Tapering and discontinuation of TNF-α blockers without disease relapse using ultrasonography as a tool to identify patients with rheumatoid arthritis in clinical and histological remission
Crystal Ball Gazing in RA: predicting who can taper down (or off) treatment, and when.
Today, I wade into the murky waters of tapering in RA, the last in my 4 crystal ball gazing series. The earlier 3 tried to answer the questions:
1) Can we predict who will develop RA, so that we can pre-emptively treat so as to prevent it?
2) Can we predict which RA patients will do badly (rapid joint damage resulting in disability), so as to target the effective but costly targeted therapies to them early before damage accrues?
3) Can we predict, among those who will do badly, which drug will work best for whom, so as to minimise costly trial and error?
You should refer to my earlier posts at The Rheuma Muse for the answers.
Firstly, it appears that, for those already in clinical remission, ultrasound of the joints (especially the previously inflamed ones) will help guide decision whether stopping the biologic (anti-TNF, in this study) is doable to begin with. Anti-TNF was stopped in patients without active inflammation (power Doppler negative). 31% flared after 3 months. Another 7% flared after 6 months. That means 62% remained in clinical remission 6 months after stopping the biologic. That means lots of RA patients can save lots of money by going off their biologic.
Therefore, for RA patients with bad enough disease to warrant a biologic, it is not a life sentence. Once they are in clinical AND sonographic remission, they can try going off the biologic. There is almost a 2 in 3 chance that they will remain in remission at least 6 months out. This study should be extended to see what proportion stays in remission on MTX alone at further time points.
Some may question why I should even subject up to a third of my RA patients to the "pain and suffering" of flaring by advocating stopping their anti-TNF. Because biologics are costly and money does not grow on trees. Because biologics are not risk-free, and the longer you play Russian roulette, the more likely you'll get shot. Because patients are hopeful and demand it, and the glass is two-thirds full.
Another piece of good news from this same study (this is an editorial of sorts of the previous paper) is that, for those who flare, resuming the previously successful biologic enabled all of them to regain full clinical remission.
So, you won't know whether you'll succeed unless you try. And there appears to be no harm in trying.
PS 2 minor points highlighted in this review of the earlier paper bears noting:
1) grey scale synovitis of the 2nd MCPJ & 5th MTPJ appears to predict for those who may flare after stopping the biologic;
2) even in those in clinical and sonographic remission, there remains histological inflammatory activity. This means we may never be able to get rid of the disease completely, but it does not necessarily mean it would amount to any clinically relevant problem like joint damage and disability. The next post examines this further.
In this appropriately named study (HOPEFUL2), about 60% of RA patients remained in clinical remission or low disease activity after stopping their anti-TNF (Adalimumab) for 1 year. However, the authors were quick with the rejoinder that 76% remained similarly well because they stayed on the biologic.
Is this a case of glass-half-full-vs-half-empty (it's 60% full, actually), or is a 16% difference indeed worth the additional cost and risk?
The good news is, the statistically significant slightly higher inflammation detected in the discontinuation group did not translate to clinically relevant impairment of physical functioning or structural deterioration (joint damage). At risk of sounding schadenfreude, those who chose to remain on Adalimumab suffered more adverse events, especially infections (almost double).
Of course, HOPEFUL2 was an open-label prospective observational study, and may be biased in patients who chose to discontinue the biologic downplaying their pain or disability, but the structural and infection data should not lie.
Another predictor of successful tapering came from this study, particularly HOPEFUL1: of those who were randomised to initial intensive therapy (anti-TNF + MTX) vs MTX alone in the first 6 months (everyone received Adalimumab in the second 6 months), far higher proportions remained in low disease activity at the end of 2 years, whether they chose to continue or stop Adalimumab in the second year (97% & 80% respectively). Also, of those who were found to have rapid joint damage at the end of year 2 (13% of entire cohort), ALL came from the initial MTX alone arm.
This theme of treating RA early and aggressively to prevent joint damage and to give hope for therapy tapering is borne out not only in this study, but in various other studies involving other drugs as well, including the subsequent posts.
In the AVERT trial involving Abatacept, a T-cell co-stimulation blockade biologic, after 1 year of treatment and if patient was in low disease activity or better, inflammation and joint damage was assessed at the 6-month timepoint after withdrawal of ALL drugs. It was found that clinical and MRI remission was most sustained in those who received Abatacept, compared to those who received MTX alone.
It should be noted that most of the trial participants were deemed to have "very early RA". "Early RA" is taken to mean less than 1 year from persistent symptom onset. VERA may be 3-6 months, the "window of opportunity" when RA may be more amenable to a "cure".
Take-home-message for optimising towards a drug-free remission: treat early, hit hard.
Sustained improvements in MRI outcomes with abatacept following the withdrawal of all treatments in patients with early, progressive rheumatoid arthritis
The Dutch can be so understated. I would have named this study RADICAL or DRASTIC, instead of modestly as IMPROVED.
How is it radical/drastic?
1) Even patients too early to meet classification criteria for RA ~undifferentiated (UA), were enrolled, as long as it is inflammatory. Obviously, some cases could have been Reactive Arthritis.
2) Treatment target was very stringent, aiming for deep remission with a Disease Activity Score (DAS44) of <1.6.
3) This tight control was maintained by what I call a "cruise control" strategy, pulling back (stopping even) when DAS44 <1.6, and switching/escalating/restarting treatment when DAS44 >1.6.
610 patients were enrolled, starting with MTX and tapering high-dose steroid. Those failing to achieve target on this at Week 8 were randomised into 2 arms:
1) triple csDMARDs (MTX + SSZ + HCQ) with low-dose steroid;
2) MTX with anti-TNF (Adalimumab).
Treatment was "cruise-controlled" every 4 months. At the end of 2 years, clinical (DR ~DAS44 Remission, & DFR ~Drug-Free Remission) and radiographic (SHS ~Sharp-van der Heijde score) outcomes were assessed.
1) On the average, about 50% achieved DR, 21% DFR.
2) Regardless of treatment modalities, no clinically relevant damage (SHS change) was evident, and differences in DR and DFR rates were not statistically significant, even though it numerically favored the biologic arm. Early DAS44 Remission (EDR) rates also did not differ between arms.
3) EDR had significantly higher rates of DR (62%) and DFR(29%).
4) UA had significantly higher DFR rate (34%).
How do I read this? If you are aiming for a cure (DFR):
1) aim to achieve as deep a remission as possible;
2) aim to do so as quickly as possible;
3) intervene as early as possible, at the UA stage if possible;
4) use whatever it takes to do so, high or low dose steroid or biologic, combination or single csDMARD, choose the tools to fit the patient's acceptance and budget. Black cat, white cat, whichever catches the rat.
I have just one criticism of this excellent study (as good as the landmark BeSt trial in informing clinical practice, in my opinion): despite the tight control in a real-world mimicking trial in relatively early disease, only half of the cohort achieved remission. I can suggest 2 reasons for this:
1) The bar was set really high, at deep remission of DAS44 <1.6;
2) The "cruise control" therapy titration, especially therapy stoppage, is too drastic and does not reflect real-world practice (we typically make no big moves in either direction).
This brings me to another predictor for successful tapering: the longer the remission, the more likely you'll stay in remission upon tapering or drug withdrawal. Now to find a study to support this.