Hormone-Secreting Pituitary Tumors

When these tumors occur in children, the manifestations are often different than in adults.

Andrew Norris, MD PhDPost by
Andrew Norris, MD PhD
Director, Pediatric Endocrinology & Diabetes
University of Iowa Stead Family Children’s Hospital

A concise review of hormone-secreting pituitary tumors and their clinical syndromes appears in today’s New England Journal of Medicine. The article starts by noting that hormone secreting pituitary adenomas account for ~15% of all intercranial tumors. Although the article is informative and well written, it largely omits the characteristics of these disorders in childhood. When these tumors occur in children, the manifestations are often different than in adults. Below I have tabulated the anterior pituitary hormones that can be oversecreted by pituitary adenomas, and their common related childhood syndromes / symptoms. The table is listed in order of prevalence, from occasional to exceedingly rare (just a few case reports ever). Some of the symptoms of these conditions are common and non-specific (e.g. headache) and usually do not indicate a pituitary adenoma. Other symptoms almost always warrant an endocrine workup, especially growth failure, galactorrhea, precocious puberty, pubertal failure, gigantism, and acromegaly. On the flip side of hormone-secreting adenomas are pituitary adenomas that do not secrete hormones. Even though such adenomas do not secrete hormones, they eventually can lead to symptoms once their size impinges on local function. These manifestations can include visual field defects, headache, deficiency of pituitary hormones though prolactin can be modestly elevated due to pituitary stalk compression. Importantly, hormone secreting adenomas can also lead to these size-related effects as well.

Hormone oversecretedChildhood manifestations
Prolactin
(prevalence ~1/10,000)

Menstrual disturbance (girls)
Galactorrhea (girls)
Gynecomastia (boys)
Pubertal delay/failure (boys)
ACTH
(incidence <1/million/yr)
Weight gain
Growth failure
Striae
Hypertension
Amenorrhea (girls)
Hirsutism (girls)
Growth hormone (rare)Gigantism
Acromegaly
TSH (exceedingly rare)Hyperthyroidism
Headache
LH, FSH (exceedingly rare)Precocious puberty

Can insulin degludec reduce ketoacidosis rates? Exciting new data suggests “maybe”.

Andrew Norris, MD PhDPost by
Andrew Norris, MD PhD
Director, Pediatric Endocrinology & Diabetes
University of Iowa Stead Family Children’s Hospital

Diabetic ketoacidosis (DKA) is a diabetes emergency that can result in death when not detected quickly and treated in a timely fashion.  DKA is most commonly caused by taking insufficient insulin, especially forgetting to take long acting insulin, or taking insufficient extra insulin during illness. Children and adolescents with diabetes are at particular risk to develop DKA. For reasons that are not fully understood, rates of DKA are increasing (see this 2018 commentary in Diabetes Care). Insulin degludec is an ultra-long-acting insulin analog. Its duration of action exceeds 30 hours, which is longer than other current long acting insulin types.  For this reason, it has been postulated that use of insulin degludec might reduce DKA risk when compared to other long-acting insulin analogs, especially among those who occasionally forget to take their long acting insulin. However, evidence has mixed. A study in 2015 Pediatric Diabetes found a reduction in ketosis when comparing children on degludec versus insulin detemir (note: the study was funded by Novo Nordisk, the maker of both degludec and detemir).  This is the least meaningful of all possible comparisons since insulin detemir is the shortest acting of current long acting insulins. This 2015 European regulatory document presents an analysis of DKA rates from a company trial comparing insulin degludec versus detemir, finding no differences. A study published in 2018 Diabetes Therapy examined 42 adults who switched to insulin degludec found fewer DKA events after the switch, though the study was not powered for statistical conclusions (again this study was funded by Novo Nordisk). This month, work published in Feb 2020 Hormone Research in Paediatrics reports a retrospective study of 35 adolescents with DKA who switched from insulin glargine (in its most common “U100” formulation) to insulin degludec. The adolescents experienced significantly fewer DKA episodes after the switch. This exciting data suggests that degludec may indeed help reduce risks of DKA in youth at risk.  Although these are encouraging, the retrospective nature of the study and lack of a control group prevents firm conclusions. In general, DKA rates are expected to subside with time in adolescents as they mature and better learn to prevent this unpleasant complication. A control group of adolescents who did not switch to degludec would have helped interpret the results. Degludec has other benefits, especially less hypoglycemia compared to other long acting insulins (see this 2018 meta-analysis).

A lower safe threshold for treating hypoglycemia in newborns? An endocrine perspective.

Andrew Norris, MD PhDPost by
Andrew Norris, MD PhD
Director, Pediatric Endocrinology & Diabetes
University of Iowa Stead Family Children’s Hospital

There has long been some controversy regarding what blood glucose threshold should be used in newborns to guide when to initiate therapeutic intervention. This is an important concern, because sufficiently severe hypoglycemia can cause damage to the brain. However, it has been difficult to ascertain what degrees of hypoglycemia induce risk. This week the New England Journal of Medicine has published results from a new study focused on this question. Infants were randomized to receive intervention once blood glucose was under 47 (conventional cutoff) or 36 mg/dL. Psychomotor development was assessed at 18 months of age. The study found that outcomes in the more liberal cutoff group (36 mg/dL) were not any different from the conventional group. This suggests that the more liberal cutoff may be reasonable to use in clinical practice. An important caveat from the perspective of pediatric endocrinology is that this study pertains to healthy infants who do not have any specific endocrinologic or metabolic disorder. In fact the authors “emphasize the need for a higher target glucose concentration in newborns who have persistent hypoglycemia due to endocrine or metabolic disorders“. –Andrew Norris

Impaired glucagon responses in patients with cystic fibrosis and hypoglycemia

“Thus, these data implicate that cystic fibrosis induces defects in glucagon secretion leading to hypoglycemia risk. The association with pancreatic insufficiency suggests a link to pancreatic exocrine disease.” –Andrew Norris

Andrew Norris, MD PhDPost by
Andrew Norris, MD PhD
Director, Pediatric Endocrinology & Diabetes
University of Iowa Stead Family Children’s Hospital

Results from an important recent clinical study of hypoglycemia in patients with cystic fibrosis (CF) are now available. The study was conducted at the University of Washington and headed by preeminent diabetes physiologist Dr. Steven Kahn. Non-diabetic adults with cystic fibrosis were challenged with a 3-hour 75-gram frequently-sampled oral glucose tolerance test.

Half (14/27) of the subjects experienced hypoglycemia during the test. The glycemic pattern during the test was similar between these two groups until 135 minutes and thereafter, when the hypoglycemia group diverged downward. Those who exhibited hypoglycemia were more likely to be pancreatic insufficient. Importantly, those with hypoglycemia had lower plasma insulin & C-peptide levels. Modeling suggest that insulin sensitivity was greater in the subjects with hypoglycemia, accounting for the lower insulin levels. Importantly though, the “oral disposition index” did not differ between the two groups. This indicates that the amount of insulin secreted between the two groups would expected to have the same actions on lowering glucose when accounting for insulin sensitivity. Perhaps one could argue that the “oral disposition index” should have been lower in the hypoglycemia group, but in fact there was a trend towards lower values in this group (P=0.16). Plasma GLP-1 and GIP did not differ between the groups. In the hypoglycemia group, plasma epinephrine rose in the hypoglycemia group, albeit modestly. Cortisol did not rise, though it is not clear how many patients reached the lower blood glucose thresholds needed to trigger cortisol secretion. Growth hormone rose in some but not all patients. Perhaps most importantly, glucagon did not rise with hypoglycemia and did not differ at any point between the two groups.

These results suggest an impairment in counterregulatory response in patients with CF and hypoglycemia. Typically, the threshold for glucagon secretion occurs at less severe degrees of hypoglycemia than for other counterregulatory responses (see this nice review from Elizabeth Seaquist). Furthermore, the hypoglycemia in these subjects was mild. Thus, these data implicate that cystic fibrosis induces defects in glucagon secretion leading to hypoglycemia risk. The association with pancreatic insufficiency suggests a link to pancreatic exocrine disease. However, the mechanisms responsible remain to be determined. The manuscript describing these results is now published in the prestigious journal Diabetologia (link to article).

Might continuous glucose monitoring discern diabetes pathogenesis?

Andrew Norris, MD PhDPost by
Andrew Norris, MD PhD
Director, Pediatric Endocrinology & Diabetes
University of Iowa Stead Family Children’s Hospital

It can be challenging to distinguish type 1 from type 2 diabetes, especially in overweight adolescents and other populations. An upcoming report in the scientific journal Diabetes Care (link here) presents provocative data suggesting that continuous glucose monitoring (CGM) may distinguish forms of diabetes that occur because of loss of insulin secretion capacity (such as type 1) from forms due to insulin resistance (type 2). In particular, when CGM was performed when diabetes was very mild, insulin secretion defects were marked by increased blood glucose variability but normal fasting glucoses, whereas insulin resistance was marked by an overall increase in baseline (i.e. fasting) blood glucose levels.

–Andrew Norris