LP, 4) levator ani muscles were measured bilaterally in mid vagina and close to the rectum. Using 3D EVUS, a levator ani defect (LAD) score between 0 and 3 (Fig.1) was determined for the left and right puborectalis and pubococcygeus. A correlation analysis with R2, adjusted R2 (to account for outliers), and P-value was done against non-injured and injured patients per visit based on combined LAD score, left muscular region LAD, and right muscular region LAD scores – all based on v2 scores.

Results: 422 patients were approached, 46 enrolled over a two-year study. 20/46 completed so far. 8/20 patients had sustained LAM injury indicated by at least one of the 4 postpartum LAD scores being 2 or above. No difference in SWE measurement was found between the two groups at the corresponding visit. Two of the three regions of the bladder (rhabdomyosphincter and sub-urethra) and perineum (EAS and close to rectum) showed statistically significant differences when comparing between v1 and v3 in non-injured patients, the perineum in both views between v2 and v3, and between v1 and v3 in non-injured patients (Table 1). No correlations were found between LAD score (total, left or right regions). Figure 2 shows a SWE correlation of the EAS region (for non-injured- v2-v3: 0.0062, v1-v3: 0.0225; for injured- v1-v3: 0.0396 ) at postpartum of the combined (panel A: p =0.703; R2 = 0.009; R2 adjusted = -0.053), left (panel B: p=0.399; R2 = 0.045; R2 adjusted = -0.015) and right (panel C: p = 0.998, R2 < 0.001; R2 adjusted = -0.059), where no correlation is observed.

Conclusions: The PF undergoes architectural and elasticity changes. The tissue elasticity seems to be independent of changes that occur in the PF due to VB related trauma.