Disappointed but Undaunted: Pursuing New Avenues in Lupus Drug Development
The march of scientific progress usually goes as follows: failure, failure, failure, failure, success, failure, failure, ad infinitum. Progress toward effective drug therapies often follows a similar pattern; yet even among the host of difficult to treat autoimmune diseases, systemic lupus erythematosus has received a disproportionate share of therapeutic setbacks and disappointing clinical trials, particularly in recent years.
Over the last 50 years, systemic lupus erythematosus has garnered only a single FDA-approved therapy — belimumab (Benlysta, GlaxoSmithKline) in 2011 — and only narrowly, as its multicenter phase 2 trial failed to meet its primary endpoints. Since then, roughly 20 trials of various therapies have achieved suboptimal results, hampered by the much-chronicled heterogeneity of SLE, which frequently makes it a moving target for both clinicians and researchers.
However, the complications of the disease run deeper, with poor clinical trial design and ineffective inclusion criteria often cited as factors. From the outside, it may appear as though research has stagnated.
Joan T. Merrill, MD, director of clinical projects in the Arthritis & Clinical Immunology Program at the Oklahoma Medical Research Foundation and chief advisor for clinical development at the Lupus Foundation of America, suggested that it is not stagnation, per se. “There have been many phase 2 studies and several completed phase 3 trials of other agents since 2011 representing an investment in lupus of more than a billion dollars and a huge input of heart,” she said.
Current treatments for lupus are largely unapproved, inadequately studied, and poorly understood, Merrill noted. “They all have toxicities, either immediately or over time,” she said. “We know that sometimes these treatments work, but often they do not. Sometimes, even after these treatments work, they then stop working later on. We think perhaps the immune system has a way of ‘revving itself up’ again and getting around whatever roadblock a given treatment may put up to temporarily stop the inflammatory process. All of this needs to be better understood.”
Saira Sheikh, MD, director of the Clinical Trials Program at the Thurston Arthritis Research Center of the University of North Carolina at Chapel Hill, added more questions to the list. “The development of lupus is likely a complex interplay of environmental, hormonal and epigenetic factors in a genetically-susceptible individual that lead to the disease,” she said. “But since we don’t know what exactly causes it, we don’t know how to prevent it or cure it.”
Sheikh also suggested that there is a necessity to better understand the concept of preclinical autoimmunity. “This will help us identify individuals early who are at risk for the development of lupus,” she said in an interview. “We also need to understand why some patients experience disease progression while others do not.”
Despite the uncertainty, Susan M. Manzi, MD, MPH, medical director of the Lupus Foundation of America, and director of the Lupus Center of Excellence at Allegheny Health Network, is hopeful for the future. “There has been more activity in the lupus space than ever before, which is somewhat surprising considering all of the failures,” she said. “You would think people would get cold feet, but the truth is there are probably close to 50 different companies in this space, with more than 30 agents in the pipeline. That one win for belimumab has inspired people that they could win as well.”
However, the track record of failed clinical trials for lupus continues to undermine even “big pharma” endeavors; in August, AstraZeneca and MedImmune announced that anifrolumab — an experimental monoclonal antibody touted as the next contender in the lupus therapy market — failed to meet the primary endpoint of a statistically-significant reduction in disease activity among patients with SLE in the TULIP study.
Heterogeneity of Disease
The heterogeneity of SLE and failed trials go hand-in-hand, according to Richard A. Furie, MD, of the Division of Rheumatology at Northwell Health and the Zucker School of Medicine at Hofstra/Northwell. “Studies are segregated into the broad categories of lupus nephritis and extra-renal disease,” he said. “The study designs are different because the background medications and endpoints are different.”
The main issue is that “extra-renal disease” includes a wide variety of organ systems, each with their own complications, according to Daniel J. Wallace, MD, associate director of the Rheumatology Fellowship Program at Cedars Sinai Medical Center in Los Angeles. “Few of the failed trials were really organ-specific, which is problematic,” he said.
In an effort to address the lack of lupus therapies, in 2005, the FDA released a draft guidance with recommendations for the measurement of lupus disease activity and clinical outcomes, necessary reduction in disease activity and flares, treatment of organ-specific disease, surrogate markers as endpoints and risk-benefit assessment. Yet, even with specific guidelines to confront the challenges of trial design, trial failures continued.
There are a few ways to deal with these challenges and conduct a successful trial, according to Furie. One is to “take all comers” and use the same endpoints for everyone. “That is how the phase 3 belimumab program was designed,” he said. The alternative is to focus on specific disease domains, such as the mucocutaneous or musculoskeletal domains. “The latter, albeit more restrictive, is the logical approach from a mechanistic perspective.”
But Manzi stressed that scientists are not the only players in the game. “When you have a treatment only for skin symptoms, there is concern about the narrowness of marketing for such a small patient population,” she said. “But when the business side becomes involved in the decision-making, the narrow market becomes part of the problem.”
Furie noted another challenge to this focused approach. “Homogeneity is not always as homogeneous as we may think,” he said. “There are several different types of arthritis, and there is a range of cutaneous manifestations of lupus from acute cutaneous to chronic cutaneous.”
All of this makes it difficult to draw conclusions, even from successful studies, noted Furie. “In the phase 2 anifrolumab study I previously reported, the skin response was amazing, but there was no breakdown of the types of skin disease,” he said.
There may be hope, particularly due to increased understanding of B-cell and IFN pathways, according to Sheikh. “It is intuitive to think of lupus as a primarily B-cell mediated disease, but there is more to the story,” she said. “We know that B-cells produce autoantibodies, which are a hallmark of the disease and result in tissue injury in lupus, so that has been the focus. While targeting B-cell activating factor (BAFF) has been an effective therapeutic strategy, B-cell depleting therapies were tested and did not yield positive results in large clinical trials.”
Sheikh noted that it is now understood that both the innate and adaptive immune system play a role in lupus pathogenesis. “It is likely that the complex interplay between these two arms of the immune system results in the development of disease,” she said. “Distinct signaling pathways have been identified and we recognize the role of co-stimulatory molecules that promote activation and collaboration of T-cells and B-cells. We have an improved understanding of the role of dendritic cells in the development of lupus, and that immune complexes induce dendritic cells to produce proinflammatory cytokines such as type 1 interferons, which are recognized as key players in lupus pathogenesis.”
One Treatment Does Not Fit All
In spite of these advances in understanding, for Manzi, a wholesale rethinking of the approach to lupus may be necessary, with the cancer community seen as a prime example. “We currently treat lupus as lupus, but that is silly when you think about it,” she said. “You don’t treat cancer as cancer. You treat for lung or breast cancer, but it actually goes deeper than that. Not only do you treat by organ system, you treat by the molecular signature of the tumor within that organ system. Not all lupus is the same, just like not all breast cancer is the same.”
For Merrill, the fact that more and more people in the SLE community are thinking this way bodes well. “A hundred years from now, we will not be talking about which drug is better or which drug has a higher response rate,” she said. “We will be using good clean data to figure out which drug is best for an individual patient. But so far, all we have managed to accomplish with our increasingly sophisticated knowledge is cure a lot of mice.”
Deeper investigation of the failed trials may also be useful, noted Manzi. “If you do post-hoc analyses, you see that certain populations respond beautifully, while others don’t,” she said. “We need to understand that molecular signatures and ethnicity can play huge roles. We can be so much more targeted than we are currently.”
Sheikh added another component to the discussion. “It is likely that each of these pathways is involved in different phases of the disease, so perhaps the key will be a combination of targeted therapies at different time points in the disease,” she said.
Clinical Trial Design
Although casting too wide a net for treatment efficacy — especially in a disease known for its widespread symptoms — is often cited as the biggest hindrance to true progress in SLE research, the flawed design of clinical trials has also hampered the production of meaningful results, Sheikh said.
“Trials are often too long or perhaps, in some cases, not long enough to detect a difference between treatment groups, and we are still trying to figure out if we have the correct endpoints and outcome measures,” she said. “Lupus is a very heterogeneous disease, and patients with severe disease, particularly affecting the kidneys and central nervous system, are often excluded from studies.”
Merrill and colleagues echoed these sentiments in a white paper published in Lupus Science and Medicine, in which they noted that the current standard of lupus disease control is unacceptable, and smaller and shorter drug trials are required to develop more effective treatments. In their proposal to the FDA, which includes recommendations for overcoming barriers obstructing lupus treatment development, the researchers identified notable issues such as the “unwieldy size” of current trials, combined with global competition for patients “who meet the stringent entry criteria at limited numbers of adequately trained trial sites.”
“The main themes address the fact that lupus is a complicated, heterogenous disease so we need robust trial designs that strongly discriminate between an effective treatment and placebo,” Merrill told Healio Rheumatology. “Some of the major proposals are to design trials that lower placebo group responses either by including only the more acutely ill patients or else decreasing, or at least simplifying, background medications. Outcome measures must also be more robust. Single organ outcomes, achievement of very low disease activity and measurements of sustained response are all approaches that can increase the difference between effective treatments and placebo.”
Allen Anandarajah, MD, MS, associate professor of allergy, immunology and rheumatology at the University of Rochester Medical Center, agreed. “Ethical reasons prevent one from using a true placebo arm,” he said in an interview. “This, however, makes it hard to assess the benefits of the trial medication vs. the background medications. Patients also often need courses of steroids for flares during clinical trials which, again, can make interpretation of results confusing.”
Selection of treatments, and consideration of how treatments interact with each other, is the next component, according to Merrill. “If you put two immune-modulating drugs together with no rationale for it — which is our current standard of care — they may have added benefits, but there is also a very good chance they will be redundant to each other or even inhibit each other,” she said. “There may also be added toxicities that, sadly, we have the technology, but not the research funding, to predict.”
Anandarajah suggested that one way of improving clinical trial design is to reassess, if not overhaul, trial endpoints. “Until the last decade, the lack of well-defined outcome measures made it hard to ascertain the benefits of some medications,” he said. “The choice of a primary endpoint remains a source of contention, even in recent trials.”
Furie and colleagues conducted a post-hoc analysis of pooled data from two phase 2b trials of sifalimumab (MedImmune) and anifrolumab to further determine the clinical meaningfulness of achieving an SRI (SLE Responder Index-4 (SRI-4)) response. Results supported findings from a previously published analysis of the phase 3 belimumab data; significantly greater improvements in global clinical benefit were associated with an SRI-4 response.
“We don’t do a good job in terms of defining endpoints,” Furie said. “Patients can be described as complete responder, partial responder or non-responder, but the definition of what that means can vary from study to study.”
Morand and colleagues aimed to validate lupus low disease activity state (LLDAS) as an endpoint in a post-hoc analysis of data from a cohort of 102 patients treated with anifrolumab or placebo. Results at 52 weeks showed that LLDAS was associated with an SRI-4 and British Isles Lupus Assessment Group-based Composite Lupus Assessment (BICLA) response (P < .001).
There is hope that all of these outcome metrics could become a thing of the past, Wallace told Healio Rheumatology. “Committees are working on putting together better metrics for studying lupus in clinical trials,” he said, and referenced his work with the Lupus Multivariate Outcome Score (LuMOS), which aims to create criteria similar to ACR 20/50/70 criteria. “For the first time, pharma is sharing documents from failed phase 2 and 3 trials which may be helpful to create more useful prospective measures.”
Anifrolumab and the IFN Pathway
Data from the phase 2 anifrolumab study boosted anifrolumab to the status of the most promising agent in the pipeline for extra-renal disease. In their study published in Arthritis & Rheumatology, Furie and colleagues studied 305 patients treated with IV anifrolumab at 300 mg or 1,000 mg, or placebo, plus standard of care, for 48 weeks. Results showed that in addition to besting placebo in terms of the primary endpoint of SLEDAI 2000 score, patients with a high IFN signature at baseline demonstrated an improved response (13.2% for placebo vs. 36% [P = .004] for 300 mg and 28.2% [P = .029] for 1,000 mg anifrolumab).
“However, the recent failure of one of the two phase 3 studies — TULIP 1 — once again brings into question, was it the drug, the study design, or the study execution that was responsible for its failure?” Furie said.
He added that it may soon be possible for practitioners to stratify patients by high or low IFN signature at baseline. “We will be able to tell what patients are likely to respond to a particular therapy that targets that pathway,” he said.
However, it may not be that easy, Manzi told Healio Rheumatology. “Most of us recognize that anifrolumab is effective in patients with high interferon signatures,” she said. “But we are going to have to prove to insurance companies that the patient has a high interferon signature before we can use the drug, and this will present another set of challenges.”
Regarding JAK inhibitors, Wallace and colleagues conducted a phase 2 trial in 105 patients treated with baricitinib (Olumiant, Lilly) 2 mg, 104 treated with baricitinib 4 mg, and 105 patients treated with placebo. Results at 24 weeks showed that 67% of patients in the 4-mg arm experienced resolution of SLEDAI-2000 arthritis or rash (P = .0414 vs. placebo), while 58% of those in the 2-mg arm reached this result (P = .39). “The baricitinib 4-mg dose, but not the 2-mg dose, significantly improved the signs and symptoms of active systemic lupus erythematosus in patients who were not adequately controlled despite standard of care therapy,” the researchers concluded.
“This is more good news for lupus,” Wallace said. “This small-molecule agent showed that a higher dose could lead to improved outcomes.”
“Targeting kinases to interfere with interferon is another route of interest,” Anandarajah said, adding that tofacitinib (Xeljanz, Pfizer) is also working through phase 2 in SLE.
Deeper Into the Pipeline
Van Vollenhoven and colleagues looked at the IL-12/23 pathway in a phase 2 study of ustekinumab (Stelara, Janssen). The analysis included 102 patients randomly assigned the study drug or placebo. Results at 24 weeks showed a 60% SRI-4 response rate among patients treated with ustekinumab, compared with 31% for placebo (P = .0046).
“This is an agent with very exciting preliminary results in phase 2 trials, as well,” Manzi said. “They are moving forward with phase 3 studies.”
Anandarajah said that the depletion of B cells is a key area of research in SLE. “Phase 3 clinical trials have been completed for the following medications, but their use in SLE is still not clear: rituximab (Rituxan, Genentech); ocrelizumab (Ocrevus, Genentech), an anti-CD20 antibody; and epratuzumab (LymphoCide), an anti-CD22 humanized antibody.”
“Conventional therapies do not affect memory plasma cells, which therefore continue to secrete antibodies that maintain disease state,” Anandarajah said. “Recently, therapies that inhibit proteasomes have been shown to deplete the level of these memory plasma cells and are therefore of interest in the treatment of SLE. These medications have been shown to be effective for treatment of other diseases such as multiple myeloma. Ixazomib (Ninlaro, Takeda) is a drug in this class undergoing phase 1 clinical trials.”
Therapies that target the crosstalk between T-cells and B-cells are another target for management of SLE, according to Anandarajah, who noted that abatacept (Orencia, Bristol-Myers Squibb) is one such drug that has completed phase 2 trials.
In Best Practice & Research: Clinical Rheumatology, Mok wrote an overview of calcineurin inhibitors (CNIs), suggesting that tacrolimus may be preferable to cyclosporine due to reduced cosmetic, hypertensive and dyslipidemic events. He suggested that tacrolimus has demonstrated non-inferiority to mycophenolate mofetil (MMF) or cyclophosphamide as an induction regimen for lupus nephritis, while low-dose tacrolimus plus MMF may also be preferable to cyclosporine in this patient population.
“However, the therapeutic window of [tacrolimus] is narrow, and drug level monitoring is required to ensure drug exposure and minimize toxicities,” he wrote, suggesting that safety beyond 6 months remains uncertain. “Newer chemical analogues of CNIs, such as voclosporin, with less variable plasma concentrations are being tested in lupus nephritis.”
“The biggest unmet need in SLE is lupus nephritis,” Furie said. “We don’t do a good job in this disease, with a complete response rate around 35%. We haven’t been able to climb above that with any experimental therapies.”
Furie suggested that the phase 2 study with voclosporin provides hope that the complete response rate can be challenged.
For Manzi, moving the ball forward is a matter of getting the clinical, research, pharma and patient communities on the same page. “The industry needs to be involved because they’re developing the drugs,” she said. “Those of us in the biomarker world need to share our findings about segmenting patients in terms of genetic and molecular signatures. We need to come together with patients and clinicians and the FDA. We have a crisis on our hands.”
With this in mind, the field demands a transition in thinking, particularly in terms of clinical design and execution, according to Furie. “The path of least resistance is to continue doing what we have done, because changing will require a slowdown in drug development until we re-engineer our design,” he said. “It will take time, and we probably will end up with a hybrid of slowing down our current approach and re-engineering a new approach, but this is necessary.”
Merrill echoed these sentiments but is hopeful that enough people recognize the issues at hand and the need to rethink the approach to SLE research. “We aren’t there yet, but that is where it is going,” she said. – by Rob Volansky
- Furie R, et al. Arthritis Rheumatol. 2017;doi: 10.1002/art.39962
- Furie R, et al. Arthritis Rheumatol. 2018;doi:10.1177/0961203318758506.
- Merrill JT, et al. Lupus Sci Med. 2018; doi: 10.1136/lupus-2018-000258.
- Mok CC. Best Pract Res Clin Rheumatol. 2017;doi: 10.1016/j.berh.2017.09.010.
- Morand EF, et al. Ann Rheum Dis. 2018;doi:10.1136/annrheumdis-2017-212504.
- Van Vollenhoven R, et al. Abstract #6L. Presented at: ACR 2017.
- Wallace DJ, et al. Lancet. 2018;doi: 10.1016/S0140-6736(18)31363-1.
- For more information:
- Allen Anandarajah, MD, MS, can be reached at 601 Elmwood Avenue, Box 695, Rochester, NY 14642; email: Allen_Anandarajah@URMC.Rochester.edu.
- Richard A. Furie, MD, can be reached at 865 Northern Boulevard, Suite 302, Great Neck, NY 11021; email: RFurie@northwell.edu.
- Susan M. Manzi, MD, MPH, can be reached at Highmark Health, Attn: Candace Herrington, 120 Fifth Avenue, Suite 1122, Pittsburgh, PA 15222; email: Candace.Herrington@highmarkhealth.org.
- Joan T. Merrill, MD, can be reached at 825 NE 13th St., Oklahoma City, OK 73104; email: Joan-Merrill@omrf.org.
- Saira Sheikh, MD, can be reached at 3300 Thurston Building, CB 7280, Chapel Hill, NC 27599-7280.
- Daniel J. Wallace, MD, can be reached at 8750 Wilshire Boulevard, Suite 350, Beverly Hills, CA 90211; email: email@example.com.
Disclosures: Anandarajah, Manzi and Sheikh report no relevant financial disclosures. Furie reports consulting for AstraZeneca/MedImmune and GlaxoSmithKline. Merrill reports consulting relationships with Amgen, Astellas, AstraZeneca, Bristol-Myers Squibb, Celgene, EMD Serono, GlaxoSmithKline, Glenmark, Idorsia, Iltoo, ImmuPharma, Incyte, Janssen, Lilly, OncoMed, Pfizer, RemeGen and Sanofi. Wallace reports consulting for GlaxoSmithKline, Lilly, Merck, Pfizer, and Serono.