Goal Directed Perfusion: Genuinely Beneficial or Smoke and Mirrors? By Gary Grist RN CCP Emeritus
- History of Goal Directed Therapy.
Goal Directed Perfusion (GDP) is a concept inspired by Goal Directed Therapy (GDT). GDT was first described in 2001 by Rivers (1). The technique uses rigorous monitoring and intensive management of hemodynamics in high-risk perioperative patients as a means to improve outcomes (2). Adults admitted to the ICU for different conditions needing aggressive volume resuscitation and treated by GDT appear to have reduced mortality, shorter length of stay (LOS) in the ICU as well as less time on mechanical ventilation (3). However, this is not necessarily true of post-cardiotomy patients. In one recent meta-analysis, GDT used to prevent or reverse hypoperfusion after cardiotomy provided no improved outcome over conventionally treated patients in terms of total mortality, mechanical intervention, or LOS (4).
- GDP proposed.
Despite these ambivalent findings, some well-respected perfusionists think that the concept of GDT might still be beneficial to CPB patients (5). Utilizing special methods different from, but akin to GDT, GDP might reduce the common complication of acute kidney injury (AKI) often seen after CPB (6). In one small study, increasing oxygen delivery (DO2) on CPB to 300+ mL O2/min/M2 and maintaining a mean arterial pressure of 60+ mmHg throughout CPB reduced the incidence of AKI from 23.9% to 9.1% (7).
- Greatest proponent of GDP.
Perhaps the greatest proponent of GDP is the group headed by Marco Rannuci, MD. He is the principal in a clinical trial which plans to enroll 700 CPB patients by October 2018 (8). The premise of the trial is that DO2 is an important factor in preventing AKI with a minimum DO2 value of 262-272 mL O2/min/M2 during CPB. The Rannuci group focuses on pump flow, perfusion pressure and hemodilution during CPB as contributing causes of AKI. They nevertheless acknowledge that as many as 15 other factors outside of the perfusionist’s control (EUROSCORE modifiers) can play a significant role as well (9). He adds a new modifier in the form of carbon dioxide production (VCO2) during CPB with a critical ratio of DO2/VCO2 less than 5.3. In other words, if the DO2 is 270 mL O2/min/M2, then the CO2 produced should be less than 50.9 mls/min/M2. Elevated CO2 production is thought to either correlate with or directly cause an increase in lactate; a marker for inadequate oxygenation.
- Evolution of clear prime.
The adoption of clear primes had a significant impact on DO2 during and after CPB. Back in “my day” in 1968, (Oh boy! Here it comes!), we originally primed our adult circuits with either fresh heparinized or citrated whole blood. The concept of using a clear prime was still being tested by individual programs cautious about the risks of hemodilution such as low hemoglobin, low albumin as well as the dilution or destruction of coagulation factors like fibrinogen and platelets. This was despite the reported use of hemodilutional prime eight years earlier in 1960 (10). As disposable oxygenators and circuits with smaller prime volumes started coming to market, the trend towards hemodilution became more common. It was not unusual to cut the hemoglobin/hematocrit by 50% with hemodilution during CPB. However the backstop to this was the common use of hypothermia which greatly reduced oxygen consumption and the need for high DO2 which was difficult to achieve at normothermia with hemodilution. As long as the patient was cool, it was thought there was no harm.
- Salvaging the RBC mass.
The problem came when it was time to rewarm and wean the patient. Half of the patient’s red blood cell (RBC) mass was in the pump after weaning. Fresh blood was often started by IV as soon as the pump stopped. The problem was salvaging the circuit RBCs for return to the patient. We had no auto-transfusion equipment to re-concentrate the leftover circuit volume. We had no hemoconcentrators. We could slowly transfuse the volume (2-3 liters) into the patient, but that slowed down decannulation and messed up the protamine heparin reversal, not to mention possible problems with fluid overload. We could bottle the leftover circuit volume (we were still using glass IV bottles because plastic IV bags were yet to become widely available) and give it to anesthesia. They would administer it as rapidly as was safe, letting the patients’ kidneys discard the excess fluid. But the heparinized circuit blood would sometimes overtake the protamine reversal and the patients often started bleeding again in the ICU. The consensus back then was that circuit blood was too dangerous to give back to the patient for a variety of reasons. (It was usually flushed down the hopper in the ante-room….at a time when ORs were permitted to have hoppers.) I always felt that the normothermic delay in returning the hemoglobin to near normal played some role in post-cardiotomy complications (duh!) in the 1960s and that is still true today, 50 years hence, despite all our modern tools for RBC salvage and reduced circuit prime (11). Back then if an allogenic blood transfusion was given to speed the hemoglobin normalization, even more pulmonary, renal, neurologic, inflammatory, and immunologic complications (which people did not necessarily associate with transfusions at the time) could ensue. During that period in medical history, blood transfusions did not carry the stigma that they do today.
- The need to calculate DO2.
What does all that have to do with GDP? During CPB back in the 1960s, I was constantly calculating DO2 with pencil, paper, and my own math skills (there were no portable electronic calculators back then, let alone a computer with “fill in the blank” formulas). This became doubly important when hemodilution came into common use so we could document for the surgeon (and any other interested parties) that the patient was being adequately perfused during the case. This was also about the time that a surgeon told me, in all seriousness, that any “monkey” could run a heart pump (12). I have yet to see a monkey perform the calculations needed for the Fick formula (although I may have seen a few operate a heart pump).
- Method to calculate DO2 and oxygen consumption.
The pump RPM tachometer gave me the pump speed. I multiplied that by the stoke volume of the raceway to get the total blood flow. Then I would divide the total blood flow by the body surface area that I calculated beforehand from a nomogram. That would give me the blood flow per square meter (the cardiac index). Then I would calculate the oxygen delivery based on the hemoglobin carrying capacity and the hemoglobin from lab results. Once the venous oxygen saturation results were called into the room from the lab (we did not have SVO2 monitors or point-of-care testing then, either), I would calculate the oxygen consumption
- Calculating for body mass.
At normothermia my goal would be to provide a DO2 between 250 and 300 mlsO2/min/M2 depending on lean body mass. I used another nomogram to determine the ideal body weight for the height, sex and age of the patient. If the patient weighed 20%+ over the ideal, I would consider the patient chubby; lots of fat which needed less oxygen. The lean patients got the higher DO2 (if I could pump that fast with a low hemoglobin) and the fat patients got the lower DO2. I don’t know if this was the right thing to do, but at the time it was our practice.
- Recalculating for temperature and blood flow changes.
As I reduced the patient’s temperature, I would recalculate DO2 and oxygen consumption. My goal of perfusion then was focused on achieving a suitable oxygen delivery based on the body temperature and the reduced blood flow, if any, ordered by the surgeon. There was a lot of guesswork that math couldn’t help me with. For example, what was the safe DO2 at 32 or 28 or even 18 degrees centigrade? This was primarily a guess based on oxygen consumption being halved with each 10-degree drop in temperature (known as the Q10). I pretty much ignored the oxyhemoglobin dissociation curve which inhibits the detachment of oxygen from oxyhemoglobin as the temperature decreases. However, I used what is now called pH stat gas control which helped to detach oxygen from the cold oxyhemoglobin. I did not learn until 30 years later that brain oxygen consumption decreased only half as much as other organs with a temperature reduction and that cold brains utilize mostly dissolved oxygen (13). In hindsight I should have pumped a little faster to protect those brains than what my calculations indicated, but fortunately, the high FiO2 used on my disk and bubble oxygenators provided an abundance of dissolved oxygen.
- How the importance of DO2 faded over the years.
Those were some of the goals I would use at a time when CPB was more of a “fly by the seat of your pants” method than a formalized process. As the years passed and circuit primes became smaller, perfusionists became accustomed to clear primes and hemodilution. Real-time physiologic monitoring also came into being. As a result, the need to calculate DO2 during a case seemed to go by the wayside. Decades later when I began working with students, I noticed that they were being trained to simply select a cardiac index, commonly ranging from 1.8 to 2.2 L/min/M2, as a target blood flow without really thinking much about DO2 except to maintain the SVO2 above 70%. I always taught them to use a minimum of 2.5 L/min/M2. I don’t know where the students’ lower figures came from. Mine came from clinical experience. Many of the students were also learning from other clinical rotation sites to use vasopressors rather than blood flow or sweep gas CO2 manipulation to control perfusion pressure. Mean arterial pressure normalized by vasoconstriction of the arterioles with drugs is no substitute for adequate blood flow.
- The real reason GDP may be beneficial.
So, what do I think of the new GDP concept? Firstly, it is not a new concept, as some believe (5). It is just something that perfusionists stopped doing long ago when circuits became smaller and clear primes were not a potential threat. (I include myself in that neglect, mea culpa!) Also, the premise that GDP must meet higher DO2 levels may be incorrect. Marginally increasing oxygen delivery may not be all that important. Rather, the increased flow required to meet specific DO2 goals may instead be needed to restore or maintain adequate perfused capillary density to prevent the formation of a hypercapnic or anoxic lethal corner which can lead to AKI, organ failure, or even death. Perfused capillary density and the lethal corner are concepts explained by the oxygen pressure field theory (OPFT). Unfortunately, I do not have the space in this article to explain the OPFT in detail. But if you are interested you can find information about the OPFT on my website; <Perfusiontheory.com>.
- Final answer.
So, is GDP genuinely beneficial or just “smoke and mirrors”. I would say it is beneficial but not because of marginally increased DO2. That, I believe, is a misleading or irrelevant assumption that obscures the truth of why patients may do better. However, keeping the DO2/VCO2 ratio low to prevent the formation of a hypercapnic lethal corner may be closer to the reality of why GDP works.