Link Search Menu Expand Document

Analysis of Cardiovascular and Thromboembolic Toxicity with High Dose Parenteral Polyestradiol Phosphate in the Treatment of Prostate Cancer

By Sam S. | First published April 6, 2020 | Last modified October 5, 2020

In the late 1980s, there was a renewed interest in the Nordic countries for estrogen therapy as a treatment for prostate cancer. The reason for this was that, in small pilot studies, high doses of polyestradiol phosphate (PEP) administered parenterally did not seem to result in a greater incidence of cardiovascular complications (Henriksson et al., 1988; Stege et al., 1989; Carlström et al., 1989; Henriksson et al., 1999). This finding was in direct contrast to earlier studies that had reported significantly greater incidences of cardiovascular morbidity and mortality in elderly male prostate cancer patients treated with oral ethinylestradiol (EE) and diethylstilbestrol (DES) (Bailar & Byar, 1972; Henriksson & Edhag, 1986).

At sufficient exposure, systemic estrogenic activity has been associated with a greater synthesis of liver proteins, lipids and coagulation factors such as sex-hormone binding globulin (SHBG) and triglycerides (Ottosson et al., 1986). In turn, many of these unphysiologic homeostatic changes are correlated with increased cardiovascular and thromboembolic toxicity (Scarabin, 2014; Seal, 2019; Khan et al., 2019). In addition to pure estrogens, raloxifene usage also has been shown to confer increased risk of venous thromboembolism (Vinogradova, Coupland & Hippisley-Cox, 2019). This is of significance as it was also found in the pilot studies that, while high dose EE therapy greatly increased the synthesis of SHBG and coagulation factor VIII, the effect was significantly less marked with parenteral PEP (Stege et al., 1988; Henriksson et al., 1990). However, both have been shown to induce a significant decrease in antithrombin III; a liver protein that prevents the formation of abnormal blood clots (Aro et al., 1990). By contrast, LHRH agonists do not appear to alter the production of this protein (Varenhorst et al., 1986). These differences were attributed to the disproportionate estrogenic exposure that occurs within the liver relative to other tissues via the oral route (von Schoultz et al., 1989; Ockrim, Lalani & Abel, 2006). Moreover, synthetic estrogens such as EE are resistant to enzymatic liver metabolism due to their augmented or non steroidal chemical structures and cause further disproportionate hepatic receptor activation (Kuhl, 2005). For this reason, intramuscular therapy with PEP; a prodrug for estradiol, was predicted to have much lesser cardiovascular and thromboembolic toxicity than with conventional estrogen therapy (Hedlund, 1999).

During the 1990s and 2000s, several multicenter clinical trials were conducted to establish the efficacy and safety of high dose parenteral estradiol therapy; as compared to bilateral orchiectomy or LHRH agonists (Lycette et al., 2006). The results of these clinical trials are of obvious importance to gender affirming hormone therapy as transgender women usually require long term use of estrogen formulations and a good number choose parenteral forms. It may not be possible to extrapolate these findings to all doses and types of estrogen esters. Nonetheless, these clinical trials give us some idea of the cardiovascular and thromboembolic toxicity of parenteral estradiol. For reference, a 160 to 240 mg/month dose of PEP has been found to produce estradiol levels of between approximately 300 to 500 pg/mL at steady state (Graph). It is obvious from findings that parenteral PEP is associated with a much lesser incidence of adverse cardiovascular and thromboembolic events than was oral EE (Mikkola et al., 1998; Hedlund & Henriksson, 2000; Hedlund et al., 2002). However, it is much less clear if toxicity is eliminated completely. In most of these studies there was inadequate sample power to show if the incidence was changed significantly (Bishop et al., 1989; Lukkarinen & Kontturi, 1994). Nevertheless, one of the largest studies did conclude that parenteral PEP ought not to be used as primary therapy for prostate cancer because of a significant increase in the incidence of such complications (Mikkola et al., 2005; Mikkola et al., 2007). The results of this clinical trial are probably one factor in why the use of high dose estrogen therapy in the treatment of prostate cancer diminished thereafter in many countries.

I thought it would be interesting to bring together all of the data in a pooled meta-analysis. In total, there were six different clinical trials evaluating polyestradiol phosphate. I noted the number of cardiovascular and thromboembolic events in each study for the treatment group exposed to high dose parenteral estradiol and for the comparison group (which had no exposure to estrogens). Unfortunately, one of these clinical trials (FinnProstate-4) did not differentiate between which complications were cardiovascular and which were thromboembolic and so it was not possible to include. There was patient follow-up for a total of 1,890 patients. From this information, I was able to calculate the odds of cardiovascular and thromboembolic complications in patients exposed to therapy for each individual study and for the overall sample. In addition, I calculated a 95% confidence interval. I may turn these findings into part of a more comprehensive article in the future. For now, I have presented them here in the form of some (hopefully useful) infographics (Sources for reference values in Fig. 4 were Vinogradova, Coupland & Hippisley-Cox, 2019 and Langley et al., 2013).

The sources for each study can be found here.

Pooling together the findings of six randomised controlled trials reveals that parenteral estrogen therapy is associated with an increased incidence of cardiovascular and thromboembolic complications; both separately and combined. Nevertheless, due to poor study precision, the exact risk with high doses of parenteral estrogens remains to be accurately quantified.

Figure 1: Odds of CVD with parenteral PEP vs. orchiectomy or LHRH agonist (p = 0.0017; significant at p<0.01).
Figure 2: Odds of VTE with parenteral PEP vs. orchiectomy or LHRH agonist (p = 0.052; significant at p<0.10).
Figure 3: Combined odds of VTE and CVD with parenteral PEP vs. orchiectomy or LHRH agonist (p = 0.0003).
Figure 4: Odds of VTE with parenteral PEP compared to other formulations and doses.