Ever since a French study linked growth hormone therapy to strokes two years ago, the debate has raged on in the endocrinology community. While most agree that more studies are needed, researchers disagree on how to proceed.
Two years ago, shockwaves shook the endocrinology science community when a study purported a link between growth hormone (GH) therapy during childhood to stroke risk later in life.
The epicenter from which these shockwaves emanated was the journal Neurology when researchers led by Amelie Poidvin, MD, MSc concluded that there was a possible association between GH therapy administered during childhood to “low risk” patients, i.e., those with isolated GH deficiency or idiopathic short stature — or kids who were simply small for their age — and an increased risk of stroke when these patients became young adults. Out of the 6,874 recipients who received GH, 11 had a fatal or non-fatal stroke.
Naturally, this set off some alarms, especially for pediatric endocrinologists who were treating low risk patients with GH, and it prompted a response from the Endocrine Society, which wrote in a statement: “While this raises some concerns, there are several significant limitations to this study identified by experts from the Society, Pediatric Endocrine Society, and Growth Hormone Research Society that necessitate caution in interpreting the conclusions.” The statement goes on to point out several of those limitations, such as data for the cause of death being obtained from French Death Certificates; a lack of adjustment for other risk factors like diabetes, hypertension, and smoking; and the fact that one of the two control cohorts was from another country.
The Society concluded that it “believes that until rigorously performed studies are done which confirm the Poidvin et al’s observations, GH therapy can continue to be safely administered to children who would benefit from it.”
Since then, the debate has continued, even culminating in a session at ENDO 2016 in Boston, which saw Jean-Claude Carel, MD, head of the Department of Pediatric Endocrinology at Robert Debré Hospital in Paris, France, and a co-author of the Neurology paper, make his case that the risks of GH therapy are indeed plausible. Ron Rosenfeld, MD, a professor and chair of Pediatrics (emeritus) at Oregon Health & Science University, Portland, again pointed to the limitations of the French study and echoed the Society’s statement that more studies are needed.
“Most endocrinologists take every case at face value and try to do a thorough as diagnostic evaluation as possible and come up with some reasonable estimate of both the benefits and the risks and discuss that with the family as they’re trying to determine whether or not they want to go ahead with therapy.” – Ron Rosenfeld, MD, pediatric clinical scientist, Oregon Health & Science University, Portland
Both researchers, of course, presented their cases well, each bringing up excellent points on the matter and both looking to similar futures for GH therapy and its study, albeit with different ideas of how those futures play out. The debate itself not only underscores the ever-evolving treatment and study of GH therapy, but is also a fascinating example of the nature of science itself.
Growth hormone therapy has been around for over 50 years. The use of GH for GH deficiency (GHD) started in the 1950s, when the supply was extracted from human cadaver pituitary glands, but that practice was halted in the 1980s when it was discovered that human GH (hGH) from an actual human body was associated with the devastating Creutzfeldt-Jakob disease, the incurable brain disorder that can lead to dementia, coma, or even death. Fortunately, around this time, recombinant hGH (rhGH) was already in clinical trials, so the FDA moved to expedite its approval to treat children with GHD.
“Over the ensuing 20 years many other pediatric disorders associated with short stature received approval also,” Rosenfeld says. “Although generally speaking the benefit of GH therapy for those disorders isn’t as great as it is with treatment for GHD. In children with GHD, if you make the diagnosis early in life and initiate proper therapy early, you can often restore the child to his or her genetically destined height. But for many other disorders like Turner syndrome or Prader–Willi syndrome or Intrauterine growth restriction or idiopathic short stature (ISS), the response to GH may be less robust.”
So the risk/benefit ratios depend on the patient and what disorder is being treated. The benefits and response are greatest for GH therapy in those who have an actual deficiency in GH, but Rosenfeld says that this is even further complicated because it can sometimes be difficult to diagnose GHD. “We can easily diagnose the child with total GHD, but there’s a spectrum and the pharmacological or chemical tests they use to diagnose GHD are imperfect,” he says. “So the borderline between GHD with a robust response and ISS with a more modest response is blurred. So it makes the benefit/risk ratio difficult because the diagnosis is often equivocal. Having said that, I think most endocrinologists take every case at face value and try to do as thorough a diagnostic evaluation as possible and come up with some reasonable estimate of both the benefits and the risks and discuss that with the family as they’re trying to determine whether or not they want to go ahead with therapy.”
Another thing to consider with these low risk patients is exactly how much the GH treatments will add to their height in the long run. Alan D. Rogol, MD, a professor of pediatrics at the University of Virginia in Charlottesville, Va., who moderated the ENDO debate, says that most endocrinologists look for a two-to-three-centimeter increase in growth in the first year. If the patient doesn’t reach that, the therapy may be futile. “So the issue is, is it really helpful,” he says. “It’s helpful in terms of putting the top of your head taller, but does it make a difference in the kid’s quality of life. That’s what the real issue is. If I took all the shots and not much happens, is that a success? It’s a big biopsy out of someone’s wallet, but I’m not so sure that’s a success.”
Angela Delaney, MD, a pediatric endocrinologist with the National Institutes of Health in Bethesda, Md., agrees, saying that it would be riskier to not treat a GHD patient with GH, but when it comes to patients with something like ISS, those risk/benefit ratios come into play because that patient is, by definition, not GH deficient. She says that there’s a lot to be considered in deciding whether to use GH to treat ISS: what the patient’s final height is predicted to be, which is based in part on how delayed their skeletal maturation is as well as their age and their pubertal status; how much longer they have left for growth; and so on.
“So to get an idea of the potential benefit, you have to consider all of those things as well as in other conditions that aren’t idiopathic short stature, you have to consider what’s the cause of the short stature, what impact will that have on the outcome,” Delaney says. “Then you have to pair that with the risks. The more well established risks include increased intracranial pressure, worsening scoliosis in kids who already have scoliosis due to the rapid growth, and slipped capital femoral epiphysis, which is a surgical emergency. Particularly in non GH-deficient patients there is a risk of increasing hyperglycemia, but that I think is patient population dependent. Then there are the theoretical risks of increased malignancies, as well as these more recent theoretical risks of vascular events.”
Worth the Risk?
And it’s these more recent theoretical vascular risks that are the subject of the most recent debates in GH therapy, especially in these low risk patients. To date, the French study published in Neurology is really the only one to reach this conclusion, but Carel says he believes that, while the study did indeed have its limitations, it’s the best the scientific community can do at the moment, and the findings shouldn’t be dismissed outright.
“My opinion is that we should, first of all, recognize that there is a potential risk and that the findings of the Neurology paper are potentially true findings,” Carel says, “and I do think that maybe this is not demonstrated, but to me it is really plausible that there is a link between GH treatment and stroke in these series of patients.”
Not surprisingly, Carel disagrees with the Endocrine Society’s assessment because he feels conducting a more rigorous study would be extremely difficult. “We should acknowledge that [the original study’s findings] could be true and keep working on it,” he says, “which is really not what the Endocrine Society’s reaction was, because the Endocrine Society’s reaction was ‘until rigorously performed studies are done, GH therapy can continue to be used as usual.’ Which, to me, is not appropriate, because the Society does not define what a rigorously performed study is in the matter. I think that a learned society should be open to new information and should be proactive in the advancement of science.”
But while Rosenfeld says that the study Carel worked on was an ambitious undertaking and was carefully performed, he points out that so far, no independent investigator has been able to corroborate the findings. In fact, parallel studies that were done in other countries that were part of the SAGhE (Safety and Appropriateness of Growth Hormone Treatments in Europe) study consortium so far have not at all confirmed any of the results that Carel and his team reported. Rosenfeld says that even if the French study turns out to be true, it could be possible that these children or some of these children had underlying conditions that predisposed them to cardiovascular events.
“You can say there’s an association,” Rosenfeld says, “but you can’t necessarily demonstrate causality, and that’s a difficult problem, and it’s a problem that’s almost impossible to tackle, through no fault of any of the investigators because nobody is going to do a controlled study where you take thousands of children who are GH deficient and randomize half to treatment and half to no treatment and follow up for life.”
“[Further studies] would take lots and lots of people,” says Rogol. “They would have to be placebo controlled. So you’re talking about giving someone 365 placebo injections a year. I don’t think it’s practical to really do them properly. The difficulty is not in designing the study. The difficulty is in the execution of the study. It’s going to take a lot of kids who would be getting therapy that may not be so helpful and maybe also include a number of kids getting a placebo for a long time.”
Again, Carel acknowledges the limitations of the French study, but he reiterates that the way he and his team performed the study and interpreted the data is the only option at present and that no controlled study can and will be performed to address these issues. “Going further, I think [we need to] repeat similar studies in other settings, other countries or continents, or find a more rigorous way – to use the word from the Endocrine Society – to do these kinds of studies,” he says. “You say, well I take a bunch of patients. I try to select a rather homogenous population to try to get away from as many confounding factors as I can and follow them over a long term, compared to the general population because that’s my best comparator, and then try to interpret your data, which is what we did and that’s to me the only way right now we can do. Other ways to do this are not realistic.”
For now, the best way for endocrinologists to navigate this with patients is to manage expectations and keep the patients and their families informed of while these studies do exist, they have limitations that haven’t been worked out yet. “In talking with families,” Delaney says, “knowing that they might get their hands on that information in one way or another, you want to make sure that they understand that you know there’s a potential that that’s a risk, but there are limitations to the data and you don’t really know what to say about it.”
“Over the last 25 years, every few years a paper appears citing a cancer risk, Hodgkins disease, colon cancer, whatever,” Rosenfeld says. “And none of those studies has ever been corroborated. So there’s a period of alarm, then the pendulum swings the other way. No drug is going to be without risk, and even if you are giving it as a replacement therapy as you do in GHD, we’re not entirely mimicking the normal physiology of growth hormone secretion and so we have to always be observant and open to the possibility that there may be adverse effects.”
Long-Term Surveillance Needed
Still, everyone says they see the need and even potential avenues for long-term surveillance of children treated with GH and what possible outcomes and events they may experience in adulthood, and all agree that more information is needed. Carel says that he can see getting closer to proving causality by enrolling larger numbers of patients or even devising animal models to show the effects of GH on the cardiovascular system. “We could enlarge the number of patients so we can have more information, maybe not on causality but the relationship between the [GH] dose and these events [stroke], which we do not have in that study and are actually difficult to obtain in a small number of patients,” he says.
Delaney says that, in the absence of more long-term studies, physicians need to continue to be as diligent as they can with reporting to databases. “How often does someone who got GH treatment for five years when they were a kid mention it to their GP when they’re 50,” she says. “There are a lot of limitations even to the long-term surveillance outside of the strict research system.”
“How often does someone who got GH treatment for five years when they were a kid mention it to their GP when they’re 50,” she says. “There are a lot of limitations even to the long-term surveillance outside of the strict research system.” – Angela Delaney, MD, pediatric endocrinologist, National Institutes of Health, Bethesda, Md.
Rosenfeld sees opportunity in the rise of long-acting GH treatments, which will not only be more beneficial to the patients themselves – most would rather take a shot every week or four than every day – but also need to be approved by the FDA and European authorities, which will require long-term surveillance. “There will be an opportunity if the endocrine community grabs it to demand that with the approval of long-acting GH, proper lifetime surveillance registries be constructed,” he says. “I personally would very much like to see lifetime registries constructed and organized, supported by the FDA, NIH, industry, under independent supervision with a team of people including endocrinologists and statisticians, that can properly address some of the long-term concerns both about benefits and risks, otherwise, you’re going to be calling up some other endocrinologist 10 years from now asking the same questions you’re asking me.”
Bagley is the associate editor of Endocrine News. He writes about the challenges of pediatric patients transitioning to adult care elsewhere in this issue.
- Growth hormone treatment in low risk patients (those without growth hormone deficiency) has been the subject of debate for the last two years, after a French study concluded the treatment is associated with risk of stroke later in life, while the Endocrine Society pointed to limitations in that study.
- The call for “more rigorous studies” could be problematic, since researchers would have to give placebo treatments to many children, among other concerns.
- While there is debate on whether growth hormone treatment is associated with adverse events later in life, most agree more work needs to be done, although there are differing ideas on execution.