Effect of phlebotomy versus oral contraceptives containing cyproterone acetate on the clinical and biochemical parameters in women with polycystic ovary syndrome: a randomized controlled trial

Polycystic ovarian syndrome (PCOS) is the most common ovarian endocrinopathy affecting 6–14% of reproductive aged women [1]. It is characterized by ovulatory dysfunction, clinical or para-clinical hyperandrogenism, and polycystic ovary morphology [2]. Insulin resistance (IR) and obesity are closely associated with PCOS and further reinforce the severity of clinical manifestations [3‐5]. Treatment of PCOS is symptom-based and aims to reduce the metabolic risk associated with HA, obesity, and insulin resistance [6‐9].

Combined oral contraceptive pills (OCs) have been the key component of relieving PCOS symptoms specifically menstrual irregularity, hirsutism, and acne through improving androgen excess and regulating menstruation [10, 11]. However, there are doubts concerning their impacts on the cardio-metabolism outcomes, which may be aggravated or even triggered by the use of OCs [12, 13]. Nevertheless, the benefits of OCs outweigh related risks in women with PCOS. Since patients with PCOS mostly use these drugs for several years [11], safer therapeutic options are considered suitable for the treatment of PCOS symptoms.

PCOS is associated with iron overload and, which can be associated with IR and further glucose intolerance [14‐17]. Factors contributing to potential iron overload in women with PCOS are (i) the iron saved secondary to chronic infrequent menstrual bleeding and (ii) compensatory hyperinsulinemia following insulin resistance. It is believed that insulin favors the intestinal and tissue deposition of iron and can decrease hepcidin, that consequently leads to elevated iron absorption [14, 17, 18]. In addition, reduction of body iron stores can improve hyperandrogenemia and adverse cardio-metabolic effects [19‐21]. In earlier studies, iron-chelating agents and blood donation can prevent the development of diabetes due to iron overload [22, 23]. Other studies also have shown that depleting iron stores in diabetes through phlebotomy and blood donation can decrease insulin resistance [24, 25].

The effects of iron reductive therapy in the treatment of PCOS symptoms have not been previously addressed. Therefore, this randomized controlled trial was conducted to investigate the effect of phlebotomy on clinical and biochemical parameters in women with PCOS compared to oral contraceptives containing cyproterone acetate.

This parallel non-inferiority randomized controlled trial was conducted on women with PCOS to compare the usual PCOS therapy using OCs containing cyproterone acetate and phlebotomy as the assumed potential modality.

This clinical trial was approved by the ethics committee of the Research Institute of Endocrine Sciences, Shahid Beheshti University of Medical Sciences (decree code: 17ECRIES91/12/08). The research protocol was registered on the Iranian Registry of Clinical Trials at www.​irct.​ir (decree code: IRCT2013080514277N1). The written informed consent form was signed by the women prior to conducting the study.

In the phlebotomy group, 27 (84.3%) and in the OCs group 30 (93.7%) of the women completed the 3-month follow-up. Five women in the phlebotomy group and 2 women in the OCs group withdrew from the study (Fig. 1). However, the basic characteristics of those women who lost to follow were not significantly different from remaining ones.

The demographic characteristics of the women were shown in Table 1. The mean age and BMI in the OCs vs. the phlebotomy group were 24 (4.06) vs. 24.2 (5.2), years and 27.1 (5.04) vs. 25.4 (5.3), kg/m², respectively at the initiation of the study.

At baseline, all parameters were comparable between the groups and no statistically significant differences between the groups with regard to age, BMI, androgenic profile, and other carbohydrate or metabolic profile were reported.

Effects of both treatments on HOMA-IR were presented in Table 2. A statistically significant decline of HOMA-IR from the baseline compared to the end of the 3th month of the treatment in both groups [3.5 (1.3–4.5) vs. 2.7 (1.2–4.3), P = 0.047 and 3.1 (1.4–3.5) vs. 2.8 (1.8–3.9), P = 0.032, in the phlebotomy and the OCs groups, respectively] was reported, but it was not statistically significant (P = 0.516).

Both treatments significantly decreased insulin levels throughout the study (9.5 (6.8–20) vs. 8.1 (6.5–15.7) P = 0.025 and 10.8 (7.6–14.9) vs. 9.8 (7.8–16.1), P = 0.015 in the phlebotomy and the OCs group, respectively). Moreover, mean differences were not statistically significant between the groups.

All androgenic parameters including testosterone, androstenedione and FAI decreased in both groups. However, no statistically significant differences between the groups were reported (Table 2). Compared to the baseline, SHBG showed a rising trend during the treatment in both groups, but the mean score of changes in the OCs was significantly higher than of the phlebotomy groups (P = 0.009).

Despite no statistically significant difference in lipid profiles of both groups at the initiation of study (Table 1), there was a statistically significant rise in triglycerides from the baseline in the OCs (114.2 (48) vs. 152 (89.9), P = 0.004) at the end of the 3th month of the treatment. Accordingly, the mean score changes of TG between the groups was statistically significant (8.3 (83.9) vs. 37.7 (65.1), P = 0.019). Changes in LDL-C, HDL-C and TC had no statistically significant differences within and between the groups.

At the baseline, the mean FG scores in the phlebotomy and the OCs groups were 16.8 (6) and 14.3 (7), respectively as decreased to 13.3 (7.4), P = 0.001 and 9.8 (7.6), P = 0.028, respectively, However, the changes were not statistically significant between the groups (Table 2). AMH and PSA remained unchanged in both groups throughout the study.

This is the first study on the endocrine, clinical and metabolic effects of treatment using phlebotomy in women with PCOS and comparison of the beneficial effects of two different treatments using the OCs containing cyproterone acetate and phlebotomy. As a result, both treatments had similar beneficial effects on insulin resistance as well as on androgenic profiles. However, OCs containing cyproterone acetate was found to be more effective in the treatment of menstrual irregularity and phlebotomy and had less adverse effects.

Ethinyl estradiol and cyproterone acetate in combination (cyproterone compound) have been widely used as reliable contraceptive pills in women with PCOS suffering from hyperanderogenism [31]. Such a combination can improve various PCOS manifestations through several potential mechanisms including the inhibition of LH secretion from the pituitary, reduction of androgen production from ovaries, reduction of free androgens, increase of circulating SHBG levels and thereby limiting peripheral androgen exposure. Cyproterone acetate can block peripheral androgen receptors at target organs and can reduce the ovarian androgen production and decrease the plasma levels of free testosterone [32].

However, OCs are known to have adverse effects on lipid metabolism and carbohydrate intolerance. Estrogen has pro-thrombotic effects and increases the cardiovascular venous thromboembolism [33].

The increased potential adverse cardio-metabolic effects of these drugs encourage the use of non-pharmacological treatments for relieving PCOS symptoms. Surprisingly, a lower prevalence of diabetes has been observed among blood donors [34]. Recently, it has been shown that hyperandrogenemia and insulin resistance are associated with ferritin [35, 36].

Emerging evidence hence suggests that in the general population, body iron stores and iron intake are positively associated with glucose intolerance and diabetes [21, 37, 38]. Although, potential underlying mechanisms are not fully understood, iron influences insulin production and sensitivity, and together they influence on iron metabolism. Therefore, iron interferes with the insulin prohibition of hepatic glucose secretion, and hepatic iron stores reduce the production of insulin leading to systemic hyperinsulinemia [14]. In this respect, previous studies found the beneficial effects of phlebotomy and blood donation in patients with type 2 diabetes [24, 25, 39], demonstrating that iron reduction by phlebotomy improved insulin sensitivity.

Women with PCOS may suffer from iron overload due to chronic infrequent menstrual bleeding, that in turn reduces menstrual blood loss along with reduced serum hepcidin levels as an iron-regulatory hormone [40]. It is also affected by insulin resistance and hyperandrogenism that may improve iron absorption and reduce iron secretion from macrophages, and possibly leads to mild iron overload in some patients with PCOS [17, 35, 41]. Such an overload can contribute to glucose intolerance in women with PCOS [35] independent of chronic inflammation [42].

Consistent with this hypothesis, a few studies are available on the effects of therapeutic interventions on iron exess in women with PCOS. In this respect, Luque-Ramírez et al. (2011) in a randomized clinical trial compared the effect of metformin as an insulin sensitizer with an antiandrogenic contraceptive pill for 6 months and found that patients with PCOS decreased hepcidin levels that might contribute to iron overload through facilating the intestinal absorption of iron [40].

However, our study as a pioneer investigation confirmed that phlebotomy enhanced insulin sensitivity and improved the state of hyperanderogenism in women with PCOS. However, more studies with longer follow-ups are still needed to confirm these thraputic effects.

In the present study, the side effects of phlebotomy were rare. The appropriate selection of vein and careful procedure of venipuncture helped us to minimize probable vasovagal reactions.

As the limitation of this study the short study duration might not have shown the long terms effect of this treatment, because the improvement of some PCOS manifestations might need long term follow-ups. In addition, some PCOS confounders including the women lifestyle were not assessed. No enogh power was present to performe the subgroup analysis based on various PCOS phenotypes or obesity status. Also, the HOMA-IR has been used as a surrogate marker for the assessment of IR, despite the good correlation between HOMA-IR and gold standard clamp methods [43], which might be inaccurate in women with PCOS [44]. Also, this study did not measure free testosterone due to a lack of access to the proper method. It has been shown that FAI has a good correlation with free testosterone [45]. Since ferritin was not measured at the end of the study follow-up, assessment of changes in iron overload was impossible. Moreover, the number of lost to follow up subjects in the phlebotomy group was higher than those in the OCs group, but our study results may not be affected as there would be no statistically significant differences in basic characteristics between those who lost to follow up and the remained participants in each group. In addition, we cross validated our data and found that the internal validity of the results was not affected by lost to follow up. Finally, the sample size was calculated based on the HOMA-IR, and there was not adequate number of subjects for the precise comparison of other clinical or para-clinical manifestations of PCOS. Therefore, observiation of no significant differences on these items could be due to the small sample size, which should be interpreted with cautious.

This study suggest that the phlebotomy therapy is associated with a tendency to decrease insulin resistance and hyperanderogenism, that merits further investigations. In addition, large prospective, randomized controlled trials with longer treatment periods are recommended to confirm the study findings.