How a pancreatic tumour revealed growth-hormone-releasing hormone
Until 1982 nobody had isolated the hypothalamic hormone that stimulates growth hormone secretion — although its existence had been postulated for two decades. Roger Guillemin and, in parallel, Michael Thorner isolated it from an unexpected place: the pancreatic tumours of two patients with ectopic acromegaly. The substance became sermorelin, later CJC-1295, and eventually the entire GHRH analog family.
The gap after 1977
When Roger Guillemin shared the Nobel Prize in 1977 for the isolation of TRH and somatostatin, an obvious gap remained in the hypothalamic hormone system: growth-hormone-releasing hormone (GHRH), the hypothetical counterpart to somatostatin. Somatostatin inhibits growth hormone secretion; a releasing hormone stimulating secretion logically had to exist. But the substance had resisted hypothalamic isolation — even from hundreds of thousands of sheep and pig brains, GHRH could not be obtained in sufficient purity.
Through the late 1970s and early 1980s, several groups — Guillemin at the Salk Institute, Wylie Vale at the same place, Michael Thorner at the University of Virginia, research groups in France and Sweden — worked on GHRH isolation. Nobody got through.
The unexpected source: ectopic acromegaly
In 1980 and 1981 two clinical cases became known that became the decisive source: two patients with acromegaly without typical pituitary enlargement. On precise diagnosis, endocrinologists found that both had pancreatic tumours — neuroendocrine tumours ectopically secreting a pituitary-stimulating hormone and thereby triggering excessive growth hormone production in an otherwise healthy pituitary. It was one of the rare cases in which a tumour brought research closer to a hypothalamic hormone than the hypothalamus itself — the tumour produced the sought GHRH in hundred- to thousand-fold higher concentrations than the normal hypothalamus.
In 1982 Guillemin and co-authors published in Science the isolation and sequencing of a 44-amino-acid peptide from one of these pancreatic tumours — they called it 'somatocrinin'. Practically simultaneously, Thorner published in Nature the isolation of a related 40-amino-acid peptide from the second tumour. The two sequences differed only in C-terminal length; the bioactive fragment was GHRH(1-29) — the N-terminal 29 amino acids sufficed for full biological activity.
From tumour hormone to medicine
The GHRH(1-29) sequence was immediately made synthetically accessible — via solid-phase peptide synthesis (Merrifield 1963, see separate article), a 29-amino-acid peptide was no longer an insurmountable task in the 1980s. This synthetic form was designated sermorelin (chemical: GRF 1-29). In 1990 sermorelin received FDA approval in the US under the name Geref for diagnostic use — assessment of pituitary secretion capacity — and later for therapeutic use in growth hormone deficiency in children.
Sermorelin's own clinical career was limited. Because of the very short plasma half-life (about 10-30 minutes), therapy required daily injection and competed against recombinant human growth hormone, available after 1985, which produced a direct — albeit supraphysiological — effect. Sermorelin disappeared from the US market in 2008; recombinant growth hormone has dominated GH replacement since.
Iteration: tesamorelin and CJC-1295
The GHRH line did not end with sermorelin. Theratechnologies in Montréal developed tesamorelin (see separate article) — an N-terminally stabilised form that received FDA approval for HIV lipodystrophy in 2010. CJC-1295 is a further stabilisation variant with Drug Affinity Complex (DAC) modification for serum albumin binding, extending half-life from minutes to about one week. CJC-1295 is unapproved and stands in the black-market discourse as a non-recombinant GH alternative.
This iteration line shows a reproducible logic: the originally isolated hormone (GHRH 44/40 amino acids) is reduced to the active fragment (29 amino acids), then pharmacokinetically stabilised (N-terminal modification, albumin binding). The same logic is found in octreotide (vs. somatostatin), in leuprolide (vs. GnRH) and in semaglutide (vs. GLP-1).
„What we couldn't get out of 250,000 pig brains was delivered by a single pancreatic tumour in one gram of tissue. The pathology was more valuable to our research than the normal physiology."
What the GHRH story methodologically shows
Three structural lessons from this line. First: the source for a hypothalamic hormone need not be the hypothalamus. Ectopic secretion by tumours is a known endocrinological phenomenon that became a methodological lever here. Second: the temporal delay from first postulation (TRH/somatostatin identification late 1960s to 1973) to GHRH isolation (1982) shows that not all hypothalamic hormones are equally accessible — concentration in native tissue determines methodological feasibility. Third: the competition between Guillemin and Thorner ran, unlike the Schally-Guillemin GnRH competition of the 1970s, largely collegially — both teams indirectly used the same patient population and could publish their results nearly simultaneously.
Open questions
- Why have GHRH analogs never been able to compete clinically with recombinant growth hormone — physiological, commercial or regulatory reasons?
- Would modern long-acting GHRH analogs (e.g. once-weekly) be superior in paediatric growth-hormone deficiency to daily rHGH injections?
- Which ectopic-producing tumours still deliver unidentified hypothalamic hormones — are there potentially unexploited substance sources in tumour registries?
- How does the availability of large recombinant protein production (yeast, CHO cells) change the logic of peptide reduction — when is a shorter analog still worth it if the full protein can be produced recombinantly?