Cancer Risk Profile of the Levonorgestrel-Releasing Intrauterine Device: A Narrative Review

This article presents a structured narrative review to synthesise and critically appraise the available evidence on the association between LNG-IUD and cancer risk. The review was developed by a multidisciplinary group of experts, including gynaecologic oncologists affiliated with the European Society of Gynaecological Oncology (ESGO) Prevention Committee and specialists in gynaecological endoscopy from the European Society for Gynaecological Endoscopy (ESGE), as well as clinical researchers experienced in evidence synthesis. The manuscript was structured in accordance with recommendations for narrative reviews and follows principles outlined in the Scale for the Assessment of Narrative Review Articles (SANRA) framework to enhance methodological transparency and reporting quality [ 10 ]. A literature search was conducted in December 2024 using the major biomedical databases PubMed/MEDLINE, Embase, and Scopus, without date restrictions. The search strategy combined terminology related to LNG-IUDs with cancer-related outcomes to maximise sensitivity and minimise the risk of missing relevant studies. The core search strategy included combinations of the following terms: (“levonorgestrel-releasing intrauterine system” OR “levonorgestrel intrauterine device” OR “LNG-IUS” OR “LNG-IUD” OR “hormonal intrauterine device” OR “Mirena”) AND (“cancer” OR “neoplasm” OR “malignancy” OR “carcinoma” OR “tumour” OR “tumour”). Where appropriate, additional site-specific cancer terms (e.g., “breast cancer,” “endometrial cancer,” “ovarian cancer,” “cervical cancer,” “colorectal cancer,” “lung cancer,” “pancreatic cancer,” “thyroid cancer,” and “melanoma”) were incorporated into the search strategy to identify studies evaluating the association between LNG-IUD use and gynaecologic and non-gynaecologic malignancies. During the study selection process, priority was given to original research with robust study designs, including large population-based cohort studies and case–control studies, as well as relevant systematic reviews and meta-analyses addressing cancer outcomes associated with LNG-IUD use. Case reports and studies with insufficient outcome data or inadequate follow-up were excluded. To minimise the risk of missing relevant evidence, the reference lists of included articles were manually screened, and citations of key publications were tracked to identify additional relevant studies not captured by the initial database search. Generative artificial intelligence (GenAI) was used to assist in generating and refining the graphical layout of Figure 1 . GenAI was not used for data collection, data analysis, evidence interpretation, or clinical recommendations. All scientific content, figure labels, and conclusions were critically reviewed and approved by the authors.

Levonorgestrel is a synthetic second-generation progestin with sixfold higher binding affinity for progesterone receptors than natural progesterone [ 11 ]. After inserting an LNG-IUD containing 52 mg of levonorgestrel, the initial levonorgestrel release into the uterine cavity is 20 μg per day. A stable plasma level of 150–200 pg/mL is typically established within the first few weeks. These plasma concentrations are lower than those observed with levonorgestrel implants and oral contraceptives [ 12 ]. Individual variations in plasma levonorgestrel levels and systemic effects may be due to differences in levonorgestrel binding to sex hormone-binding globulin (SHBG) and metabolic clearance rates [ 13 ]. Levonorgestrel primarily exerts its progestogenic effects by binding to progesterone receptors and also interacts with other steroid receptors. It demonstrates a weak agonist effect at glucocorticoid receptors and may exhibit mild androgenic effects due to its moderate affinity for androgen receptors. Additionally, its ability to antagonise mineralocorticoid receptors might help decrease sodium and water retention [ 14 ]. The pharmacokinetic profile of the LNG-IUD—characterised by very high local intrauterine levonorgestrel concentrations, reported to be approximately 200–800 times higher than those achieved with daily oral administration, together with low systemic plasma concentrations of approximately 150–200 pg/mL—provides a biologically plausible mechanistic basis for its organ-specific effects on carcinogenesis [ 11 , 12 ]. Within the uterus, sustained local progestogenic exposure suppresses oestrogen-driven endometrial proliferation, induces glandular atrophy and stromal decidualisation, and promotes endometrial thinning [ 11 ]. These changes provide a coherent explanation for the consistently observed reduction in endometrial hyperplasia and cancer risk. In contrast, the low systemic levonorgestrel concentrations result in limited progestogenic exposure in extrauterine hormone-sensitive tissues [ 12 , 13 ], which may partly explain the generally neutral or only modest associations reported for breast and other extrauterine cancers. This marked compartmentalisation of levonorgestrel exposure—high within the endometrium but low in the systemic circulation—provides a useful biological framework for interpreting the differing associations observed across cancer sites.

A direct carcinogenic or mutagenic mechanism specific to the LNG-IUD has not been demonstrated. Any potential influence on cancer risk is more likely to reflect tissue-specific hormonal and local biological effects than a generalised carcinogenic action. These effects may differ across organs because the LNG-IUD produces very high local levonorgestrel exposure within the uterus while maintaining substantially lower, yet still measurable, systemic concentrations [ 11 ]. The principal biological concern relates to breast tissue because intrauterine administration does not completely eliminate systemic levonorgestrel exposure. Levonorgestrel is detectable in both serum and breast tissue, although breast tissue concentrations are substantially lower than those observed with oral levonorgestrel-containing contraceptives [ 15 ]. Experimental models suggest that androgenic progestins, including levonorgestrel, may stimulate the proliferation of hormone-responsive breast epithelial cells through androgen receptor–related signalling [ 16 ]. Therefore, if the modest epidemiological association with breast cancer is causal, hormonal stimulation of pre-existing susceptible epithelial cell populations may be more biologically plausible than direct mutagenic initiation. However, the clinical relevance of these experimental findings at the low systemic exposure levels achieved with the LNG-IUD remains uncertain. Moreover, limited human breast tissue data have not consistently demonstrated increased cellular proliferation, and causality has not been established [ 17 ]. The mechanisms potentially underlying the reported ovarian cancer association remain uncertain. Reduced menstrual bleeding and amenorrhoea may decrease retrograde menstruation, while local foreign-body-related inflammatory responses and anti-oestrogenic effects within the reproductive tract have also been proposed. Partial suppression of ovulation may contribute in some users; however, ovulation frequently continues during LNG-IUD use, making this an incomplete explanation. Overall, these mechanisms do not establish a protective causal effect [ 18 ]. For cervical cancer, device-related local inflammation and altered epithelial immunosurveillance, including possible changes in Langerhans cell activity, have been proposed [ 18 ]. Conversely, hormonal modulation of innate antiviral factors could theoretically influence HPV persistence [ 19 ]. Given these potentially opposing effects and the limited LNG-IUD-specific evidence, no direct mechanism linking LNG-IUD use to either cervical carcinogenesis or cervical cancer prevention has been established. Mechanistic evidence for non-gynaecologic cancers is even more limited, and no consistent causal pathway has been identified.

Progesterone-only hormonal intrauterine devices have become some of the most successful methods for preventing pregnancy due to their high efficacy, minimal side effects, and patient satisfaction. LNG-IUDs are particularly effective as a contraceptive option for women unable to use oestrogen-based methods [ 20 ]. LNG-IUDs exert therapeutic effects on heavy menstrual bleeding by acting on the endometrium. This effect has been demonstrated in multiple studies, which have consistently yielded similar results [ 21 ]. Comparative studies have shown that LNG-IUDs are more effective than various oral progestins in managing heavy menstrual bleeding [ 22 , 23 ]. In women scheduled for hysterectomy due to heavy menstrual bleeding, the use of an LNG-IUD significantly reduces the need for surgery, with rates reported at 64–80%. However, among groups receiving only medical treatment, the rate of hysterectomy avoidance remained 9–14% [ 24 , 25 ]. The LNG-IUD provides clinical advantages for women with hereditary bleeding disorders by markedly decreasing menstrual blood loss. This decline correlates with improved anaemia markers, with an average increase of 1.4 g/dL in haemoglobin and 19.8 ng/mL in ferritin. Furthermore, its use has been associated with improved quality of life [ 26 , 27 , 28 ]. Various studies have shown that the LNG-IUD is effective in managing endometriosis-related pelvic pain and secondary dysmenorrhoea, as well as primary dysmenorrhoea [ 29 ]. It has also been reported to reduce the risk of symptom recurrence when used after endometriosis surgery [ 30 , 31 ]. In patients with adenomyosis, the device provides therapeutic benefits by relieving pain and controlling excessive menstrual bleeding [ 32 , 33 ]. An additional benefit of LNG-IUD is that it can be used in obese patients without contraindications. The local effects of progestins within the uterus provide effective contraception and protect the endometrium in obese women, all while reducing systemic side effects compared with systemic hormonal contraceptive methods [ 34 ].

Evidence suggesting a modest increase in breast cancer risk has mainly emerged from large registry-based studies. In a nationwide Finnish cohort including 93,843 women, LNG-IUD use was associated with a statistically significant increase in breast cancer incidence compared with the general female population, with a standardised incidence ratio (SIR) of 1.19 (95% CI: 1.13–1.25). The association appeared more pronounced among long-term users, particularly those who had used the LNG-IUD for more than five years (SIR 1.40; 95% CI: 1.24–1.57) [ 6 ]. A subsequent histology-specific analysis from the same cohort supported this finding and reported a stronger association for lobular carcinoma (SIR 1.73; 95% CI: 1.37–2.15) [ 4 ]. Similarly, the nationwide Danish registry study by Mørch et al. [ 7 ], which included approximately 1.8 million women aged 15–49 years and more than 20 million person-years of follow-up, reported a modestly increased risk of invasive breast cancer among current or recent users of contemporary hormonal contraception. For LNG-IUD users, the relative risk was 1.21 (95% CI: 1.11–1.33), comparable to that observed with other hormonal contraceptive methods. However, as with other registry-based observational studies, residual confounding, differences in screening behaviour, and lack of detailed information on individual risk factors should be considered when interpreting this association [ 7 ]. In a Finnish population-based analysis by Heikkinen et al. [ 35 ], including 13,265 breast cancer cases, menopausal status appeared to modify the association between LNG-IUD use and breast cancer risk. An increased risk was observed among postmenopausal LNG-IUD users (OR 1.48; 95% CI: 1.10–1.99), whereas no increased risk was found among premenopausal users (OR 0.77; 95% CI: 0.52–1.13) [ 35 ]. Importantly, LNG-IUD use in postmenopausal women often differs clinically from contraceptive use in premenopausal women. In this setting, menopausal hormone therapy typically consists of systemic oestrogen therapy; because unopposed oestrogen increases the risk of endometrial hyperplasia and endometrial cancer in women with an intact uterus, a progestogen is required for endometrial protection. The LNG-IUD may provide this progestogenic component by delivering levonorgestrel locally, thereby avoiding the need for systemic progestin therapy. Consequently, the increased breast cancer risk observed among postmenopausal LNG-IUD users may partly reflect concomitant systemic oestrogen exposure within the context of menopausal hormone therapy rather than an independent effect of the LNG-IUD alone [ 36 ]. In contrast, several studies have reported neutral associations. In the Norwegian Women and Cancer (NOWAC) cohort, which included 104,318 women, LNG-IUD use was not associated with an increased risk of breast cancer after multivariable adjustment for age, BMI, parity, physical activity, and oral contraceptive use (RR 1.03; 95% CI: 0.91–1.17) [ 9 ]. Similarly, an Israeli cohort study including 13,354 LNG-IUD users aged 40–50 years and 27,324 age-matched controls found no overall increase in breast cancer risk, although a small absolute difference was observed among women aged 40–45 years (0.88% vs. 0.69%; p = 0.014) [ 37 ]. A Finnish post-marketing study of 17,360 LNG-IUD users aged 30–54 years also found no significant difference in breast cancer incidence compared with the general female population across five-year age groups [ 38 ]. Furthermore, a retrospective case–control study conducted in Finland and Germany, including 5113 breast cancer cases and 20,452 matched controls, found no significant association between LNG-IUD use and breast cancer risk. Compared with cu-IUD use, ever-use of LNG-IUD was not associated with increased breast cancer risk (adjusted OR 0.99; 95% CI: 0.88–1.12), and no significant increase was observed among current LNG-IUD users either (adjusted OR 0.85; 95% CI: 0.52–1.39) [ 39 ]. More recently, the 2024 Swedish nationwide cohort study by Yi et al. [ 40 ] reported a modest increase in breast cancer risk among LNG-IUD users (adjusted HR 1.13; 95% CI: 1.10–1.17). The association appeared stronger among women with a family history of breast cancer, in whom concomitant LNG-IUD exposure was associated with a higher risk estimate (HR 2.07; 95% CI: 1.71–2.51), compared with women without a family history (HR 1.09; 95% CI: 1.04–1.15). Age-stratified analyses also suggested slightly higher estimates among postmenopausal women than among premenopausal women. These findings support the need for individualised counselling in women with additional breast cancer risk factors, although residual confounding and limited duration-specific exposure data remain important considerations [ 40 ]. Overall, evidence on LNG-IUD use and breast cancer risk remains heterogeneous. While some large registry-based studies suggest a modest increase in risk, other cohort and case–control studies have shown no significant association. The studies that contributed to this evidence base are summarised in Table 1 . These inconsistencies likely reflect differences in menopausal status, baseline risk profile, indication for use, duration of exposure, comparator groups, and adjustment for key confounders. Accordingly, LNG-IUD use should not be considered contraindicated in the general population solely based on current breast cancer data; however, individualised counselling is warranted for women with elevated baseline breast cancer risk, particularly those with advanced age, postmenopausal status, obesity, or a family history of breast cancer.

Several articles have examined the risk of cervical cancer associated with intrauterine device use [ 41 , 42 , 43 ]. However, in most cases, the specific type of intrauterine device is not reported. The first review identified four case–control studies investigating the risk of cervical cancer among intrauterine device users and found either a similar risk or even a decreased risk compared with those who did not report using an intrauterine device [ 44 ]. A pooled analysis of 26 epidemiological studies reported that intrauterine device use (predominantly copper intrauterine device) substantially reduces the risk of cervical cancer [ 41 ], an outcome consistent with a more recent meta-analysis reporting that, among women who have used an intrauterine device, invasive cervical cancer might be one-third less frequent than among those who have not [ 42 ]. In a case–control study of women in the Kaiser Permanente Northern California Healthcare System, 1657 intrauterine device users with incident CIN2+ were compared with 7925 intrauterine device users as controls, and LNG-IUD use was not associated with CIN3+. In particular, regarding the occurrence of cervical cancer, the authors calculated a relative risk (RR) of 1.47 (95% CI: 0.71–3.06, p = 0.3) [ 43 ]. Another cohort study of 789 intrauterine device users, 354 of whom were LNG-IUD users, compared with intrauterine device non-users and those with an abnormal Pap test, reported no increased risk of cervical cancer among the former group [ 45 ]. More robust data come from four registry-based cohort studies—the first of which compared cancer risk in women using an LNG-IUD with that of the general population in Finland. The study reported a standardised incidence ratio of 0.90 (95% CI: 0.69–1.15) among women aged 30–49 who purchased an LNG-IUD at least once over 13 years, compared with women who did not purchase such a device [ 6 ]. The second study, from Denmark, compared hormonal intrauterine device users, Cu-intrauterine device users, and oral contraceptive users, directly assessing the incidence of CIN3+. The study reported an adjusted RR for CIN3 or cancer of 1.08 (95% CI, 0.94–1.22), indicating no statistically significant difference between users of hormonal intrauterine devices and users of Cu-intrauterine devices. This suggests that hormonal intrauterine device use would also confer no additional cervical cancer risk, given that multiple studies have previously shown a decreased cervical cancer risk with Cu-intrauterine device use [ 46 ]. The third cohort study, from Denmark, investigated data on contemporary hormonal contraception and cervical cancer risk in women aged 15 to 49 years from 1995 to 2014. Among the more than 20 million person-years of the total study population, the study found that LNG-IUD users’ RR for the occurrence of cervical cancer, compared to women who had never used hormonal contraceptives, was 0.76 (0.52–1.11) [ 47 ]. The last and most recent study examining the effect of LNG-IUD use on cervical cancer risk was conducted in Sweden, among 16,181 women aged 30–49 years old, and found a slightly increased adjusted odds ratio of 1.45 (95% CI: 1.02–2.06) for the occurrence of histologically confirmed CIN2+ [ 48 ]. However, this study included only 297 cases of CIN2+ and did not report separately on CIN3 or cervical cancer, so this result cannot be generalised and interpreted as increased cervical cancer risk. The characteristics and principal findings of the available studies are summarised in Table 2 .

The effect of Cu intrauterine devices on ovarian cancer has been investigated for a long time, and most studies have reported neutral or protective results [ 8 ]. In contrast, early data on LNG-IUDs were limited, still raising concerns about potential risks. However, large-scale studies conducted in recent years have begun to clarify the picture. One of the most comprehensive cohort studies addressing this issue is the Norwegian NOWAC study, which examined the risk of ovarian cancer among LNG-IUD users in a cohort of 104,380 pre- and postmenopausal women, including 9146 LNG-IUD users, totalling 1,305,435 person-years. Results showed a significant reduction in ovarian cancer risk among LNG-IUD users, with an incidence rate of 16.7 per 100,000 person-years compared with 38.1 among non-users. The age-adjusted relative risk (RR) was 0.49 (95% CI: 0.30, 0.82), and the multivariable-adjusted RR was 0.53 (95% CI: 0.32, 0.88) after adjusting for confounders, including age, contraceptive use, parity, and menopausal status. Although duration-response analysis was limited by case numbers, most ovarian cancer cases among users occurred within the first seven years of use [ 9 ]. Soini et al. provided additional insights from two studies involving premenopausal women, in which the use of an LNG-IUD was associated with a lower incidence of ovarian cancer, with an SIR of 0.60 in their first study, which encompassed 855,324 person-years [ 6 ]. Their second study, which covered one million person-years, examined the impact of the device on the risk of various histological types of ovarian and primary fallopian tube cancers. The findings were consistent with those of the initial study, with an SIR of 0.59. The protective effect was observed across different histological subtypes, particularly mucinous (SIR 0.49) and serous (SIR 0.75) carcinomas. Conversely, the risk of fallopian tube cancer did not differ significantly between users of the LNG-IUD and the general population [ 49 ]. Iversen et al. [ 50 ] studied the impact of modern hormonal contraceptives, including levonorgestrel-releasing intrauterine devices, on ovarian cancer risk in a Danish cohort of 1.9 million women (21 million person-years), identifying 1249 ovarian cancer cases. The study found no significant risk reduction in ovarian cancer among users of progestogen-only products, including levonorgestrel-releasing intrauterine devices, in contrast to combined hormonal contraceptives, which showed a protective effect [ 50 ]. Balayla et al. [ 51 ] pooled nine observational studies and reported a protective association between intrauterine device use and ovarian cancer risk irrespective of device type (OR 0.67; 95% CI 0.60–0.74; p < 0.001), with a levonorgestrel-specific subgroup estimate derived from two studies (SIR 0.58; 95% CI 0.47–0.71) [ 51 ]. Conversely, D’Alessandro et al. [ 52 ] restricted their meta-analysis to levonorgestrel-releasing intrauterine system studies and found no significant risk reduction (OR 0.66; 95% CI 0.41–1.08) [ 52 ]. The apparent discrepancy between these two meta-analyses is more likely methodological than biological. Balayla et al. [ 51 ] primarily evaluated ever-use of IUDs, with most included studies not distinguishing between copper and levonorgestrel-releasing devices, whereas D’Alessandro et al. [ 52 ] specifically focused on LNG-IUD use. Furthermore, adjustment for important confounders, particularly prior oral contraceptive use, was inconsistent across the included studies, and some registry-based cohorts compared LNG-IUD users with the general population rather than an internal unexposed cohort. In addition, D’Alessandro et al. [ 52 ] included the large Danish cohort by Iversen et al. [ 50 ], which reported no significant protective association for progestogen-only contraceptives, potentially attenuating the pooled estimate. Consequently, although both meta-analyses demonstrated point estimates suggestive of a protective association, the current LNG-IUD-specific evidence remains insufficient to establish a definitive protective effect against ovarian cancer. The available evidence evaluating the association between LNG-IUD use and ovarian cancer risk is summarised in Table 3 . Koskela-Niska et al. [ 53 ] assessed the impact of postmenopausal hormone therapies on the risk of ovarian and fallopian tube cancers. They included 3958 ovarian cancer cases and 11,325 controls. They found that the combination of oestradiol and LNG-IUD had no significant effect on ovarian cancer risk, with an odds ratio of 1.02 (95% CI, 0.63–1.66) [ 53 ]. In another study, they evaluated the risk of fallopian tube carcinoma and included 360 cases of primary fallopian tube carcinoma and 3442 age-matched controls. They found that use of an LNG-IUD for more than five years was associated with a twofold increase in the risk of primary fallopian tube carcinoma (OR 2.84, 95% CI 1.10–7.38), p = 0.032. However, due to the limited number of cases, particularly among long-term users, statistical power was limited [ 54 ].

Hormonal imbalances play an essential role in the etiopathogenesis of endometrial cancer, especially oestrogen-dominant hormonal environments with insufficient progesterone. The local progestogenic effect of the LNG-IUD may protect against this mechanism. The NOWAC study reported a large and statistically significant cancer risk reduction for endometrial cancer with a multivariable-adjusted RR of 0.22 (95% CI: 0.13–0.40), with an especially pronounced risk reduction among women who never used oral contraceptives (RR: 0.08, 95% CI: 0.02–0.34) (p-heterogeneity = 0.18) compared to never-users. Adjustments for age, BMI, physical activity, and reproductive factors were included, and the findings suggest a protective role for the LNG-IUD against endometrial cancer, likely due to its localised progestogenic effects [ 9 ]. Soini et al. [ 6 ] investigated the use of LNG-IUDs in a cohort of 93,843 Finnish women (855,324 person-years). They found a 54% reduction in the incidence of endometrial adenocarcinoma among users (SIR: 0.46, 95% CI: 0.33–0.64). Among women with multiple LNG-IUD purchases, the risk reduction was even greater (SIR: 0.25, 95% CI: 0.05–0.73) [ 6 ]. In a case–control study, Jaakkola et al. [ 55 ] assessed the effects of different oestradiol-progestin therapy (EPT) regimens on postmenopausal endometrial cancer risk in Finnish women. This included 7261 endometrial cancer cases and 19,490 age-matched controls. They reported an OR of 0.39 (95% CI: 0.17–0.88), p = 0.023, for women using an LNG-IUD for less than 5 years, with continued reductions observed with longer durations of use, OR: 0.16 (95% CI: 0.37–0.68), p = 0.013, for durations between 5 and 10 years. The protective effect persisted among users of continuous oestradiol-progestogen therapy. In contrast, prolonged sequential or long-cycle regimens were associated with an increased risk, underscoring the importance of regimen type and duration. Overall, these studies suggest that the LNG-IUD offers endometrial protection, particularly with continuous progestin exposure [ 55 ]. Curtis et al.’s [ 44 ] systematic review, which analysed ten studies, reinforced this evidence, showing a 40% lower risk of endometrial cancer among intrauterine device users (pooled OR: 0.6, 95% CI: 0.4–0.7), regardless of the type of intrauterine device. The anti-carcinogenic effect of LNG-IUD is attributed to its influence on endometrial atrophy, which mitigates hyperplasia, a precursor to endometrial cancer [ 44 ]. In young patients who want to preserve their fertility, an LNG-IUD is used as an option in the treatment of both endometrial hyperplasia and early-stage endometrial cancer in appropriately selected cases [ 56 ]. The highest response rate was observed in women under 45 years of age with a BMI of less than 30 kg/m 2 (84.6%) [ 57 ]. According to the European Society of Gynaecological Oncology (ESGO)–European Society for Medical Oncology (ESMO)–European Society of Gynaecological Endoscopy (ESGE) joint recommendations, a combined approach consisting of hysteroscopic tumour resection followed by oral progestins and/or an LNG-IUD represents the most effective fertility-sparing treatment, achieving the highest complete response and live birth rates compared with other conservative options [ 3 ].

There are very few published studies on the incidence of non-gynaecological cancers among users of LNG-IUDs compared with non-users. A Finnish cohort study [ 6 ] used registry data from 1994 to 2007 and included women aged 30–49 who were coded as LNG-IUD users for menorrhagia ( n = 93,843) and compared cancer incidence in this cohort with that in the general population. They reported data for several non-gynaecological cancers, including lung and colon cancers, tentatively associated with sex hormones, such as melanoma and other cancers [ 58 ]. Lung cancer incidence was lower than expected among those with one or more LNG-IUD purchases compared with the general population, with a standardised incidence ratio of 0.68 (95% CI 0.49–0.91). Pancreatic cancer incidence was also reduced among those with one or more purchases of a levonorgestrel-releasing intrauterine device; the SIR was 0.50 (95% CI, 0.28–0.81). Colon and rectal cancer incidence was slightly increased among LNG-IUD users (SIR 1.17), although this finding was not statistically significant (95% CI 0.99–1.36). Similarly, thyroid: SIR 1.09 (95% CI 0.92–1.28) and melanoma of the skin: SIR 1.08 (95% CI 0.90–1.27). As shown in Table 4 , the incidence of stomach, liver, gallbladder and biliary tract, vulvar, vaginal, kidney, bladder and urinary tract, and brain and nervous system cancers as well as non-Hodgkin lymphoma, Hodgkin lymphoma, multiple myeloma, and leukaemia, did not differ significantly among LNG-IUD users. A similar study used registry data to investigate cancer incidence in women in Shanghai, China ( n = 66,661) [ 58 ]. This study compared the incidence of all and site-specific cancers between ‘ever’ and ‘never’ users of different contraceptive methods, including intrauterine devices. However, it was not specified whether these were exclusively levonorgestrel-releasing, and no 7.5-year follow-up was reported. In this study, no individual cancer type showed a statistically significant difference in incidence between ever- and never-users of intrauterine devices. Similarly to the Finnish cohort, these authors found a lower incidence of lung cancer among ever-users of intrauterine devices compared with never-users, with an HR of 0.80. However, this was not statistically significant (95% CI, 0.56–1.16), and a lower incidence of pancreatic cancer was observed, yielding an HR of 0.87 (95% CI, 0.46–1.63) among ever-users of intrauterine devices. Colon cancer and rectal cancer were not statistically significantly higher among intrauterine device users (colon cancer HR 1.01 (95% CI 0.70–1.47); rectal cancer HR 1.27 (95% CI 0.81–1.98). In contrast to Soini et al. [ 6 ], this study found non-statistically significant lower rates of thyroid cancer among intrauterine device users: HR 0.64 (95% CI 0.38–1.07). In addition to gynaecological cancers, including breast, liver, gallbladder, and stomach, incidences were reported, and these three cancers were not statistically significantly different in intrauterine device ever-users compared with never-users. Figure 1 summarises the overall benefit–risk profile of LNG-IUD use in relation to gynaecologic and non-gynaecologic cancer risk, highlighting substantial endometrial protection, possible ovarian protection, neutral cervical cancer findings, limited evidence for non-gynaecologic malignancies, and the need for individualised counselling in women with elevated baseline breast cancer risk.

The available evidence regarding the association between the LNG-IUD and cancer risk is derived almost exclusively from observational studies, including registry-based cohorts, case–control studies, and meta-analyses. Although these studies provide valuable long-term data, they are inherently susceptible to several methodological limitations that should be considered when interpreting the reported associations. Residual confounding remains one of the principal challenges. Important determinants of cancer risk—including family history, body mass index (BMI), parity, menopausal status, smoking, screening behaviour, and previous oral contraceptive use—are not consistently available or uniformly adjusted for across studies. In particular, differences in adjustment for prior oral contraceptive exposure may partially contribute to the variability observed in ovarian cancer risk estimates. Furthermore, women receiving an LNG-IUD often differ systematically from comparison groups because of the clinical indication for device insertion, introducing the potential for indication and selection bias. Additional variability arises from differences in exposure and comparator definitions. Several studies evaluated ever-use of any intrauterine device rather than LNG-IUD-specific exposure, while others compared LNG-IUD users with the general population instead of an internal unexposed cohort. Information regarding duration of use, cumulative exposure, timing of removal, and use of the LNG-IUD as part of menopausal hormone therapy is also frequently unavailable. These differences limit direct comparison between studies and may contribute to between-study heterogeneity. Moreover, the relatively short follow-up of many cohorts, in which cancers among users tend to cluster in the early years of use, may cause late-emerging risks to be missed or early protective signals to be overstated. Outcome assessment also differs considerably across the literature. Cervical studies have variably reported HPV infection, CIN2+, CIN3+, HSIL, or invasive cervical cancer, whereas ovarian cancer studies have not consistently distinguished histological subtypes. In addition, more frequent gynaecological surveillance among LNG-IUD users may increase the detection of premalignant lesions without necessarily influencing the incidence of invasive malignancies. Future prospective studies should therefore employ standardised definitions of LNG-IUD exposure, collect comprehensive information on major confounding variables—including prior oral contraceptive use—and ensure adequate follow-up to evaluate long-term cancer outcomes. Reporting both relative and absolute risk estimates, together with analyses stratified by cancer type and histological subtype, will provide more robust evidence and improve individualised clinical counselling.

The levonorgestrel-releasing intrauterine device (LNG-IUD) is one of the most widely used long-acting reversible contraceptive methods worldwide [ 1 ]. Beyond contraception, the LNG-IUD is widely used to treat heavy menstrual bleeding, dysmenorrhoea, endometriosis-associated pelvic pain, and adenomyosis; it provides significant symptom relief and improves quality of life for affected women [ 2 ]. In premenopausal and postmenopausal women, the LNG-IUD is also utilised for the treatment of endometrial hyperplasia and, in carefully selected cases, early-stage endometrioid endometrial carcinoma as part of fertility-sparing treatments [ 3 ]. Given its widespread use, the measurable systemic absorption of levonorgestrel despite the LNG-IUD’s predominantly local endometrial effects has raised important questions about its potential hormonal influence on cancer development [ 4 ]. In parallel, increased awareness of the association between combined oral contraceptives and breast cancer has intensified concerns about the safety of hormonal contraceptives and highlighted the need for evidence to guide counselling on progestin-only drugs or devices, including the LNG-IUD [ 5 , 6 ]. The existing literature on the oncological safety of the LNG-IUD presents a complex and sometimes conflicting landscape. In the context of breast cancer, research remains notably heterogeneous. While some large-scale cohorts and meta-analyses report a 20% to 30% increase in incidence, these findings are often countered by pooled analyses that demonstrate neutral associations. This discrepancy has fuelled a persistent debate over whether observed risk estimates reflect a true biological induction or are a byproduct of methodological differences, such as variations in study design and the extent of adjustment for critical confounders like family history and body mass index (BMI) [ 7 ]. In contrast, LNG-IUD use has consistently been associated with a reduced risk of endometrial cancer, likely reflecting the local progestogenic effects of levonorgestrel on the endometrium [ 6 ]. Observational studies have suggested a possible protective association between LNG-IUD use and ovarian cancer risk [ 8 ]. However, findings remain inconsistent, partly due to differences in adjustment for prior oral contraceptive use and the limited evidence available across histological subtypes of ovarian cancer [ 6 , 9 ]. For cervical cancer, available evidence specific to LNG-IUD exposure does not indicate a clear increase in risk. Nevertheless, interpretation remains challenging due to the frequent pooling of different intrauterine device subtypes in registry-based datasets and potential differences in screening practices among device users [ 7 ]. Evidence regarding non-gynaecologic malignancies, including colorectal and lung cancers, remains limited, and currently available studies generally suggest neutral or uncertain associations [ 4 , 6 ]. In this context, a synthesis of the available evidence is necessary to better define the oncological safety profile of the LNG-IUD. This narrative review aims to critically appraise the association between LNG-IUD use and cancer risk—encompassing breast, gynaecological (endometrial, ovarian, and cervical), and non-gynaecological malignancies (including pancreatic, lung, colorectal, gastric, hepatobiliary, thyroid, melanoma, urinary tract, neurological, and haematologic cancers)—to provide a balanced framework for evidence-based clinical counselling.

Current evidence suggests that the LNG-IUD has a generally favourable oncological safety profile. The most consistent data support a substantial reduction in endometrial cancer risk, plausibly attributable to the endometrium’s strong local progestogenic effect. Available evidence also suggests a possible protective association with ovarian cancer, whereas cervical cancer risk appears neutral, with no clear increase in CIN3+ or invasive disease. Evidence regarding non-gynaecologic cancers remains limited and does not currently support a definitive association. In contrast, findings for breast cancer remain heterogeneous. Some large registry-based studies suggest a modest increase in risk, particularly among women with additional predisposing factors (advanced age, postmenopausal status, obesity, or a family history of breast cancer). In contrast, other cohort and case–control studies have reported no significant increase in breast cancer risk. Therefore, LNG-IUD use should not be considered contraindicated in the general population solely based on current cancer risk data. However, individualised counselling is warranted, particularly in women with elevated baseline breast cancer risk, advanced age, postmenopausal status, obesity, or a family history of breast cancer. Further well-designed prospective studies with detailed information on duration of use, indication, menopausal status, comparator groups, and baseline breast cancer risk factors are needed to better define the long-term oncological safety of the LNG-IUD. Reporting both relative and absolute risks across cancer sites and histological subtypes would further support individualised counselling.