RCT of Rituximab and cost-effectiveness analysis in treating Sjogren's syndrome

[Murph]

Bad news in a treatment that had made it all the way to a multi-center, randomised, placebo-controlled trial. A reminder that results don't always replicate, I guess.

https://www.ncbi.nlm.nih.gov/pubmed/28296257

Randomized Controlled Trial of Rituximab and cost-effectiveness analysis in treating fatigue and oral dryness in primary Sjogren's Syndrome.

Abstract
OBJECTIVE:
We investigated whether rituximab, an anti-B-cell therapy, improved symptoms of fatigue and oral dryness in patients with Primary Sjögren's Syndrome (PSS).

METHODS:
Multicentre, randomised, double-blind, parallel-group placebo-controlled trial, including Health Economic Analysis. Anti-Ro positive patients with PSS, symptomatic fatigue and oral dryness were recruited from 25 UK rheumatology clinics from August 2011 to January 2014. Patients were centrally-randomised to either placebo IV or rituximab IV (1000mg in 250mL) at weeks 0, 2, 24 and 26, with pre-and post-infusion medication including corticosteroids. Primary endpoint was the proportion of patients achieving 30% reduction in either fatigue or oral dryness at 48 weeks, measured by Visual Analogue Scale. Other outcomes included salivary and lachrymal flow rates, quality of life, ESSDAI and ESSPRI, symptoms of ocular and overall dryness, pain, global disease assessment and cost-effectiveness. ISRCTN 65360827 Results: All patients (n=133) randomised to placebo (n=66) and to rituximab (n=67) were included in the primary analysis. Among complete cases, 21/56 placebo and 24/61 rituximab patients achieved primary endpoint. After multiple imputation of missing outcomes, placebo and rituximab response rates were 36·8% and 39·8%, respectively (adjusted odds ratio 1·13 95% CI 0·50-2·55). There were no significant improvements in any outcome measure, except unstimulated salivary flow. Mean (SD) costs for rituximab and placebo were £10,752 (SD 264·75) and £2,672 (SD 241·71). There were slightly more adverse events reported in total for rituximab, but no difference in serious adverse events (ten in each group).

CONCLUSIONS:
Rituximab is neither clinically or cost-effective in this patient population

 

[Woolie]

Oh, how disappointing! You would think this disease would one of those most likely to respond. I wonder what @Jonathan Edwards thinks of the method they used?

 

[user9876]

Interesting cost data in that 4 infusions seemed to cost 11k with the placebo costing 2.5k. I assume the placbo cost represents the cost of giving the drug.

 

[Valentijn]

Full text at http://onlinelibrary.wiley.com.sci-hub.cc/doi/10.1002/art.40093/pdf

Participants received either rituximab IV (1000mg in 250mL saline) or placebo IV (250mL saline) in 2 courses at weeks 0, 2, 24 and 26. To reduce risk of infusion reactions, patients received pre-infusion medication of methylprednisolone, acetaminophen and chlorphenamine, and post-infusion oral prednisolone, reducing from 60mg to 15mg over 7 days post infusion.

The amount of rituximab was actually slightly lower, since some was removed to make it look the same as the plain saline bags.

The main outcomes were a fatigue questionnaire and self-reported oral dryness at week 48, with a 30% reduction being a success. Data for week 48 was imputed where it was missing for some patients. 12% of patients had incomplete fatigue or oral dryness data at the start or at 48 weeks, but there's really no excuse to lack that data at the start, and it's not specified how many lacked it at the start.

We did observe a difference between the arms in unstimulated salivary flow. Over the duration of follow-up, we found that the log-transformed USF values seemed to hold constant for rituximab patients, and to deteriorate for placebo patients.

So an objective secondary outcome did show a difference. There's no explanation for the why the primary outcomes were both subjective.

3 placebo controls were on a drug used to stimulate saliva, before and during the trial, versus 11 who were in the rituximab arm. So any effect of rituximab on dry mouth was likely less noticable due to that.

 

[Jonathan Edwards]

Oh, how disappointing! You would think this disease would one of those most likely to respond. I wonder what @Jonathan Edwards thinks of the method they used?

Most of the authors are colleagues I know well. A study of this sort has been in planning for about 15 years and it is useful to see it completed. I think the methodology is fine. With blinding, subjective measures are appropriate, since in this case they are the direct problems for the patients - fatigue and dryness.

Unfortunately we were pretty sure this would be the result by about 2002 because we had established that autoantibodies like Ro (the Sjogren's one) do not decline after rituximab. Antibodies to intracellular and nuclear protein antigens tend not to decline - presumably because they are made by long lived plasma cells. There is also the problem in Sjogren's that dryness is probably due to fairly irreversible changes in salivary glands once the problem is established. But clearly the fatigue side is most relevant to ME and there I think the issues is the long lived plasma cells.

We also had some clinical evidence from treating individual patients from around 2000 that there was no obvious response.

In RA, where autoantibodies go down, fatigue improves along with other things, so we still know that rituximab can improve fatigue in an autoantibody mediated disease.

I was slightly baffled by the costing of the placebo. I cannot quite see why you should cost a procedure that is only performed in a trial. It is not as if one could subtract this cost from the rituximab cost to get an idea of cost effectiveness because in the real world the alternative to using a drug is not to use anything.

 

[Snow Leopard]

The Meijer 2010 (30 patients) study found positive results, but notably no significant results for the Schirmer/lacrimal gland test.

Dass 2008 (17 patients), Devauchelle-Pensec (120 patients) 2014 likewise found no difference on the lacrimal test.

So perhaps not as much of a suprise that Bowman 2017 failed to find results on lacrimal test. Notable was the subjective differences found on the prior studies disappeared too. (note that the effect on subjective differences was never a strong effect)

Part of this message of the failure to replicate is the unreliability of subjective self report measures and a lesson not to cherry pick the results. Additional edit after reading JE's reply, it seems to me the lesson is to choose appropriate (a prior) outcome measures that show an impact on the disease mechanism. Still, I think it's worthwhile to conduct the studies to their full conclusion, even if to simply settle any doubt about whether the drug works.

 

[charles shepherd]

Some conference coverage from Rheumatology News last year below

Interesting to note that Arthritis Research UK seems to have vast sums of money to spend on research

I was in contact with ARUK last year

They are certainly interested in some of the overlapping features involving ME/CFS and arthritis - fatigue and pain in particular

Looks as though this is the end of the road for Rituximab and Sjogren's syndrome

CS

TRACTISS ends rituximab hopes for Sjögren’s syndrome

Rituximab is unlikely to work as a treatment for Sjögren’s syndrome, according to a sneak peak of results from a multicenter study.

The study, the Trial of Anti–B Cell Therapy in Patients With Primary Sjögren’s Syndrome (TRACTISS), showed that there were no significant differences between the active and placebo arms in terms of alleviating oral dryness or fatigue, Dr. Simon J. Bowman, a principal investigator for the trial, observed at the British Society for Rheumatology annual conference. In addition, no difference was found in the EULAR Sjögren’s syndrome disease activity index (ESSDAI).

Dr. Bowman, a consultant rheumatologist and honorary professor of rheumatology at the Queen Elizabeth Hospital, Birmingham, England, noted that there was some indication of an effect of rituximab on fatigue and oral dryness in another large, randomized, controlled trial – the Tolerance and Efficacy of Rituximab in Sjögren’s syndrome (TEARS) study – a couple of years ago (Ann Rheum Dis. 2013;72[6]:1026-31), but these were not the primary endpoints of the study, so those results were essentially negative.In TRACTISS, 110 patients with primary Sjögren’s syndrome received two courses of rituximab (Rituxan) or placebo in addition to standard therapy, and they were followed up for up to 48 weeks (BMC Musculoskelet Disord. 2014;15:21).

Calling the TRACTISS findings something of disappointment, Dr. Bowman said: “I think we’ll have to wait and see, but having had two big formal trials, I have to say that it doesn’t look like rituximab on its own is going to be any [great] success.”

Ann Rheum Dis. 2015;74[3]:526-31) showed there were some improvements in the ESSDAI and the EULAR Sjögren’s Syndrome Patient-Reported Index (ESSPRI). The improvements were great enough to justify proceeding to a larger clinical trial.

Perhaps the real hope for biologic therapy lies in targeting the T cells for which there is a very good rationale, said Wan-Fai Ng, Ph.D., professor of rheumatology at Newcastle University, Newcastle upon Tyne, England. Various approaches are available, and in addition, there have been some “encouraging” data from early-phase clinical trials.

Abatacept (Orencia), for example, has been looked at in a few trials, with positive drops in ESSDAI and other endpoints, and currently, there is an ongoing phase III trial in 88 patients with primary Sjögren’s syndrome. There are a couple of ongoing phase II trials, one with the anti-CD40 molecule CFZ533 and another with the anti-B7RP1 molecule AMG 557.

Dr. Ng noted that a trial of efalizumab (Raptiva) had terminated because of an increased risk for progressive multifocal leukoencephalopathy. In addition, a couple of other molecules, rhIL-2 and alefacept (Amevive), are under investigation, he said.

Four additional therapeutic targets

Other novel approaches to developing biologic therapy for Sjögren’s were summarized by Dr. Francesca Barone, a senior lecturer at the Queen Elizabeth Hospital in Birmingham, England. She described four strategies, the first of which was targeting the cross talk between antigen-presenting dendritic cells and T cells using a novel molecule RO5459072 that targets a protein called cathepsin S. Cathepsin S is involved in the assembly of the major histocompatibility complex II protein, and by interfering with this process, the theory is that the effector T-helper cell response will be muted and T-cell interaction with B cells will be decreased. A phase II “proof-of-concept” trial is underway with RO5459072 and will recruit 70 patients, she said.

A second strategy is targeting intracellular B cell signaling by targeting P13 kinase delta with UCB 5857. P13 kinase delta catalyzes B cell activation and “is really at the core of the biology of the B cell,” Dr. Barone observed. A phase II trial with UCB 5857 is planned in 52 patients.

A third approach is to target lymphoneogenesis more generally by interfering with the production of the chemokines CXCL13 and CXCL12. This might be achieved via interleukin (IL)–17 and IL-22 blockade. A trial with the novel agent baminercept, a lymphotoxin beta receptor fusion protein. However, that trial was stopped in 2014 for technical reasons.

A fourth strategy is to use combination “sandwich” treatment, consisting of belimumab, rituximab, and then belimumab again. “This is what is going to come next, putting two drugs together,” Dr. Barone said. A randomized, double-blind, controlled trial now open to recruitment aims to investigate whether there is value in this approach.

“We’ve bombarded you with a lot of targets,” Dr. Barone observed. “What do we believe is going to be the best one is very, very unclear.”

TRACTISS was funded by Arthritis Research UK with rituximab provided by Roche. Dr. Bowman did not provide any disclosures but previously has been a consultant for Roche, UCB Pharma, and Merck-Serono. Dr. Ng reported no conflicts of interest, and Dr. Barone disclosed consultancy or collaboration with UCB Pharma, GlaxoSmithKline, Celgene, Eli Lilly, and


http://www.mdedge.com/rheumatologyn...tissue-diseases/tractiss-ends-rituximab-hopes

 

[Woolie]

I do find it reassuring, though, that you can get null findings in these kinds of studies. We can know right away what's probably not worth our future efforts.

Unlike psychotherapy/behavioral treatment studies where success is pretty much a certainty, whatever you do, just because patients are hopeful. So we get stuck in an endless battle of is-it-isn't-it-good evidence, and can never move on.

 

[Jonathan Edwards]

I do find it reassuring, though, that you can get null findings in these kinds of studies. We can know right away what's probably not worth our future efforts.

Unlike psychotherapy/behavioral treatment studies where success is pretty much a certainty, whatever you do, just because patients are hopeful. So we get stuck in an endless battle of is-it-isn't-it-good evidence, and can never move on.

Yes, I think this sort of study restores faith in the fact that proper trials can be done as long as the basic rules are adhered. The result was the expected on, on several grounds, but with a blinded study (and a negative expectation) that does not matter.

What is disappointing is that the blurb from ARC is focusing on completely inappropriate strategies that we have no reason to think will work. The failure of rituximab in Sjogren's is almost certainly due to a failure to remove plasma cells. So we need a plasma cell targeting strategy. Andreas Radbruch is about the only person I know who understands this and is pursuing it. But he has a substantial programme and may well make headway.

So I too am reassured. The data all make sense. We know where to go next (or Andreas does) and the technology cannot be impossible. Plasma cells are just rather obstinate.