In the 1950s and 1960s Geritol tonic was heavily advertised on television as a fast cure-all if “you’re feeling weak and rundown, tired or nervous.” It was lampooned in a more recent sitcom as a magic elixir capable of transforming a friendless geek into the most popular kid at school. While the recent and numerous television advertisements for the anti-tumor necrosis factor (TNF) agents currently on the market do not make such outlandish claims as this, one cannot help but be awed by the overwhelming presence that these intravenously infused prescription drugs have so quickly established in the public sphere.
TNF-α generates a vast array of biological properties, including cellular differentiation, proliferation, and apoptosis1–4. This variability is attributed to the binding of TNF to 2 distinct transmembrane receptors that can mediate TNF-induced inflammation and cell death, the promotion of proliferative responses in T lymphocytes and other hematopoietic cells5–7, and the induction of apoptosis in mature activated T cells8. Thus, TNF can serve dually as a potent proinflammatory mediator and a key immune modulator.
Due to the multifarious effects of TNF, the results of either TNF administration or blockade have varied significantly from disease to disease and from animal models to human patients. In terms of TNF blockade, currently 3 anti-TNF agents are licensed for clinical use (the 2 monoclonal antibodies adalimumab and infliximab and a soluble TNF receptor, etanercept) and are prescribed for the treatment of rheumatoid arthritis (RA), juvenile RA, psoriatic arthritis, chronic inflammatory bowel disease (IBD), and ankylosing spondylitis9,10.
Contrasting the beneficial effects of TNF neutralization, TNF administration in animal models of spontaneous insulin-dependent diabetes mellitus reduces the incidence of the disease. Similarly, models of systemic lupus erythematosus (SLE) have clearly demonstrated the benefits of early administration of recombinant TNF or TNF-inducing agents on inhibition of lupus nephritis11–13. Indeed, autoimmunity and nephritis are accelerated in lupus-prone mice deficient in TNF14.
Taken together, these disparate findings suggest that the clinically relevant pathophysiological role for TNF in certain disorders is that of a proinflammatory agent. Hence, neutralization or elimination of TNF is beneficial. In other disorders, however, the role for TNF as an immune regulator seems to be more important clinically, therefore administration or induction of TNF is ameliorative (while its blockade may be detrimental).
Accordingly, the complexity of the divergent effects of TNF is further exemplified by reports that, rather than improving their condition, use of anti-TNF in patients with multiple sclerosis has resulted in immune activation and disease exacerbation15,16. Moreover, anti-TNF treatments in some patients with RA or chronic IBD incite the development of autoantibodies, neuroinflammatory disease, or SLE-like features17–19. Nevertheless, it has been suggested that anti-TNF agents can be therapeutically administered to SLE patients20,21, emphasizing the pressing need for a greater understanding of the multiple (and often divergent) roles of TNF in various conditions.
The current report by Soforo, et al in this issue of The Journal presents 6 patients who, during anti-TNF therapy, developed active SLE, fulfilling 4 or more American College of Rheumatology criteria for diagnosis22. Although lupus-like symptoms are not uncommon after TNF blockade, these cases are unique in that they are associated with major organ involvement and life-threatening manifestations. It might be difficult to argue against the possibility that some of these subjects had underlying predisposition to SLE (or undiagnosed mild SLE), and therefore, these cases represent exacerbation of SLE under anti-TNF treatment rather than induction of de novo SLE. The major point of Soforo, et al’s observations, however, is that TNF blockade for SLE may be inappropriate, a view that we support regardless of whether or not these individuals represent genuine de novo cases of SLE, or are cases of underlying predisposition to SLE (or undiagnosed mild SLE) exacerbated by anti-TNF. The induction of SLE following TNF blockade reported in this study is further highlighted by the recent results of Jacob, et al23 in the lupus-prone NZM 2328 mouse model, which show that abrogating the effects of TNF by deleting both TNF receptors leads to a heightened and distinct inflammatory pathway that accelerates onset of disease, supporting the notion that anti-TNF may be contraindicated in the treatment of SLE.
Another consideration regarding these case reports is whether they represent the latest class of drug-induced lupus; or rather, represent a distinct syndrome of “anti-TNF–induced lupus.” We believe that the term drug-induced SLE should remain associated with the classical presentation due to procainamide, which has a milder disease course than classical SLE, and in which anti-dsDNA antibodies generally do not develop. We tend to agree with Williams, et al24 that anti-TNF–induced SLE is quite distinct from classical drug-induced lupus in that it seems to be associated with a more severe disease presentation that includes cutaneous, renal, and central nervous system involvement and anti-dsDNA antibodies, and that necessitates additional treatment beyond cessation of the anti-TNF regimen. In the cases presented by Soforo, et al, discontinuation of anti-TNF agents alone was insufficient, supporting their classification as anti-TNF–induced SLE, rather than drug-induced lupus.
The significant increase in the occurrence of antinuclear antibody and anti-dsDNA antibodies with anti-TNF treatment is well documented25. However, the incidence of SLE with the use of TNF blockade is very low and the incidence of anti-dsDNA antibodies in such patients does not predict the development of full-blown SLE. Given the important immunoregulatory effects of TNF, the breakdown of which presumably can result in adverse disease manifestations, why are there not more cases/reports of systemic autoimmunity resulting from this loss of TNF-induced immune regulation? Moreover, how are we to explain some of the anecdotal reports of the benefit following anti-TNF treatment in a few patients with SLE20?
One potential answer is that the current anti-TNF agents provide an incomplete blockade of TNF. That is, there is still free TNF available to provide certain protective effects. Thus, some proinflammatory effects of TNF are mitigated while some of the immunoregulatory functions remain intact. A non-mutually exclusive possibility is that most patients treated so far who have benefited from anti-TNF are Caucasians possessing HLA class II genotype of DR3 and/or DR4, who have been shown to be capable of producing higher levels of TNF in response to activation26. Consequently, reducing TNF levels in such patients might mitigate the inflammatory response while preserving sufficient TNF levels to provide protective regulation. Thus, this could explain the low frequency of development of systemic autoimmunity in patients treated with anti-TNF agents. Four out of 6 patients presented by Soforo, et al are Caucasians who still developed these manifestations, but we do not know their HLA class II genotype or their TNF inducibility. However, the fact that they are Caucasian and might have DR3 or DR4 genotype, and concomitant high TNF inducibility, and still develop this adverse effect underscores the caution that must be applied with these agents.
One should also keep in mind that DR2/DQw1-positive subjects (mostly non-Caucasian) have lower TNF inducibility26, and therefore might be more prone to the harmful consequences of TNF blockade. Thus, for this population of patients even more circumspection is suggested before using TNF antagonists.
There is no question that anti-TNF agents have improved the lives of many individuals suffering from RA and are beneficial in other conditions as well. With this benefit in mind, however, TNF antagonists should not be viewed as wonder tonics or magical elixirs capable of curing all rheumatologic conditions.