This phase 2, randomized, double-blind and placebo-controlled trial «Hirschberg AL, Sánchez-Rovira P, Presa-Lorite J, y...»1 assessed the efficacy and safety of 12 weeks of 0.005% (50 ug/g) vaginal estriol gel or placebo gel in postmenopausal women with a previous hormone-receptor positive breast cancer (BC) (stages I-IIIa, any HER2 status) and current aromatase inhibitor (AI) treatment. From 86 participants screened, a total of 61 (median age 59 years, range 45-77) women were randomized at a 4:1 ratio to receive either study treatment once daily for 3 weeks, then twice weekly for the remaining 8 weeks. Primary safety outcome was variation in the level of follicle-stimulating hormone (FSH) from baseline to week 12, and efficacy was assessed by using subjective and objective measures. The levels of estrogens (estradiol, estriol and estrone) were determined with liquid chromatography tandem mass spectrometry at baseline and at weeks 1, 3, 8, and 12. Nine participants in the active treatment group and two in the placebo group discontinued the trial, but analyses were done intention-to-treat.
Treatment with estriol gel did not confer significant changes in FSH levels at 12 weeks (median difference from baseline -2.8 [interquartile range {IQR} -13.1 to7.4]; P.0.104) and LH levels (median -0.8 [IGR -5.3 to 2.9]; P.0.116). In the active group, serum estriol was transiently elevated within the first 3 weeks, with the highest levels observed at week 1 after treatment start. However, the estriol levels (median, 3.9 pg/mL; IQR, 0.5 to 12.1) remained low. The levels of estrone and estradiolwere persistently below the limit of quantification at all study visits. During a total of 16 weeks of follow-up including the posttreatment visit, no deaths or serious adverse events related to the treatment were reported.
Comment: Short-term use of low-dose vaginal estriol in current AI users does not significantly increase the levels of FSH or systemic estrogen. Safety of long-term estriol in BC survivors using AI remains to be determined.
General comment: Estriol is a final metabolite in estradiol metabolism, which is unable to convert back its' other estrogenic precursors, and is considered a less potent hormone with lower estrogenic potential. Estriol also displays preferential affinity to urogenital (beta) rather than breast (alpha) estrogen receptors. Very low-dose estriol may not be effective enough for the treatment of GSM in all users.
In a Danish nationwide, prospective register-based cohort study (n=8461) «Cold S, Cold F, Jensen MB, ym. Systemic or Vaginal...»2 gathered data from the Danish Breast Cancer Group (DBCG) clinical database, the Danish National Patient Registry, the Danish Civil Registration System, and the Danish National Prescription Database. The study subjects had had BC surgery between 1997-2004 to treat an early stage, estrogen-receptor positive BC (43% node positive, 34% tumor size >2 cm). All women received either 5 years of tamoxifen (TMX), an aromatase inhibitor (AI), or both treatments in sequence, or no endocrine treatment, in accordance with national treatment guidelines. Use of MHT or VE was classified as a time-varying dichotomous variable updated follow-up. The clinical follow-up continued until recurrence of BC, secondary malignancy, death, or a maximum of 10 years. Overall survival was determined until date of death, emigration, or December 31, 2016. Analyses were adjusted for relevant variables, e.g. age at surgery, tumor size, nodal status, histology, grade.
The final cohort consisted of 8461 women (mean age 61 years) out of whom 133 (2%) had used MHT (concomitant VE use was allowed) and 1957 (23%) women had used VE during follow-up. No adjuvant BC treatment was given to 3112 women (VE use n=790). Adjuvant treatment with TMX was given to 2007 women (VE users n=345), and AI to 403 women (VE users n=93). A total of 2939 women received a sequence of TMX and AI (VE use n=729). During an estimated median of 9.8 years of potential follow-up, 1333 (16%) women had a recurrent BC (111 in VE and 16 MHT users). The adjusted risk of recurrence was similar in VE users (HR = 1.08, 95% CI 0.89-1.32) compared to never-users. Sub-group analyses showed increased recurrence risk in women who had used VE concurrent with AI (n=93), or TMX in sequence with AI (n=729) (adjusted HR 1.39, 95% CI 1.04-1.85). Overall survival was not affected by VE use (adjusted HR 0.78, 95% CI 0.71-0.87). Neither was any effect seen in analyses stratified by adjuvant therapy (adjusted HR 0.94, 95% CI 0.70-1.26), as compared with never-users.
Comment: The study did not differentiate between types or doses of VE.
A population-based cohort study «McVicker L, Labeit AM, Coupland CAC, ym. Vaginal E...»3 on women with recent BC (n=49 237) were identified from national cancer registry reports (Scotland 2010-2017, Wales 2010-2016). The use of estrogen creams and tablets was ascertained from pharmacy dispensing records and general practice prescription records. BC-related deaths (n=5 795) were identified from national mortality records and BC-specific mortality in VE users was compared to non-users. During follow-up until 2020 (Scotland 5 years [3-7], Wales 8 years [5-12], data as median [IQR]), a total 2 551 (5%) of the women had used VE and 556 (1%) systemic HRT (with or without VE]. BC-specific deaths occurred in 120 women (follow-up person-years 11 437) who had used VE, and 51 HRT users (follow-up person-years 3 894), and in 5 624 non-users (follow-up person-years 285 342).
No increase in BC-specific mortality was seen in fully adjusted (e.g. age, year, exposure, cancer stage, grade and treatment, adjuvant therapy, hysterectomy, oophorectomy) pooled analyses on VE users compared with non-users (HR 0.77, 95% CI 0.63-0.94). A statistically significant decrease in mortality was seen for women with both 1-4 and ≥5 prescriptions (HR 0.81, 95%CI 0.67-0.99) and HR 0.57, 95% CI 0.34-0.96], respectively). When the use of VE was stratified according to dose, no significant mortality reduction was seen (higher-dose VE, 25ug tablets [HR 0.81, 95% CI 0.55-1.21], lower-dose VE HR 0.77, 95%CI [0.56-1.07]). In AI users, the use of VE did not confer an increased mortality risk compared with no VE use (pooled fully adjusted HR 0.72, 95%CI 0.58-0.91). No increased mortality risks were observed in VE users with estrogen-receptor positive BC (HR 0.88, 95%CI 0.62-1.25, available for Scotland only).
Comment: Systemic MHT users contributed to the data with a very low number of person-years (3 894) and fatalities (n=51) and very wide confidence intervals, and as such are not reliably translatable to a clinical setting.
General comment:
No randomized trials exist on the risk of recurrence in BC survivors using vaginally administered hormonal treatments for the treatment of genitourinary syndrome of menopause (GSM). Thus, the safety of different vaginal treatments has often been assessed indirectly by studying possible changes in sex steroids during the use of vaginal estriol, estradiol, testosterone or dehydroepiandrostendione. Of note, the measurement of sex steroids with earlier methods, such as radioimmunoassay, may have yielded different results than newer more sensitive assays using either gas or liquid chromatography-tandem mass spectrometry. Observational data support the notion of no increased risk recurrent BC with VE use. Subgroup analyses on users of adjuvant endocrine therapies, such as AI or tamoxifen, point towards increased recurrence risk in users of AI and concurrent VE, but the studies are limited in sample sizes. Studies designed to assess mortality in BC survivors did not find an increased death risk associated with the use of VE.