An Empty-Looking Sella and Labs That Are Almost, But Not Quite, Normal
A single patient, an incidentally found empty sella with two borderline hormone axes, neither clearly failed nor clearly normal. The disagreement is whether borderline results in a structurally abnormal gland should be treated as deficiency now or watched until they resolve into a clearer answer.
F.R., a 52-year-old man, has owned a small hardware store for twenty years, and came in originally for an unrelated headache workup that led, almost as an aside, to the finding driving today's visit: an MRI showing a markedly flattened, CSF-filled sella with minimal residual pituitary tissue draped along its floor — primary empty sella syndrome, most likely from longstanding, previously undiagnosed idiopathic intracranial pressure rather than any prior surgery or infarction, since he has no history of either. He describes fatigue and a decline in libido over roughly two years, changes gradual enough that he'd mostly attributed them to normal aging and the demands of running a business alone since his business partner retired.
His hormonal workup, drawn to determine whether the empty sella has actually compromised function or is an anatomic finding with normally functioning residual tissue, landed in a genuinely ambiguous zone rather than a clear answer either way. His morning total testosterone is 280 ng/dL — below the typical lower reference limit but not dramatically so, with LH and FSH both low-normal rather than clearly suppressed, a pattern more consistent with mild central hypogonadism than with primary testicular failure, but not unambiguous. His IGF-1 is at the lower edge of the age-adjusted normal range, and a subsequent GHRH-arginine stimulation test produced a peak GH response just above the diagnostic cutoff for severe deficiency — technically a pass, but close enough to the line that a repeat test on a different day, or a different provocative agent, could plausibly land on the other side of it. Cortisol and thyroid axes are both clearly normal, unaffected by whatever degree of compression his residual pituitary tissue has experienced — a detail worth reading directly rather than passing over, since a gland compromised enough to affect two axes but not the other two argues for genuine, if partial, functional loss rather than an incidental anatomic finding sitting on top of an otherwise fully intact pituitary.
How low is low enough to treat, when the numbers sit right on the line
I'd start testosterone replacement but hold off on growth hormone. His testosterone and gonadotropin pattern, taken together with two years of genuinely matching symptoms, crosses into treatable territory even without dramatically abnormal numbers — a mildly low testosterone with an inappropriately low-normal LH/FSH, rather than an appropriately elevated one, is itself evidence of central hypogonadism, not just a borderline lab value in isolation.
His GH axis is a different situation — a technical pass on stimulation testing, even a close one, isn't the same finding as his testosterone pattern, and I'd rather repeat GH testing on a separate occasion than treat a single borderline result as deficiency.
I'd apply the same caution to testosterone that you're applying to GH. A single testosterone level with low-normal, not clearly suppressed, gonadotropins is genuinely ambiguous on its own — normal aging can produce a similar pattern, and I'd want at least one repeat morning testosterone before committing him to lifelong replacement, the same standard we'd apply to any borderline androgen deficiency workup regardless of what the sella looks like on imaging.
The structurally abnormal gland is relevant context, but it isn't itself proof that this particular axis has failed — an empty sella with adequate residual tissue can still produce normal hormone function, and I don't think the anatomic finding should lower our evidentiary bar for treating one axis while we're holding a higher bar for the other.
Fair, and consistent — I'd apply the same repeat-testing standard to both axes rather than treating one as sufficiently proven and the other as not. Repeat morning testosterone and repeat GH stimulation testing, ideally with a different provocative agent for the second GH test given how close his first result sat to the cutoff, before starting either replacement. If the repeat testosterone confirms the pattern, that's the one I'd expect to cross into clearly treatable territory; the GH result is genuinely a coin flip until repeated.
Agreed: repeat morning testosterone with LH/FSH, and repeat GH stimulation testing using a different provocative agent, before starting any replacement therapy — applying the same evidentiary standard to both axes rather than treating one borderline result as sufficient and the other as requiring confirmation.
Not fully agreed: how much weight his structurally abnormal gland and two years of matching symptoms should carry against a single borderline lab value while awaiting repeat testing. The endocrinologist initially leaned toward starting testosterone now given the symptom match; the pharmacologist's consistency argument shifted the plan toward confirming both axes first, though the underlying read of how strongly the empty sella itself should shift the diagnostic threshold was not something both voices fully settled.