Intrauterine levonorgestrel (LNG-IUD) is used to treat patients with endometrial adenocarcinoma (EAC) and endometrial hyperplasia with atypia (EHA) but limited evidence is available on its effectiveness. The study determined the extent to which LNG-IUD with or without metformin (M) or weight loss (WL) achieves a pathological complete response (pCR) in patients with EAC or EHA. This phase II randomized controlled clinical trial enrolled patients with histologically confirmed, clinically stage 1 FIGO grade 1 EAC or EHA; a body mass index (BMI) >30 kg/m2; a depth of myometrial invasion of less than 50% on MRI; a serum CA125 ≤ 30 U/mL. All patients received LNG-IUD and were randomized to observation (OBS), M (500 mg orally twice daily), or WL (pooled analysis). The primary outcome measure was the proportion of patients developing a pCR (defined as absence of any evidence of EAC or EHA) after 6 months. From December 2012 to October 2019, 165 patients were enrolled and 154 completed the 6-months follow up. Women were on average 53 years of age, with BMI 48kg/m2. A total of 96 patients were diagnosed with EAC (58%) and 69 patients with EHA (42%). A total of 35 participants were randomized to OBS, 36 to WL and 47 to M (10 patients were withdrawn). After 6 months, the rate of pCR was 61% (95% CI 42% to 77%) for OBS, 67% (95% CI 48% to 82%) for WL and 57% (95% CI 41% to 72%) for M. Across the 3 treatment groups, the pCR was 82% and 43% for EHA and EAC, respectively. Complete response rates at 6 months were encouraging for patients with EAC and EHA across the three groups.