To the Editor:

We thank Dr. Pan et al for their interest in our recent article. They cite several studies of circulating APRIL levels in SLE and conclude that the available data seem to be controversial. Based on a critical analysis of the studies in question, we conclude that the preponderance of evidence points against the notion that APRIL has a contributory role in SLE.

In the largest study on this topic, our group (1) longitudinally studied 68 SLE patients for a median of 369 days, during which a median of 8 blood samples were collected from each patient. Serum levels of APRIL were significantly inversely correlated with disease activity. In a smaller study by Morel et al (2), the investigators also documented a significant inverse relationship between disease activity and serum APRIL levels among the 27 SLE patients who were evaluated twice (rather than just once).

In contrast to the above-cited longitudinal studies, the studies that suggested a positive relationship between disease activity and APRIL levels involved only single collections of blood samples (3, 4). Unfortunately, the findings were not validated either through repeat blood samples from the same patients at a later time or through testing in a second independent cohort. Indeed, one of these studies (3) was a retrospective study in which disease activity was apparently scored through review of the medical charts, a method well known to be fraught with imprecision.

Pan et al also state that genetics studies indicate that APRIL polymorphisms may contribute to the pathogenesis of SLE. One study of 148 Japanese SLE patients demonstrated an association between the G67R polymorphism in the APRIL gene and development of SLE (5), but this study was too small to reach reliable conclusions without validation in independent cohorts. In fact, a second independent study failed to demonstrate an association between the G67R APRIL polymorphism and SLE among European American subjects, and the number of African American and Hispanic subjects was deemed to be too low to reach any meaningful conclusions (6). Furthermore, genome-wide association studies and large case–control studies to date have failed to document a genetic association between APRIL and SLE, suggesting that if an association does exist, it is very weak at best.

Contrary to the implication by Pan and colleagues, we did not claim that APRIL is dispensable for the development of SLE in all circumstances. What we did claim, with very compelling supporting data, is that APRIL is dispensable for development of SLE in NZM 2328 mice. The equally mild renal immunopathology and equally absent clinical disease in single-deficient NZM.Baff−/− mice and double-deficient NZM.Baff−/−.April−/− mice suggest that the elimination (neutralization) of APRIL superimposed upon the elimination of BAFF may have limited benefit. A similar lack of incremental clinical benefit in NZM 2410 mice (closely related to our NZM 2328 mice) through the use of a therapeutic agent that inhibits both BAFF and APRIL to a level beyond that achieved with an agent that inhibits only BAFF has been previously documented by Ramanujam and colleagues (7). That is, regardless of whether BAFF and APRIL were negated genetically or pharmacologically, disease at both the clinical and immunopathologic level was ameliorated to the same degree as when only BAFF was negated. Given these findings, coupled with the premature termination of a clinical trial of atacicept (which neutralizes both BAFF and APRIL) in combination with mycophenolate mofetil due to an increase in serious infections among SLE patients ( identifier NCT00573157) (which has not been observed in any trials of BAFF-only inhibitors in human SLE), we stand by our assertion that the wisdom of antagonizing both BAFF and APRIL rather than antagonizing BAFF alone should be questioned, and that circumspection is certainly warranted. We never claimed that the case for therapeutic targeting of APRIL was closed; we, however, did (and continue to) raise a cautionary flag.

  • 1
    Stohl W, Metyas S, Tan SM, Cheema GS, Oamar B, Roschke V, et al. Inverse association between circulating APRIL levels and serological and clinical disease activity in patients with systemic lupus erythematosus. Ann Rheum Dis 2004; 63: 1096103.
  • 2
    Morel J, Roubille C, Planelles L, Rocha C, Fernandez L, Lukas C, et al. Serum levels of tumour necrosis factor family members a proliferation-inducing ligand (APRIL) and B lymphocyte stimulator (BLyS) are inversely correlated in systemic lupus erythematosus. Ann Rheum Dis 2009; 68: 9971002.
  • 3
    Koyama T, Tsukamoto H, Miyagi Y, Himeji D, Otsuka J, Miyagawa H, et al. Raised serum APRIL levels in patients with systemic lupus erythematosus. Ann Rheum Dis 2005; 64: 10657.
  • 4
    Hegazy M, Darwish H, Darweesh H, El-Shehaby A, Emad Y. Raised serum level of APRIL in patients with systemic lupus erythematosus: correlations with disease activity indices. Clin Immunol 2010; 135: 11824.
  • 5
    Koyama T, Tsukamoto H, Masumoto K, Himeji D, Hayashi K, Harada M, et al. A novel polymorphism of the human APRIL gene is associated with systemic lupus erythematosus. Rheumatology (Oxford) 2003; 42: 9805.
  • 6
    Lee YH, Ota F, Kim-Howard X, Kaufman KM, Nath SK. APRIL polymorphism and systemic lupus erythematosus (SLE) susceptibility. Rheumatology (Oxford) 2007; 46: 12746.
  • 7
    Ramanujam M, Wang X, Huang W, Liu Z, Schiffer L, Tao H, et al. Similarities and differences between selective and nonselective BAFF blockade in murine SLE. J Clin Invest 2006; 116: 72434.

Chaim O. Jacob MD, PhD*, William Stohl MD, PhD*, * University of Southern California, Los Angeles, CA.