An alternative route for the administration of progestins is via the intrauterine one. Intrauterine progestin delivery, in particular levonorgestrel (LNG), is much more effective in suppressing the endometrium than oral treatment. This is due to the uniform suppression of the endometrium throughout the whole thickness of the mucosa caused by the high tissue concentrations when the hormone is applied locally [1]. This form of treatment should, therefore, be preferred over oral treatment for precancerous lesions of the endometrium and early stage invasive endometrial cancer. Surgery is nowadays still the method of choice for the management of these lesions. However, many patients are not eligible (unfit) for surgery and younger women want to preserve their fertility. For these reasons, local delivery of a potent progestin should be the alternative method of choice in these women if it could be shown to be effective. An intrauterine system also provides optimal patient compliance.

In May 2008, a 67-year-old woman residing in the United States of America presented with a history of postmenopausal bleeding since January 2008. She had been using local estrogen treatment for vaginal dryness without any systemic hormonal therapy. She had been diagnosed with an endometrial carcinoma and was advised to undergo surgery. As she wanted to avoid surgery, she sought an alternative treatment and consulted DW. The clinical examination was normal, with normal body weight, and a transvaginal ultrasound scan showed normal appearance of the myometrium and endometrium. An outpatient endometrial pipelle biopsy was performed which revealed a moderately differentiated adenocarcinoma (Figure 1) with minimal myometrial invasion (Clinical FIGO stage I). An LNG-IUS (Femilis, Contrel Research, Ghent, Belgium) releasing 20μg of LNG/d was inserted (Figure 2), and advice was given to request a repeat biopsy within the next 3 to 6 months. Spotting continued for several weeks and then stopped. As the patient was completely free of symptoms 6 months after insertion of the LNG-IUS, a pelvic transvaginal ultrasound was performed, including a 3D-ultrasound and a Duplex Doppler with Color Flow Mapping. The uterus appeared completely normal, and there was no evidence of any pathology. The endometrium showed normal thickness, and there was no evidence of any endometrial abnormality or myometrial invasion (Figures ​(Figures33 and ​and4).4). The LNG-IUS was identified in situ, as expected. Three months later, in order to ascertain complete remission, the LNG-IUS was removed, and a full D&C was performed. The uterine sound length was 6cm. The whole cavity was explored, and very scant tissue was removed. Histological examination of the specimen revealed a secretory endometrium without signs of hyperplasia or atypia (Figure 5). A mini Femilis LNG-IUS (crossarm 24mm long and 28mm long and 2.0 wide drug compartment) was inserted as a precaution.

We previously reported on the successful management of nonatypical and atypical endometrial hyperplasia with an LNG-IUS [6, 7]. Endometrial cancer is a very common disorder. Over 40,000 cases are diagnosed in the United States annually with almost 75% of cases presenting with FIGO stage I disease (Cancer fact & figures, 2004). The underlying mechanism for the development of some types of endometrial cancer is exogenous or endogenous hormone stimulation [8]. Therefore, progesterone or progestin treatment, by counteracting the effect of estrogen, may reverse the neoplastic process. Levonorgestrel, delivered locally with a drug delivery system, is probably the best choice as the tissue concentrations are many times higher than when the hormone is administered orally. “Progestasert,” a progesterone-releasing IUS (Alza Corp., Palo Alto, CA), has been used with good results. However, a levonorgestrel-releasing IUS should be preferred as it is much more potent [9]. Histological regression of early endometrial cancer has also been obtained with orally administered progestins but the results cannot be expected to be as good as local treatment [10, 11]. In addition, with an IUS there is less risk of systemic side effects, often leading to poor compliance. Intrauterine delivery of levonorgestrel causes a more uniform response throughout the whole thickness of the endometrium, including the basal layer, compared to intrauterine progesterone delivery or orally administered progestins [12]. The Femilis, currently still an experimental IUS, releases the same amount of LNG as Mirena (BayerSchering Inc., Berlin, Germany) but the initial release is approximately three times higher than the release of LNG of Mirena during the first weeks. This could be advantageous as it increases the impact on the neoplastic endometrial cells. It is possible that the fast remission in this patient is due to the strong suppression of the endometrium immediately after insertion of the LNG-IUS. Several cases of endometrial carcinoma treated with an LNG-IUS (Mirena) have been published [11, 13, 14]. The LNG-IUS should be high up in the fundus to guarantee proper release of the hormone in the fundal area of the uterus. We question if the LNG-IUS was properly located in the uterine cavity in some of these unsuccessful cases. A frameless LNG-IUS should be preferred in case of displacement, partial or total expulsion of a framed LNG-IUS. This LNG-IUS is anchored to the uterine fundus and is unlikely to become displaced or expelled [15]. We would also recommend to remain cautious in women presenting with irregular bleeding after 6 to 12 months following the insertion of an LNG-IUS and to evaluate the endometrium thoroughly.