Case-Control-Study on the Breast Cancer Risk of Mirena® Compared With Copper IUDs

• ClinicalTrials.gov Identifier: NCT00461253
• Recruitment Status: Completed
• First Posted: April 17, 2007
• Results First Posted: January 25, 2016
• Last Update Posted: January 25, 2016
• Sponsor: Center for Epidemiology and Health Research, Germany
• Collaborator: Bayer
• Information provided by (Responsible Party): Juergen Dinger, MD, PhD, Center for Epidemiology and Health Research, Germany

• Study Type: Observational
• Study Design: Observational Model: Case-Control
• Condition: Breast Cancer
• Interventions: Device: Levonorgestrel-releasing IUD; Device: Copper IUD
• Enrollment: 25565

Participant Flow

Recruitment Details
Pre-assignment Details

Arm/Group Title	Cases	Controls
Arm/Group Description	Cases were defined as women who were diagnosed with breast cancer (invasive carcinoma or carcinoma in situ) between 1/2000 and 12/2007 and aged <50 years at diagnosis.	Controls were selected randomly from the national population registry (Finland) or via neighborhood controls by a controlled random route method (Germany).
Period Title: Overall Study
Started	10093 [1]	40250 [2]
Completed	5113 [3]	20452 [4]
Not Completed	4980	19798
Reason Not Completed
Death	985	0
Protocol Violation	613	236
no response	2609	11789
administrative	773	7773
[1] Identified cases; [2] Identified controls; [3] Analyzed cases; [4] Analyzed controls

Baseline Characteristics

Arm/Group Title		Cases	Controls	Total
Arm/Group Description		Cases were defined as women who were diagnosed with breast cancer (invasive carcinoma or carcinoma in situ) between 1/2000 and 12/2007 and aged <50 years at diagnosis.	Controls were selected randomly from the national population registry (Finland) or via neighborhood controls by a controlled random route method (Germany).	Total of all reporting groups
Overall Number of Baseline Participants		5113	20452	25565
Baseline Analysis Population Description		Definitions: cases were defined as women who were diagnised with breast cancer (invasive carcinoma or carcinoma in situ) between 1/2000 and 12/2007 and aged <50 years at diagnosis. Controls were selected randomly from the national population registry (Finland) or via neighborhood controls by a controlled random route method (Germany).
Age, Continuous; Mean (Standard Deviation); Unit of measure: Years
	Number Analyzed	5113 participants	20452 participants	25565 participants
		44.5 (5.1)	44.2 (5.1)	44.3 (5.1)
Gender; Measure Type: Number; Unit of measure: Participants	Number Analyzed	5113 participants	20452 participants	25565 participants
Female		5,113	20,452	25565
Male		0	0	0
Region of Enrollment; Measure Type: Number; Unit of measure: Participants	Number Analyzed	5113 participants	20452 participants	25565 participants
Germany		1815	7260	9075
Finland		3298	13192	16490

Outcome Measures

1. Primary Outcome

Title	Breast Cancer Risk
Description	Breast cancer (invasive carcinoma or carcinoma in situ) in women aged <50 years at diagnosis. Cases were excluded if they had died before study start or had a history of malignancy.
Time Frame	retrospective, January 2000 to December 2007

Outcome Measure Data

Analysis Population Description

Frequency of breast cancer (in situ and invasive) in LNG-IUD users and Cu-IUD users

Arm/Group Title	LNG IUD	Cu-IUD
Arm/Group Description:	Women who currently or ever used a LNG IUDs (Mirena) at time of breast cancer diagnosis	Women who currently or ever used a copper IUDs at time of breast cancer diagnosis
Overall Number of Participants Analyzed	5346	5428
Measure Type: Number; Unit of Measure: participants
	1076	1068

Statistical Analysis 1

Statistical Analysis Overview	Comparison Group Selection	LNG IUD, Cu-IUD
	Comments	In this matched case-control study, conditional logistic regression was used to investigate the relationship between a woman being a breast cancer case or a control (dichotomous dependent variable), and a set of actual or potential prognostic factors (covariates) for breast cancer (incl. the use of Mirena). Conditional logistic regression uses a maximum likelihood approach for estimating the covariates; data are stratified and the likelihoods are computed relative to each stratum.
	Type of Statistical Test	Non-Inferiority or Equivalence
	Comments	Assuming a prevalence of current Mirena use among fertile women of 8% in Finland and 2% in Germany, it was estimated that 3,500 cases and 14,000 controls would be needed to exclude a 1.5-fold breast cancer risk for Mirena users compared with users of copper IUDs.
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	0.99
	Confidence Interval	(2-Sided) 95%; 0.88 to 1.12
	Estimation Comments	Breast cancer OR for ever use of LNG-IUD vs. Cu-IUD; adjusted for BMI, family history of breast cancer, age at first birth, age at menarche and physical activity (primary analysis)

Statistical Analysis 2

Statistical Analysis Overview	Comparison Group Selection	LNG IUD, Cu-IUD
	Comments	In this matched case-control study, conditional logistic regression was used to investigate the relationship between a woman being a breast cancer case or a control (dichotomous dependent variable), and a set of actual or potential prognostic factors (covariates) for breast cancer (incl. the use of Mirena). Conditional logistic regression uses a maximum likelihood approach for estimating the covariates; data are stratified and the likelihoods are computed relative to each stratum.
	Type of Statistical Test	Non-Inferiority or Equivalence
	Comments	Assuming a prevalence of current Mirena use among fertile women of 8% in Finland and 2% in Germany, it was estimated that 3,500 cases and 14,000 controls would be needed to exclude a 1.5-fold breast cancer risk for Mirena users compared with users of copper IUDs.
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	1.04
	Confidence Interval	(2-Sided) 95%; 0.93 to 1.17
	Estimation Comments	Breast cancer OR for ever use of LNG-IUD vs. Cu-IUD, crude

Statistical Analysis 3

Statistical Analysis Overview	Comparison Group Selection	LNG IUD, Cu-IUD
	Comments	In this matched case-control study, conditional logistic regression was used to investigate the relationship between a woman being a breast cancer case or a control (dichotomous dependent variable), and a set of actual or potential prognostic factors (covariates) for breast cancer (incl. the use of Mirena). Conditional logistic regression uses a maximum likelihood approach for estimating the covariates; data are stratified and the likelihoods are computed relative to each stratum.
	Type of Statistical Test	Non-Inferiority or Equivalence
	Comments	Assuming a prevalence of current Mirena use among fertile women of 8% in Finland and 2% in Germany, it was estimated that 3,500 cases and 14,000 controls would be needed to exclude a 1.5-fold breast cancer risk for Mirena users compared with users of copper IUDs.
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	0.85
	Confidence Interval	(2-Sided) 95%; 0.52 to 1.39
	Estimation Comments	Breast cancer OR for "current use" of LNG-IUD vs. Cu-IUD at time of breast cancer diagnosis; adjusted for BMI, family history of breast cancer, age at first birth, age at first menarche, physical activity (primary analysis)

Statistical Analysis 4

Statistical Analysis Overview	Comparison Group Selection	LNG IUD, Cu-IUD
	Comments	In this matched case-control study, conditional logistic regression was used to investigate the relationship between a woman being a breast cancer case or a control (dichotomous dependent variable), and a set of actual or potential prognostic factors (covariates) for breast cancer (incl. the use of Mirena). Conditional logistic regression uses a maximum likelihood approach for estimating the covariates; data are stratified and the likelihoods are computed relative to each stratum.
	Type of Statistical Test	Non-Inferiority or Equivalence
	Comments	Assuming a prevalence of current Mirena use among fertile women of 8% in Finland and 2% in Germany, it was estimated that 3,500 cases and 14,000 controls would be needed to exclude a 1.5-fold breast cancer risk for Mirena users compared with users of copper IUDs.
Method of Estimation	Estimation Parameter	Odds Ratio (OR)
	Estimated Value	0.90
	Confidence Interval	(2-Sided) 95%; 0.58 to 1.41
	Estimation Comments	Breast cancer OR for "current use" of LNG-IUD vs. Cu-IUD at time of breast cancer diagnosis, crude

Adverse Events

Time Frame	Retrospective population based case-control study identifying breast cancer cases in the between January 2000 and December 2007 in women younger than 50 years of age at time of diagnosis.
Adverse Event Reporting Description	[Not Specified]
Arm/Group Title	LNG IUD		Cu-IUD
Arm/Group Description	Users of LNG IUDs (Mirena)		Users of copper IUDs
All-Cause Mortality
	LNG IUD		Cu-IUD
	Affected / at Risk (%)		Affected / at Risk (%)
Total	--/--		--/--
Serious Adverse Events
	LNG IUD		Cu-IUD
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	1085/5346 (20.30%)		1130/5428 (20.82%)
Neoplasms benign, malignant and unspecified (incl cysts and polyps)
Breast cancer † [1]	1085/5346 (20.30%)	1085	1130/5428 (20.82%)	1130
† Indicates events were collected by systematic assessment; [1] Breast cancer (in situ and invasive)
Other (Not Including Serious) Adverse Events
Frequency Threshold for Reporting Other Adverse Events	5%
	LNG IUD		Cu-IUD
	Affected / at Risk (%)	# Events	Affected / at Risk (%)	# Events
Total	0/5346 (0.00%)		0/5428 (0.00%)

Limitations and Caveats

Limitations include information-bias, which cannot be entirely eliminated in retrospective studies. However, in this study recall is likely not to be differential between cases and controls.

More Information

Certain Agreements

All Principal Investigators ARE employed by the organization sponsoring the study.

Results Point of Contact

• Name/Title: Dr. Jürgen Dinger
• Organization: Center for Epidemiology and Health Research, Germany
• Phone: 0049 (0) 30 945 101 20
• EMail: dinger@zeg-berlin.de

Publications of Results:

Dinger J, Bardenheuer K, Minh TD. Levonorgestrel-releasing and copper intrauterine devices and the risk of breast cancer. Contraception. 2011 Mar;83(3):211-7. doi: 10.1016/j.contraception.2010.11.009. Epub 2011 Jan 7.

• Responsible Party: Juergen Dinger, MD, PhD, Center for Epidemiology and Health Research, Germany
• ClinicalTrials.gov Identifier: NCT00461253
• Other Study ID Numbers: ZEG2006_02
• First Submitted: April 16, 2007
• First Posted: April 17, 2007
• Results First Submitted: December 17, 2015
• Results First Posted: January 25, 2016
• Last Update Posted: January 25, 2016