Escalating onto a rising curve

We have looked for a randomised comparison of escalation intervals — same molecule, same target dose, four-week steps against two-week steps or six-week steps — and we cannot find one of any size. There are protocol amendments, there are post-hoc tolerability analyses, and there is a great deal of clinical convention. What there is not, thirty million prescriptions into this class, is a trial that answers the single most frequently asked practical question about it. The Journal regards that as the most striking evidence gap in the field, and we intend to keep saying so.

The tirzepatide ladder is a different shape

Tirzepatide begins at 2.5 mg weekly for four weeks, moves to 5 mg, and thereafter increases in 2.5 mg increments at intervals of not less than four weeks, to a maximum of 15 mg. The structural difference from semaglutide is important: after the first doubling the increments are fixed in absolute terms, which means the ratio falls steadily — 1.5-fold, then 1.33, then 1.25, then 1.20.

The practical consequence is that the upper half of the tirzepatide ladder is unusually gentle in proportional terms, and the first step from 2.5 mg to 5 mg is by some distance the most demanding thing the schedule asks. Clinicians we spoke to described the 2.5-to-5 transition as the point at which most early attrition occurs, which is what the ratios predict.

The label also states, in language that repays attention, that 5 mg is a therapeutic dose in its own right and that escalation beyond it should reflect response and tolerability. That is a materially different instruction from a ladder with a fixed destination, and it is closer to how the drug is actually used.¹

Why four weeks, precisely

The elimination rate constant of a drug is 0.693 divided by its half-life. Fractional approach to steady state after time t is 1 minus e to the power of minus k times t. For a seven-day half-life this yields about seventy-five per cent of steady state at two weeks, eighty-eight per cent at three, ninety-four per cent at four and ninety-seven per cent at five.

Four weeks is therefore the point at which a once-weekly dose has essentially finished getting stronger. Escalate at two weeks and the person receives the increment written on the pen plus roughly a further quarter of the previous rung still accumulating underneath it. That is not dangerous in any dramatic sense, but it does mean the symptom burden attributed to the new dose is partly the tail of the old one, and it makes the escalation harder to interpret.

For tirzepatide, with a half-life closer to five days, four weeks corresponds to more than five half-lives and the previous rung is fully settled. The same interval is therefore slightly conservative for one molecule and exactly adequate for the other, which is a small illustration of how a shared convention can be right for different reasons.²

The shape of the curve, stated numerically

Across the programmes the exposure-response relationship for weight is approximately log-linear over the lower and middle range and flattens above it, while the relationship for gastrointestinal adverse events is closer to linear in dose and does not flatten in the same range. That divergence is the ceiling.

In glycaemic endpoints the flattening is even more pronounced. HbA1c reduction in the tirzepatide diabetes programme differed by roughly a quarter of a percentage point between the 5 mg and 15 mg arms in the head-to-head against semaglutide, which is a small difference against a threefold dose range.³ For a person whose objective is glycaemic control rather than weight, the argument for the upper rungs is correspondingly weaker.

The general point is that the class has two dose-response curves running in parallel and only one of them flattens. Any discussion of escalation that quotes the efficacy curve without the tolerability curve is quoting half a graph, which is how the top of the ladder came to be treated as an obvious destination.

A short glossary, because the terms get swapped

Initiation dose: the first rung, chosen for tolerability and generally sub-therapeutic. Not a low treatment dose. Target dose: the dose a protocol or prescriber intends to reach. Maintenance dose: the dose continued once the intended effect is achieved. Maximum approved dose: the highest dose in the label, set by the studied range and the tolerability ceiling.

Escalation interval: the time between increments. Hold: deliberately remaining at a rung beyond the standard interval. Re-titration: re-ascending after exposure has been substantially cleared. Dose-limiting: describing an effect severe enough to prevent escalation, which is a property of the person and the dose jointly, not of the drug alone.

Steady state: the condition in which drug entering the body equals drug leaving it. Accumulation ratio: steady-state average concentration divided by first-dose average concentration. Precision here matters more than it sounds: a large share of the correspondence this desk receives about titration turns out on inspection to be a disagreement about which of these words the writer meant.

Five things about titration nobody can currently answer

First, the optimal escalation interval. No adequately powered randomised comparison of intervals at a fixed target dose exists for any molecule in this class.

Second, the optimal hold duration for a person who has not adapted at four weeks. Practice ranges from four to twelve weeks on no comparative evidence at all.

Third, the lowest maintenance dose that preserves a result. The withdrawal trials compared full dose with nothing.

Fourth, whether tolerability at one rung predicts tolerability at the next. Clinicians assume it does, plausibly, and the published dose-ranging data is not analysed in a way that answers the question.

Fifth, whether any measurable baseline characteristic predicts the ceiling. Nothing published does so usefully, which mirrors the situation for efficacy: mean behaviour in this class is well characterised and individual variation is not.⁴

The Journal lists these not as a complaint about researchers but as a map of where confident advice is currently outrunning its evidence. Anyone offering a precise answer to any of the five is offering an opinion, and should be read as doing so.

Readers who think a paragraph above has outrun its evidence should write to the standards desk. Titration is a subject on which practically everybody has an opinion and practically nobody has a trial, and our correction log for this file is longer than we would like. That is the correct outcome of publishing numbers in a field where the numbers keep being checked.