Why, why, why de-lupus (does so badly in clinical trials)

There is abundant evidence that B-lymphocytes play a major role in the development of lupus (reviewed in [1]). It thus seemed logical with the introduction of rituximab which blocks the CD20 molecule on many, but not all, B-lymphocytes, for the treatment of non-Hodgkin’s lymphoma, that this partly humanized monoclonal antibody should be useful for lupus patients. Frustratingly, in spite of a host of reports (summarized in [2]) from many different centers describing significant success using B-cell depletion to treat a wide range of lupus features, the two key clinical trials (LUNAR [3] and EXPLORER [4]) did not meet their primary end points. As has often been the case, however, post-hoc analyses did give some encouragement indicating that improvements in clinical findings, and pertinent biomarkers, including anti-dsDNA antibody and serum C3 levels had occurred.

Blocking another B-cell-associated molecule CD22 using the monoclonal antibody epratuzumab gave encouraging results in a Phase II trial.[5] In spite of this encouragement, it has recently been reported that epratuzumab failed to meet its end points in two large-scale Phase III clinical trials (http://www.reuters.com/article/2015/07/28/ucb-lupus-idUSL5N1080XI20150728).The data from these trials as well as any exploratory insights are not yet in the public domain.

Much better news was evident when both the BLISS 52 [6] and BLISS 76 [7] trials of an antibody, belimumab, that blocks the B-cell activating factor known as BLyS or BAFF, had met their primary end points and at least some of the secondary end points. An important caveat is that the most common features of patients in large Phase III trials are skin and joint manifestations. We await confirmation that this approach is effective for less common aspects of lupus. It has, however, become evident that this drug does not work rapidly and is thus of little immediate use in the acutely ill lupus patient.

In contrast to the success of belimumab, it was disappointing that Eli-Lilly have withdrawn tabalumab (another anti-BLyS antibody) from further studies after two international pivotal trials [8] were not deemed sufficiently successful to warrant further development.[9] More encouragingly, atacicept which blocks two B-cell activating factors (BLyS and APRIL) has shown significant promise in a flare prevention study.[10] Unfortunately, the study Safety Committee recommended the withdrawal of the high-dose arm of this trial because two patients in this arm died of infection. However, an intention to treat analysis on those patients in the high-dose arm of the study who did complete follow-up showed that the drug both increased the time to first flare and reduced number of flares compared to the placebo and lower-dose arm. Given the limited options for very ill patients with SLE, and the already high risk for serious infections in this population, further study of this dose in carefully monitored patients may be warranted.

Excitement in recent years about the potential importance of the Interferon-alpha signature in patients with lupus supported the prospect that inhibitory antibodies to interferon-alpha would be a promising approach. Sifalimumab has provided an encouraging result in a Phase IIb trial.[11] In contrast, rontalizumab did not meet its primary end point [12] and, paradoxically, patients with lower Interferon-alpha gene signature did better than those with the higher signature. Genentech discontinued development of rontalizumab (http://adisinsight.springer.com/drugs/800026240). MedImmune/Astra Zeneca has announced that it will not continue to develop sifalimumab, but will focus on a Type I interferon receptor inhibitor, anifrolumab, which met its primary end point in a Phase 2 trial, although the details of these results are not yet released (http://www.healio.com/rheumatology/lupus/news/online/%7B43eccff7-f738-4569-9c95-).

It is thus clear on balance that we are a long way from the green, green grass of homecare delivery of biologic drugs widely available for patients with lupus, analogous to the situation for patients with rheumatoid arthritis.

We and others have previously considered some aspects of why lupus clinical trials fail.[13,14] Lupus is clearly most unlikely to arise by a single pathogenic mechanism and it seems equally likely that different factors are at play in the “conspiracy” that gives rise to disease in the central nervous system as opposed to the skin or kidney. Furthermore, we have no reason to believe that every biologic drug will be equally effective against the diversity of pathogenic mechanisms. Other contributory factors include the concomitant prescribing of excessive doses of corticosteroids and/or immunosuppressive drugs which confuse interpretation in trials; inadequate appreciation by clinical investigators who are scoring disease activity that not every clinical feature in a patient with lupus is due to active disease (it may be due to damage or a comorbid condition) and inadequate training of those individuals participating in the trials in the use of the well-known disease activity assessment tools such as SLEDAI-2000 and BILAG-2004 so that even if there is real lupus activity the severity may not be assessed properly.

Given the bleak clinical trial results, should the lupus community (and by inference the biotechnology companies) conclude that there is no real hope of a biological drug being meaningfully effective in lupus? Should our main target be simply to try and improve the use of conventional steroids and our range of unproven and unapproved immunosuppressive therapies?

We beg to differ. While we accept that the challenges involved in setting up expensive clinical trials remain daunting, the unmet clinical need and potential market for these drugs is equally formidable. Data exist that indicate that even the most skillful optimization of steroids and immunosuppressive drugs in renal lupus [15] is accompanied by the persistence of unacceptable morbidity and mortality figures. Systemic lupus is a disease that primarily affects young women, and approximately 15% of lupus patients die within 15 years of the diagnosis which is comparable to Stage II breast cancer. The numerous harmful side effects of steroids and immunosuppressive including osteoporosis, infection and infertility have a destructive effect on careers, family relationships and quality of life over decades, and remain powerful arguments in favour of trying to do better.

How then can we try to improve the chances of successful clinical trials and a better knowledge of how to use biologics more successfully for our patients? One possibility is that combined treatment will be required in some subsets of patients to be effective. For example, some lupus patients given Rituximab develop increasingly high levels of DNA-antibodies linked to the presence of rising levels of BLyS.[16] These data have encouraged the development of two ongoing clinical trials that will study the effects of a combination of rituximab and cyclophosphamide followed by belimumab.[17] Ongoing trials of blisibimod [an anti-BLyS monoclonal] and atacicept following early encouraging results from Phase II trials, may also yet bear fruit. The blisibimod trial is restricting entry to more active subset of patients, a strategy that is supported by exploratory results from other trials. Furthermore, the early use of B-cell depletion at the time of diagnosis in lupus nephritis enabled 48 out of 50 patients to avoid the use of regular oral steroids for two years in a group of patients treated in London.[18] This result has led directly to a non-inferiority trial now recruiting, Rituxilup, which is comparing mycophenolate and steroids with mycophenolate and rituximab http://clinicaltrials.gov/ct2/show/NCT01773616?term=Rituxiliupαrank=1).

In addition to these suggestions, other relatively simple ideas might also improve the chances that a clinical trial will be unequivocally effective in lupus. Placebo group response rates would certainly be decreased by minimizing concomitant corticosteroids and immunosuppressive drugs at least for a portion of time during studies! The heterogeneity of lupus is challenging, but focussing on particular aspects of the disease, e.g. those with just skin or joint disease using separate assessment tools, may be useful. Improving data quality might also be useful by adjudicating rigorous standards for active disease in potential participants and using better trained monitors to review data collection in clinical trials. Perhaps most importantly, it will be essential not to weigh down trials for certain subsets of patients with primary end points that are almost impossible to obtain. See below for a summary table of biologic treatments in lupus.

Financial & competing interests disclosure

Professor David Isenberg has acted as a consultant for a number of companies including Merck Serono, GlaxoSmithKline, Eli Lilly and Celltech. The honoraria paid are transferred to a local arthritis charity. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Overview of biologic treatment in lupus.

Biologic	Trial results	Current “real world” use
B-cell blocking Rituximab (anti-CD20)	Explorer – failed (non-renal) Lunar – failed (renal) (some clinical and serological benefits noted in post-hoc analyses)	Widely used, most physicians have found to be of benefit
Epratuzumab (anti-CD22)	Embody 1 – Failed Embody 2 – Failed	Company has released information that these trials did not meet primary end points. Data are pending
APC – T-cell blocking Abatacept	MMF + abatacept – Failed Access – Failed (cyclophos + abatacept)	Another trial of abatacept and MMF is underway
Blocking B-cell activating factors Belimumab (anti BAFF/BLyS) Tabalumab (anti-BAFF)	BLISS-52 met primary end point with both doses used. BLISS-76 met primary end point with higher dose used. Illuminate 1 – Failed Illuminate 2 – Met its end point	Increasingly used in USA and Europe, but not UK where NICE have not approved it. Eli-Lilly have decided not to proceed to any further studies
Blisibimod (anti-BAFF)	In PEARL-SC (phase II) post-doc analysis of SR1-8 and high disease activity the drug 200 mg weekly showed benefit	In Phase III trial
Atacicept (anti-BAFF and APRIL)	APRIL-SLE safety committee prematurely terminated the higher dose arm, but ITT showed this reduced the numbers of flares and time to first flare	Ongoing Phase III trial of patients with active lupus
Blocking interferon α Sifalimumab	Primary end point (SR1-4 at 1 year) was achieved in a Phase IIb study	Company has announced that it will not develop sifalimumab further now, but will focus on anifrolumab, an interferon receptor anagonist which also met its primary end point in a Phase II study (data for anifrolumab pending release)
Rontalizumab	Rose – Failed (Some improvement in disease activity and decreased steroid use in those with low interferon signature gene expression)	Genentech discontinued plans for further studies