Hot off the press– my raw notes from the American Diabetes Association 2013 on immunotherapy trials in type 1 diabetes. My apologies that these are unedited and not guaranteed– but let me know if anything you see piques your interest, and I can elaborate!
Il2/Rapamycin treatment for T1D
Alice Long
- Can T1D be treated by addressing both Tregs and Teffs at the same time?
- Rapamycin addresses Teff, IL2 preferentially boosts Treg proliferation
- Combo txt has been shown to cure t1d in NOD
- Study design
- 20 – 36 yro
- t1d for 3 – 48 mo
- C-peptide >= .4/pmol/ml
- Published last year in Diabetes. Highlights here.
- For all patients, a decline in C-peptide.
- In fact, C-pep decreased more than control arm
- But some rebound of C-peptide with withdrawal of therapy
- Trial indicated that beta cell function decreased with therapy.
- Why?
- Tregs were increased with therapy.
- Absolute numbers increased
- p-Stat5 increased
- But confirm that t1d Tregs have decreased responsiveness to Il2 in vitro
- But that defect is repaired with therapy
- So Tregs were doing what we wanted; what went wrong?
- Teffs?
- Same ratios of CD8s, NK, etc. with Rapamycin alone
- But Il2 increases NK cells and Eosinophils
- Combo therapy enhanced Teff responses to Il2
- So, we boosted Treg responses. But, pleitropic effects of IL2 + Rapa in sum had pro-inflammatory effects via Teffs, NK cells, eosinophils
- What is behind the rescue of IL2 signaling in Tregs in t1ds?
- Isolate t1d T cells, culture ex vivo
- Without cytokine milieu, how do they respond to Il2?
- At day 1, and ever after stimulation, a decreased response to Il2
- But high levels of Il2 in the culture normalize t1d and control responses
- In conclusion, the presence of Il2 hae an effect on a cell-intrinsic deficiency of t1d Tregs
- Were the effects of therapy specific to t1d? That is, do non-D also see increase of inflammation with Il2+Rapa? The effect seems less prevalent in non-diabetics.
- Importantly, we see that response to IL2 in t1d can be enhanced, and it’s possible we might leverage that in other ways
- Q: Did you look at function? How did you characterize your Tregs?
- No functional studies, but all the classic markers. They don’t produce IFNg or IL7
- Questioner is doubtful; can you be sure these are Tregs?
- Q: How do you know this wasn’t just a toxic effect of rapa on beta cells?
- Not sure– still working to rule that out
- But recent animal studies imply this is unlikely
- Q: Have you looked earlier? Pre-diabetic? Is that decreased responsiveness to IL2 already there?
- We would love to do that, but have not been able to yet.
Low dose IL2 in T1D
David Klatzmann
- Started with interest in Treg deficiency in HCV-induced vasculitis. A quantitative defect in Tregs in these patients.
- Continued to study these patients, including treatments given for vasculitis
- Treg recovery correlated with response to anti-viral
- Similarly, Treg recovery correlated with responsiveness to Retuximab
- Chicken or the egg? Is Treg deficiency a cause of autoimmunity or a response to it?
- Try major Treg induction with Il2 in cancer patients. Failure as a trial. But noticed dramatic increases in Treg counts.
- So can this same treatment be used for HCV-induced vasculitis? (NEJM 2011)
- Multiple doses at 1.5-3mio UI/day
- Well tolerated
- Good Treg induction
- And, clinical improvement in 8/10 patients!
- Similar observation with low-dose IL2 in GVHD (NEJM 2011)
- Next turn to T1D. Previously established that IL2 increases Tregs, cures t1d in NOD.
- Effect in t1d humans? DF-IL2 trial.
- A dose-finding, double blind, placebo controlled, randomized trial
- First question: what dose will stimulate Tregs but not Teffs?
- Treated 5 day course, 0.33, 1, or 3 mio UI/day
- 24 patients, >= 18 yro
- Not expecting clinical improvement– just looking for Treg response.
- Safety? Very well tolerated. No serious adverse events at any dose. Some local reactions (not detected in placebo, only in some patients). Flu-like syndrome with higher doses. Considering all adverse events, more reported in placebo group overall.
- Primary criteria: increase in Tregs. Follow for 60 days, but best response at day 6 – 10. Increase detected for all doses, dose-dependently.
- Patient variability; some patients didn’t respond much at all, others responded quite well.
- No marked increase in T cells across doses. At highest dose, some increase in NK cells (though NS, technically)
- A dose-dependent effect of Il2 on Tregs at <= 1.0 MUI a day. No response in NK cells at that dose.
- Evaluate a variety of Treg parameters– CD25, GITR, CTLA4. Seem to be increases. Increase in regulatory cytokines.
- Notably, highest dose increases IL17 slightly. Il5 increased at early timepoint with highest dose. IFNg slightly increased with two higher doses.
- Look at T cells response to specific beta cell antigens in vitro. Tendency for increased responses to beta cell antigens (..maybe)
- A dose-dependent tipping of Treg/Teff balance
- Change during treatment of C-peptide? Nothing, as for A1c. But here we’re just looking for negative effect– just a short term test.
- Why such differences between this trial and the Rapa+IL2 trial?
- Rapa works against anti-CD3 and Il2 in vivo
- Rapamycin reverses curing of NOD mice with anti-CD3 or low-dose IL2
- Mechanism still TBD. Beta cell toxicity?
- What’s next? Multiple trials. In US:
- Recruiting for clinical trial (NCT01862120)
- Recently diagnosed
- 1 year treatment
- Q: A narrow therapeutic window. Even at a low dose, you’re dealing with a range of patients. How do you individualize this therapy so that you hit patients in exactly the right dose in exactly the right window to increase Tregs and not Teffs?
- Even the dose of .3 is quite good in terms of inducing Tregs. But effect is shortlived.
- But highlights that dosage is key issue. In next trial, we start lower, move up.
- Trying to find biomarkers of Tregs and pathology is a very important effort, and we are seeking grants.
- Q: Combine an anti-inflammatory strategy? Giving immune therapy in the presence of inflammatory milieu can have unpredictable effects.
- We published previously that Il2 has anti-inflammatory effects on whole PBMC transcriptome
- But supplementing is something we are considering; will have to see how this trial goes
T1DAL
Mark Rigby
- Hypothesis: depleting peripheral mem T cells will ameliorate disease progressions, preserve beta cell mass at diagnosis.
- Strategy: target CD2, which is preferentially expressed on Tems
- Using Alefacept, LFA3-Ig, which binds to CD2+ T cells and deactivates, depletes
- Developed and approved for psoriasis. Used in GVHD, some others.
- Study design
- 66 patients, with 2:1 drug:placebo.
- 2 year follow-up
- C-peptide > .25pmol/ml
- 12 week course of alefacept, 12 weeks off, another 12 weeks on. Same as used do demonstrate durable remission in psoriasis.
- Primary endpoint: AUC of C-peptide after 1 year
- Recruitment
- 19 participating centers, 14 with patients enrolled
- January 10 2010 start, Mar 2012 last patient enrolled
- Unfortunately, in Dec 2011, manufacturer cut off production for business reasons (not safety), so only 75% participant enrollment (49 instead of 66)
- Groups were demographically well-matched, except that alefacept group was more evenly age distributed than control, and control was skewed male
- Safety and tolerability
- No serious adverse events in either arm
- Same adverse in txt and control groups
- Hypoglycemic event rate was statistically higher in control than txt
- Efficacy
- After 52 weeks, Alefacept does not have same trend down in AUC as control.
- Change in AUC over 1yr? +0.015 in Alefacept, -0.115 in placebo.
- Looks good on a plot, but p = .065. Not significant
- 4h MMT AUC looks similar, but passes the threshold of statistical significance.
- A1c not sig different
- But insulin requirements did not increase as much over the year with Alefacept versus control (p = 0.02)
- Mechanistic effects
- Total CD4s: txt decreases total number, but re-increases by 52 weeks
- Ditto for CD8s
- Looking at percentage of naive, memory (central and eff) with flow, relative increase in naive cells versus memory in txt arm, for both CD4 and CD8
- Tregs seem better retained in txt versus control arms.
- Conclusions?
- Seems to have effects, but unclear if that is because of statistical power rather than biology for primary endpoints
- But two secondaries met–> much promise for preserving beta cell function
- Q: Compare your results to other studies. How does it look?
- We have done so. Our preservation looks better.
- Also, our placebo groups seem to be declining less than others–> could affect our ability to differentiate
Anti-thymocite gloculin in New Onset T1D (START Trial)
Stephen Gitelman
- Anti-thymocyte gloculin (ATG).
- mAbs have had some success directed at T1d. ATG is a polyclonal approach.
- ATG is used in a number of therapies. Create rabbit polyclonal Abs to human thymocytes
- Induces lasting remission in t1d
- Useful in inducing “partial” tolerance in transplantation
- Early studies in t1d? Pilot by George Eisenbarth in 1985, follow up by Saudek in 2004
- Mechanisms?
- T cell depletion. But this would be transient.
- Modulation and/or anergy of remaining T cells?
- Induction of Tregs?
- Study design
- 66 subjects, 12 – 35 yro
- Within 100 days of t1d
- Peak C-peptide > 0.4 pmol/ml
- Had to have prior evidence of EBV infections
- 4 day infusion. 6.5mg/kg total dose.
- Txt group only has pre-txt with steroids. (Why not control?)
- 58 enrolled, 38 in txt, 20 to placebo
- Adverse events
- Every subject reported at least one, txt and placebo.
- But more in txt group.
- 37/38 had cytokine release syndrome, all 38 had serum sickness
- Increase in infections and infestations in txt
- Notably, txt group received antimicrobials for 3months until cd4s were replenished.
- Efficiency
- Nope. The same endpoint
- But maybe biphasic? Steady decline with placebo, but fast down then steadying with txt?
- But why this change in response? Maybe cytokine release syndrome? Abrupt increase in IL6, etc. increases inflammation, disease progression, but then this normalizes by 1 month?
- Comparing T cell subsets
- Overall, initial acute depletion followed by reconstitution. Not normal yet at 12mo
- CD8 slightly faster than CD4 at reconstitution
- Tem not affected significantly by ATG
- Tregs? A profound initial decline and very slow recovery, especially as compared to other T cells.
- Slice and dice patient groups?
- Maybe effective in age 22 – 35? Stabilization from baseline without initial decline?
- Seems that young patients decrease C-pep rapidly first then stabilize, but older stay stable.
- Why might this be? Age? Our ability to regenerate immune cells decreases with age.
- Indeed, it seems that the young populations are recovering CD4, CD8, B cells, etc more rapidly
- Conclusions
- A warning about extrapolating from NOD to man.
- Response appears to be dependent on age. This is different from other recent onset studies where younger subjects seem to fare better.
- Subjects will be followed out to month 24. How long will that stabilization last?
- Mechanisms behind decline of beta cell functions initially?
- Nonspecific activation because of serum response, cytokines
- Persistence of Tem following ATG
- What’s next?
- Try to identify responders versus non-responders
- Compare changes to related trials to evaluate differing responses
- Combo therapies? ATG + IL6 blockade, or with GCSF?
- Q: The effect on Tem was unexpected. Is there any other data that supports those findings? A: Yes, there seems to be corroboration from transplant studies.
- Q: How much steroids did you use? Effect of that?
- 1mg/kg prednisone per day
- Plus others, for 7 days
- WHy not in control? Regulatory agencies guided them to not doing so.
Combo ATG-GCSF in Established and New Onset T1D
Mike Haller
- Given the effect of ATG just talked about, our rationale for using low-dose ATG plus GCSF
- An ongoing clinical trial, so no efficacy/patient data
- But some mechanistic findings so far
- Current trials in established t1ds; planning new onset
- To date, no mono or combo therapy that provides long-term preservation of beta cell function.
- Anti-CD3 (Teplizumab) study shows early differences, but then slope is the same; a delay, but then the txt arm catches up eventually.
- What amount of risk are we willing to tolerate to generate the benefit we want to see?
- Extreme combo therapy: the Brazil approach
- Cyclophosphamide + GCSF + stem cell harvesting. Deplete and replace.
- 20 of 23 subjects became insulin free for 1 month, 12 for 14 – 52 months (average 2.5 years)
- A1c < 7% and AUC c-pep increases after 24 months
- But that’s major. Brazil light?
- ATG + GCSF cure in NOD
- GCSF?
- Many mechanisms
- Mobilizes DCs (tolerogenic)
- Induces Tregs (?)
- Skewing cytokine profile to tolerogenic
- But doesn’t work as a monotherapy to preserve c-peptide. But is tolerated well.
- Funded entirely by the Helmsley trust. Unique in that it was intentionally designed to addressed established disease.
- Combo trial performs better along many mechanistic endpoints
- Total CD4+CD8 T cell counts go lower with ATG alone
- ATG alone reduces Tregs; combo maintains Tregs
- ATG depletes naive; combo maintains naive
- Combo induces transient increase in Treg/Teff ratio
- Very preliminary data! But not the same as high-dose ATG therapy.
- New onset trial
- Randomized, 1:1, with 80 subjects.
- Low dose (2.5mg/kg) ATG
- 80% power to detect 60% duff in AUC of c-pep
- Over-enroll by 10 people to mitigate loss due to dropout
- Results to come. But actively discussing the parameters for the next clinical trial.
