Lupus Highlights

Michelle Petri, M.D.

Abstract #562: Low Mycophenolic Acid Area under the Curve is a Major Parameter of Systemic Lupus Erythematosus Activity

Authors:

Laurent Arnaud, Noël Zahr, Julien Haroche, Jean-Sébastien Hulot, Pierre Marquet, Christian Funck-Brentano, Jean-Charles Piette and Zahir Amoura

Background: 

Mycophenolate mofetil (MMF) is equal to IV cytoxan for renal lupus in Caucasians, and superior to IV cytoxan in African-American (Ginzler EM, et al. N Engl J Med. 353:2219, 2005; Appel GB, et al. J Am Soc Nephrol. 20:1103, 2009).  It is widely being used for other lupus organ manifestations, such as cutaneous lupus and serositis.  However, complete renal response rates in the Ginzler trial were very low at 6 months – 25%.  Can some of the non-response be explained by inadequate dosing? 

Methods: 

A Bayesian estimate was used to determine mycophenolic acid AUC with only 3 time points.  Using ROC, an optimum AUC was determined to be 35. 

Results:

Patients with high activity by BILAG or SLEDAI were below this level.

Discussion: 

This group, from Paris, previously demonstrated that only 50% of their patients had measurable Plaquenil levels (Costedoat-Chalumeau N, et al. Arthritis Rheum. 54:3284, 2006).  Jill Buyon’s group has shown that we often underdose azathioprine (Askanase AD, et al. J Rheumatol. 36:89, 2009). 

  1. In this study, the average dose of MMF was below 2000.  To us in the U.S., that seems very low.  Most of our patients with disease activity would be on 2000 to 3000.  We think that African-Americans need higher doses.  The French group did not find that ethnicity made any difference. 
  2. Dr. Urowitz asked how this could be translated into practice, since labs are not set up to do mycophenolic acid AUC (and the data were not presented as “peak” or “trough” levels). 
  3. The authors, in a poster, addressed the issue that low serum albumin and proteinuria may affect binding and clearance of MMF in renal lupus 
  4. The study definitely questions the current practice of never going above 3000 grams daily: but if we increase dosing, there may be a price in terms of toxicity. 

Abstract #565: Expressing Urinary Protein Excretion in Lupus Nephritis

Authors:

Debra Dye-Torrington, Emily Siu, Dominique Ibañez, Murray B. Urowitz and Dafna D. Gladman

Background: 

The SLICC group has developed a Renal Activity Score to follow lupus renal activity over time (Petri M, et al. Arthritis Rheum. 58:1784, 2008).  By far the most important component of this score is urine protein.  Twenty-four hour urines, though, are cumbersome, and often under-or over-collected.  However, a urine protein/creatinine ratio, done on a 24 hour collection, correlates very highly with the TRUE 24 hour protein (even if the collection was under or over-collected:  Christopher-Stine L, et al. J Rheumatol. 31:1557, 2004).  This led us to do a study of timed collection, to see if 6 hour collections would do as well (Ziegenbein M, et al. Kidney Int. 2009;in press). 

Methods: 

SLE patients were recruited with < 0.5, 0.5 – 2, and > 2 grams proteinuria.  One spot urine/protein was done in a paired fashion with a 24 hour urine. 

Results: 

A good correlation was found between the spot urine protein/creatine and the 24 hour urine protein ONLY for the group with < 0.500 grams. 

Discussion: 

There is no doubt that a protein/creatinine ratio is better than a dipstick urine protein estimate.  This study raises disquiet though about whether spot urine protein/creatinines could be used longitudinally in assessing treatment response and need for medication adjustment. 

  1. The study did not look at timed (6 hour) collections, which we think is an alternative.  It did not consider the circadian rhythm of proteinuria at all, in terms of the timing of the spot urine protein/creatinine.
  2. Dr. Hahn raised the point that there were few patients in the > 2 gram group, which may have made that estimate less precise. 

Ultimately, my conclusion is that if the patient does a spot urine protein/creatinine in the morning at every visit, AND there are periodic 24 hour urine protein/creatinine ratios to ensure good agreement, that is sufficient for routine clinical care.  The standard for research studies, though, may need to be different:  the urine protein/creatinine done on a 24 hour collection. 

Abstract #566: Impact of Clinical and Demographic Features on Influenza Vaccination Responses in Human Systemic Lupus Erythematosus.

Authors:

Sherry R. Crowe, Jourdan R. Anderson, Amy B. Dedeke, Virginia C. Roberts, Gillian M. Air, Linda F. Thompson and Judith A. James

Background: 

Several past clinical trials have shown that both influenza and pneumococcal vaccines are safe in SLE:  i.e., no increase in flares. 

Methods: 

This was a prospective study of 72 SLE patients who received seasonal influenza vaccine. 

Results: 

Some patients were low responders to influenza vaccine.  These were more likely to be Caucasian and to be on prednisone.  They were more likely to have a flare (mild/moderate, but 4 had severe flares) in the next 6 weeks.  High responders to the vaccine were more likely to make new autoantibodies or higher levels of autoantibodies. 

Discussion: 

Given the H1N1 epidemic, this study created great interest and many questions were asked!

  1. It is not a controlled study:  this is the most important criticism, because the patients may have been destined to flare at about this rate anyway.  However, the temporal association – within 6 weeks of the vaccine – remains concerning. 
  2. Dr. Bonnie Bermas made the most important scientific criticism:  the “low responders” may be different a priori.  They may have had, pre-vaccine, a more aberrant immune system, and therefore more likely to flare. 
  3. Dr. Murray Urowitz brought up the issue of adjuvant – in other countries, the H1N1 vaccine comes with or without adjuvant.  Would adjuvant be good in SLE patients (i.e., more high responders) or bad (i.e., more new autoantibodies)?  No one knows!
  4. Dr. Angela Tincani commented that normal children make antiphospholipid antibodies after vaccination. 

Thus, new or higher autoantibodies might not be a concern in SLE, if they were transient.
My personal comment is that I have been vaccinating lupus patients for over 26 years and have not seen any increase in flares! 

Abstract #564: Urine Cyto/Chemokines Correlate with Renal Histopathology in Systemic Lupus Erythematosus.

Authors:

C. Landolt-Marticorena, H. Reich, S. Morrison, E. Aghdassi, C. A. Pineau, J. Scholey, D. Gladman, M.B. Urowitz, A. Herzenberg, P. R. Fortin, J. E. Wither and CaNIOS LuNNET Investigators

Background: 

The gold standard in diagnosing the ISN class, judging activity, and determining need for treatment, remains the renal biopsy.  Because of its cost and risk, renal biopsy cannot be done routinely in follow-up.  Thus, a biomarker is needed that correlates highly with the renal biopsy so that clinicians could use it in longitudinal follow up.  Urine biomarkers would be preferable, since serum biomarkers might not “catch” the intrarenal inflammation. 

Methods: 

22 SLE patients underwent renal biopsy and had urine collected.  Urine cytokine/chemokines were compared vs. controls and with and without renal lupus.  Very few patients had elevated urine VCAM, so it was not studied further.  The authors combined the ones most frequently abnormal (MCP-1, IP-10, liponectin) into a composite score. 

Results: 

This composite score correlated highly with proliferative nephritis by biopsy, identifying 87% .  There was one patient with membranous nephritis who also had a high score. 

Discussion: 

There is much greater interest in urine biomarkers of lupus nephritis than ever before.  The difficulty is in finding sufficient patients to study at one institution! 

  1. Procedural issues are important in determining if the assay results are valid.  In this study, protease inhibitors were not added to the urine; the urine, though, was processed and frozen within 2 hours.  The authors did normalize the results to urine creatinine. 
  2. Dr. Oates asked whether treatment with steroids could affect the results.  The response was that no one had received IV solumedrol, but we were not shown how many were on high dose prednisone. 
  3. Dr. Bauer asked whether MCP-1 alone would have performed as well. 
  4. One question was on using the NIH Activity Score instead of just asking whether the composite score was associated with proliferative lesions.  The authors had not done this. 
  5. Ultimately, longitudinal data are required to show that the composite scroe is “sensitive to change” and whether it predicts long-term outcomes. 

Abstract #2069: Four-Year Experience of Belimumab, a BLyS-Specific Inhibitor, in Systemic Lupus Erythematosus (SLE) Patients

Authors:

Michelle Petri, R. Furie, Joan Merrill, D. J. Wallace, E. M. Ginzler, W. Stohl, W. Chatham, J. McCune, A. Weinstein, L. Pineda, Z. J. Zhong, J. Klein, D. Hough, W. Freimuth

Background: 

Belimumab, a monoclonal antibody directed against the B lymphocyte stimulator protein (BLyS), allows B cells to undergo apoptosis.  

Methods: 

SLE patients in the Phase 2 trial of belimumab were allowed to enter an open label continuation. 

Results: 

  1. Severe flares were reduced to very low levels.  All flares were reduced from 62% to 16%.  BILAG A or 2B flares reduced from 23% to 5%. 
  2. Autoantibody titers (anti-dsDNA, anticardiolipin, anti-Sm) were reduced. 
  3. There was no increase in serious adverse events, serious infections, or neoplasms. 

Discussion: 

  1. The Phase 2 trial itself did not meet the primary endpoint.  In the subset that was ANA or anti-DNA positive, the SLE responder index (a 4 point SELENA SLEDAI reduction with no BILAG A or 2 B flare and no 0.3 worsening in Physician Global Assessment) was met (46% vs 29%, p < 0.05). 
  2. The open label continuation does not have a placebo group.  However, the patients who entered the open-label did not differ from those who entered the Phase 2 trial in any important demographic or treatment variable.  Thus, it is hard to explain away the long-term benefit seen as a selection bias. 
  3. Dr. David Isenberg asked for more information about the important finding of autoantibody reduction. Human Genome Sciences does not have data on anti-Ro or La.  Anti-RNP was NOT reduced, but that may be a problem with the dynamic range of the assay. 

Abstract #LB1: Belimumab, a BLyS-Specific Inhibitor, Reduced Disease Activity, Flares and Prednisone Use in Patients with Active SLE:  Efficacy and Safety Results from the Phase 3 BLISS-52 Study.

Authors:

S. Navarra, R. Guzman, A. Gallacher, R.A. Levy, E.K. Li, M. Thomas, R. Jimenez, M. Leon, S. Hall, J.L. Lan, E. Nasonov, C. Tanasescu, H.-Y. Kim, L. Pineda, Z.J. Zhong, W. Freimuth14, M. A. Petri and BLISS-52 Study Group.

Background: 

This is the first successful Phase III trial of a biologic in SLE!  BLyS is the B Lymphocyte Stimulator protein – it prevents programmed cell death of autoimmune B cells.  Belimumab is a fully humanized monoclonal antibody against BLyS.

Methods: 

SLE patients had to be ANA or anti-DNA positive (in the Phase II study, this subgroup showed efficacy).  An entry SLEDAI score of 6 or more was required, representing moderate to severe activity.  Active Renal or CNS lupus was excluded.

Patients were randomized to placebo, 1 mg/kg and 10 mg/kg Belimumab on a background of standard of care (prednisone, hydroxychloroquine, and/or immunosuppressive drugs).

The outcome measure was the SLE Responder Index.

  1. a 4 point reduction in the SELENA SLEDAI
    and
  2. no worsening on the BILAG
  3. no worsening on the Physician Global Assessment

In addition, treatment failurs (need for more prednisone or immunosuppressive) were considered non-responders. 

Results: 

  1. The trial met the “primary” outcome, the SLE Responder Index, for both 10 mg/kg and 1 mg/kg.
  2. Additional data were presented to show prolongation of time to flare.
  3. There was a suggestion of steroid-sparing, using multiple analyses, such as 50% reduction and 25% reduction to below 7.5 mg/day
  4. Quality-of-life improvement was shown for fatigue reduction and the physical component of the SF36.
  5. There was no increase in serious infections or malignancies

Discussion: 

Never before have so many SLE investigators attended the Late Breaking Abstract Session!  This was truly a groundbreaking clinical trial. 

  1. The first question was actually from James O’Dell, wondering how medications could be managed in SLE clinical trials. This trial set a new standard for SLE trials, because both the outcome measure (SLE Responder Index) and the study design worked.  The sponsor had success in managing background medication, allowing some flexibility initially, but then by week 48, no changes were allowed. 
  2. Many of the analyses showed benefit by 16 weeks, contradicting the previous impression that Belimumab was “slow, but safe”. 
  3. The great safety was shown in analyses of infections, deaths, malignancies and common adverse events. 

Abstract #563: Validation of Interferon-Regulated Chemokines as Predictors of Lupus Flare.

Authors:

Jason W. Bauer, Michelle Petri, Franak Batliwalla, Joseph C. Wilson, Thearith Koeuth, Marlene Hyer Kern, Sukhminder Singh, Hanna Tesfasyone, Peter K. Gregersen, Timothy W. Behrens and Emily C. Baechler

Background: 

Laboratory tests do not predict SLE flares well.  Although it has been assumed that rising anti-dsDNA means a flare will occur, anti-dsDNA actually goes DOWN on the day of the flare (Ho et al, Arthritis Rheum 44:2342, 2001), likely reflecting tissue deposition of anti-dsDNA immune complexes.  Although the interferon gene signature is important in SLE, two major groups – our own (Petri et al, Lupus) and the University of Toronto (Annals Rheum Dis) – have shown that it does not predict flares.  Jason Bauer has previously reported on a chemokine profile (IFNgamma-inducible 10-kd protein [IP-10], MCP-1 and MIP-3beta) as a possible predictor of SLE flares (Bauer et al, Arthritis Rheum 60: 3098, 2009). 

Methods: 

The chemokine profile – a set of 3 interferon-induced chemokines – was tested on an independent set of 373 patients.  A flare was defined as a change in the SELENA SLEDAI of 3 points over the next year. 

Results: 

The chemokine profile predicted flares in this independent data set.  One chemokine in particular – IP10 – was the best. 

Discussion: 

  1. The presentation clearly showed that existing laboratory tests:  anti-dsDNA, complement, ESR – do not predict flares well.  It is important that this now be accepted!
  2. IP-10 appeared to work well alone.  This might make it more cost-effective.  It was the only variable significant in multivariate analyses. 
  3. Patients may have interferon-driven chemokines without the interferon gene signature. 
  4. Questions asked after the presentation were the predictive value for flares in the next 3 months and for flares in different organ systems.  These analyses are underway. 

AddThis Social Bookmark Button