Though a distinction is being made between diabetic ketoacidosis and HONK, the two really form a part of the same disease spectrum. Some ketoacidosis is present in HONK, and some hyperosmolarity is present in DKA. However, different mechanisms are at play. HONK is distinct form DKA, and the distinction is not entirely arbitrary, at least from the management point of view. For instance, even though the conditions co-exist 30% of the time, it is possible to treat pure HONK without any supplemental insulin (because there is a satisfactory amount of it in circulation already).DKA is 3 times more common, but HONK has 3 times greater mortality. The chapter on DKA presents a table of discriminating features to help distinguish HONK from DKA.
Past CICM SAQs involving HONk have included the following:
Similarly to DKA, a stress response which mobilises metabolic substrates in a Type 2 diabetic will result in HONK.
Mismanagement of diabetes
Drugs which trigger HONK
The key distinction between DKA and HONK seems to be the fact that in HONk, there is still enough insulin to overcome the ketogenic effects of glucagon.
Glucagon inhibits acetyl-CoA carboxylase, which normally converts acetyl-CoA into malonyl-CoA. Malonyl CoA inhibits acyl-carnitine synthesis; if this is uninhibited, it results in a stream of fatty acids being sucked up into the mitochondria to be converted into ketones.
Thus, we have a hyperglycaemic patient who remains reasonably asymptomatic because in them acidosis fails to develop (and thus, they are not short of breath). They remain hyperglycaemic for some time. As a result, they subject themselves to osmotic diuresis for a prolonged period, which allows them to become progressively more and more dehydrated.
The result is the hyperosmolar state which is usually associated with HONK.
This hyperosmolar hyperglycaemia is an intensely proinflammatory and prothrombotic state, which gives rise to the various complications of HONK.
In order to qualify as a HONK, you must strive towards the following diagnostic criteria:
One can summarise the presenting features of HONK in the form of a table:
Hypotension due to hypovolemia
High anion gap metabolic acidosis (with lactate and uremia rather than ketones)
Hypokalemia (due to vomiting)
Obtundation and coma
Acute renal failure
Naturally, having your body fluid gradually concentrated up to a high osmolarity is not a consequence-free process. One must recall that there are various rheological changes which take place as the result of hyperosmolarity. The following list of complications of HHS is a combination of several sources, including local resources as well as he college answers to Question 18.1 from the second paper of 2008 and Question 13 from the first paper of 2002.
The key issue is hyperosmolarity, and naturally the first instinct is to dilute the concentrated body fluids with some nice clean crystalloid. Indeed, Oh's Manual suggests that there is no specific difference between the fluid management in DKA and in HONK. On the basis of this consensus statement, they recommend the following fluid resuscitation schedule for both:
With this regimen, for a 70kg DKA/HONK patient, one ends up giving about 1.5-3L in the first 3 hours. The total deficit in HONK may be around 200ml/kg of free water (that might be 20 litres for a big person).
The college touched on this issue in Question 24 from the first paper of 2017. Specifically, the candidates were asked for "risk factors for all patients that predispose to the development of cerebral oedema" in HONK/HHS. Turns out, in adults this is a fairly uncommon complication (Matz, 1999).
The college give the following risk factors in their model answer:
The precise origin of these risk factors is unclear.
Looking through the literature, one finds such authoritative sources as the American Diabetes Association (Kitabchi et al, 2009) making a commend regarding bicarbonate being one of the causes of cerebral oedema. In the same article one may find the statistic that the majority of deaths from HHS in children are due to cerebral oedema, and a reference to an article by EC Quintana (2004) where severity of dehydration is mentioned as one of the predictors of mortality. Lawrence et al (2005) offer a population-based study of Canadian children (n=13!) where risk of cerebral oedema was associated with a lower bicarbonate, higher urea and higher glucose on presentation. Bialo et al (2015) scraped together a few other risk factors, including new onset Type 1 diabetes and a low PaCO2. Oh'Manual (p. 635 of the 7th edition) also mentions the failure of serum sodium to rise with treatment. Following a trail of references from Oh's one comes across the 2011 paper from Glaser et al, who also had low CO2 and high BUN. Then there is Rosenbloom (1990), where age under 3 years and Down syndrome were mentioned. Where the college got the rest of these risk factors, that is anybody's guess. If one had to put them all together into a list, it would look like this:
How does one avoid cerebral oedema in HHS? Though there is a real risk of cerebral oedema with vigorous fluid resuscitation, there is only some vague mention of the need to be careful. Gouveia et al (2013) and Dhatariya (2014) both authored review articles which warn of cerebral oedema and central myelinolysis. A recent Diabetes UK Position Statement (Scott et al, 2015) recommends we reduce osmolality by 3–8 mOsm/kg/h, as a safe rate.
There is little evidence to demonstrate the advantage of one form of crystalloid fluid over another. In HONK, there is no acidosis to defeat, which robs Hartmanns and Plasmalyte of their advantage.
However, don't let this fool you into thinking that the choice of fluid does not matter. One will be infusing an awful lot of crystalloid into this patient, and in context of that perhaps a little bit of intellectual processing is in order. I suppose one might make a comment about the chloride content of 20 litres of normal saline, and whether the patient requires this much chloride. In short, when in doubt one cannot go wrong with a "physiologically balanced solution".
Table 58.1 on page 631 of the new Ohs Manual presents a list of electrolyte deficits which develop in DKA and HONK. A 2003 article contains a similar table. I will summarise the relevant features:
Thus, the 70kg HONK patient stands to be infused with up to 14 litres of water, 910mmol of sodium, 1050mmol of potassium, 490mmol of chloride, and 140mmol each of phosphate, magnesium and calcium. In reality, the total infusion requirement may be greater due to diuresis.
"Describe your specific treatment", the examiner cackles malignantly.
The stereotypical approach is listed below:
Key issues of "specific therapy:
Mortality is higher for HONK than for DKA, and the patients tend to be older. One study sheds some light on poor prognostic indicators: