AVAILABLE FOR PUBLIC DISCLOSURE WITHOUT REDACTION
Genelabs Technologies, Inc.
505 Penobscot Drive
Redwood City, CA 94063

        The   Table of Contents

1   Table of Contents *
2    Executive Summary *
2.1 Background *
2.2 Efficacy *
2.3 Safety *
2.4 Conclusions *
3    Background *
3.1 Rationale for DHEA as a Therapy in SLE *
4    Development of GL701 for Systemic Lupus Erythematosus *
4.1 Clinical Studies Supporting Efficacy and Safety of GL701 in SLE *
4.2 Overview of Clinical Development Program for GL701 *
4.3 Pharmacology *
4.3.1 Background *
4.3.2 Pharmacokinetic Studies *
4.3.3 DHEA-S Levels from GL701 Controlled Clinical Studies *
5      Organization of Clinical Results *
6     Well controlled Studies *
6.1  Study GL94-01 (Corticosteroid Reduction Study) *
6.1.1 Protocol Development *
6.1.2 Entry Criteria *
6.1.3 Study Activities *
6.1.4 Patient Populations for Analysis *
6.1.5 Efficacy Variables *
6.1.6 Study Patient Population and Demography Results *
6.1.7 Dosing Information *
6.1.8 Efficacy Results *
6.1.9 SLE Scoring Instruments *
6.1.10 Study GL94-01 Conclusion *
6.2 Study GL95-02 (Improvement in SLE) *
6.2.1 Protocol Development *
6.2.2 Entry Criteria *
6.2.3 Study Activities *
6.2.4 Patient Populations for Analysis *
6.2.5 Primary Efficacy Variables *
6.2.6 Secondary Efficacy Variables *
6.2.7 Study Patient Population and Demography Results *
6.2.8 Patient Disposition *
6.2.9 Dosing Information *
6.2.10 Efficacy Results *
6.2.11 Study 95-02 Conclusion *
6.3 Well Controlled Studies, Foreign Source (Non-US IND) *
6.3.1 Study GBL96-01 (Disease improvement in SLE – Taiwan Study) *
6.3.2 Results *
6.3.3 Efficacy Results *
6.4 Published Studies Investigating DHEA in SLE (Stanford University Studies) *
6.5 Discussion of Efficacy *
6.5.1 Population Subsets *
6.5.2 Overview of Efficacy of GL701 in SLE *
7 Safety *
7.1 Organization of Safety Analyses *
7.1.1 Extent of Exposure *
7.1.2 Demographics *
7.2 Adverse Events *
7.2.1 All Adverse Events *
7.2.2 Severe Adverse Events *
7.2.3 Relationship of Adverse Events and Duration of Exposure to GL701 *
7.2.4 Early Termination from Study Drug *
7.2.5 Deaths *
7.2.6 Other Serious Adverse Events *
7.3 Clinical Laboratory Evaluation *
7.3.1 Hematology *
7.3.2 Liver Function Tests *
7.3.3 Renal Function Tests *
7.3.4 Serum Glucose *
7.3.5 Urinalysis *
7.3.6 Serum Complement, Anti DS DNA *
7.3.7 Serum Lipids *
7.3.8 Serum Hormone Levels *
7.4 Safety Issues of Potential Concern *
7.4.1 Acne and Hirsutism *
7.4.2 Hypertension *
7.4.3 Abdominal Pain *
7.4.4 Decreases in Serum Complement *
7.4.5 Serum Lipids *
7.4.6 Effects of GL701 on Sex Hormones *
7.5 Special populations *
7.5.1 Race *
7.5.2 Age, Sex, and Menopausal Status *
7.5.3 Pregnancy *
7.5.4 Drug Interactions *
7.5.5 Potential Interaction with Hydroxychloroquine for Serum Lipids *
7.6 Relationship of Dose to Safety *
8    Benefit/Risk Assessment *
8.1 Benefit *
8.2 Risks *
9    Bibliography *

1. Executive Summary
1. Background

Systemic lupus erythematosus (SLE) is a serious chronic, autoimmune, inflammatory disease that may affect the skin and joints, as well as internal organs and serous membranes. Approximately 65% of patients develop SLE between 16 and 55 years of age, and it is 8 to 10 times more common in women than in men.

Current therapies for active SLE are limited and include only hydroxychloroquine, corticosteroids, and immunosuppressive/cytotoxic drugs. Patients are exposed to multiple toxicities, many serious, during treatment with these drugs. Since patients are often dependent on steroids and immunosuppressive drugs, discontinuing them or reducing their dose may cause serious flares while continued use leads to multiple cumulative toxicities.

Although the etiology of lupus is unknown, hormonal influences seem to play a key role in disease development and progression. Dehydroepiandrosterone (DHEA) and its sulfated metabolite, DHEA-S, are the most abundant circulating steroid hormones in humans and are the principal androgens secreted by the adrenal gland. Circulating levels of DHEA and DHEA-S are reduced by approximately 50% in female SLE patients with active disease.

Two clinical studies conducted at Stanford University – one open-label and one double-blind, placebo-controlled - initially suggested that orally administered DHEA may improve manifestations of disease in mild to moderate SLE, reduce steroid requirements, reduce flares, and improve patient’s overall self-assessment of the status of her SLE.

The New Drug Application for GL701 (Genelabs’ formulation of prasterone [dehydroepiandrosterone, DHEA]) is based primarily on two double-blind placebo-controlled studies in women with SLE: Study GL94-01, which assessed the ability of GL701 to enable steroid reduction to 7.5 mg/day in steroid-dependent patients; and Study GL95-02, which assessed the effects of GL701 on overall signs and symptoms of SLE in patients with active disease. Long term safety data are provided from Study GL95-01, an open-label 1-year extension study. The application is also supported by a double-blind, placebo-controlled, non-US IND study conducted by a licensee in Taiwan (GBL96-01).

The GL701 clinical program has focused primarily on adult women with SLE inasmuch as approximately 90% of affected patients with this disease are women in their childbearing years.

Back to Table of Contents

2. Efficacy

In Study GL94-01, women with mild to moderate SLE were randomized to receive GL701 100 mg/day, GL701 200 mg/day or placebo for 7 to 9 months to determine whether GL701 would allow reduction of prednisone to < 7.5 mg/day for 2 consecutive months including the last visit while improving or maintaining disease activity (responders). Study patients had been treated with prednisone 10 - 30 mg/day and either a) in the last 12 months attempted to taper prednisone dose but failed and had a stable prednisone dose for at least 6 weeks preceding the study, or b) in whom there had been no attempt to taper in the last 12 months and had been on a stable prednisone dose for at least 3 months preceding the study.

Patients returned at monthly visits, at which time corticosteroid dose was reduced by algorithm: prednisone dose (or its equivalent in other corticosteroids) was reduced if disease activity was stable or improved as assessed by the SLE Disease Activity Index (SLEDAI), a composite score which measures overall disease activity in multiple organ systems.

GL94-01 was the first study of its type to investigate steroid sparing as an endpoint in a therapeutic trial for SLE. Given that there was no precedent for design of steroid sparing studies in SLE, investigator experts could not determine during the planning of this study whether patients with low SLEDAI scores (indicative of low disease activity) should be enrolled, since a low SLEDAI score could either reflect suppression of disease activity by exogenous corticosteroid treatment or relatively inactive disease. If the latter were true, patients might be unnecessarily treated with doses of corticosteroids higher than necessary for maintenance of disease suppression.

Because this issue was impossible to resolve without any prior SLE investigation precedent, an assessment of overall responders by baseline SLEDAI score was conducted while the study was blinded to treatment assignments and showed that patients with low SLEDAI scores (e.g., SLEDAI < 2) had much higher responder rates than patients who entered with higher SLEDAI scores. As a consequence, FDA agreed to defining a subgroup of patients with baseline SLEDAI >2 as part of the efficacy analyses.

The study population consisted of 191 patients. The percent of patients who achieved sustained reduction to 7.5 mg/d for at least the last 2 months of the study ("responders") were as follows: for analysis of all randomized patients (N=191) 40.6% for placebo, 44.4% for GL701 100 mg, and 54.7% for GL701 200 mg, (P=0.110, GL701 200 mg vs. placebo). For the patient group with baseline SLEDAI >2 (N=137), responders were 28.9% for placebo, 38.3% for GL701 100 mg, and 51.1% for GL701 200 mg (P=0.031, GL701 200 mg vs. placebo). In this group, there was a dose response relationship (P=0.033 for linear trend).

Additionally, a treatment effect was evident for the total number of days over the entire 7 month period during which daily prednisone dose was < 7.5 mg/day with mean (median) days for all patients showing placebo 71.7 (66.5) vs. GL701 200 mg 92.1 (111.5 days), P <0.069; and for patients with baseline SLEDAI >2, placebo 59.7 (28.0) days vs. GL701 200 mg 93.4 (111.0) days, P <0.013 GL701 200 mg vs. placebo.

In summary, in GL94-01 more GL701 patients than placebo patients were responders: i.e., achieved sustained prednisone reduction without worsening of signs and symptoms of SLE. This difference approached significance in the intent-to-treat patients and achieved significance in those with baseline SLEDAI > 2. There was a dose-response relationship, with the numbers of responders in the GL701 group being intermediate between GL701 200 mg and placebo. The difference in responders appeared early in the study and was maintained for the duration of the study. The mean number of days with a daily dose of prednisone < 7.5 mg was significantly greater in the GL701 200 mg group.

In the second study, Study GL95-02, women with mild to moderate active SLE were randomized to placebo or GL701 200 mg/day for 12 months. The primary efficacy objective of the study was to demonstrate improvement or stabilization in the disease and/or its symptoms in women with active SLE. The study population was mild to moderate SLE, defined as patients receiving either no prednisone or up to 10 mg/day (or its equivalent of other corticosteroids). Active disease was initially defined by a criterion of a Systemic Lupus Activity Measure (SLAM) score of ³ 7, but following the findings from the earlier GL94-01 study, the protocol was amended while ongoing and blinded to also require baseline SLEDAI > 2.

Concomitant medications, including NSAIDs, glucocorticoids, anti-malarials, and immunosuppressives were required to be stable for at least 6 weeks prior to enrollment and were to be held stable throughout the duration of the study.

The primary efficacy variable or endpoint was "response." The response was a per-patient endpoint that integrated the three domains of SLE including disease activity, as measured by the SLEDAI and SLAM; quality of life or constitutional symptoms, as measured by the Patient Visual Analog Scale (VAS) and Krupp Fatigue Severity Score (KFSS); and organ damage, as assessed by clinical deterioration. A responder had to demonstrate improvement or stabilization of each of the above four scoring instruments, and also not experience "clinical deterioration."

Because there was no precedent for this composite responder definition, which had not been previously validated in a clinical trial, it became clear that it would be difficult to quantify "stabilization" of disease. As a consequence, allowance for minor variability in the scoring instruments was identified and proposed for the definition of responders. Therefore, prior to unblinding the study, a "tolerance window" was defined to ensure that minor variability in these instruments would not confound the primary efficacy analysis for determining improvement and/or stabilization.

An additional efficacy variable defined toward the end of Study GL95-02 was "definite flare" which was defined to be consistent with flare descriptors for the ongoing NIH sponsored SELENA (The Safety of Estrogen in Lupus Erythematosus National Assessment ) study in women with lupus, which commenced after the start date of this ongoing study.

Bone mineral density at baseline and at 12 months was measured at 8 study sites on selected patients who had been on corticosteroids for at least 6 months prior to study entry.

Three hundred and eighty-one (381) patients were randomized to Study GL95-02, of whom 346 were in the per-protocol population, which was defined in the statistical analysis plan and which consisted of patients who had been treated for at least 60 days and had at least 1 post baseline assessment. Of these, 265 patients also had baseline SLEDAI >2. Of the 35 patients not meeting the criteria for the per-protocol population, 32 had no post-baseline assessments. The other 3 were excluded because of a major protocol violation (one patient) or receiving less than 60 days study drug (2 patients).

Study GL95-02 confirmed the findings of the earlier study, GL94-01, in that patients with no or minimal SLE activity (defined as SLEDAI < 2) should be viewed separately from those with baseline SLEDAI > 2 since a significant (P < 0.001) treatment interaction with baseline SLEDAI > 2 (yes/no) was noted.

Among the per-protocol population (N = 346), the percent of patients who were responders were 45.5% of 176 placebo patients vs. 58.2% 170 GL701 200 mg patients, P=0.018; in those with baseline SLEDAI >2, the percent of patients who were responders were 48.0% of placebo vs. 65.9% of GL701 200 mg, P=0.005. In the intent-to-treat population (N = 381), there were also more responders in the GL701 group compared to the placebo group, 51.3% of 189 GL701 patients were responders compared to 42.2% of 192 placebo patients (p=0.074); in those with baseline SLEDAI > 2 group, 58.5% of 147 GL701 patients were responders compared to 44.5% of 146 placebo patients (P=0.017)

Both treatment groups showed improvement in individual scoring instruments, but the mean improvement for the GL701 group was greater than the placebo group for each instrument. In particular, for patients with baseline SLEDAI >2, the improvement in Patient VAS was greater in the GL701 group (P=0.057 GL701 vs. placebo).

Fewer patients in the GL701 group with baseline SLEDAI >2 experienced a definite flare: 31 (23.5%) of 132 GL701 patients had at least one definite flare compared to versus 41 (30.8%) of 133 placebo patients, a trend favoring GL701, but the difference was not statistically significant (P=0.201).

At eight investigator sites, thirty-seven of the patients who had been receiving chronic corticosteroids (> 6 months) prior to enrollment underwent bone mineral density measurements by DEXA scanning at baseline and at 12 months. In these patients receiving chronic corticosteroid treatment, there was a significant decrease in bone density in the placebo group. By contrast, bone density increased in the GL701 group. The differences at 12 months between GL701 and placebo were seen most prominently in the spine, where the mean decrease of 1.78% in bone density in the placebo group is compared to a mean increase of 1.83% in the GL701 group (P=0.004).

With respect to the objective of the GL95-02 study, the GL701 group had approximately a 35% increase in the proportion of responders, and a 24% decrease in proportion of patients with definite flares .

Supportive data for efficacy are provided from a non-US IND study conducted in Taiwan by a licensee. This study was similar in design except that it was a 6-month study. Ninety-seven percent (97%) of the patients entered with baseline SLEDAI >2; also, 97% of the patients were receiving corticosteroids at study entry. The primary endpoint was percent improvement in SLAM.

The GL701 group demonstrated a greater reduction from baseline for SLAM score in both mean and median scores as compared to the placebo group, but the difference was not significant. The difference in mean change in patient VAS was statistically significant. The GL701 group improved, while the placebo group worsened (P=0.005). Utilizing the same flare definition as in Study GL95-02, the GL701 group in the Taiwan study had fewer patients with at least one definite flare. The number of patients with definite flares in the GL701 group was decreased 46.0% compared to placebo (18.3% vs. 33.9%, p = 0.044 based on time to first flare).

Thus, the overall results of the primary and secondary efficacy analyses for the Taiwan study consistently showed benefit for GL701 in the treatment of SLE.

In summary, consistent efficacy findings have been observed across all studies and have demonstrated improvement in all three domains of SLE in GL701-treated patients. Reduction in manifestations of SLE or improvement in the area of SLE disease activity was demonstrated by the increase in proportion of responders and decrease in proportion of patients with flares. Achievement of sustained reduction of corticosteroids in corticosteroid-dependent patients, without worsening of disease, as well as improvement in bone mineral density in patients receiving corticosteroids chronically, are important benefits for the domain of SLE damage. Finally, a benefit in the patients’ overall assessment of quality of life or constitutional symptoms was demonstrated by improvement in the Patient VAS, a finding which was consistently observed in these studies.

Back to Table of Contents

3. Safety

In the double-blind, placebo-controlled trials and the open-label extension study which followed completion of the double-blind studies, 387 women have received GL701 for at least 6 months; 242 for > 12 months, and 138 for > 18 months, and 36 for >24 months. The principal adverse events and biochemical changes associated with treatment with GL701 were principally related to its androgenic properties. No new unexpected safety findings emerged with continued treatment with GL701 for up to 2 years.

Adverse events that were each statistically significantly more common in GL701 200 mg-treated patients compared to placebo were acne (36.0% vs. 15.2%) and hirsutism (14.2% vs. 2.3%), as well as hypertension (7.9% vs. 2.7%) , hematuria (3.6% vs. 0.4%), and increased creatinine (2.4% vs. 0%). Acne and hirsutism were expected androgenic events; and the difference between placebo and GL701 was clinically meaningful as well as statistically significant. Hematuria, hypertension and creatinine increase occurred in only small numbers of patients. Their relationship to GL701 and their clinical significance wass less clear. These events were viewed in greater detail and are discussed in Section 7.4.

For hypertension, no differences between treatment groups were observed in a composite measure which integrated the number of patients experiencing new onset hypertension, or experiencing an increase in hypertension, or requiring an additional anti-hypertensive medication or an increase in dose of existing anti-hypertensive medications.

For the adverse events of "hematuria" or "creatinine increase," there was no consistent association with renal dysfunction.

Serious adverse events occurred in 38 GL701 200 mg, 7 GL701 100 mg, and 39 placebo patients. Only 3 of these events were assessed by the investigators as related to study drug: 2 in placebo and 1 in GL701.

For clinical laboratory assessments, there were no meaningful or significant differences between treatment groups in hematology or liver function parameters. BUN and creatinine levels did not change during study and were similar within or between treatment groups. Mean changes in 24-hour urine protein excretion increased in all treatment groups, but to a greater extent in GL701 patients. However, a few patients with very high values impacted 24-hour urine protein; and median changes were only slightly higher in the GL701 groups.

Decrease in serum complement changes, particularly C3, were more common among patients treated with GL701. Most patients remained within the normal range; 16% of patients on GL701 200 mg went from normal to low C3 (< 85 mg/dl), compared to 6% of placebo patients at last visit. Few of the patients who reduced complement exhibited other manifestations of renal lupus including worsening of proteinuria, new hematuria, or addition of new immunosuppressives; those that did occur were equally distributed between the GL701 and placebo-treated groups. An androgenic effect on hepatic complement synthesis is postulated as the cause for the reduced C3 complement rather than an effect on inflammation-related consumption.

Changes in serum lipids were consistent with changes associated with androgenic steroids. Total cholesterol, HDL-C, and triglycerides were decreased in GL701-treated patients in comparison to placebo patients (Figure 2-1). These changes were seen by three months with relatively little further progressive decrease. The mechanism of this androgenic-induced decrease in serum lipids is believed to be enhanced hepatic clearance of HDL-C and triglycerides.

Figure –1: Lipid Changes

The hormonal effects of GL701 were primarily androgenic in nature with significant increases in total testosterone, but primarily to levels at the upper range of normal for women. Serum estradiol did not change in pre-menopausal women, while in post-menopausal women, increases in serum estradiol were primarily to levels consistent with hormone replacement therapy. Only a few post-menopausal SLE patients achieved levels in the pre-menopausal range.

There were few reports of menstrual bleeding abnormalities either in pre- or post-menopausal patients suggesting there were no significant estrogenic effects on the endometrium. Additionally, mammography and uterine ultrasound measurements on some of the post-menopausal patients who had participated in these studies did not show evidence of neoplasia of the breast or hyperplasia or neoplasia of the endometrium.

There were 3 (1.5%) cases of breast cancer in 206 GL701-treated patients older than 45 years and 1 (4.2%) case in 24 placebo patients (never received GL701) of the same age group. Expressed as a rate per-patient year of observation for patients 45 years or older, these rates are 4.6/1000 for GL701 and 6.5/1000 for placebo. These are similar to breast cancer rates recently reported in the medical literature for post-menopausal women.

Taking this data as a whole, it would appear that increased risk, if any, of breast cancer associated with GL701 therapy should be low. GL701 therapy produces estradiol levels lower or similar to those with HRT and GL701 therapy is not accompanied by progestin therapy. The data from the GL701 clinical trials so far do not suggest an increased rate of breast cancer.

4. Conclusions

1. Background

Systemic lupus erythematosus (SLE) is a chronic, autoimmune, inflammatory disease that may affect the skin and joints, as well as internal organs, such as the heart, lungs, kidneys, spleen, nervous system and serous membranes lining the lungs, heart and abdominal cavity. Approximately 65% of patients develop SLE between 16 and 55 years of age, and it is 8 to 10 times more common in women than in men (American College of Rheumatology Ad Hoc Committee, 1999).

Although the etiology of lupus is unknown, hormonal influences seem to play a key role in disease development and progression. Beyond the increased incidence in women, several studies have noted that alterations in estrogen and androgen metabolism occur in patients with lupus. Decreased levels of androgens (androstenedione, dehydroepiandrosterone [DHEA], DHEA-S, and testosterone) have been observed in female lupus patients, especially in those with active disease (Lahita, 1987; Jungers, 1983).

Patients with mild to moderate symptoms are usually managed with administration of analgesics, nonsteroidal anti-inflammatory drugs (NSAIDs) and sunscreens. However, NSAIDs may reduce glomerular filtration rates and renal blood flow, cause gastrointestinal bleeding and can be associated with hepatotoxicity. If symptoms are not well controlled by these therapies, the patient's treatment may be augmented by the addition of an antimalarial drug, such as hydroxychloroquine (Plaquenil), although chloroquine and quinacrine are used less frequently. Toxicities with these agents include retinopathy with hydroxychloroquine and chloroquine, aplastic anemia with quinacrine, skin pigmentation changes, as well as development of peripheral neuropathy and myopathy with hydroxychloroquine only.

Most lupus patients do not respond to conservative therapy and require glucocorticoids for control of disease activity. In the Johns Hopkins Lupus Cohort, comprising of 539 patients, 89% had used prednisone, 21% of whom had been treated on one or more occasions with doses over 60 mg/day for at least 2 months. Most patients had used prednisone at doses above 10 mg/day (Zonana-Nacach, 2000).

Despite the fact that they are the mainstay of treatment for most lupus patients, glucocorticoids are well-known to be associated with significant toxicity including ischemic cardiovascular disease, serious infection, hyperglycemia, hypertension, osteoporosis, muscle wasting, avascular necrosis of bone, and cataracts (Zonana-Nacach, 2000). Female SLE patients are at a 5-fold risk of osteoporotic fractures (Ramsey-Goldman, 1999), and a 1.6-fold risk of ischemic necrosis (Zonana-Nacach, 2000). Patients dying early from lupus succumb to complications of lupus and/or infection while those dying after 5 years of disease often succumb to atherosclerotic complications, which may be related in part to chronic corticosteroid therapy (Urowitz, 1999, 2000).

Flares occur commonly among SLE patients. In the Johns Hopkins Lupus Cohort, the incidence of flare was 0.65 per patient-year of follow-up with the median time from first study visit to a flare being 12 months (Petri, 1991). While most flares involve "minor" organ systems, i.e. constitutional (fatigue) musculoskeletal, and cutaneous, it is of interest that in the Hopkins cohort, prednisone dose was increased in 39.7% of the flares (Petri, 1991). Furthermore, of 261 patients, 56.3% were hospitalized over a 2 year period from 1989 to 1990 (Petri, 1992). Prednisone dose over 10 mg/day was one of the risk factors for infection requiring hospitalization (P=0.04), as was use of immunosuppressive drugs (P=0.003).

Thus, an extremely important goal in the treatment of SLE is to reduce glucocorticoids to the lowest dose required to maintain suppression of SLE activity. However, during glucocorticoid taper, patients may experience symptoms of steroid withdrawal (e.g., joint pain, malaise) and the underlying disease may flare, thus perpetuating the need for treatment with glucocorticoids to control flare in this disease.

More aggressive management is often warranted in the treatment of lupus. In the Johns Hopkins lupus cohort alone, 34% of patients were using cytotoxic drugs (Zonana-Nacach, 2000). Immunosuppressive agents such as azathioprine (Imuran) and cyclophosphamide (Cytoxan) are used for patients with life-threatening or major organ system involvement. Toxicities associated with administration of azathioprine include leukopenia, hepatitis, pancreatitis, nausea and vomiting, and infections. Cyclophosphamide can result in urinary bladder toxicity, sterility, teratogenic effects, infections, and cancer.

Progression of SLE is highly variable and is difficult to predict from one individual to another. Lupus remains a serious disease with a 10-year mortality rate of approximately 10 to 20% (Uramoto, 1999).

Back to Table of Contents

1. Rationale for DHEA as a Therapy in SLE

DHEA is a naturally occurring steroid produced by the adrenal glands, testes, ovaries and brain. Its metabolite, DHEA sulfate (DHEA-S), is the most abundant circulating adrenal steroid in the human body and is converted by peripheral tissues containing DHEA sulfatases, including lymphocytes and macrophages, to DHEA. DHEA is in turn metabolized to androstenedione as well as other potent androgens: testosterone, dihydrotestosterone and estrogens, including estrone and estradiol. As a biologically inactive precursor, selective conversion of DHEA to other androgenic and estrogenic steroids enables metabolism on a tissue-specific basis, a concept known as "intracrinology" (Labrie, 1995).

Figure –1

 

Figure –1: Downstream Metabolites of DHEA

 

As described above, endogenous levels of androgens (androstenedione, DHEA, DHEA-S and testosterone) are decreased (approximately 50%) in women with SLE, especially in those with active disease (Lahita, 1987; Jungers, 1983).

The abnormalities contributing to low DHEA-S concentrations in this disease are poorly understood, but may be related to defects in the hypothalamic pituitary adrenal axis or direct suppression of adrenal androgen secretion by inflammatory cytokines including IL-1 and IL-6, the latter of which is significantly elevated in active SLE (Brink, 1999; Straub, 2000). DHEA has been shown in vitro to directly suppress release of IL-6, IL-1 and TNF-a from human mononuclear cells (Straub, 1998; Padgett, 1998) as well as IL-6 from bone marrow stromal cells (Gordon, 2000); whether these effects occur in vivo during treatment with DHEA, however, has not been determined.

 

 

Circulating levels of endogenous DHEA and DHEA-S, already low in active SLE disease, are further reduced during administration of glucocorticoids (Hedman, 1989).

PreclinicallyIn animal models of SLE, administration of androgens, including DHEA, resulted in delayed formation of anti double-stranded DNA antibodies (Siiteri, 1980; Roubinian, 1977; Roubinian, 1979a) and increased survival in hybrid NZB/NZW mice (Melez, 1980; Roubinian, 1979b, Lucas, 1985). Additionally, DHEA may be an important regulator of the immune system by up-regulating IL-2 secretion by activated T cells as demonstrated in both murine and human in vitro assays (Daynes, 1990a; Suzuki, 1991).

Two clinical studies conducted at Stanford University - one open-label and one double-blind, placebo-controlled¾ initially suggested that orally administered DHEA may improve manifestations of disease in mild to moderate SLE (van Vollenhoven, 1994, 1995). In the placebo-controlled study, concomitant doses of glucocorticoids were decreased, while disease activity as assessed by SLE Disease Activity Index (SLEDAI) score and patient and physician assessment (Visual Analog Scale [VAS]) stabilized or improved. Patient assessment of disease activity improved significantly in the DHEA group compared with placebo, and the number of disease flares (determined clinically by the investigator) experienced by the DHEA treatment group was significantly less than in the placebo treatment group (van Vollenhoven, 1995).

Based on the pre-clinical and clinical observations noted earlier, including the pilot studies at Stanford University, the development of GL701 (Genelabs’ formulation of prasterone [dehydroepiandrosterone, DHEA]) in SLE was initiated. GL701 is a pharmaceutical preparation of prasterone, and should be distinguished from DHEA currently marketed as dietary supplements, which are unregulated as to purity or potency, and may vary substantially in content (as much as 0 to 150% of labeled amount) (Parasampuria, 1998).

Back to Table of Contents

2. Development of GL701 for Systemic Lupus Erythematosus
1. Clinical Studies Supporting Efficacy and Safety of GL701 in SLE

1. GL701 vs placebo, Wilcoxon Rank-Sum test

1. Secondary Efficacy VariablesSLE Scoring Instruments

The secondary efficacy variables included change from baseline in the measurements of the following: SLEDAI, SF36, KFSS, Physician VAS and Patient VAS. In both groups measures of SLE activity remained stable. There was no significant change in any of the secondary variablescores, as well as no statistically significant differences between treatment differences groups for changes in any of these variables from baseline. As noted earlier, this was expected, as the main purpose of the trial was to reduce prednisone dose without causing worsening of disease. This was to be accomplished by a forced titration, based on the SLEDAI score at each visit. Therefore, only stabilization of disease rather than improvement, as measured by the different scores, would be expected; and the critical difference between treatment groups would be expected to be found in the proportion of patients successfully achieving steroid reduction.

Back to Table of Contents

2. Study GL94-01 Conclusion

1. Study GL95-02 (Improvement in SLE)

Study GL95-02 was a Phase III study which compared the proportion of patients achieving improvement or stabilization of the signs and symptoms of SLE. The treatment phase was 12 months.

Back to Table of Contents

1. Protocol Development

1. Primary Population and Handling Missing Values

The original protocol (January 31, 1996) had specified that "All randomized patients will be analyzed in an intent-to-treat analysis." However, how patients with missing values should be treated was not specified. This made the analysis of the primary endpoint, responder, ambiguous because the responder definition specified that for each of the 4 scoring instruments "the value of interest for each patient" would be the difference between the baseline means and "the mean of all values during on-treatment scheduled visits". If a patient did not return for at least her first visit, (not scheduled until after 90 days on treatment), and consequently had no post-baseline assessments, the protocol did not specify how she should be evaluated as a responder or non-responder. Accordingly, in an analysis plan submitted to FDA in October 1998, and discussed at a November 1998 meeting, Genelabs proposed that the population for analysis be defined as all randomized patients with at least one post-baseline measurement (per protocol population) and at least 60 days of study drug treatment.

Per FDA’s request, a more detailed analysis plan was submitted as an IND amendment in July 1999. The analysis plan specified that the primary population for analysis of efficacy would be the per-protocol population: those patients who had been on the study drug for more than 60 days and had had at least one post-baseline measurement beyond 60 days. An intent-to-treat population was defined as all randomized patients. The justification for requiring a minimum of > 60 days exposure to study drug for the primary analysis of efficacy was based on the fact that the first scheduled visit was at 90 days and the need for a minimal treatment period required for effectiveness. This minimal treatment period was discussed at a protocol planning meeting in September 1995, and eventually was included as part of the definition of clinical deterioration. In this proposal, those patients without post-baseline assessments would be excluded from the primary analysis of efficacy, consistent with the three criteria described in the final FDA guidance document "E9 Statistical Principles for Clinical Trials," Per-Protocol Set: (i) completion of certain pre-specified minimal exposure to treatment regimen; (ii) availability of the primary variable(s), and (iii) absence of any major protocol violations. However, in the intent-to-treat population, patients without post-baseline assessments would, by default, be considered as non-responders.

2. Handling Inter-Intrarater Variability in the Composite Endpoint

1. Well Controlled Studies, Foreign Source (Non-US IND)
1. Study GBL96-01 (Disease improvement in SLE – Taiwan Study)

This non-US IND study was conducted by a Genelabs licensee in Taiwan (Genelabs Biotechnologies Co., Ltd.). The study drug was designated by the licensee as "GL701", though it differed slightly from the Genelabs GL701. The active substance, prasterone, was identical to Genelabs’ GL701: the formulation differed in the amount of one of the excipients. The primary efficacy objective of the study was to demonstrate improvement in women with active SLE. The treatment phase was 6 months. Study GBL96-01 provides foreign source of supportive data from a double-blind placebo-controlled trial, and was conducted in accordance with Good Clinical Practice guidelines.

Back to Table of Contents

1. Study Design

GBL96-01 was a multicenter, randomized, parallel group, double-blind, placebo-controlled study. Patients were randomized to receive 200 mg/day GL701 or placebo for 6 months. The study population was restricted to women with mild to moderately active SLE, defined as patients receiving no prednisone or up to 10 mg/day (or its equivalent of other corticosteroids).

The study was composed of two periods: A Screening/Qualifying period and a double-blind placebo-controlled treatment period. The Screening/Qualifying period was a 7- to 10-day period in which patients had SLAM and SLEDAI measurements taken. SLAM determinations at both baseline visits had to be ³ 7 (excluding points assessed for ESR). Additionally, SLEDAI determination qualifying visit had to be > 2. The mean of both measurements was used as the baseline value.

Back to Table of Contents

2. Population for Analysis

The primary patient population for efficacy analysis was the intent-to-treat population defined in this study as all randomized patients with at least one post-baseline visit.

3. Primary and Secondary Efficacy Variables
1. Primary Endpoint

The primary endpoint was mean change from baseline in the SLAM at 24 weeks of therapy.

2. Secondary Endpoints

Secondary endpoints were similar to those in Study GL95-02 and included:

Definite flare, defined identically to definite flare in GL95-02,

Change in SLEDAI , Physician VAS, or Patient VAS score after 6 months

*All patients

**Per-protocol patients

 

¹A responder in GL94-01 was a patient who achieved sustained (at least 2 months) corticosteroid reduction to < 7.5 mg/day prednisone; a responder in GL95-02 was a patient whose SLEDAI, SLAM, KFSS and Patient VAS scores improved or stabilized over the study, and who had no clinical deterioration.