Cardiac surgery-associated acute kidney injury

"Cardiac Surgery", the L1 topic c from Section 2.1.18 in the second edition of the CICM Syllabus for the Second Part Examination, stretches in all directions from the central premise, and one of these directions is post-operative renal complications. This was the subject of Question 2 from the second paper of 2024, where the examiners asked for the mechanisms and preventative measures.

Googling the string "Cardiac surgery-associated acute kidney injury" tends to produce about one million search results, of which the topmost are identically titled "Cardiac surgery-associated acute kidney injury", making it difficult to give reading recommendations. Helpfully, these are all largely interchangeable and contain mostly the same material. To support a local business, one would have to give a shout-out to Wang & Bellomo (2017), but Cheruku et al (2023) is more recent and frankly better, so considering that in either case the reader will have to resort to piracy for full-text content, there is probably no reason to opt for the older article. On the other hand, Harky et al (2020) is very detailed, sufficiently recent, and free.

Epidemiology of cardiac surgery-associated AKI

Depending on how you measure AKI,

• 20-30% of post-op cardiac surgical patients develop some level of KDIGO-identifiable AKI, but in the majority it will be a clinically insignificant "creatinitis", as a meaningless number changes in the bloods with no impact on survival or other outcomes
• Still, 1-2% of these will be sufficiently severe to require dialysis
• The 30-day mortality is ~ 3.2% (Wang et al, 2024) which might seem low but is about 3-8 times higher than the usual mortality from elective cardiac surgery (depending on how clumsy your surgeons are) and prolongs the length of stay from an average of 2.2 to 5.4 days. 

Risk factors for cardiac surgery-associated AKI

Cheruku et al (2023)  lists several scoring systems, for example:

• Cleveland clinic score (Thakar et al, 2005)
• Mehta score (Mehta et al, 2006)
• Simplified Renal Index (Wijeysundera et al, 2007)

These all share the following factors:

• Baseline high creatinine
• High risk surgery
• Emergency surgery
• Re-do surgery
• Diabetes
• Poor LVEF (<35-40%)

There are also other factors:

• Perioperative use of IABP
• COPD
• Age > 55
• Female sex

Of these, some must clearly act as surrogate markers of increased surgical complexity and influence outcome by prolonging bypass time, whereas others are more about the baseline level of circulatory fragility and renal frailty. There are a whole range of other metrics as well. For example Fan et al (2023) and Wang (2024), looking back on several hundred patients, found factors like the total dose of cryoprecipitate, on-pump versus off-pump, left main disease, and many others. For purposes of regurgitating these features into a CICM exam paper, they should probably be separated into some kind of neat cognitive niches, eg.

Or, after Yu et al (2023), attempting to amalgamate the entire range,

Alternatively one might want to divide these functionally into "modifiable" and "nonmodifiable", except it seems that there are very few of the former. Karkouti et al (2009) reported that only preoperative anemia, perioperative red blood cell transfusions, and surgical
exploration were "modifiable". Some might argue that surely nephrotoxins contrast and bypass time are also within the range of things the clinicians can control (eg. deciding not to do the double valve "while you're there"). 

Mechanisms of cardiac surgery-associated AKI

From looking even briefly over the list of mechanisms suggested as explanations of this phenomenon, one might come to the conclusion that it might be easier to enumerate the aspects of bypass which are safe for the kidneys, as the list would be shorter. Pretty much everything about the combination of needing cardiac surgery and having cardiac surgery is a massive risk. From Bellomo et al (2008):

• Hypoperfusion: the low cardiac output state obviously hurts kidneys in the conventional way
• Nephrotoxins are an uncontroversial thing to mention in this space, as they will damage the kidneys whether or not one is going to go on bypass
• Ischaemia–reperfusion injury: the reperfusion of previously flow-starved kidneys brings oxygen, which overwhelms the local free radical scavenger systems and creates oxidative stress. Bellomo et al note that this may be proportional to the ischaemia time and inversely proportional to the temperature at which the patient was while ischaemic.
• Oxidative stress is occasionally listed as a separate issue, mostly because it is not only related to the ischaemia and reperfusion, but also drugs like contrast and gentamicin, and endogenous oxidants like haem.
• Neurohumoral activation: this is mentioned by many authors without  overmuch specific discussion, and even Bellomo et al sidestep the issue by waving hands at "activated renin-angiotensin-aldosterone activity" and "generalized vasoconstriction" the net outcome of which is less “renal reserve”. Del Rio et al (2017) is a more considered take on this, and the reader is left to explore this themselves (as to regurgitate "neurohormonal activation" is probably enough for exam purposes). On the other hand:
• Inflammation is a broad enough statement that to blurt it randomly in the exam without any qualifying remarks would not gain marks in any SAQ, which means one may need to elaborate, eg. "renal tubular dysfunction induced by cytokine release" or something similar. It is perhaps worth noting that nothing worth mentioning seems to ever have been published on this matter, and all trials of antiinflammatory agents have failed to demonstrate a difference.
• Mechanical factors: these are mostly related to renal emboli, eg.
  • atheroembolism from aortic manipulation
  • intra-aortic balloon pump use
  • cross-clamping
  • air emboli
• Renal vasoconstriction:  bypass grinds red cells, haemolysis leads to the release of free haemoglobin, and free haemoglobin is a terrible thing that promotes vasoconstriction by scavenging nitric oxide

Prevention of cardiac surgery-associated AKI

It would be fair to say that nothing has really been demonstrated to have a uniformly protective effect on this outcome, on the basis of wide-scope reviews such as Jacob & Leaf (2019) and Martín-Fernández et al (2024). The poor authors scoured thousands of papers and ended up quoting stuff like the 1964 McDonald paper on the renal effects of dopamine. Brushing aside the nagging sensation that one's time is being wasted, the reading of these systematic reviews suggest that, while there is no proven effective therapy, there are promising ideas that might have applications in individualised patients, such as:

• Corticosteroids
• N-acetylcysteine
• Remote ischaemic preconditioning
• Statins
• Sodium bicarbonate
• Leukocyte filtration
• Renal vasodilators (eg. fenoldopam)
• Nitric oxide
• Balanced crystalloid (mostly extrapolating from data that does not involve cardiac surgical patients, and which is already unclear)
• Tight glucose control
• Use of vasopressin instead of (and in addition to) noradrenaline

One pitches these ideas with nervous side glance at the pile of contradictory literature that produced them, acknowledging that beneath the surface a lot of these are either supported by absolutely nothing, or are in fact refuted by the available data. The only reason these are listed here is because the college rubric for Question 2 from the second paper of 2024 suggested that a candidate who "mentions novel interventions under investigation" is to be upheld as "above expected standard".  At least, using the list of known risk factors, one could construct a series of lame predictable statements about things not to do. These, on close inspection, would seem to resemble the KDIGO Bundle of Care from the original 2012 KDIGO AKI guideline.

• Avoid excessive fluid resuscitation
• Avoid excessive diuresis
• Avoid copious amounts of normal saline
• Watch the renal function carefully, monitoring the urine output and biochemistry
• Avoid haemodynamic insufficiency (i.e. one should aim for a  more normal cardiac output)
• Avoid nephrotoxins, eg. ACE-inhibitors, ARBs, NSAIDs, contrast and aminoglycosides
• Avoid hyperglycaemia (aim BSL ~6-8 mmol/L)
• Avoid weird colloids (eg, starch, except why would anyone ever, etc)