Combined oral contraceptives and VTE: Which pill conveys the highest risk in young women?

Combined oral contraceptives (COCs) that contain gestodene confer a fivefold increased risk of venous thromboembolism (VTE) in young women as compared with those that contain levonorgestrel and the lowest doses of ethinylestradiol, as shown in a study.

Meanwhile, the risk increase is twofold for COCs containing chlormadinone, cyproterone, desogestrel, dienogest and drospirenone, according to the study authors, Dr Tania Schink and colleagues, from Leibniz Institute for Prevention Research and Epidemiology – BIPS, Bremen, Germany. [BJOG 2022;doi:10.1111/1471-0528.17268]

“For norgestimate and nomegestrol, the point estimates suggested a twofold and a 40-percent increased risk, respectively, but confidence intervals [CIs] overlapped the null value of 1, and for norgestimate the [estimate] was close to 1 when the ethinylestradiol content was considered,” the team said.

“The risk of VTE is not only influenced by the type of progestogen, but also by the level of ethinylestradiol. Comparison of different progestogens without taking the ethinylestradiol content into account might bias the results if one progestogen is usually combined with higher ethinylestradiol as the reference,” they added.

The analysis was based on 1,166 cases of VTE matched to 11,660 non-VTE controls nested in a cohort of 677,331 girls and young women aged 10–19 years in Germany who had one or more COCs dispensed after a 1-year period without any COCs.

Using levonorgestrel with low ethinylestradiol (<50 μg) as the reference, the adjusted odds ratios for VTE risk were 5.05 (95 percent CI, 1.23–20.74) with gestodene, 2.23 (95 percent CI, 1.77–2.80) with dienogest, 2.15 (95 percent CI, 1.43–3.25) with cyproterone, 2.06 (95 percent CI, 1.58–2.68) with chlormadinone, 1.93 (95 percent CI, 1.44–2.61) with desogestrel, 1.89 (95 percent CI, 1.41–2.55) with drospirenone, 1.90 (95 percent CI, 0.62–5.81) with norgestimate, and 1.41 (95 percent CI, 0.52–3.81) with nomegestrol.

Stratified analyses did not detect any effect modification by age, diagnosis of an acute infection, or immobilisation/recent trauma.

“In the interpretation of this study, it must be noted that we could only include girls and young women up to the age of 19 years, as COCs were only reimbursable up to the 20th birthday during the study period,” Schink and colleagues said. “The risk estimates observed in our study thus refer to this young population and it is not clear whether they can be extrapolated to women aged ≥20 years, where incidences of VTE are increasing and the aetiology of VTE might differ.”

Broadly used by women of childbearing age, COCs are known to be associated with an increased risk of VTE and the present data confirm that levonorgestrel with low-doses of ethinylestradiol (<50 μg) is the COC that conveys the lowest risk, according to the authors.