Efficacy of estrogen plus progestin on menopausal symptoms in women with systemic lupus erythematosus: A randomized, double-blind, controlled trial

Authors


  • ClinicalTrials.gov identifier: NCT00392093.

Abstract

Objective

To define the effects of continuous sequential estrogen plus progestin therapy on menopausal symptoms in women with systemic lupus erythematosus (SLE).

Methods

We performed a randomized, double-blind, 24-month clinical trial involving 106 women with SLE who were in the menopausal transition or early or late postmenopause. Patients received continuous sequential estrogen plus progestin (n = 52) or placebo (n = 54). Menopausal symptoms were assessed using the Greene Climacteric Scale at 0, 1, 2, 3, 6, 9, 12, 15, 18, 21, and 24 months. A new factor analysis of the scale reduced 21 items to 5 factors. The primary outcome was improvement of menopausal symptoms throughout the followup period. Results were analyzed by the intent-to-treat principle.

Results

At baseline, demographic and disease characteristics were similar in both groups. Fifteen of 21 menopausal symptoms had a prevalence of ≥50%, with a similar distribution between groups. Vasomotor factor scores decreased over time in both groups (P = 0.002), but in the estrogen plus progestin group the reduction was more pronounced than in the placebo group (1.5–2.0 versus 0.35–0.8 points on a scale of 0–6; P = 0.03). Maximum effects were observed among the most symptomatic women. Psychological, subjective–somatic, and organic–somatic factors scores also improved along time (P < 0.001), but the treatment and placebo arms improved to a similar degree. Thromboses occurred in 3 patients receiving estrogen plus progestin and in 1 patient receiving placebo.

Conclusion

Menopausal symptoms are highly prevalent in peri- and postmenopausal lupus patients. Estrogen plus progestin improved vasomotor symptoms at a clinically significant level, but not other menopausal symptoms. Given the thrombotic risks of menopausal hormone therapy, this should be used only in women with significant vasomotor symptoms.

INTRODUCTION

Systemic lupus erythematosus is an autoimmune disease primarily affecting young women. Nonetheless, many patients reach menopause due to an improved survival rate (1, 2) and increased frequency of premature menopause (3, 4). Although menopausal symptoms are highly prevalent in women with lupus (5), their nature and severity have not been explored. Furthermore, the benefits of menopausal hormone therapy in lupus patients have not been assessed, and clinicians' decisions should rely on what is known about the effects on women who do not have lupus (6–8). To better estimate the risk/benefit ratio while prescribing hormone therapy to lupus patients, it is necessary to define its effects on the affected population. This study was designed to assess the effects of a continuous–sequential estrogen plus progestin regimen on the severity of menopausal symptoms in women with lupus who were at the menopausal transition or early or late postmenopausal stages.

Significance & Innovations

  • Through this study it has been possible to learn about the symptoms presented by peri- and postmenopausal lupus patients, including their type and frequency.

  • This randomized, double-blind, 24-month clinical trial constitutes the first direct evidence on the benefits of hormone therapy on menopausal symptoms in women with systemic lupus erythematosus. Such information was lacking and is relevant to help women to decide about treatment.

  • The study demonstrates that hormone therapy is clinically effective only for alleviation of vasomotor symptoms, but not for the treatment of other menopause-associated symptoms.

  • Overall, the study provides crucial information to assess the risk/benefit ratio of hormone therapy in women with systemic lupus erythematosus.

PATIENTS AND METHODS

We conducted a randomized, double-blind, placebo-controlled, 24-month clinical trial in women with lupus. The study was approved by the hospital's institutional review board, and all patients provided written informed consent. The methodologic details of this clinical trial have been published previously, with the results of the effects of estrogen plus progestin on lupus activity (9). Briefly, 1,981 women with a diagnosis of systemic lupus erythematosus (10) were assessed for eligibility. Eligible women were those having any 2 of the following criteria: amenorrhea for ≥6 months (except for women who had a hysterectomy), serum follicle-stimulating hormone levels of ≥30 IU/liter, menopausal symptoms, and age ≥48 years. Main exclusion criteria were age >65 years, Systemic Lupus Erythematosus Disease Activity Index (11) score of >30 at baseline, use of estrogen within 3 months of the screening visit, thrombosis within 6 months, and hormone-dependent malignancies. Women who had presented with endometrial bleeding within the 12 months prior to the study initiation were considered to be in the menopausal transition, whereas those with amenorrhea of a duration of at least 12 months were postmenopausal. Early postmenopausal women included those with ≤5 years of amenorrhea, whereas late postmenopausal were those with >5 years of amenorrhea (12). Premature menopause was diagnosed when it occurred at age ≤40 years (6).

Protocol.

A computer-generated randomization list was used to assign the women to either menopause hormonal therapy (a continuous–sequential estrogen plus progestin regimen of conjugated equine estrogen 0.625 mg/day plus 5 mg/day of medroxyprogesterone acetate by mouth for the first 10 days per month; Wyeth Pharmaceuticals) or a biologically inert placebo identical in appearance and packing size to the medication in the active treatment regimen. The study coordinator assigned the next available number to each patient upon her entry into the trial, and the pharmacy personnel dispensed the study medications according to the randomization list. To maintain blinding, the investigators were instructed not to provide care for gynecologic or menopausal symptoms and the women were asked to avoid discussing any related issue with them. All of the women were told during the first visit that they might experience uterine bleeding independently of the treatment assigned. The randomization code was broken at the end of the study.

Clinical assessments.

At baseline, data on demographic and clinical characteristics were collected by the investigators. The menopausal symptoms were assessed in each visit through a face-to-face interview, utilizing the Greene Climacteric Scale (13). Followup evaluations, including menopausal symptoms assessment and gynecologic evaluations, were undertaken at 1, 2, 3, 6, 9, 12, 15, 18, 21, and 24 months of treatment. Gynecologists were not allowed to modify the treatment for menopausal symptoms. Adherence to treatment was assessed primarily by self-reported medication intake, and confirmed by measurement of serum levels of estradiol at baseline and at 1, 2, 3, 6, and 15 months. Study treatment was discontinued in women who developed any severe complication or whenever they desired (9). The date and primary reason for discontinuation were coded.

The Greene Climacteric Scale is a standardized instrument widely used for the evaluation of menopausal symptoms in epidemiologic (14–16) and clinical studies (17–22). It comprises 21 questions that measure physical and psychological symptoms associated with menopause, further grouped in 3 independent subscales called factors: psychological, somatic, and vasomotor. Symptom 21 is a probe for sexual dysfunction. Each symptom is rated according to its severity, using a 4-point scale: not at all (0), a little (1), quite a bit (2), and extremely (3). The overall score ranges from 0 (best) to 63 (worst) (13). Since the Greene Climacteric Scale was not standardized for women with a chronic disease, a new factor analysis was undertaken.

Outcome measures.

The primary outcome was severity of menopausal symptoms throughout the followup period. Secondary outcomes included treatment discontinuation rates and reasons and side effects.

Statistical analyses.

Factor analysis was done using principal components extraction and varimax rotation. Factors with eigenvalues >1 were extracted and a factor-loading cutoff of 0.40 was used. This analysis was conducted on the entire sample (n = 106) at baseline, using the severity scores (from 0–3) on the 21 items of the Greene Climacteric Scale. Reliability of the factor solution adopted was assessed by repeating the factor analysis in the placebo group at baseline and at followup visits 1, 2, and 3 months (for details, see Supplementary Appendix A and Supplementary Tables 1 and 2, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658).

Comparison of baseline characteristics between treatment groups was performed with one-way analysis of variance for continuous variables and chi-square tests for categorical variables.

Primary analysis used repeated-measures analysis for modeling changes along time in menopausal symptom factor scores, controlling for menopause cause (natural versus hysterectomy) and ovarian stage (menopausal transition plus early postmenopause versus late postmenopause), and on the basis of the intent-to-treat principle. This repeated-measures analysis was conducted for the entire sample and for women with at least moderate symptoms at baseline. Because the 2 groups (treated/not treated) were similar at baseline, any systematic treatment-related effect could spuriously be represented as a “time by treatment” interaction; therefore, the analysis was also conducted excluding baseline measurements (see Supplementary Tables 3 and 4, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658).

Discontinuation rates were compared between treatment groups using Kaplan-Meier statistics and log rank tests. Separate analyses were conducted for “all reasons” (excluding end of study), “medical reasons,” and “nonmedical reasons.”

Only 1 woman was considered nonadherent to treatment, so a per-protocol analysis was not done. The SPSS statistical package, version 11.5, was used for all analyses, with 2-tailed tests and P values of 0.05 or less considered significant.

Sample size calculations were based on the disease activity end point (9). Assuming that 60% of symptomatic women at baseline became asymptomatic at 3 months in the hormone-treated group, and that the corresponding figure in the placebo group was <25%, a total sample of 72 symptomatic women, one-half in the treated group and one-half in the placebo group, would be required to detect this difference as significant, with a significance level of 95% and power of 80%. From previous studies we know that the prevalence of symptomatic women is approximately 70%, so our sample of 106 women should generally be able to provide the 72 symptomatic women required. In fact, in our sample we observed that more than 70% of all subjects had at least 5 (of the 21) symptoms present at baseline.

RESULTS

As reported previously (9), from the 1,981 patients who underwent screening, 149 remained eligible for the study, 43 declined to participate, and 106 were randomly assigned to menopausal hormone therapy (n = 52) or placebo (n = 54). Women who declined to participate in the study and those who underwent randomization did not differ in most demographic and lupus characteristics, including previous use of hormone therapy. Nevertheless, those who underwent randomization were younger (P = 0.004), developed menopause at an earlier age (P = 0.04), had a higher body mass index (P = 0.002), and reported more frequent smoking and alcohol consumption (data not shown). Among the 106 randomized patients, 22% were transitioning to menopause and 78% were postmenopausal, including 28% with premature menopause. The baseline demographic and clinical features, including menopause and lupus characteristics, were similar between both treatment groups (Table 1).

Table 1. Selected baseline demographic features and menopause characteristics of the 106 women, according to the assigned treatment*
 Hormone therapy (n = 52)Placebo (n = 54)P
  • *

    SLEDAI = Systemic Lupus Erythematosus Disease Activity Index; aPL = antiphospholipid antibodies; anti-β2GPI = anti–β2-glycoprotein I.

  • Menopause characteristics estimated only for those patients who were at the early or late postmenopausal stages.

  • SLEDAI scores range from 0–105, with higher scores indicating more severe disease.

  • §

    Low-titer aPL were considered positive.

Demographic features   
 Age, mean ± SD (range) years47.4 ± 7.4 (28–65)50.1 ± 7.7 (28–63)0.08
 Education, mean ± SD years10.5 ± 4.910.3 ± 4.70.83
 Body mass index, mean ± SD kg/m227.5 ± 6.226.3 ± 4.40.25
  ≤25, no. (%)22 (42)23 (43)0.93
  >25 to 27, no. (%)6 (11)8 (15) 
  >27 to 30, no. (%)7 (13)8 (15) 
  >30, no. (%)17 (33)15 (28) 
 Current smoker, no. (%)10 (19.2)12 (22.2)0.86
Menopause characteristics   
 Stages of menopause, no. (%)   
  Menopausal transition12 (23)11 (20)0.40
  Early postmenopause13 (25)20 (37) 
  Late postmenopause27 (52)23 (43) 
 Years since menopause, mean ± SD8.1 ± 5.78.3 ± 6.70.86
 Premature menopause, no. (%)17 (42.5)13 (30.2)0.26
 Surgically-induced menopause, no. (%)10 (25)6 (14)0.16
 Age at menopause, mean ± SD years40.0 ± 7.542.9 ± 7.80.09
 Previous use of hormone therapy, no. (%)18 (35)22 (41)0.55
Systemic lupus erythematosus characteristics   
 Disease duration, mean ± SD years9.6 ± 8.98.9 ± 7.20.62
 SLEDAI score, mean ± SD (range)3.5 ± 3.3 (0–10)3.1 ± 3.4 (0–15)0.54
 Positive aPL test result, no. (%)§   
  Anticardiolipin antibodies13 (25)18 (33)0.35
   Anti-β2GPI8 (15)8 (15)0.94
   Lupus anticoagulant6 (12)5 (9)0.76
   Any aPL20 (38)23 (43)0.67

There were 2,095 patient-months of followup (82.3% of the possible 2,544 total months of followup), corresponding to 1,043 and 1,052 patient-months for the hormone therapy and placebo groups, respectively. Mean ± SD serum estradiol levels at baseline were 32 ± 35 pg/ml and 37 ± 73 pg/ml (P = 0.62) in the hormone therapy and placebo groups, respectively. During the followup, serum estradiol mean levels ranged between 56 and 81 pg/ml in the hormone therapy group and between 28 and 45 pg/ml in the placebo group. Compared with baseline, estradiol levels at each followup assessment were higher in the hormone therapy group (P < 0.001) but not in the placebo group. Between-group comparison also showed higher estradiol levels in the hormone therapy group than in the placebo group (P < 0.001).

Factor structure.

The factor analysis in lupus patients initially produced 6 factors, which explained 61% of the variability in the original 21 Greene Climacteric Scale items (see Supplementary Table 2, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). Hot flashes and night sweats were grouped in a single factor, as it is in the standardized Greene Climacteric Scale. However, the other symptoms followed a different pathway of distribution, leading to 4 new factors in which the original psychological and somatic factor components were recombined. These new lupus factors were renamed as: psychological (score 0–18), subjective–somatic (score 0–18), organic–somatic (score 0–9), and sensory–somatic (score 0–6) (Table 2). Attacks of panic and loss of interest in sex were not grouped with other symptoms and were excluded, so our final factor analysis solution had only 5 factors.

Table 2. Menopausal symptom factors in the Greene Climacteric Scale and in SLE*
 Greene Climacteric Scale factorsSLE factors
12345
  • *

    Numbers denote systemic lupus erythematosus (SLE) factors: 1 = psychological factor, 2 = subjective–somatic factor, 3 = organic–somatic factor, 4 = vasomotor factor, and 5 = sensory–somatic factor. Attacks of panic and loss of interest in sex did not group with other symptoms among lupus patients.

1. Heart beating quickly and stronglyPsychological  X  
2. Feeling tense or nervousPsychologicalX    
3. Difficulty in sleepingPsychological X   
4. ExcitablePsychologicalX    
5. Attacks of panicPsychological     
6. Difficulty in concentratingPsychological X   
7. Feeling tired or lacking in energyPsychological X   
8. Loss of interest in most thingsPsychologicalX    
9. Feeling unhappy or depressedPsychologicalX    
10. Crying spellsPsychologicalX    
11. IrritabilityPsychologicalX    
12. Feeling dizzy or faintSomatic X   
13. Pressure or tightness in head or bodySomatic X   
14. Parts of body feeling numb or tinglingSomatic    X
15. HeadachesSomatic X   
16. Muscle or joint painsSomatic  X  
17. Loss of feeling in hands or feetSomatic    X
18. Breathing difficultiesSomatic  X  
19. Hot flashesVasomotor   X 
20. Sweating at nightVasomotor   X 
21. Loss of interest in sex      

Menopausal symptoms.

Fifteen of the 21 evaluated symptoms had a prevalence of ≥50%, with a similar distribution between treatment groups. After 3 months of treatment, 60% of treated women with vasomotor symptoms at baseline became asymptomatic in contrast with 25.8% of untreated women. At baseline, mean ± SD scores of the new factors were similar in the study groups (Table 3). No association was found between factor scores and any of the following variables: age, age at menopause, previous hysterectomy, body mass index, previous use of hormone therapy, and lupus activity score.

Table 3. Baseline menopausal symptoms and menopausal factor scores of the 106 women, according to the treatment assigned*
 Hormone therapy (n = 52)Placebo (n = 54)P
  • *

    Psychological and subjective–somatic factors scores range from 0–18, organic–somatic factor scores range from 0–9, and vasomotor and sensory–somatic factors scores range from 0–6. Higher scores indicate higher severity of symptoms.

  • Chi-square test corrected for continuity.

Menopausal symptoms, %   
 1. Heart beating quickly and strongly42.347.20.71
 2. Feeling tense or nervous82.769.80.19
 3. Difficulty in sleeping67.358.50.46
 4. Excitable51.960.40.50
 5. Attacks of panic13.524.50.22
 6. Difficulty in concentrating67.375.50.48
 7. Feeling tired or lacking in energy82.771.70.27
 8. Loss of interest in most things38.550.90.28
 9. Feeling unhappy or depressed69.260.40.46
 10. Crying spells44.241.50.93
 11. Irritability67.358.50.47
 12. Feeling dizzy or faint59.660.41.0
 13. Pressure or tightness in head or body69.254.70.18
 14. Parts of body feeling numb or tingling63.549.10.20
 15. Headaches75.064.20.32
 16. Muscle or joint pains76.973.60.87
 17. Loss of feeling in hands or feet19.220.81.0
 18. Breathing difficulties25.032.10.56
 19. Hot flashes71.262.30.45
 20. Sweating at night69.260.40.46
 21. Loss of interest in sex80.873.60.52
Menopausal factor scores, mean ± SD   
 Factor 1: psychological5.4 ± 3.65.3 ± 3.90.81
 Factor 2: subjective–somatic7.0 ± 4.16.0 ± 3.60.16
 Factor 3: organic–somatic2.2 ± 1.62.4 ± 1.90.63
 Factor 4: vasomotor2.6 ± 1.91.9 ± 1.70.06
 Factor 5: sensory–somatic1.2 ± 1.11.1 ± 1.40.58

Effects of menopausal hormone therapy.

Figure 1 depicts the menopausal factor scores at baseline and at each followup visit throughout the study. From the repeated-measures analysis included in Supplementary Table 3 (available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658), it can be seen that the vasomotor factor decreased significantly over time (P = 0.002) with differential patterns in relation to treatment group (P = 0.027); in the hormone therapy group, the reduction was more pronounced than in the placebo group, at between 1.5 and 2.0 versus between 0.35 and 0.8 points, respectively (on a scale of 0–6). The score reductions in both groups were observed since the first month of followup. Psychological, subjective–somatic, and organic–somatic factors also showed significant reductions along time (P < 0.001), but their patterns were similar with respect to treatment group (0.123 < P < 0.727). In these 3 factors, baseline scores decreased in both treatment groups since the first month of followup, but a tendency for returning to basal values was observed after 1 year. The sensory–somatic factor scores did not change significantly over time (P = 0.065), nor did the pattern differ between the treatment groups (P = 0.968). During the 2-year followup period, global mean scores for all of the factors except for subjective–somatic tended to be smaller in the hormone therapy than in the placebo group; however, the effect size of this treatment did not reach significance in any of the 5 factors. As mentioned in the previous section, in order to eliminate a possible spurious statistically significant time-by-treatment interaction, we repeated the above analysis excluding 1) baseline measurements, 2) baseline and first month measurements, 3) baseline and first 2 months measurements, and 4) baseline and first 3 months measurements. In the case of the vasomotor factor it was observed that the treatment effect (differences “treated” minus “placebo”) stabilized at approximately −0.4 scores but never reached statistical significance (0.077 < P < 0.221), i.e., the average score in the treated group for the vasomotor factor was 0.4 points below (not statistically significant) the corresponding average in the placebo group (see Supplementary Table 3, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658). A similar analysis undertaken in a subgroup of 48 patients with moderate to severe vasomotor symptoms at baseline (score ≥3) demonstrated that in this case, the hormone treatment produced a higher decrease in symptom severity, now with a significant difference between groups (hormone therapy minus placebo = −0.96, P = 0.02) (see Supplementary Table 4, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)2151-4658).

Figure 1.

Menopausal symptom factor scores (mean ± SD) in patients with systemic lupus erythematosus at baseline and throughout 24 months of treatment with either hormone therapy or placebo. No significant differences were observed between the 2 study groups in the factor scores at baseline. Vasomotor factor scores range from 0–6, psychological and subjective–somatic factors from 0–18, organic–somatic factors from 0–9, and sensory–somatic factors from 0–6. Higher scores indicate higher severity of symptoms.

Discontinuation rates and reasons.

Cumulative discontinuation rates at 2 years for all reasons (including loss to followup), medical reasons, and nonmedical reasons are shown in Table 4. Three women receiving hormone therapy and 1 receiving placebo discontinued the study due to thrombosis. One patient in each group died due to sepsis. However, neither this medical reason nor other withdrawal causes were significantly different between the study groups (Figure 2).

Table 4. Cumulative discontinuation rates and reasons by treatment group in the 106 women during 24 months of followup*
 Number of eventsRatesP
Hormone therapy (n = 52)Placebo (n = 54)Hormone therapy (n = 52)Placebo (n = 54)
  • *

    NA = not applicable.

All medical reasons1170.220.150.34
All nonmedical reasons480.090.160.26
Lost to followup01NA0.020.33
Total terminations15160.290.300.88
Total completing study3738   
Figure 2.

Cumulative discontinuation rates for all reasons and medical reasons in 106 patients with systemic lupus erythematosus at baseline and throughout 24 months of treatment with either hormone therapy (HT) or placebo. No significant differences were seen between the study groups for any withdrawal cause.

Side effects.

In addition to the medical reasons for premature termination of the study, few women reported adverse symptoms during the trial. Headaches, nausea, melasma, galactorrhea, and dysmenorrhea were presented in the 2 treatment groups, intermittently and at low frequency (≤6.0%). Mastalgia had a higher frequency among hormone users versus the placebo group at 1 and 6 months of treatment (10.2% and 13.33%, respectively; P < 0.03).

DISCUSSION

We evaluated the effects of a continuous–sequential estrogen plus progestin formulation in comparison to placebo on menopausal symptoms in women with lupus who were at the menopausal transition or early or late postmenopause. The menopausal symptoms assessed were those included in the Greene Climacteric Scale, a standardized instrument built on the results of 7 factor analytic studies undertaken in menopause clinics and in the general population (13). Since in lupus patients an overlap of symptoms presents during disease flare and menopause may occur, we considered it convenient to undertake a new factor analysis because of the possibility that symptom distribution was different than in the general population. Indeed, symptoms belonging to the psychological and somatic factors of the original scale were clustered differently in lupus patients. Interestingly, some symptoms that may be present in lupus flares, e.g., sleep disturbances, fatigue, difficulty in concentrating, and headaches, were clustered in a single factor, but such symptoms by themselves would not be confused with an increase in lupus disease activity by most rheumatologists. On the other hand, hot flashes and night sweats formed a small cluster on their own, totally independent from all other symptoms, as has been demonstrated in the general population (13) and in women with physical disabilities (23). To assess the effects of hormone therapy we measured each factor separately from each other and did not total them to yield a single score, since factorial analyses assume that factors have different etiology. Therefore, we believe that our study assessments were appropriate.

Our study population was highly symptomatic, more so than the Hispanic postmenopausal women enrolled in the Study of Women's Health Across the Nation (24). It can be explained because one-half of the women in our study were at the menopausal transition or early postmenopause, where menopausal symptoms are more prevalent (25), but other factors, i.e., ethnicity, health status, etc., cannot be ruled out (26).

Only vasomotor symptoms showed a sustained clinically significant improvement with estrogen plus progestin in comparison to placebo. In the subgroup of patients with moderate to severe vasomotor symptoms at baseline, the hormone treatment effect size was higher and reached statistical significance, supporting the notion that the absolute reduction in hot flashes is positively correlated with their frequency (27). We did not find an additional improvement with hormone therapy in the psychological, subjective–somatic, and organic–somatic factors scores, even though all decreased significantly over time. These findings are on line with those studies reporting the strong placebo effect on menopausal symptoms (28–30).

Regarding safety, we and other investigators have reported that hormone therapy did not affect the course of lupus activity at a clinically significant level, but as in the general population, increased the risk of thromboses (9, 31). No additional severe adverse events were found; however, the study was not adequately powered to detect differences in the incidence of cardiovascular diseases, cancer, etc., between groups. Among minor symptoms, only mastalgia was significantly more frequent in the hormone therapy than in the placebo group.

Because this is the first study of its class in women with lupus, no information is available for direct comparison. The hormone therapy arm of the Safety of Estrogens in Lupus Erythematosus National Assessment trial did not seek after the effects of menopausal hormone therapy on menopausal symptoms (31). Other randomized, placebo-controlled, clinical trials conducted in the general population cannot be compared with our study because they differ in several aspects. Indeed, most clinical trials on the effects of hormone therapy on menopausal symptoms have included healthy, non-Hispanic, older women. The Postmenopausal Estrogen/Progestin Intervention (PEPI) trial included only postmenopausal women whose symptom severity was not assessed (30). In the clinical trials designed to evaluate risks and benefits of the long-term administration of hormone therapy for primary or secondary prevention of chronic diseases in postmenopausal women, those with more severe symptoms were poorly represented (32–34). Lastly, the duration of our study was longer than most clinical trials of treatments for hot flashes, which have typically been small and brief. Only 3 (35–37) of 24 studies included in the meta-analysis published by MacLennan et al (27) lasted 24 months.

Our study has several limitations. We did not assess other health-related quality of life variables; therefore, we cannot define the effects of estrogen plus progestin on these measures in women with lupus. We used a continuous–sequential estrogen plus progestin formulation containing standard doses of synthetic steroids administered orally; therefore, our results should not be extrapolated to low-dose formulations or other routes of administration, nor can they be generalized to other regimens, although the PEPI trial did not find differences on the effects on menopausal symptoms across diverse hormone regimens (30). The study was underpowered to assess the risk for cardiovascular events, cancer, and other disorders associated with the use of estrogen plus progestin. It was conducted in a single center with limited ethnic variation among the patients, so one must be circumspect about extrapolating the results to all women with lupus.

On the other hand, this study possesses several strengths. We included relatively young women with a high prevalence of menopausal symptoms. The length of the trial extended over the recommended period of treatment, the participation rate was high, the premature discontinuation rates (especially that for loss of followup) were low, and the adherence to treatment was confirmed by estradiol measurements. We offered participation in the study to all women with lupus at different ovarian aging stages, including those with variable disease activity and those who tested positive for antiphospholipid antibodies, provided they had no history of recent thrombosis.

In conclusion, hormone therapy with estrogen plus progestin did improve vasomotor symptoms at a clinically significant level, but not other menopausal symptoms, during the menopausal transition or early or late postmenopausal stages in women with systemic lupus erythematosus. Whenever hormone therapy is prescribed to lupus patients to alleviate vasomotor symptoms, such benefit should be weighted with the potential risk of developing thrombosis, a severe side effect also observed among combined oral contraceptive users (38, 39). The prescription of hormone therapy with the purpose of solely treating psychosomatic symptoms in women with lupus is not supported by our data. Overall, we provide the lacking information that according to a lupus specialist's opinion is required to help lupus patients with menopausal symptoms decide about treatment (40, 41).

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Cravioto had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Cravioto, Sánchez-Guerrero.

Acquisition of data. Cravioto, Durand-Carbajal, Jiménez-Santana, Lara-Reyes, Sánchez-Guerrero.

Analysis and interpretation of data. Cravioto, Durand-Carbajal, Jiménez-Santana, Seuc, Sánchez-Guerrero.

Ancillary