Risk factors for coexistent leiomyoma and adenomyosis compared to coexistent leiomyoma and endometriosis in patients with symptomatic leiomyoma requiring myomectomy

Abstract

To investigate the risk factors for coexistent leiomyoma with endometriosis and leiomyoma with adenomyosis in a population of women who elect for myomectomy due to symptomatic leiomyoma. Retrospective study. This was a single-center case control study to compare patients with leiomyoma only, to those that had either coexistent leiomyoma and endometriosis or leiomyoma and adenomyosis. Multinomial logistic regression models were used to compute relative-risk ratios (RRR) for the association between patients with endometriosis or adenomyosis in the presence of leiomyoma with patients with leiomyoma only. The analysis was first performed for each independent variable alone and subsequently, a best-subset selection method was used to construct a parsimonious model by selecting the model with lowest Akaike information criteria (AIC). 961 patients underwent surgery for symptomatic leiomyoma only, and either coexistent leiomyoma and adenomyosis or endometriosis (leiomyoma only (L) – 799, leiomyoma and endometriosis (LE) – 98 and leiomyoma and adenomyosis (LA) – 64 patients respectively) between 2013 and 2018 at a free-standing ambulatory surgical setting. The LE group was significantly younger (35.2 +5 years) (RR 0.95 95% C.I. 0.911-0.986) compared to the LA group (39.4 +5 .7 years) (RR 1.08 95% C. I. 1.035 -1.129) ( p<0.001). LE group (RR 0.23 95% C.I 0.08 – 0.661) (p<0.006) had higher nulliparity rate compared to LA group (RR 1.6 95% C.I. 1.193-2.148) ( p<0.002). The uterine length was significantly shorter in the LA group compared to the LE group. (L – 12.3 +3.7 cm, LE 11.7 +3.7, LA 10.4 +3.1 cm (p <0.001)). Age is an important predictor of developing adenomyosis but not endometriosis. The risk of developing adenomyosis in patients with leiomyoma increases by almost 9% for every year of age compared to the leiomyoma only group. Both nulliparity and uterine length are important predictors of both endometriosis and leiomyoma in patients compared to the leiomyoma control group.Table 1Relative-risk ratios (RRR) from most parsimonious* multinomial logistic regression modelRRR95% C.I.p-valueFibroids_only(base outcome)Fibroids + Endo Age0.9600.912, 1.0100.107 Parity0.3080.109, 0.8750.027 Uterine length0.9310.864, 1.0020.058 Constant1.3680.204, 9.1630.747Fibroids + Adeno Age1.0871.032, 1.1460.002 Parity1.3890.993, 1.9430.055 Uterine length0.8250.746, 0.911<0.001 Constant0.0270.003, 0.2560.002Note: Constant estimates baseline relative risk for each outcome.* Based on model with lowest Akaike information criteria (AIC) Open table in a new tab