Dosing Levothyroxine With No TSH to Trust
A single patient, central hypothyroidism confirmed after pituitary surgery, started on levothyroxine with no reliable TSH feedback to titrate against. The disagreement is where in the normal free T4 range the dose should actually land.
S.B., a 55-year-old woman, has run the front office of a family dental practice for over twenty years, work she says depends on being sharp enough to catch a scheduling conflict before it becomes an angry phone call — a bar she's been missing lately, along with a cold intolerance she initially blamed on the practice's aggressive air conditioning. Six weeks ago she underwent transsphenoidal resection of a 22mm nonfunctioning macroadenoma; postoperative testing confirmed new central hypothyroidism (free T4 0.6 ng/dL, TSH an unhelpfully normal 1.4 mIU/L) alongside intact cortisol and gonadal axes, and her adrenal function was specifically confirmed adequate before levothyroxine was started, avoiding the real risk of unmasking undiagnosed central adrenal insufficiency by increasing cortisol clearance before glucocorticoid coverage is in place.
The dosing problem is a genuine mechanical one, not just a matter of caution: in primary hypothyroidism, TSH is both diagnostic and the whole titration target, since an intact pituitary raises or lowers TSH in direct response to circulating thyroid hormone. In central hypothyroidism, that feedback loop is precisely what's broken — her TSH will not reliably rise even if she is meaningfully underreplaced, which means dosing has to be titrated against free T4 directly, and the literature is genuinely split on where in the normal range that target should sit. Several endocrine society guidance documents recommend targeting the upper half of the normal free T4 range specifically because central hypothyroidism patients, tested against symptom scores, tend to feel and function better there than at the lower-normal levels that would be entirely adequate replacement in a primary hypothyroidism patient with an intact TSH to confirm it. Her own free T4 today, 0.6 ng/dL, sits far enough below normal that even a conservative starting dose should produce a measurable move within a few weeks, giving the team an early read on whether she's responding at all before the harder question of exactly where in the range to land ever needs deciding.
Without a TSH to chase, what number does the dose actually follow
I'd titrate her levothyroxine to the upper half of the normal free T4 range, not just anywhere inside it. Since her TSH can't confirm adequate replacement the way it would in primary hypothyroidism, the closest available proxy is symptom resolution, and the literature comparing central hypothyroidism patients across the normal free T4 range consistently finds better symptom scores in the upper half — she's describing real cognitive and cold-intolerance symptoms that fit that pattern directly. Her free T4 today, 0.6 ng/dL, is far enough below the normal range that even a conservative starting dose should move her measurably within a few weeks, giving us an early read on whether she's responding at all before we even reach the harder question of exactly where in the range to land.
I'd be more cautious about aiming deliberately high rather than letting symptoms guide a more conservative titration. Free T4 assays carry real inter-assay variability, and without TSH as a check, overshooting into a subclinically thyrotoxic range is genuinely easier to do here than in primary hypothyroidism — and she's six weeks post-surgery, still recovering, which is exactly the wrong time to introduce a new source of tachycardia or anxiety that gets misattributed to her surgery rather than her dose.
Targeting the upper-normal range as a starting point risks treating a population-level finding as an individual prescription before we've even seen how she responds to a first dose.
That's a reasonable sequencing concern, and I'd adjust the plan rather than the target: start at a conservative weight-based dose, recheck free T4 at 6 weeks rather than the standard longer interval, and only push toward the upper-normal range once we've confirmed she's tolerating the starting dose well. The target itself is still worth aiming for over time — it's the pace of getting there that should be more conservative than I originally proposed.
Agreed: start levothyroxine at a conservative weight-based dose, recheck free T4 at 6 weeks rather than waiting the longer interval used when TSH can also confirm adequacy, and titrate toward the upper half of the normal free T4 range over subsequent visits once early tolerance is confirmed — reflecting both the symptom-outcome data favoring that target and the real overshoot risk of aiming there immediately.
Both physicians converged on this plan; the disagreement was about pace, not destination, and resolved once the target was decoupled from the starting dose.