Safety Assessment of Raloxifene Over Eight Years in a Clinical Trial Setting*

By Martino, Silvana; Disch, Damon; Dowsett, Sherie A; Keech, Cheryl A; Mershon, John L

Key words: Adverse event * Raloxifene * Safety * Trial

ABSTRACT

Objective: Osteoporosis is a chronic disorder that warrants long- term therapy. If benefits are to outweigh risks, the long-term safety profiles of these therapies must be favorable. The aim of this study was to assess the safety of raloxifene over 8 years in 4011 postmenopausal women with osteoporosis in a clinical trial setting through adverse event reporting.

Methods: Data analyzed comprised all reported adverse events collected at each visit of both the Multiple Outcomes of Raloxifene Evaluation (MORE) trial, and the subsequent Continuing Outcomes Relevant to Evistat (CORE) trial. MORE was an international, 4-year doubleblind, randomized, placebo-controlled study, designed to assess the effect of raloxifene on bone mineral density and vertebral fracture incidence in 7705 (placebo, 2576; raloxifene, 5129) postmenopausal women with osteoporosis. Breast cancer was a secondary endpoint. Based on the breast cancer findings of MORE, the CORE trial, a 4-year double-blind, placebo-controlled trial of a subset of MORE participants, was subsequently conducted. CORE enrolled 4011 (placebo, 1286; raloxifene, 2725) participants and was designed to examine raloxifene’s effect on breast cancer incidence. Safety analyses were performed using the intention-to-treat principle, and comparison between therapies was analyzed using a two- sided Fisher’s exact test.

Results: Over the 8 years of follow-up of 4011 women, there was no difference in all-cause mortality or hospitalization incidence between raloxifene and placebo groups (p> 0.1). Excluding breast cancer and non-melanoma skin cancer, cancer incidence was 4.6% and 6.3% in the raloxifene and placebo group, respectively (p= 0.027). Raloxifene was associated with a 1.7-fold increase in venous thromboembolism incidence (95% confidence interval 0.93-3.14), with an absolute risk difference of 0.9 per 1000 woman-years. There was no difference in the incidence of myocardial infarction, stroke, uterine cancer, endometrial hyperplasia, ovarian cancer or postmenopausal bleeding between the raloxifene and placebo treatment groups (p > 0.5). Uterine polyps, hot flushes and muscle cramps were more common in those receiving raloxifene versus placebo (p = 0.028, p < 0.001, and p = 0.008, respectively).

Conclusion: These 8-year data support the known clinical safety profile of raloxifene, established in the MORE trial.

Introduction

Raloxifene is a selective estrogen-receptor modulator (SERM) indicated for the treatment and prevention of osteoporosis in postmenopausal women. This treatment indication is based on the 3- year results from the Multiple Outcomes of Raloxifene Evaluation (MORE) trial1, a 4-year (3 years plus 1 year extension) international, double-blind randomized placebo-controlled study, designed to assess the effect of raloxifene (60mg/day or 120mg/day) on bone mineral density and vertebral fracture incidence in 7705 postmenopausal women with osteoporosis. In the first 3 years of MORE, raloxifene 60mg/day (Evista*) was associated with a 55% relative risk reduction and a 2.4% absolute risk reduction in new vertebral fracture risk in women without pre-existing vertebral fractures2,3. In those with pre-existing vertebral fracture, raloxifene 60 mg/day was associated with a 30% relative risk reduction and a 6.1 % absolute risk reduction1-3. Assessment of bone endpoints continued into the 4th year extension phase, although use of other bone-active agents was permitted during this phase. Over the 4-year period, raloxifene 60 mg/day was associated with a 36% reduction in vertebral fracture risk overall, and an absolute risk reduction 4.6%3. A further primary objective of MORE was to establish the safety of raloxifene. secondary objectives included assessment of the possible impact of raloxifene on cognitive and neuropsychomotor function, risk of breast and endometrial cancer, and risk of cardiovascular disease.

Based on the breast cancer findings of the MORE trial, a follow- up study, the Continuing Outcomes Relevant to Evista (CORE) trial, was conducted to determine the effect of an additional 4 years of raloxifene therapy on invasive breast cancer incidence in postmenopausal women with osteoporosis. Of the 7705 women participating in MORE, 4011 chose to enroll in CORE. The breast cancer findings based on the 3- and 4-year MORE data (N = 7705) the 4-year CORE data (N = 4011), as well as the 8-year MORE plus CORE data (N= 7705), have been previously published4-6.

Raloxifene safety data from the 4-year MORE trial, a 5-year osteoporosis prevention trial (N = 328), and other trials of lesser duration have been published. Common side effects reported to be associated with raloxifene therapy were hot flushes and leg cramps3,7-9. In these studies, raloxifene (60 mg/day or 120 mg/day) was not associated with an increase in incidence of uterine cancer, postmenopausal bleeding, endometrial hyperplasia8,10,11, ovarian cancer12, cognitive decline13, or arterial thrombotic events (including myocardial infarction and stroke)14. An increase in endometrial cavity fluid has been noted in postmenopausal women with osteoporosis receiving raloxifene for 4 years3. In the venous system, a serious adverse event associated with raloxifene therapy is a twofold increase in the risk of venous thromboembolic events compared with placebo (absolute rates of 3.5 and 1.7 events per 1000 women-years in raloxifene and placebo treatment groups, respectively)10. In the MORE trial, this risk was greatest in the first year10, or, more specifically, the first 4 months of therapy2. No dose effect (120mg/ day versus 60 mg/day) has been reported for these events related to raloxifene safety5,10,12,14.

Recently, the 4-year safety data from the CORE trial and 8-year safety data from the MORE plus CORE trials were reported in brief6. The purpose of the current analysis is to present the findings of a more detailed assessment of safety of raloxifene over the 8 years of MORE and CORE in the 4011 women who participated in both trials.

Patients and methods

Study participants

Each study protocol was approved by the ethical review board at each investigative site, and all women gave written informed consent for participation in MORE and CORE separately. Eligibility criteria for enrollment in the MORE trial have been previously published1. Briefly, women were < 80 years of age, were at least 2 years postmenopausal, and had documented osteoporosis as defined by a lumbar spine or femoral neck bone mineral density T-score of ≤- 2.5, and/or the presence of radiographically-apparent vertebral fracture. Women with a history of breast cancer, invasive endometrial cancer, or a history of stroke or venous thromboembolism (VTE) during the preceding 10 years were excluded. To ensure maximal enrollment in the CORE trial, the only eligibility requirement for CORE was to have been enrolled in the MORE trial. Women who were not willing to take the study drug in CORE were still eligible for participation.

Study design

The designs of both the MORE and CORE trials have been described elsewhere3’6. MORE was a 4-year trial (3 years plus 1 year extension), and CORE was a 4-year follow-up trial to the MORE trial.

The flow of participants from MORE into CORE is shown in Figure 1. Of the 180 original MORE investigative sites, 130 chose to participate in CORE. Overall, 4011 women chose to participate in CORE. The beginning of the CORE trial did not coincide with the end of the MORE trial. The median time between the end of participation in MORE and subsequent enrollment in CORE was 10.6 months (range = 2.6-62 months) for both treatment groups. Ninety-five percent of those assigned to raloxifene and 94% of those assigned to placebo had an interval between the two trials of less than 2 years. The mean time from randomization in MORE to end of participation in CORE, including the period between trials when prospective CORE participants did not receive study drug, was 7.8 years for each treatment group.

Figure 1. Flow of participants from the MORE into the CORE trial, and follow-up of enrollees in CORE

Study participation and treatment assignment

In MORE, participants were randomly assigned to placebo (N = 2576), raloxifene 60mg/day (N = 2557) or raloxifene 120 mg/day (N = 2572). CORE participants continued in the treatment group to which they were originally randomized in MORE, except that those women randomized to raloxifene 120mg/day in MORE were assigned to raloxifene 60mg/day (dose approved for treatment and prevention of osteoporosis) in CORE. Thus, in CORE, 1286 women were assigned to placebo, and 2725 to raloxifene.

Participants and investigators in MORE and CORE were blinded to treatment assignment unless they had been unblinded for reasons of safety. MORE participants, and CORE participants taking study drug were not permitted to take cholestyramine, raloxifene (unless as part of study protocol), tamoxifen, or systemic estrogen-containing preparations. During years 1-3 of MORE, participants were also not permitted to take bone-active agents.

Participants were discontinued from the MORE trial if they developed cancer of the breast, uterus or other malignancies considered to be estrogen-dependent, or experienced a VTE event. These women were no\t followed further, unless they chose to participate in CORE, for which they were still eligible. In CORE, these women were not permitted to take the study drug.

Safety data collection

The safety of raloxifene was evaluated through analyses of all adverse event (AE) data, following standard clinical trial procedures. AEs were collected at each visit of MORE (bi-annually) and CORE (yearly). Those AEs occurring during the period between the MORE and CORE trials were collected on Visit 1 of CORE. For those women discontinuing early from the MORE trial, events occurring since their last visit of MORE and prior to starting CORE were captured on Visit 1 of CORE. Treatment-emergent AEs were defined as those medical events (e.g., sign, symptom, test result) that had developed or worsened in intensity/frequency since entry into the MORE trial. Participants were questioned at each visit of MORE and CORE on the occurrence and nature of any AE since their previous visit and responses were recorded on standardized adverse event forms. Treatment-emergent serious AEs (i.e., those resulting in death, permanent disability or inpatient hospitalization, life threatening events, cancer, drug overdose, and events significant for any other reason e.g. VTE) were reported within 24 hours to the sponsor, who subsequently requested additional confirmatory data.

All AEs were classified using the Medical Dictionary for Regulatory Activities (MedDRA, International Federation of Pharmaceutical Manufacturers Associations*) version 6.0, the international standard medical terminology for AEs. MedDRA is designed to support the classification, retrieval, presentation, and communication of medical information throughout the medical product regulatory cycle and was developed under the auspices of the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH)15. Using MedDRA, events are classified at five hierarchical levels (Table 1). In the current analysis, the Preferred Terms (describing a single medical concept), or the High Level Terms (describing a group of Preferred Terms at a more general level) are reported. In certain analyses, Preferred Terms were grouped into a medically meaningful category as a single term (e.g., cardiovascular events).

Certain specific events pertaining to safety (see below) were also solicited at each visit during MORE and CORE. In the case of these solicited events, investigators were required to directly ask the study participants a series of questions to ascertain whether such an event had occurred. If participant responses warranted further questioning, this was performed at the investigator’s own discretion. In MORE, solicited events were cardiovascular (CV) (i.e. arterial) events, e.g. myocardial infarction (MI) and stroke, endometrial cancer, and breast cancer; In CORE, breast cancer, endometrial cancer, endometrial hyperplasia, postmenopausal bleeding, and VTE events were solicited, to permit more specific assessment of gynecological safety and VTE risk.

Table 1. Hierarchical structure of the Medical Dictionary for Regulatory Activities (MedDRA)

Mortality, cancer, and hospitalization

Since death of a woman who participated in both trials was only possible during the CORE period, mortality data presented are for the 4011 CORE participants during the 4-year CORE period. To account for biases that could occur as a result of this, mortality data for all 7705 women participating in MORE, whether they participated only in MORE or they continued in CORE are also presented. Statistically significant differences (p ≤ 0.05) between treatment groups in reason for death are reported at the MedDRA Preferred Term level.

Cancer data are reported as: (1) all cancers excluding non- melanoma skin cancers (i.e. excluding skin squamous cell and basal cell carcinoma); and (2) all cancers excluding breast cancer (primary endpoint of CORE trial) and non-melanoma skin cancers. Cancer and hospitalization data are those reported from the 4011 participants of both MORE and CORE over 8 years.

Venous thromboembolic events

VTEs included but were not limited to deep vein thrombosis, pulmonary embolism and retinal vein thrombosis. In both MORE and CORE, VTE events were collected through AE reporting as serious AEs. During the MORE trial period, VTE events were not solicited but medical records were requested whenever a participant reported a possible VTE during the course of the study. VTE events were then adjudicated by a panel of three physicians expert in the clinical management of VTE, independent of the sponsor and blinded to treatment assignment who classified each event as a definite, probable, or possible VTE following the guidelines of the World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception case definitions16. Events from MORE considered in the current analyses were limited to those that were classified as definite or probable. During the CORE trial, VTE events were solicited. Where VTE events were reported, confirmatory data were requested by the sponsor, but, unlike in MORE, events were not adjudicated, so that all VTE events reported in CORE were included in the analyses.

Cardiovascular safety

In the MORE trial, CV events were solicited, while in CORE CV events were collected through AE reporting. CV events were not a primary endpoint in either trial. An MI was defined as any event captured under the MedDRA Preferred Terms ‘Myocardial infarction’, ‘Acute myocardial infarction’, ‘Myocardial ischemia’, ‘Age indeterminate myocardial infarction’, ‘Coronary artery occlusion’, and ‘Coronary artery thrombosis’. A stroke was defined as any event captured as a MedDRA Preferred Term under the High Level Term ‘Central nervous system haemorrhages and cerebrovascular accidents’ or under the MedDRA Preferred Terms ‘Lacunar infarction’, and ‘Cerebral haematoma’.

Gynecologic safety

In both MORE and CORE, postmenopausal bleeding, endometrial cancer and endometrial hyperplasia were solicited AEs. Ovarian cancer was an unsolicited AE in both trials.

In MORE, gynecological examination was performed at baseline and was optional annually thereafter. Pelvic examination and transvaginal ultrasound were performed at baseline and annually thereafter in all participants with a uterus at 1 7 of the 180 centers, as well as in all participants exhibiting uterine bleeding. Women with uterine bleeding or abnormal uterine ultrasound findings were also entered into a clinical algorithm, to standardize evaluation based on clinical findings across study sites. Endometrial biopsy was recommended for women with bleeding, an endometrial thickness of more than 8mm as revealed by ultrasound, or an increase in endometrial thickness of at least 5mm. Reports of endometrial cancer were verified by a sponsor-employed gynecologist, blinded to treatment assignment.

In CORE, gynecological examinations were not specified in the protocol, and any women reporting bleeding received the appropriate local standard of care. The investigator recorded the final diagnosis at the participant’s next visit.

Common treatment-emergent adverse events and study discontinuation

Common treatment-emergent AEs (> 2% incidence overall, or in raloxifene treatment group alone) over the 8 years of MORE and CORE are reported using the MedDRA Preferred Term if statistically significant differences existed between treatment groups (p ≤ 0.05). In addition, common treatment-emergent AEs seen in the CORE period only are reported where statistically significant differences were evident between treatment groups. Treatment-emergent AEs leading to study discontinuation in CORE were also evaluated.

Statistical analysis

The majority of analyses were pre-specified. Analyses were performed using the intention-to-treat principle with participant data analyzed according to assigned treatment regardless of whether the participant took study drug, or other post baseline factors. All AEs volunteered by or solicited from participants were included in the analyses. The number and percent of participants experiencing an AE was determined by therapy and the comparison between therapies was statistically analyzed using a two-sided Fisher’s exact test. Data from women taking raloxifene 60mg/day and 120mg/day during MORE were combined since published data do not support a dose effect for specific safety events5,10,12,14, and women receiving raloxifene 120mg/day in MORE received 60mg/day in CORE.

Since in the MORE trial, raloxifene therapy was associated with an elevated early risk for VTE beyond the overall two-fold increased risk, a Poisson change-point analysis17,18 was performed to determine if and when (month since randomization) there was a change in the VTE event rate in the placebo and raloxifene treatment groups during CORE. This analysis was carried out using the Windows Bayesian inference Using Gibbs Sampling (WmBUGS) software version 1.3 (http://www.mrc-bsu. cam.ac.uk/bugs, MRC and Imperial College of Science, Technology and Medicine, London, UK; 2000)

Results

Of the 4011 women participating both in MORE and CORE, 1286 were assigned to placebo and 2725 to raloxifene. Over a mean of 7.8 years of follow-up, compliance to assigned therapy, defined as taking at least 80% of study drug, was 95% in each treatment group during the MORE period (raloxifene versus placebo; p = 0.74) and 55% during the CORE period (55% in the raloxifene treatment group, 56% in the placebo group; p = 0.63). Of the 4011 women participating both in MORE and CORE, 20.8% in the placebo group and 19.9% in the raloxifene treatment group never resumed the study drug in CORE either because they chose not to or were not permitted to due to diagnosis of an estrogendependent neoplasia, a history of VTE, ongoing treatment with cholestyramine, prescribed raloxifene, tamoxifen, systemic \estrogen/estrogen plus progestin therapy, or other oral reproductive hormone products during the CORE trial, or unblinding to study medication as a result of a safety concern during MORE.

Baseline characteristics

Baseline characteristics at the time of randomization in MORE for those women participating both in the MORE and CORE trials (N = 4011) are shown by treatment group in Table 2. There were no significant differences between treatment groups for any tabulated baseline characteristic, including age, prior systemic estrogen- containing preparations, smoking status, and diabetic status.

All cause mortality

Overall, 76 (1.9%) deaths occurred in the 4011 participants during CORE. There was no significant difference in mortality incidence between the raloxifene (1.7%) and placebo (2.3%) treatment groups (p = 0.27) (Table 3).. No significant differences were found between treatment groups in reason for death. Whether the participant continued in CORE or participated only in MORE, there were 176 deaths reported overall, with no significant difference in mortality incidence between the raloxifene (2.2%) and placebo (2.5%) treatment groups (p = 0.33).

Table 2. MORE baseline characteristics for the 4011 participants of MORE-CORE

Table 3. Summary of adverse outcomes for the 4011 participants of MORE-CORE over 8 years

Cancers

Over the 8-year period, 266 (6.6%) cancers, excluding non- melanoma skin cancer, were reported. The incidence in the raloxifene group was 5.7%, versus 8.6% in the placebo group (p = 0.001) (Table 3). Excluding breast cancer and non-melanoma skin cancer, 207 (5.2%) cancers were reported in the 4011 women over 8 years, with an incidence of 4.6% and 6.3% in the raloxifene and placebo groups, respectively {p = 0.027). In this latter case, there were no significant differences between treatment groups in the incidence of any specific cancer, as determined using Preferred Terms and High Level Terms (p > 0.1).

Hospitalization

Of the 4011 participants of both MORE and CORE, 1583 (39.5%) were hospitalized due to an AE at some point during the 8-year period. There was no significant difference in overall incidence of hospitalization between the raloxifene (38.8%) and placebo (40.9%) treatment groups (p = 0.21) (Table 3).

Venous thromboembolic events

Sixty of the 4011 participants experienced a VTE during the 8- year period. Twenty-five of these women reported their VTE during the MORE period. VTE data are summarized in Table 4. Compared with placebo, raloxifene therapy was associated with a 1.7-fold higher incidence of VTE over 8 years of treatment (95% confidence interval 0.93-3.14; p = 0.094). The absolute VTE rates in the raloxifene and placebo treatment group were 2.2 and 1.3 per 1000 women-years, respectively.

Raloxifene has previously been associated with greatest VTE risk within the first 4 months of initiation of therapy2. During the CORE period, no VTEs were reported in the first 4 months of re- initiation of therapy. A Poisson change point analysis was also performed to determine whether if and when there was a change in VTE rate during the CORE trial. Results of this analysis revealed that the VTE rate during this period remained relatively constant, with no evidence of an early increase in VTE risk after re-initiation of raloxifene therapy.

Cardiovascular safety

Overall, 108 women reported myocardial infarction during the 8- year period, with an incidence of 2.7% in the raloxifene treatment group and 2.6% in the placebo group (p > 0.99). In the case of stroke (N =110 women), the incidence in the raloxifene and placebo treatment groups was 2.9% and 2.5%, respectively (p = 0.54).

Table 4. Summary of venous thromboembolic event data for the 4011 participants of MORE-CORE over 8 years

Table 5. Summary of gynecologic adverse event data for the 4011 participants of MORE-CORE over 8 years

Gynecologic safety

Of the 4011 CORE participants, 3193 (79.6%) had an intact uterus (placebo, 1026 [79.8%]; raloxifene, 2167 [79.5%]). Gynecologic safety data over the 8 years are presented in Table 5. There was no significant difference in the incidence of uterine cancer, endometrial hyperplasia, ovarian cancer or postmenopausal bleeding between the raloxifene and placebo treatment groups (all p > 0.66). In none of the reported cases of endometrial cancer was there evidence of advanced stage disease (seven cases were stage I or lower, for three cases staging information was not provided), and no deaths as a result of endometrial cancer were reported. One uterine sarcoma was reported. This cancer occurred during the MORE period in a woman randomized to raloxifene 120mg/day.

Overall, 89 uterine polyps were reported, and the incidence was greater in the raloxifene treatment group versus placebo (3.2% versus 1.9%, respectively; p = 0.028). Of these 89 uterine polyps, 79 were reported in the MORE period alone (3.0% versus 1.4%; p = 0.005), while nine uterine polyps were reported in the CORE period alone (0.2% versus 0.4%; p = 0.48). All of the uterine polyps were benign. None of the uterine polyps identified in the raloxifene treatment group were associated with diagnosis of endometrial cancer. One participant in the placebo group had a concomitant diagnosis of a benign uterine polyp and endometrial cancer. Endometrial biopsy was also more common in the raloxifene treatment group versus placebo over the 8 years (6.8% versus 3.9%, respectively; p < 0.001) (Table 6). Of the 188 endometrial biopsies, 181 were reported in the MORE period (6.6% versus 3.6%; p < 0.001), while six were reported in the CORE period (0.1% versus 0.3%; p = 0.39).

There was no difference between the raloxifene and placebo treatment groups in the incidence of ‘Vulvovaginal signs and symptoms’ at the High Level Term (5.0% versus 5.8%; p = 0.26) (Table 5), or ‘Vaginal discharge’ at the Preferred Term level (1.6% versus 2.3%; p = 0.16).

Common treatment-emergent adverse events and study discontinuation

Over the 8 years of observation of the 4011 women in CORE, 2815 different AEs at the Preferred Term level were observed. In each treatment group, 99% reported at least one treatment-emergent AE (Table 3). Serious treatment-emergent AEs were reported in 42.3% of women receiving raloxifene and 45.5% of those receiving placebo (p = 0.07) (Table 3). Using the Preferred Term, there were 10 common treatment-emergent AEs with an incidence that was statistically significantly different (p < 0.05) between treatment groups over the 8 years of MORE and CORE. 'Excoriation' (skin abrasion), 'Increased blood cholesterol', 'Hypercholesterolemia', and 'Breast mass' were more frequently reported in the placebo group. Those six events reported more frequently in the raloxifene treatment group versus placebo were 'Influenza-like illness', 'Muscle cramp', 'Flushing', 'Biopsy Endometrium', 'Uterine polyp NOS', and 'Pollakiuria' (i.e., increased urinary frequency) (Table 6).

Table 6. Common (> 2% incidence overall, or in raloxifene treatment group alone) treatment emergent adverse events, with statistically significant greater incidence in the raloxifene treatment group versus placebo over the 8 years of MORE-CORE (p ≤ 0.05)

Hot flush (i.e., flushing) and muscle (i.e., leg) cramps data are shown by trial period in Figure 2A,B. The incidence of hot flushes during the MORE period in women receiving placebo was 6.3%, versus 0.9% in the CORE period. For muscle cramps, the incidence for the MORE period in women receiving placebo was 8.2%, versus 3.1% in the CORE period. Both hot flushes and muscle cramps were reported with a significantly greater frequency in the raloxifene treatment group versus placebo in the MORE period alone and in the MORE-CORE period (p < 0.01), but not in the CORE period alone (p > 0.5).

In the CORE period alone, there were 29 common treatment- emergent AEs at the Preferred Term level. Of these 29 AEs, there were two with a significantly different incidence between treatment groups. ‘Pneumonia’ was more commonly reported in the placebo group versus the raloxifene treatment group (3.1% versus 2.1%; p = 0.046) while ‘Depression’ was more commonly reported in the raloxifene treatment group versus placebo (4.2% versus 2.6%; p = 0.015). For depression, there was no significant difference between raloxifene and placebo treatment groups during the CORE period at the corresponding High Level Term, i.e., HLT ‘Depressive disorders’ (4.3% versus 3.0%; p = 0.052). There are four other Preferred Terms pertaining to depression that are not included under this High Level Term. These are ‘Major depressive disorder’, ‘Dysthymic disorder’, ‘Depressed mood’, and ‘Feeling of despair”, and all were more commonly reported in the placebo group. Over the 8 years of MORE and CORE, depression was reported in 10.9% of those receiving raloxifene, and 10.5% of those receiving placebo (p = 0.74), and depressive disorders in 11.1% and 10.7%, respectively (p = 0.71).

During the CORE period, 84 participants (2.1%) discontinued from the study due to an AE. There was no significant difference in the incidence of discontinuation between the raloxifene (1.9%) and placebo (2.4%) treatment groups (p = 0.35)(Table 3).

Figure 2. Incidence of (A) hot flushes and (B) leg cramps, reported as treatment emergent adverse events, in the MORE and CORE trial periods separately and over the 8 years of MORE-CORE, in those 4011 women participating both in MORE and CORE. For the CORE period, baseline was considered the first visit of CORE. For MORE and MORECORE periods, baseline was MORE first visit

Discussion

Osteoporosis is a chronic, progressive disorder that warrants long-term preventive or treatment therapy. Consequently, if benefits are to outweigh risks, the long-term safety profiles of these therapies must be favorable. Conditions that may take several years to become clinically-apparent, such as cancer and cardiovascular disease, are important considerations \in this regard. Previously published raloxifene safety data have been based primarily on the 4- year data from the 7705 women participating in the MORE trial. The results of the MORE trial and its follow-up trial, CORE, provide 8 years of safety data from the 4011 women participating in both trials.

Over the 4 years of CORE, and the combined 8 years of MORE plus CORE, there was no increase in mortality associated with raloxifene therapy. This is in agreement with the results of the MORE trial19. Raloxifene was associated with a reduced incidence of all cancers, even when breast cancer was excluded from the analysis. This reduction in cancer incidence (with breast cancer excluded] was not due to an effect of raloxifene on any specific cancer. The clinical significance of this finding, if any, remains to be determined.

Based on the results of the MORE trial, raloxifene use is associated with a statistically-significant two-fold increased risk for VTE events compared with placebo10. The 8-year MORE-CORE trial results presented here support these previous findings. Contrary to the findings of the MORE trial, the two-fold increase in VTE risk was not statistically significant in the 4011 women participating in both MORE and CORE. This may be due to fewer events over the 8 years in the 4011 women participating both in MORE and CORE, than in all 7705 women participating in MORE (62 over 8 years versus 73 over 4 years, respectively). Raloxifene should not be prescribed in those women with active or a past history of VTE. Further, it is advised that raloxifene treatment should be discontinued at least 72 hours prior to and during prolonged immobilization (e.g., postsurgical recovery, prolonged bed rest) with resumption of therapy only after the patient is fully ambulatory. Additionally, women taking raloxifene should be advised to move periodically during prolonged travel.

Since greatest VTE risk with raloxifene is reported to occur within the first year10 or, more specifically, the first 4 months of therapy initiation2, there was a potential concern that re- initiation of raloxifene therapy may result in a similar increased early VTE risk. However, in the current analysis of those 4011 women participating both in MORE and CORE, no VTE events were reported within the first 4 months of study drug re-initiation in CORE. It can be hypothesized that those women who were at increased risk for VTE at the beginning of MORE, for example as the result of the presence of predisposing genetic or environmental risk factors, would have been more likely to have a VTE events early in MORE and discontinue from the trial. One would expect these women would be less likely to participate in CORE, and, even if they did, would not have received raloxifene treatment. Indeed, only 34% [N = 25) of the 73 women who experienced a VTE event in MORE continued in CORE10, compared with the overall 52% of MORE participants who went on to participate in CORE. In addition, in the first year of MORE, temporary cessation of study drug during periods of immobilization was not part of the study protocol. Consequently, women immobilized during this time may have been at greater risk of VTE than in the remainder of MORE and in CORE.

No statistically significant difference in the incidence of MI or stroke between raloxifene and placebo treatment groups was evident. This is in agreement with the 4-year results of the MORE trial14, although events were defined differently between this 8-year and the MORE 4-year data analysis. CV events were not primary endpoints of either trial, and in CORE were only collected through AE reporting. The Raloxifene Use for The Heart (RUTH) clinical trial is an ongoing international, randomized, double-blind, placebocontrolled study of over 10000 postmenopausal women designed specifically to determine the effects of raloxifene on risk for major coronary events, as well as invasive breast cancer, in those at increased risk for or with established coronary heart disease20.

From a gynecological perspective, the 8-year findings from MORE- CORE are similar to those reported from MORE and other studies of shorter duration8,10-12. There was no increased incidence of ovarian cancer, uterine cancer, endometrial hyperplasia, or postmenopausal bleeding associated with raloxifene treatment. These uterine data support a neutral effect of raloxifene on the uterus. Uterine polyps were more common in the raloxifene treatment group over the 8-year period. They were more frequently reported during the MORE trial period versus the CORE trial period (79 versus nine events), presumably as a result of the uterine surveillance plan in place during MORE but not CORE. None of the uterine polyps diagnosed in the raloxifene treatment group were associated with an increase in reported bleeding or diagnosis of endometrial cancer over the 8 years, and most likely would not have been identified in a normal clinical setting, as demonstrated in CORE. The increased incidence of uterine polyps in the raloxifene group versus placebo during MORE may in part be related to an increased sensitivity of ultrasound in the active treatment group due to the presence of endometrial fluid. Raloxifene has previously been associated with an increase in endometrial cavity fluid3, and it is feasible that this could have aided detection of these benign endometrial changes. Moreover, it is presumed that the greater incidence of endometrial biopsy in the raloxifene treatment group during the MORE period is at least in part the result of this increase in detection of benign endometrial changes during ultrasound in the raloxifene treatment group.

During MORE and CORE, there were significant differences between treatment groups for 10 common treatment-emergent AEs. Differences between treatment groups for some of these treatment-emergent AEs are likely by chance, as a result of the multiple comparisons made. Pollakiuria (increased urinary frequency), for example, was reported more frequently in those receiving raloxifene versus placebo. However, raloxifene did not increase the incidence of any other urinary disorder, such as urinary tract infection or incontinence, and has not been associated with urinary tract dysfunction in other clinical trials21.

Of the 29 common treatment-emergent AEs that were reported during the CORE trial period, two events occurred at a significantly different incidence between treatment groups, with pneumonia more frequently reported in women assigned to placebo, and depression more frequently reported in those women assigned to raloxifene. These finding are likely to be the result of multiple comparisons. In the case of depression, there was no significant difference between treatment groups for other MedDRA dictionary terms referring to depression, and in some cases, a significantly greater number of events were reported in the placebo group. secondly, when the data were analyzed across 8 years of MORE-CORE, there was no significant difference in reporting of any of these depression-related AEs between treatment groups. It should be noted that no depression testing or psychiatric evaluations were performed in the CORE trial. Moreover, no additional clinical information was obtained on AEs reported as depression so that diagnosis could not be confirmed. There have been two studies conducted to specifically address the possible effects of raloxifene on mood and depression, utilizing validated instruments, the Hamilton Depression Rating Scale22 and the Geriatric Depression Scale23. In the first of these studies, raloxifene treatment resulted in a statistically significant improvement in depression and anxiety scores22, while the second study showed no difference between raloxifene and placebo in the scores for the Geriatric Depression Scale23.

Previously, raloxifene has been associated with an increased risk for both muscle (i.e., leg) cramps and hot flushes3,7. In the latter case, this increased risk seems to be within the first 6 months of raloxifene therapy initiation7. In the current 8-year analysis, again both muscle cramps and hot flushes were more common in those receiving raloxifene. However, these findings were a consequence of the increased incidence of these events in the raloxifene group versus placebo during the MORE rather than the CORE period. This is perhaps not surprising since one would expect that those with severe hot flushes or muscle cramps in MORE would have discontinued from MORE and chosen not to enroll in CORE. Moreover, women were older in CORE and thus at lower risk of hot flushes.

There are several limitations of this analysis. Although we have presented safety data based on 8 years experience with raloxifene in the clinical trial setting, there was a period between the MORE and CORE trials during which study drug was not prescribed. As a result, in the CORE trial women were reinitiating rather than continuing therapy. Moreover, 20% of women participating in CORE chose not to, or were not permitted to, take the study drug. AEs occurring between MORE and CORE were, however, captured at the first visit of CORE. Results presented here are based on analysis of data from participants receiving not only the recommended dose (60mg/day) of raloxifene but also those receiving raloxifene 120mg/day during the MORE period, although a dose effect for specific safety events discussed in this paper has not previously been demonstrated5,10,12,14. In some cases, data were collected differently between the MORE and CORE trials, for example VTE events were unsolicited and adjudicated in MORE but solicited and unadjudicated in CORE, and gynecologic examinations were performed as part of the MORE but not the CORE trial protocol. The AE data presented here were classified using MedDRA. While this is a standard coding terminology for AEs and used globally, it is not without its shortcomings24. For example there is n\ot always consistency in assigning a particular term to a sign/symptom, as already discussed in the case of depression. The limitations of the multiple comparisons have already been highlighted. As in every trial, there was likely some degree of selection bias in the MORE trial. Since the CORE trial entailed a further 4 years of study participation, the degree of selection bias can be expected to be even greater. That is, one might expect CORE participants to be healthier than those participating only in MORE and in the general population of a similar age. Moreover, an AE occurring during MORE may have resulted in the participant not taking study drug in CORE, introducing further bias.

Conclusion

No new clinically relevant safety concerns arose based on the 8- year results of MORE plus CORE trials. These 8-year data support the known clinical safety profile of raloxifene, established in the MORE trial. Over 8 years, raloxifene was not found to be associated with an increased risk for uterine cancer, endometrial hyperplasia, postmenopausal bleeding, ovarian cancer, MI or stroke, but was associated with a two-fold increased risk for VTE.

Acknowledgements

Declaration of interest: This study has been supported by Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN.

We are indebted to Messan Amewou-Atisso for statistical analysis, Lisa Houterloot and Bernard Tchoula for statistical programming, Mary Jane Geiger and David Cox for critical review of the manuscript. A list of MORE and CORE trial investigators have been previously published (Ettinger et al.1; Martino et al.6). The MORE and CORE trials were Eli Lilly sponsored trials.

* Evista is a registered trade name of Eli Lilly and Company, Indianapolis, USA

* MedDRA is a registered trademark of the International Federation of Pharmaceutical Manufacturers Associations

References

1. Ettinger B, Black DM, Mitlak BH, et al. Reduction of vertebral fracture risk in postmenopausal women with osteoporosis treated with raloxifene: Results from a 3-year randomized clinical trial. JAMA 1999;282:637-45

2. US prescribing information, Eli Lilly and Company, Indianapolis, IN, 2003

3. Delmas PD, Ensrud KE, Adachi JD, et al. Efficacy of raloxifene on vertebral fracture risk reduction in postmenopausal women with osteoporosis: four-year results from a randomized clinical trial. J Clin Endocrinol Metab 2002;87:3609-17

4. Cummings SR, Eckert S, Krueger K, et al. The effect of raloxifene on risk of breast cancer in postmenopausal women. Results from the MORE randomized trial. JAMA 1999:281;2189-97. Erratum in: JAMA 1999;282:2124

5. Cauley J, Norton L, Lippman ME, et al. Continued Breast Cancer Risk Reduction in Postmenopausal Women Treated with Raloxifene: 4- year Results from the MORE Trial. Breast Cancer Res Treat 2001;65:125-34

6. Martino S, Cauley JA, Barrett-Connor E, et al. Invasive breast cancer risk reduction in postmenopausal women with osteoporosis treated with raloxifene for 8 years: Results from the Continuing Outcomes Relevant to Evista trial. J Natl Cancer Inst 2004;96:1751- 61

7. Cohen FJ, Lu Y. Characterization of hot flashes reported by healthy postmenopausal women receiving raloxifene or placebo during osteoporosis prevention trials. Maturitas 2000;34:65-73

8. Jolly EE, Bjarnason NH, Neven P, et al. Prevention of osteoporosis and uterine effects in postmenopausal women taking raloxifene for 5 years. Menopause 2003; 10:337-44

9. Lu Y, Cohen F, Lakshmanan M. Characterization of hot flashes during raloxifene therapy: analyses of adverse event reports from an integrated placebo-controlled clinical trial database. Menopause 1998;5:282

10. Grady D, Ettinger B, Moscarelli E, et al. Safety and Adverse Effects Associated with Raloxifene: Multiple Outcomes of Raloxifene Evaluation. Obstet Gynecol 2004; 104:837-44

11. Cohen FJ, Watts S, Shah A, et al. Uterine effects of three- year raloxifene therapy in postmenopausal women younger than age 60. Obstet Gynecol 2000;95:104-10

12. Neven P, Goldstein SR, Ciaccia AV, et al. The effect of raloxifene on the incidence of ovarian cancer in postmenopausal women. Gynecol Oncol 2002;85:388-90

13. Yaffe K, Krueger K, Sarkar S, et al. Cognitive Function in Postmenopausal Women Treated with Raloxifene. N Engl J Med 2001;344:1207-13

14. Barrett-Connor E, Grady D, Sashegyi A, et al. Raloxifene and cardiovascular events in osteoporotic postmenopausal women: fouryear results from the MORE (Multiple Outcomes of Raloxifene Evaluation) randomized trial. JAMA 2002;287:847-57

15. International Conference on Harmonisation. International Conference on Harmonisation. 2004

16. A multinational case-control study of cardiovascular disease and steroid hormone contraceptives. Description and validation of methods. World Health Organization Collaborative Study of Cardiovascular and Steroid Hormone Contraception. J Clin Epidemiol 1995;48:1513-47

17. Carlin B, Gelfand A, Smith A. Hierarchical Bayesian analysis of change-point problems. J Roy Statistical Society, Series C 1992;41:389-405

18. Congdon P. Normal linear regression, general linear models and log-linear models. Bayesian statistical modeling. Wiley, 2001, pp. 118-21

19. Barrett-Connor E, Cauley JA, Kulkarni PM, et al. Risk- Benefit Profile for Raloxifene: 4-Year Data From the Multiple Outcomes of Raloxifene Evaluation (MORE) Randomized Trial. J Bone Miner Res 2004; 19:1270-5

20. Mosca L, Barrett-Connor E, Wenger NK, et al. Design and methods of the Raloxifene Use for The Heart (RUTH) study. Am J Cardiol 2001;88:392-5

21. Goldstein SR, Johnson S, Watts NB, et al. Incidence of urinary incontinence in postmenopausal women treated with raloxifene or estrogen. Menopause 2005; 12:160-4

22. Jarkova NB, Martenyi F, Masanauskaite D, et al. Mood effect of raloxifene in postmenopausal women. Maturitas 2002;42:71-5

23. Nickelsen T, Lufkin EG, Riggs BL, et al. III. Raloxifene hydrochloride, a selective estrogen receptor modulator: safety assessment of effects on cognitive function and mood in postmenopausal women. Psychoneuroendocrinology 1999; 24:115-28

24. Brown EG. Methods and pitfalls in searching drug safety databases utilising the Medical Dictionary for Regulatory Activities (MedDRA). Drug Safety 2003;26:145-58

CrossRef links are available in the online published version of this paper: http://www.cmrojournal.com

Paper CMRO-3034_4, Accepted for publication: 25 July 2005

Published Online: 19 August 2005

doi: 10.1185/030079905X61839

Silvana Martino(a), Damon Disch(b), Sherie A. Dowsett(b), Cheryl A. Keech(b) and John L. Mershon(b)

a The Angeles Clinic and Research Institute, Santa Monica, CA, USA

b Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN, USA

Address for correspondence: Silvana Martine, DO, The Angeles Clinic and Research Institute, 2001 Santa Monica Boulevard, Ste 560W, Santa Monica, CA 90404, USA. Tel.: + 1-310-582-7900; Fax: +1- 310-582-7979; email: smartino@cimg.theangelesclinic.org

* Some of these data (pertaining to VTE data, gynecologic safety) have been presented previously at the North American Menopause Society Annual meeting, Washington DC, 6-9 October 2004

[dagger] Evista is a registered trade name of Eli Lilly and Company, Indianapolis, USA

Copyright Librapharm Sep 2005