Case Study 1 — The Hormone That Dissolves Bone and Builds It

The problem as it looked from inside endocrinology

For most of the twentieth century, parathyroid hormone had a settled reputation among bone specialists, and the reputation was bad.

The clinical picture of primary hyperparathyroidism established it. A parathyroid adenoma secretes PTH continuously and without regard to calcium levels. The result is elevated blood calcium, kidney stones, and — the part relevant here — progressive skeletal deterioration. In advanced, long-untreated disease, the bone findings were dramatic enough that older textbooks illustrated them. PTH was, in every clinician's working model, a hormone that took bone apart.

That model was correct. It is still correct. Nothing in what follows overturns it.

Meanwhile, osteoporosis presented the opposite problem: not enough bone, or bone of insufficient quality, with fracture as the consequence that actually matters. The therapeutic logic that developed was straightforward and productive. Bone is constantly remodeled — resorbed by osteoclasts, rebuilt by osteoblasts. In osteoporosis, resorption outpaces formation. So slow the resorption. Bisphosphonates do exactly that, they do it well, they are inexpensive, and they were supported by large trials with fracture as the endpoint. The antiresorptive approach became, and remains, the backbone of osteoporosis treatment.

Notice what that approach cannot do. Slowing the loss of bone is not the same as adding bone. An antiresorptive agent preserves what is there. For a patient whose skeleton is already badly depleted — multiple vertebral fractures, very low bone density, very high assessed fracture risk — preservation may not be enough. The field wanted an anabolic agent: something that would build.

And the obvious candidate was a hormone everyone knew destroyed bone.

The observation that did not fit

The clue had been visible in the literature for a long time, in scattered animal work and in careful reading of the human physiology, and it took a particular kind of attention to notice.

Endogenous parathyroid hormone is not secreted at a constant level. Like a great many peptide hormones, it is released in a pattern — a background level with superimposed variation, responsive to minute-by-minute calcium sensing. The pathological state of hyperparathyroidism is not simply "more PTH." It is PTH decoupled from that pattern: continuous, unmodulated, and never absent.

The question that follows is the one that makes this case study worth studying. Is the destructive effect of PTH on bone a property of the hormone, or a property of the exposure?

Those are very different hypotheses, and they make different predictions. If the effect belongs to the molecule, then any PTH exposure sufficient to activate the receptor should degrade bone, and the therapeutic idea is dead. If the effect belongs to the pattern — to continuity specifically — then a brief, self-terminating exposure might do something else entirely, and possibly something opposite.

The second hypothesis was correct.

What intermittent exposure does

Given as a once-daily subcutaneous injection, PTH(1–34) produces a pulse: a rise in circulating hormone that peaks and then clears within hours, leaving the receptor unstimulated for most of the day. Under that regimen, bone mineral density increases. Markers of bone formation rise. Fracture risk falls in the populations studied. Teriparatide was approved on the strength of randomized, placebo-controlled fracture-endpoint data in the early 2000s and has been in continuous clinical use since, positioned for severe osteoporosis and high assessed fracture risk rather than as a first-line agent for everyone.

The mechanistic explanation is not fully settled, and it is worth being honest about that rather than supplying a tidy story. The general account is that osteoblast and osteoclast populations respond to PTH signaling on different timescales, and that a signal which arrives and clears before the slower resorptive response fully engages tips the remodeling balance toward formation. Sustained signaling lets the resorptive side catch up and then dominate. Refinements of that account exist and compete in the literature.

But the mechanism is not what this case study is about. The observation is what matters, because the observation is unambiguous and the variable that produced it is exactly one. Not the molecule. Not the receptor. Not, in any straightforward sense, the amount. The shape of the exposure over time.

Why this is more than a curiosity

Three things follow, and each of them reaches beyond bone.

First, it makes pulsatility falsifiable. Chapter 3 argued that many peptide hormones carry information in their temporal pattern and that flat exposure to a normally pulsatile hormone is a categorically different signal rather than a stronger one. That argument is easy to state and easy to wave at. Teriparatide is the version of it you can point to: two exposure patterns, one molecule, opposite outcomes, in humans, with a regulatory approval attached.

Second, it reframes what a drug is. We are accustomed to thinking of a therapeutic decision as a choice of molecule, with the schedule as an implementation detail. Here the schedule is the therapy. A continuous infusion of the same drug at a comparable total exposure would not be a slightly less convenient teriparatide. It would be a different intervention with a different — and, on the evidence about continuous PTH exposure, an unwanted — effect.

Third, it cuts against a common inference in the peptide space. The reasoning "this hormone does X, therefore more of this hormone will do more X" is intuitive, ubiquitous in marketing, and here demonstrably wrong in the most vivid possible way. The same reasoning applied to PTH would have predicted that PTH therapy destroys bone. It builds it.

The complication, honestly stated

None of this makes teriparatide a simple drug or an unambiguous good.

It is an injection, given daily. It is expensive relative to a generic bisphosphonate. It carries duration-of-use considerations that most osteoporosis agents do not. And it carried, from approval, a boxed warning about osteosarcoma derived from lifetime high-exposure rat studies — a warning that was only removed roughly two decades later, after extensive human post-marketing surveillance failed to show the anticipated excess.

That labeling arc deserves attention on its own terms. It is the same structure as the GLP-1 receptor agonist thyroid C-cell question from Chapter 8: a genuine animal finding, a species whose relevant biology differs from ours, a precautionary label issued under uncertainty, and then years of human data slowly answering a question the animal study could not. The right lesson is not that animal findings are noise — plenty of them have translated. The right lesson is that the specific question was answered by evidence, and the label followed the evidence in both directions.

Discussion questions

1. State the hypothesis that had to be entertained before teriparatide could be developed, and explain why it required distinguishing a property of a molecule from a property of an exposure. What kind of evidence could have decided between those two hypotheses before any human was treated?

2. A colleague argues that teriparatide simply demonstrates the general principle that "less is more" in hormone therapy. Explain why that framing is wrong, and state what the correct generalization is.

3. Suppose a manufacturer proposed a continuous-release depot formulation of PTH(1–34), marketed on the convenience of avoiding daily injections. What would you predict, on the basis of this case study, and what evidence would you demand before accepting the manufacturer's contrary claim?

4. Teriparatide's clinical position is narrow: severe osteoporosis, established fractures, very high assessed fracture risk. Given that it builds bone while antiresorptives merely preserve it, why is it not first-line for everyone? Give at least three distinct reasons, and say which of them are about evidence and which are about practicality.

5. Compare the teriparatide osteosarcoma labeling history with the thyroid C-cell situation in Chapter 8. Identify two structural similarities and at least one important difference. Then state, as precisely as you can, what a reader should conclude about rodent carcinogenicity findings in general — without concluding either that they are decisive or that they are meaningless.

6. This case study argues that "the schedule is the therapy." Name one other compound discussed anywhere in this book where you suspect temporal pattern matters and has not been adequately studied. Describe the trial you would want, including who would be enrolled, what the comparison arms would be, and what endpoint would settle the question. Be specific about what result would change your mind.