Robotic assistance may facilitate completion of minimally invasive hysterectomy, which is the standard of care for the treatment of early-stage endometrial cancer, in patients for whom conventional laparoscopy is challenging. The aim of this systematicreview was to assess conversion to laparotomy and perioperative complications after laparoscopic (LH) and robotic hysterectomy (RH) in endometrial cancer patients with obesity (body mass index, BMI>30kg/m2).
& Eligibility Criteria: We systematically searched MEDLINE, EMBASE, and Evidence-Based Medicine Reviews (January 1, 2000 to July 18, 2018) for studies of patients with endometrial cancer and obesity (BMI>30kg/m2) undergoing primary hysterectomy.
STUDY APPRAISAL AND SYNTHESIS METHODS:
We determined the pooled proportions of conversion, organ/vessel injury, venous thromboembolism (VTE), and blood transfusion. We assessed risk of bias with the Institute of Health Economics Quality Appraisal Checklist for single-arm studies, and Newcastle-Ottawa Quality Scale for double-arm studies.
We identified 51 observational studies reporting on 10,800 endometrial cancer patients with obesity (study-level BMI: 31.0-56.3). The pooled proportions of conversion from LH and RH were 6.5% (95% CI 4.3-9.9) and 5.5% (3.3-9.1) respectively among patients with BMI>30, and 7.0% (3.2-14.5) and 3.8% (1.4-9.9) among patients with BMI>40. Inadequate exposure due to adhesions/visceral adiposity was the most common reason for conversion for both LH (32%) and RH (61%); however, intolerance of Trendelenburg caused 31% of LH conversions and 6% of RH conversions. The pooled proportions of organ/vessel injury (LH 3.5% [2.2-5.5]; RH 1.2% [0.4-3.4]), VTE (LH 0.5% [0.2-1.2]; RH 0.5% [0.1-2.0]), and blood transfusion (LH 2.8% [1.5-5.1]; RH 2.1% [1.6-3.8]) were low and not appreciably different between arms.
RH and LH have similar rates perioperative complications in endometrial cancer patients with obesity, but RH may reduce conversions due to positional intolerance in patients with morbid obesity. Existing literature is limited by selection and confounding bias, and randomized trials are needed to inform practice standards in this population.