In today’s blog we will share our thoughts about a recent NEJM article, entitled “A Phase 3 Trial of Telitacicept for Systemic Lupus Erythematosus”.[1]
Publications on SLE literally come with a book of abbreviations and definitions nowadays and are not an easy Sunday reads. Of course, there is extensive literature on the details of outcome measures which are a prerequisite to understanding the various scores and scales. No wonder that SLE specialists are needed as study consultants, esp. if they sit on Guidelines Committees and follow the design minutiae.
As a dual inhibitor of BLyS and APRIL, the ‘tacicept’ group of drugs expands on the B-cell targeting of belimumab, a mAb with only BLyS blocking activity, which made it to approval after a haranguing series of studies with borderline activity in SLE. Hence, using a drug with broader anti-B-cell activity that includes autoantibody-producing plasma cells sounds like a reasonable approach.[2] The expectation would be that results would at least confer the efficacy of belimumab, but hopefully much more.
Telitacicept had an interesting development history. It is the most advanced of the 3 ‘tacicepts’ (atacicept, telitacicept, and povetacicept); it is already approved in China for several indications. To everyone’s surprise, the drug garnered NMPA provisional approval for SLE in 2021 based on a rather small set of interim data from a Phase 2 dose-finding trial conducted in China. Once published in full 2 years later, we learned further details: with approx. 60 patients per arm, the response rate of the SLE Responder Index (SRI-4) was 68.3% in the 160 mg Telitacicept group and 33.9% in the placebo group. Of note, looking only at the SLEDAI component, the response rates were 78% for telitacicept and 50% in the placebo group. RemeGen is pursuing a few more indications for telitacicept. In 2025, it garnered additional approval for gMG.
While we don’t know the FDA’s position on telitacicept and its approvability for SLE, we can assume that the Phase 2 data would not suffice for approval, for various reasons. So, we assumed that a larger global Phase 3 trial would be a logical next step.
Surprisingly, the Phase 3 study just published in the NEJM was conducted only at Chinese centers. With a prominent European rheumatologist as lead author, and a US clinical researcher listed as advisor and co-author, we had expected a pivotal multi-center international Phase 3 trial with at least some patients from Amsterdam and Oklahoma City. This was not the case here: RemeGen, a Chinese company chose to enroll the entire trial exclusively in China.
This raises some flags about the applicability of the results to a non-Han Chinese population. On the positive side, RemeGen will not have to struggle explaining ethnic sensitivity issues to their local Authorities in China.
So, what were the contributions of the expert consultants? The article mentions that “RemeGen (the trial sponsor) and the last author designed the trial.” In that case, what was the role of the first author, Prof. Vollenhoven, in all of this? He certainly did not write the manuscript; this was done with the help of a duo of professional ghost writers.
The NEJM tells us that ‘Ronald F. van Vollenhoven and Li Wang contributed equally to this article.’ Okay, that indicates good hard intellectual work. We also are informed that
- Merrill acted as a consultant to RemeGen advising on “protocol development, materials for regulatory agencies, outcome measures”
- van Vollenberg similarly acted as “consultant” without specifics mentioned
- Both were involved in the ‘analysis and interpretation of the data and review and approval of the manuscript”
So far so good.
Just a few words about the study design and its quality markers. This was a large double-blind study with 1:1 randomization and stratification. Telitacicept or placebo were administered on top of stable standard SLE treatment, which included steroids, singly or in combination. Response at Week 52 was the primary endpoint, with response determined by the SLE Responder Index 4 (SRI-4).
A ‘responder’ was a patient who achieved a 4-point reduction from baseline on SLEDAI in the absence of worsening on other scales. This is the short version; the long version reads like this:
The primary efficacy end point at week 52 was a response on the modified SLE Responder Index 4 (SRI-4), with a response on this composite measure defined as a reduction of at least 4 points from baseline in the SELENA-SLEDAI score,
AND
no new disease activity as measured by
- an A (severe) score or more than one B (moderate) score on the British Isles Lupus Assessment Group (BILAG) index (with eight organ domains scored over a period of 4 weeks on an ordinal scale ranging from A [severe disease activity] to E [no current disease activity]),
AND - no worsening in the Physician’s Global Assessment (PGA) score (on a visual analogue scale ranging from 0 to 3, with higher scores indicating worse disease activity). No worsening in the PGA score was defined as less than a 0.3-point increase from baseline.
These are reasonable criteria and make a-priori sense. The SLEDAI is not difficult to measure, as calculators are available.[3]
Looking at baseline demographics, all patients enrolled were heavily pre-treated. Most were on steroids with supposedly stable disease on entry into the trial. Thus, one would assume that any further improvement would be hard to come by. Driving down the SLEDAI from 11 to 7 as in this trial seems impressive.
But is it really? We have some doubts, mainly for the following reasons. How should we interpret the very high placebo response rate? The authors have no good explanation, so here we offer some thoughts of our own:
- Either
compliance improved considerably in a good number of patients leading to an improvement in SLEDAI. This explanation, however, is disconcerting, as it means that many patients were not optimally treated before recruitment. With more rigorous monitoring they achieved remarkably improved disease control on their old regimen. - Or
are we dealing with a largely artefactual improvement caused by variability in clinical assessments? Improvement in just 1 or 2 criteria of the SLEDAI can bring about a 4-point score change. Some clinical categories massively impact the SLEDAI which makes us wonder how useful this tool really is for assessing disease activity over time.
We do not want to delve deeper into the discussion on the SLEDAI and other scoring system – the literature on the topic is truly immense. It reflects the shortcomings of current scales and indices, but more so our state of ignorance. The inclusion of many soft clinical criteria adds arbitrariness, as scoring becomes subjective to a large degree. Such assessment criteria are hard to standardize. Wording around definitions is open to interpretation, which may get lost in translation.
In this context, having enrollment only at Chinese centers may be a plus, as it ‘normalizes’ the linguistic issues that may otherwise have crept up.
What we take away from this muddled picture is the following:
- double-blind trials are a must – thank you Drs. van Vollenhoven / Merrill if you insisted on this design element. It makes this study in SLE patients at least semi-interpretable.
- the ever-changing scales/scores/indices keep the printing presses running but remain thoroughly inadequate for anything but epidemiological purposes. Their variability is still too high. One should strive to abolish the soft clinical assessment points and replace them with harder, measurable criteria. If we are unable to do that, we stay on the path of belimumab – anafrolumab – telitacicept replacing one borderline performing drug with another.
- the lumping-together of multiple presentations of the SLE syndrome should be resisted. There are enough lupus nephritis (LN) patients that could be studied as a subset of their own. We recognize that LN may be several disease entities, but one has to start somewhere.
- Bring newer science to bear in SLE studies. This study reads like the old belimumab development plan and has the charm of an antique hold-over as far as trial design is concerned. We missed testing of newer ideas, like characterization of T- or B-cell populations as possible disease markers, interleukin profiles or modern genomic approaches to better understanding of the disease.
We believe that RemeGen’s consultants missed an opportunity to add exploratory elements to this otherwise well-conducted Phase 3 trial. When SLE patients give their consent to participate in a clinical trial, they are endorsing research in general, not just the registration of telitacicept in SLE.
Doing the same again and again, calling it ‘challenging’ to address a ‘multi-system, protean, complex disease’, is just not enough.
ABBREVIATIONS
BAFF B-cell activating factor
BLyS = BAFF
APRIL A proliferation-inducing ligand
gMG generalized myasthenia gravis
NMPA National Medical Products Administration (China FDA)
REFERENCES
[1] Van Vollenhoven R. A Phase 3 Trial of Telitacicept for Systemic Lupus Erythematosus. N Engl J Med 2025;393:1475
[2] Chappaz S. Effects of anti-APRIL antibody treatment versus dual APRIL/BAFF inhibition and anti-BAFF antibody treatment in wild type mice. Poster presented at the 18th International Symposium on IgA Nephropathy | 17–20 September 2025 | Prague, Czech Republic
[3] https://www.thecalculator.co/health/SLE-Disease-Activity-Index-(SLEDAI)-Calculator-1119.html
