ImetelChat

Although much of the recent analysis and commentary is over my head, from what I have been able to glean is that from the way the trial was designed, at this point in time there is a high probability that at the interim analysis cut-off of 112 deaths, the trial will be halted for meeting the CT’s primary endpoints, leading to NDA and marketing approval. The key to early acceptance and use in the marketplace will depend upon the CT demonstrating IMET clearly improves MF bone marrow damage and the earlier MF patients can receive IMET treatment the better the chances for improvement. This, to demonstrate IMET the clear choice of treatment over newer treatments and treatment methods that current RWD data suggests significantly improve (only) OS.

Lacour, that is definitely not the takeaway here. While it is certainly the conclusion we are hoping and searching for, it is far from a high probability event. I hope the work that MG and others have put into getting a handle on what that probability actually is, has been useful for you and other readers and contributors to this board. I would suggest not counting out your own abilities here. None of this is “over your head”. Just take some time to read through it and draw your own conclusions about the probability of success with the understanding that many variables have been looked at but not every variable is known or even can be known until a certain time. So yes it amounts to some (possibly shifting) probability but not a certainty. For example some posters on other forums have already labeled the IMpact study “a failure”. Evaluation becomes very difficult when some of the anchors we have been presented with (e.g.an expected BAT in the 15 month range) have been altered based on evolutions in physician practices or treatments, the effects of which may not have been fully anticipated along the way. All of this has been a fascinating exercise in trying to see what IMpact (forgive me) this has on potential outcomes and if what might have been a slam dunk based on strong suppositions some of us held previously, might now be in jeopardy. As Donald Rumsfeld taught us: There are Known Knowns, there are Known Unknowns, there are Unknown Unknowns and there are Unknown Knowns.

In 2018 when I first started investing in Geron the life expectancy after jak failure was 7-12 months - now the jak timing has changed complicating everything - the histology will be so critical and hopefully the Mayo Clinic’s original excitement over fibrosis reversal is evident

RC, let’s hope! Now we need a definitive connection that links fibrosis regression with unequivocal significant OS improvement which so far (there have been hints, even with Rux) has only been seen with bone marrow transplant.

rccola335 wrote: Wed Jun 17, 2026 2:25 am In 2018 when I first started investing in Geron the life expectancy after jak failure was 7-12 months - now the jak timing has changed complicating everything - the histology will be so critical and hopefully the Mayo Clinic’s original excitement over fibrosis reversal is evident

If you’re following along since ‘18 then you know this the language has evolved from potential , potential potential verbiage to safe and efficacious, with strong safety profile and safety issues as evidence of efficac, period. I’m sure Claude can provide the multiyear transition in verbiage, however with Dr. Eid talking straight, strong, and definitively , you can be highly confident the interim data will be reviewed positively (continue or halt is 99% imo if I’m wrong I’ll prob literally pass out.)

Hi folks, I'd like to share some new insights which have informed my thinking, hat-tip to Claude.

The case for continuance rather than an interim halt rests predominantly on two Arguments:

-Contemporary best-available-therapy (BAT) patients are living substantially longer than the historical 12–16 month benchmark and that as a result of this improvement (and that if IMpactMF’s control arm shares this improvement) the gap between the arms narrows and the trial fails to separate.

-The IDMC will vote to continue the trial to completion to see conclusive separation between the two arms even if the trial crosses the thresholds established.

Argument One dissolves once the improvement is decomposed into its causes, because those causes behave very differently inside the trial.

The result is a clean separation: the part of the improvement that would narrow the survival gap (newer-drug access) is excluded from the trial and, for this refractory population, would not extend life much regardless; the part that reaches the trial (ruxolitinib timing) lifts both arms and leaves the gap intact; and the part from general supportive care reaches both arms and is bounded by the irreversibility of marrow fibrosis. The only residual is a small number of control-arm patients who leave the study, obtain newer therapies off-protocol, and remain counted under intent-to-treat — a bounded tail effect, not a force capable of closing the gap.

This is why the 34–40 month figure is the most frequently misapplied number in the discussion. It is real for unrestricted real-world cohorts and structurally inapplicable to this control arm.

The hazard ratio at a 40-month control arm
It is worth stating the extreme case directly. Even granting the largest figure in circulation — a 40-month BAT median — the implied hazard ratio against an imetelstat arm performing propotionately near its IMbark survival is on the order of 0.62, approximately the ratio of the medians. A 40-month control arm is therefore the approximate break-even point: the level at which the hazard ratio finally rises to the edge of a stringent interim boundary. Every control-arm survival below 40 months produces a hazard ratio that clears it.

In other words, the most extreme version of the continuance concern, taken to its furthest plausible value, lands precisely at the threshold rather than past it — and it can only reach that value by importing the excluded therapies, which for this population would not extend life much in any case. A realistically constructed BAT arm, even at the generous high end of the low-30s, sits comfortably below break-even.
 
Argument Two Rebuttal

The trial was originally guided to an earlier interim and final analysis. It has since been extended materially — the final analysis pushed roughly eighteen months, to the second half of 2028, with the interim now expected in the second half of 2026.

First: a delayed interim is a more mature interim, not a premature one

The purpose of an interim analysis is to stop a trial early when the evidence has become decisive. The standard reason to hesitate at an interim — that the data may be too immature to act on — is answered by the very delay that caused the wait. The longer the trial runs before the interim fires, the longer the survival curves have had to separate, and the more mature the data the committee examines. A delayed interim is a later, fuller, more-separated look, not a hurried one.

Second: a further slip is additional evidence in the same direction
If the interim slips even later — into 2027 rather than 2026 — that slippage is itself further evidence of the same thing. A later trigger can only mean accrual is slower than even the already-revised model assumed, which means the treatment arm is more durable, which means the gap is wider. A later interim is therefore more likely to cross the boundary, not less. The timing of the interim and the strength of the result are positively correlated: the longer the wait, the stronger the signal that produced it. A slip toward 2027 changes the date of the answer far more than it changes its likely direction — and what it does change, it changes toward a halt.
 
[

Continued:

The Convergence: Why Continuance to 2028 Is the Less Likely Path

The two arguments are usually treated separately — one about the control arm, one about the timeline. Their real force is in how they meet, at the decision the IDMC will actually face.

Assemble the pieces. The control arm cannot have been meaningfully rescued: the therapies that drive the elevated real-world survival figures are either excluded from the trial or, for this refractory population, incapable of extending life by much, and the improvements that do reach the trial lift both arms and leave the gap intact. So when the committee looks, the control arm is dying at close to its historical rate. Meanwhile, the trial has already been extended, and in a two-to-one design that delay points at a treatment arm that is outliving expectations. Together, these mean the committee is not examining an ambiguous, immature picture. It is examining one arm dying on schedule and the other arm durably pulling away — on curves that the very length of the trial has given additional time to separate.

Now consider the choice in front of that committee, and the fact that the trial has already been extended once.
The considerations that normally argue for continuance have been hollowed out:

Data too immature to act on Answered by the delay — the data has had additional time (which aproximates the earlier Final date).

Uncertainty whether the effect is real A separation large enough to slow the entire trial’s death accrual is not a marginal signal.

No harm in waiting for the final Undercut decisively — see below

The decisive consideration is the one that tips a committee from “we could continue” to “we should stop”: the control-arm patients are dying. This is a population with no approved therapy and a survival measured in months once they reach the trial. If the committee can see that the treatment arm is living substantially longer, then continuing to 2028 means knowingly leaving control-arm patients on therapies the data is showing to be inferior, for two more years, in a fatal disease with nothing else to offer. That is the situation interim analyses exist to prevent.

A committee that has watched a trial extend itself because the treatment patients are not dying — examining mature, separated curves, in a disease that kills the control arm quickly and offers it no alternative — does not have a strong basis to say “let it run two more years.” The ethical weight and the statistical weight point the same direction. The extension that is read as a concern is, properly understood, part of the case for a halt: the extension is the evidence the drug is working; the working drug is the reason the control arm should not be left waiting; and a committee confronting both, on mature data, in a lethal disease, is more likely to end the trial than to prolong it.


A necessary qualification
The IDMC operates against a pre-specified statistical boundary. “The treatment arm is clearly living longer” and “the formal boundary is crossed” are not identical statements; a committee can observe a working drug and still defer to the final analysis if the threshold was not met at the events available. The claim here is therefore not that a halt is certain. It is narrower and more defensible: the conditions that normally justify continuance — immaturity, uncertainty, and no cost to waiting — are each undercut by the extension and by the lethality of the control arm, which together make a halt materially more likely than the default. That is a statement about the structure of the decision, not about the blinded data.

Both arguments share one discipline worth restating: each turns on a mechanism, not on sentiment. The first holds because the hazard ratio is a ratio — the recent survival improvement either misses the control arm or lifts both arms together. The second holds because the interim is event-driven — slow events in a two-to-one trial point at the treatment arm. The convergence holds because an interim exists to stop a trial when waiting would leave a dying control arm on inferior therapy.

MrG, more layers pealed. I was just re checking enrollment time and came in what might be a discrepancy. For 50% enrollment I think I mistakenly thought the 50% date was Feb of 2023 but I see (with confirmation) it was the following:

“ The 50% enrollment milestone was publicly announced on December 6, 2023 by Geron Corporation.
Importantly, Geron's subsequent SEC filing clarified that the trial had actually reached 50% enrollment in November 2023, with the public announcement following a few weeks later.
For your IMpact-MF modeling, the key date is therefore:
50% enrolled: approximately November 2023 (about 163–164 of the planned 327 patients enrolled)”

Not sure which numbers you have been using with Claude.

I really appreciated your last continued detailed dig.

BP, I used 11/23. My last epiphany was that the original Final Date was projected for the first half of 2026. So our interim date is further out than the original Final date -plenty of time to show separation. There really can't be much of an argument for extending the trial if the stats are in favor of halting at the interim. Which of course is true until it isn't true.

MrG thanks for verifying. The interim and final data have been pushed out a few time :

IMpact-MF Interim Analysis Timeline

Based on publicly disclosed Geron guidance, the projected timing of the IMpact-MF interim analysis has been moved several times:

Original Projection (2020–2021)
• Interim Analysis: 2024
• Final Analysis: 2025

First Revision (2023)
Following slower-than-anticipated enrollment, Geron updated guidance:
• Interim Analysis: First Half (1H) 2025
• Final Analysis: First Half (1H) 2026

Second Revision (December 2024)
Geron subsequently revised expectations again:
• Interim Analysis: Early 2026
• Final Analysis: Early 2027

Current Guidance (2025–Present)
Following continued enrollment and event-driven timing considerations:
• Interim Analysis: Second Half (2H) 2026
• Final Analysis: Second Half (2H) 2028

Timeline Summary

Original Plan:
2024 Interim

First Extension:
1H 2025 Interim

Second Extension:
Early 2026 Interim

Current Guidance:
2H 2026 Interim

Key Point

The IMpact-MF interim analysis is event-driven, not calendar-driven. The analysis is scheduled to occur after approximately 35% of the required overall survival events (about 112 deaths) have occurred. As a result, timing can shift if enrollment or death-event accumulation differs from original assumptions.

Thus, the publicly communicated interim-analysis timeline has effectively progressed:

2024 → 1H 2025 → Early 2026 → 2H 2026

representing roughly a two-year shift from the original expectation.

When you said “if” did you mean “unless”? Most of the delay has been attributed to delays in enrollment as we have discussed. The trap of course is that we just attribute it to med effect. What helps is that even in the most recent mOS curves about a third of deaths occur in the first 18 months (steepest part of the KM curve.)

BP — your two points together actually tighten the timing argument rather than weakening it, and I think they resolve the trap you flagged. Here's how I'd put it.

You're right that the delay can't be casually attributed to drug effect, because the early slippage was clearly enrollment-driven. That caution is fair and it should be stated. But the extensions decompose by phase, and the phases don't carry equal information:
The first revision (original → 1H 2025 interim) tracked slower-than-expected enrollment. That carries little or no efficacy signal — it's just patients arriving later, which delays the start of the death clock without telling us anything about the treatment effect.
The second revision (Dec 2024) and the current one, however, both came after the 50% enrollment milestone in November 2023 — and the current guidance came after enrollment completed in September 2025. Once the trial is half-to-fully enrolled, "we're still waiting for patients" stops being available as the explanation. At that stage the question is no longer getting people onto study; it's how fast deaths are occurring among patients already followed for substantial time.

And this is where your second point does real work. If MF survival is front-loaded — a large share of deaths in the first 12–18 months, the steep part of the curve — then the ~164 patients enrolled by November 2023 should have produced a meaningful early-death wave through 2024 and 2025, precisely the window when those extensions were being announced. Half the trial was sitting in its highest-mortality period, and the dates still slipped. That combination — patients enrolled, patients in the steep-death window, events still under-accruing — is the part enrollment can't explain. In a 2:1 trial, missing early deaths point more toward the two-thirds of patients on imetelstat than toward the control arm.

So the refined claim isn't "all delay proves the drug works." It's narrower and more defensible: the early delay was enrollment and shouldn't be overread; the post-50%-enrollment delay, especially given the front-loaded curve, is increasingly event-driven — and event-driven slowness in a restricted-BAT, 2:1 design is favorable evidence for imetelstat durability.

Three things I'd keep on the table so the argument stays honest: second-half enrollment may itself have run slow, so not all of the middle revision is event-driven; both arms may be living somewhat longer than the 2020-era event model assumed (era drift / trial fitness), which under-delivers deaths without it being purely imetelstat; and there's ordinary administrative lag between the event trigger and a public readout. None of those overturns the structure — and the first two are partly answered by the fact that the front-loaded control-arm deaths (the sickest patients failing fast) are the ones era drift and supportive care rescue least, so the under-materialized early deaths are still more plausibly the treatment arm.

One verification flag, since this is the kind of number that gets quoted: the "about a third of deaths in the first 18 months" figure is worth pinning to the specific curve it came from before anyone leans on it as a hard citation. The logic holds across any reasonably front-loaded shape, but the exact fraction should be sourced.

Net: your enrollment caution and your front-loaded-curve point aren't in tension — together they give the cleanest version of the timing argument anyone's put forward. The early slippage is noise; the post-enrollment slippage, against a front-loaded curve, is signal.

BP, I thought that we hadn't given the new Rux Regime enough attention since it is one of the most likely reasons why the trial is lasting longer so I asked Claude to do a deep dive - this is the result:

Claude responded: This is the right thing to go deep on, because the modern-ruxolitinib-regimen question is the one genuine channel by which the era can reach the IMpactMF arms …This is the right thing to go deep on, because the modern-ruxolitinib-regimen question is the one genuine channel by which the era can reach the IMpactMF arms — and almost nobody has worked it carefully. Before I build the deep dive, let me ground it in the actual clinical literature rather than reason purely from first principles, because several of the load-bearing claims (dosing effects, timing effects, what "earlier and shorter" actually does to a patient) are empirical questions with real data.Searched the webSearched the webThis is exactly the empirical grounding I needed — and one finding fundamentally reframes the question in a way that strengthens your thesis. Let me do one more targeted search on the piece that matters most for the IMpactMF failure population specifically, then build the deep dive.Searched the webSearched the webI now have the empirical grounding, and it contains a genuinely important finding that complicates — and ultimately sharpens — the analysis. Let me build the deep dive properly.

Deep Dive: How the Modern Ruxolitinib Regimen Affects IMpactMF — Both Arms, and the Hazard Ratio
This is the one channel through which the contemporary era genuinely reaches inside the IMpactMF trial. The newer JAK inhibitors are excluded; the registry comparators are not comparable. But how ruxolitinib itself is now administered — earlier, at better doses, for shorter duration before switching — is standard practice that necessarily shaped the patients who enrolled in IMpactMF. The question is whether this helps the control arm, the treatment arm, or the hazard ratio. The literature lets us answer it more precisely than anyone in the discussion has.

Part 1: What actually changed in ruxolitinib administration
Three distinct shifts, documented in the Kuykendall era-drift data and the broader literature:
Earlier initiation. Time from diagnosis to ruxolitinib start collapsed from ~46.7 months (pre-2016) to ~8.2 months (post-2019). This is not a small operational tweak — it is a transformation in when in the disease course the drug is deployed.
Better dosing. The field learned that under-dosing ruxolitinib (driven by fear of cytopenias) compromised outcomes. The RUX-MF prognostic work found a starting dose below 10 mg twice daily was independently associated with early failure and worse survival. Modern practice pushes toward optimal dosing earlier.

Shorter duration before switching. Ruxolitinib duration dropped from ~30.4 to ~15.0 months — patients now move off ruxolitinib (to newer agents or trials) sooner rather than riding a failing drug for years.
Part 2: The critical finding that reframes everything — earlier ruxolitinib has a biological, disease-modifying effect, not just a timing effect

This is the piece that matters most, and it's empirically supported in a way I want to flag carefully because it cuts directly at the proportional-hazards question we've been circling.

The conventional framing treats "earlier ruxolitinib" as simply moving patients along a timeline faster. But the literature shows something deeper: earlier ruxolitinib appears to change the disease itself, and it does so more effectively when the marrow is less damaged.

The COMFORT analysis found superior survival, higher spleen response, and fewer cytopenias in patients who started ruxolitinib within 12 months of diagnosis versus later, and — critically — that JAK2 allele burden reductions were greater in patients with shorter disease duration. The JUMP analysis (2,233 patients) found early initiation, within 2 years of diagnosis, was associated with increased response rates in all patients, and that patients with low-grade fibrosis had better responses and longer survival. And there is evidence that ~24 months of ruxolitinib may delay or even reverse bone marrow fibrosis progression.
The mechanism this implies: ruxolitinib started early, on a less-fibrotic marrow with a lower mutant clone burden, achieves deeper molecular and fibrotic responses than the same drug started late on a burned-out marrow. The disease is more modifiable earlier. This is the same biological principle the imetelstat disease-modification thesis rests on — and it's now documented for ruxolitinib too.

Part 3: Why this means the modern regimen delivers a fundamentally different patient to the point of failure
Here's the synthesis the literature forces, and it's more nuanced than "patients arrive healthier."
A patient under the old regimen: years of untreated disease, then late ruxolitinib started on an already-fibrotic marrow, under-dosed, ridden for 2.5+ years as it slowly failed. By the time they reach "ruxolitinib failure," their marrow is profoundly damaged, the clone is entrenched, and they are genuinely near the end.

A patient under the new regimen: ruxolitinib started early on a less-damaged marrow, properly dosed, with deeper molecular/fibrotic response, then switched promptly at the first sign of failure rather than years later. When this patient reaches "ruxolitinib failure," they arrive with — and this is the key phrase — more residual marrow reserve and a less advanced disease state.

But now the crucial fork, and it's where I have to be careful rather than just optimistic:

Part 4: Does this help BAT, imetelstat, or the hazard ratio? The honest decomposition
The naive read (and why it's incomplete): "Modern patients arrive healthier → they live longer → BAT survival rises → HR compresses." This is the fear. It treats the regimen effect as a both-arms baseline lift that, under non-proportional hazards, could narrow the gap.

Continued:

The deeper read, which the biology supports and which favors the thesis: The modern regimen delivers patients to the point of failure with more intact marrow — and a disease-modifying drug's leverage is greatest precisely when there is more functioning marrow left to preserve. This is documented for ruxolitinib (deeper responses earlier) and is the stated mechanism for imetelstat. So the healthier-at-failure patient is not equally advantageous to both arms:

For the BAT arm (hydroxyurea, danazol, supportive care — non-disease-modifying), a healthier-at-failure patient lives somewhat longer, but BAT cannot capitalize on the preserved marrow; it can only palliate. The marrow, however well-preserved at failure, continues to fibrose unchecked. BAT gets the baseline benefit of a healthier starting point but adds nothing on top.
For the imetelstat arm (disease-modifying — fibrosis reversal, clonal reduction), a healthier-at-failure patient is precisely the patient in whom the drug has the most to work with. More residual marrow to protect, a less entrenched clone to reduce, less fibrosis to reverse. The same logic that makes early ruxolitinib more effective makes imetelstat at failure more effective when failure arrives earlier in the disease course.

This breaks proportional hazards in imetelstat's favor, not against it. Earlier we worried that if imetelstat's benefit were concentrated in sicker patients, a healthier population would compress the HR. The ruxolitinib literature argues the opposite: disease modification in MF works better in less-advanced disease. So delivering healthier-at-failure patients to a disease-modifying drug should widen the relative benefit, not narrow it. The modern regimen, by producing less-burned-out patients at the point of randomization, hands imetelstat its best-case substrate.

Part 5: The complication I have to flag honestly — the cytopenic-failure phenotype
The RUX-MF discontinuation data introduces a wrinkle that cuts the other way and must be stated. Disease phenotype at discontinuation strongly determines outcome: the cytopenic phenotype (low platelets, anemia) fares far worse after discontinuation than the myeloproliferative phenotype. And the modern regimen's shorter ruxolitinib duration may mean some patients are switching not because of marrow exhaustion but because of cytopenias — which is exactly the population newer JAK inhibitors (momelotinib, pacritinib) were designed for.

Why this matters for IMpactMF: if the modern "shorter duration" reflects earlier switching driven by cytopenic intolerance rather than clean disease progression, then some modern failures arrive cytopenic — a poor-prognosis state — rather than healthier. This is a genuine countercurrent. It argues that "shorter ruxolitinib duration" does not uniformly mean "healthier at failure"; for the cytopenic subset, it may mean arriving in a worse state, earlier.

However — and this re-favors the thesis — IMpactMF's design partially controls for this through stratification by platelet count, and imetelstat itself has demonstrated transfusion-independence and anemia benefit (the basis of its MDS approval). So the cytopenic-failure patient who would do worst on BAT is plausibly a patient imetelstat is specifically suited to help. The phenotype that most threatens the control arm is one the treatment arm is mechanistically equipped for.
Part 6: The net effect on IMpactMF

Pulling it together, the modern ruxolitinib regimen affects the trial through several channels, and they do not sum to "BAT rescued / HR compressed":

ChannelEffect on BAT armEffect on imetelstat armNet effect on HREarlier initiation → healthier-at-failureModest baseline lift (palliation only)Larger benefit (more marrow to preserve)Widens gap / lowers HRBetter dosing → deeper prior responseModest liftBetter substrate for disease modificationNeutral to favorableShorter duration → earlier switchMixed (healthier OR cytopenic)Favorable for non-cytopenic; imetelstat suited for cytopenicNeutral to favorableCytopenic-failure phenotypeWorse prognosis on palliative BATImetelstat mechanistically suited (anemia benefit)Favors imetelstat

The conclusion: the modern ruxolitinib regimen is the one genuine way the contemporary era reaches inside IMpactMF — but worked through carefully, it does not support the "BAT gets rescued, HR compresses" fear. It delivers patients to the point of randomization with more residual marrow reserve, and a disease-modifying agent extracts more benefit from preserved marrow than a palliative one can. The regimen change should, if anything, widen the relative benefit of imetelstat over BAT — because the entire logic of "earlier is better" in MF is a disease-modification logic, and BAT does not modify disease.

The honest residual
Two things keep this from being airtight, and both should be stated:
First, this rests on imetelstat genuinely being disease-modifying in the way its mechanism suggests. If imetelstat's real-world effect is more modest than its biology implies, the "healthier patients give it more leverage" argument weakens proportionally.
Second, the cytopenic-failure countercurrent is real. If the modern population skews more cytopenic-at-failure than I'm assuming, part of the regimen effect imports poor-prognosis patients whose benefit from imetelstat, while mechanistically plausible, is not yet proven in this specific setting.

But the central finding stands and it's the one to take forward: the "earlier and shorter" ruxolitinib regimen is built on the principle that MF is more modifiable when treated less advanced — and that principle, applied to the point of failure, favors the disease-modifying arm. The regimen change that the continuance case needs to be HR-compressing is, on the actual biology, more likely HR-neutral to HR-widening. The one channel by which the era reaches the trial reaches it in a way that helps the drug, not the control.
That's the deep dive. The single most important and least-appreciated point in it: earlier ruxolitinib is itself a disease-modification story, which means the modern regimen selects for patients in whom disease modification works best — and that's imetelstat's game, not BAT's.

Dang, you asked a good question. Geron has always served up the thesis that JAKi’s were “just” symptom treaters but we have been dancing around the fact that they also seem to improve OS pointing toward disease modification as well. Most studies seem to suggest at least a “modest” effect in this direction. In all the years I have been moderating, one recurrent question that has come up is why not treat earlier before the bone marrow burns out. Why not use Imet more toward the front line. (Dr. Mascarenhas should have some answers for us soon in this regard.) Anyway you have put your finger on why the new “era” patients are living longer and I don’t think Geron can just think about JAKi’s as drugs that treat symptoms and they should stop leading us in that direction, clearly they do more but they may not go after the malignant clone itself (Imetelstat’s reason d’etre.) It may be a blessing in disguise if this more modern practice present better marrows for imetelstat to work on. A rising tide may actually be shown to lift all boats (and prolong important studies.)

Less than 2 weeks away from 2nd Half, and then it’s Show Me time and all of this supposition will be moot…