Evidence in Support of the Validity of the TNF Brake Hypothesis

Walter P. Maksymowych, Physician,
August 20, 2013

Dear Editor,

Re: Continuous long-term anti-TNF therapy does not lead to an increase in the rate of new bone formation over 8 years in patients with ankylosing spondylitis. Baraliakos X, Haibel H, Listing J, Sieper J,Braun J. 1doi:10.1136/annrheumdis-2012-202698

I read with interest the findings of an observational cohort study in which patients with ankylosing spondylitis (AS) receiving infliximab (IFX) (n = 22) were compared for radiographic progression over 8 years with a historical cohort (n = 34) naive to anti-tumor necrosis factor alpha therapy (anti-TNF). After adjusting for baseline damage there was no difference between treatment groups over the 0-4 year time frame but a significantly greater rate of progression was evident in the 4-8 year time frame in the historical cohort. The authors then conclude that since there was less bone formation in the IFX treated group, these data argue against a major role for the TNF-brake hypothesis.

The basis for this concluding statement is that the authors interpret the TNF brake hypothesis as implying that the use of anti-TNF agents will accelerate the development of new bone since TNF upregulates dickkopf-1 which, in turn, downregulates Wingless (Wnt) pathway signaling for new bone formation [1]. So by removing TNF, dickkopf-1 also decreases which then allows signaling for new bone formation through the Wnt pathway [2]. This hypothesis was put forward to explain the observation that new syndesmophytes are more likely to develop at sites where inflammation has resolved (low TNF, low dickkopf-1, high Wnt) as opposed to sites of persistent inflammation (high TNF, high dickkopf-1, low Wnt) [3].

But what Baraliakos et al misinterpret about the TNF brake hypothesis is that this mechanism was proposed to explain the sequence of events in an established inflammatory lesion where bone formation pathways have already been triggered. Proinflammatory cytokines such as TNF have been shown to stimulate expression of bone forming factors, such bone morphogenetic proteins (BMPs) and Wnt proteins [4]. In a more detailed elaboration of the hypothesis, it was proposed that early inflammatory lesions resolve without sequelae, such as new bone, if effective therapy is instituted and inflammation resolves prior to activation of bone formation pathways by triggers such as TNF [5,6]. If the lesion is advanced and bone formation pathways are entrenched, TNF may well act as a brake on new bone formation acting through dickkopf-1.

We have reported prospective data to support this hypothesis demonstrating that lesions demonstrating inflammation at vertebral corners using short tau inversion recovery (STIR) MRI resolve completely with anti -TNF therapy while more complicated inflammatory lesions, marked by fat metaplasia using T1-weighted MRI, are predisposed to development of new bone even if the inflammation resolves [7]. The average patient with AS recruited to clinical trials of anti-TNF agents will have a mixture of early inflammatory and more advanced inflammatory lesions so the net effect of using an anti-TNF will be no impact on new bone formation when assessed over time frames as short as 2 years. This would explain the lack of impact of anti-TNF therapies on radiographic progression over 2 year time frames [8]. But it was predicted by the TNF brake hypothesis that over longer time frames anti-TNF should reduce new bone formation because treatment will prevent the ongoing development of new anti-inflammatory lesions. Following institution of treatment, all the advanced lesions resolve and then develop new bone. But since new inflammatory lesions are prevented by ongoing anti-TNF therapy, the hypothesis predicts that over longer time frames there should be a divergence in radiographic progression between anti-TNF treated and control patients. This is what Baraliakos et al have now shown.

Consequently, rather than arguing against a role for the TNF brake hypothesis, the data presented by Baraliakos et al actually reinforce its validity. Furthermore, recent data further reinforces its validity by demonstrating that new bone formation is indeed inhibited by anti-TNF therapy provided it is used early in the disease course [9].


1. Diarra D, Stolina M, Polzer K, et al. Dickkopf-1 is a master regulator of joint remodeling. Nat Med 2007;13:156-63.

2. Maksymowych WP, Chiowchanwisawakit P, Clare T, et al. Inflammatory lesions of the spine on magnetic resonance imaging predict the development of new syndesmophytes in ankylosing spondylitis: evidence of a relationship between inflammation and new bone formation. Arthritis Rheum 2009;60:93-102.

3. Pedersen SJ, Chiowchanwisawakit P, Lambert RG, et al. Resolution of inflammation following treatment of ankylosing spondylitis is associated with new bone formation. J Rheumatol 2011;38:1349-54.

4. Lories RJ, Luyten FP. Bone morphogenetic proteins in destructive and remodeling arthritis. Arthritis Res Ther 2007;9:207-15.

5. Maksymowych WP. What do biomarkers tell us about the pathogenesis of ankylosing spondylitis? Arthritis Res Ther 2009;11:101-2.

6. Maksymowych WP. Advances in pathogenesis through animal models and imaging. Nature Rev Rheumatol 2013;9:72-74.

7. Maksymowych WP, Morency N, Conner-Spady B, Lambert RG. Suppession of inflammation and effects on new bone formation in ankylosing spondylitis: evidence for a window of opportunity in disease modification. Ann Rheum Dis 2013;72:23-28.

8. van der Heijde D, Salonen D, Weissman BN, et al. Assessment of radiographic progression in the spines of patients with ankylosing spondylitis treated with adalimumab for up to 2 years. Arthritis Res Ther 2009;11:R127.

9. Haroon N, Inman RD, Learch TJ, Weisman MH, Lee MJ, Rahbar MH, et al. The impact of TNF-inhibitors on radiographic progression in ankylosing spondylitis. Arthritis Rheum published online DOI 10.1002/art.38070

Conflict of Interest:

None declared

Conflict of Interest

None declared