‘Fit for surgery’: the relationship between cardiorespiratory fitness and postoperative outcomes

New Findings What is the topic of this review? The relationships and physiological mechanisms underlying the clinical benefits of cardiorespiratory fitness (CRF) in patients undergoing major intra‐abdominal surgery. What advances does it highlight? Elevated CRF reduces postoperative morbidity/mortality, thus highlighting the importance of CRF as an independent risk factor. The vascular protection afforded by exercise prehabilitation can further improve surgical risk stratification and postoperative outcomes. Abstract Surgery accounts for 7.7% of all deaths globally and the number of procedures is increasing annually. A patient's ‘fitness for surgery’ describes the ability to tolerate a physiological insult, fundamental to risk assessment and care planning. We have evolved as obligate aerobes that rely on oxygen (O2). Systemic O2 consumption can be measured via cardiopulmonary exercise testing (CPET) providing objective metrics of cardiorespiratory fitness (CRF). Impaired CRF is an independent risk factor for mortality and morbidity. The perioperative period is associated with increased O2 demand, which if not met leads to O2 deficit, the magnitude and duration of which dictates organ failure and ultimately death. CRF is by far the greatest modifiable risk factor, and optimal exercise interventions are currently under investigation in patient prehabilitation programmes. However, current practice demonstrates potential for up to 60% of patients, who undergo preoperative CPET, to have their fitness incorrectly stratified. To optimise this work we must improve the detection of CRF and reduce potential for interpretive error that may misinform risk classification and subsequent patient care, better quantify risk by expressing the power of CRF to predict mortality and morbidity compared to traditional cardiovascular risk factors, and improve patient interventions with the capacity to further enhance vascular adaptation. Thus, a better understanding of CRF, used to determine fitness for surgery, will enable both clinicians and exercise physiologists to further refine patient care and management to improve survival.


INTRODUCTION
Surgery is amongst the leading risk factors for mortality and has been estimated to account for 7.7% of all deaths globally (Nepogodiev et al., 2019). By 2030, it is estimated that one-fifth of people aged 75 years and older in the United Kingdom alone will undergo surgery (Fowler et al., 2019). Therefore, to better understand and mitigate this risk, we need to consider not just the disease or surgical procedure, but also the phenotypical response and ability to cope with the physiological insult posed by major surgery. Furthermore, prophylactic intervention targeting modifiable risk factors prior to surgery, a process known as 'prehabilitation' , requires investigation and optimisation. This review explores our relationship with oxygen (O 2 ), the elixir of life, and how its transport and use in the human body determines 'fitness for surgery' .

ORIGIN OF O AND OUR DEPENDENCY ON OXIDATIVE METABOLISM
When the solar system emerged 4.6 billion years ago (Dickerson, 1978), Earth's atmosphere was devoid of O 2 , a vast difference compared with the modern day atmospheric inspired fraction of 20.93%. The emergence of life, likely originating in alkaline thermal vents at the bottom of the oceans, initially gave rise to the domains of archaea and bacteria (Miller & Bada, 1988). Approximately 1.5 billion years ago, cyanobacteria began to release O 2 into the atmosphere (Nisbet & Sleep, 2001). The organic compounds that emerged from the 'primordial soup' were photosynthetic, capturing solar radiation and creating the organic molecule glucose. In turn, the O 2 released into the atmosphere signalled a major evolutionary event, arguably described by two oxidation 'pulses' , the Great Oxidation Event and the Neoproterozoic Event, or as a progressive evolution, the Great Oxidation Transition (Lyons et al., 2014). This gave rise to atmospheric O 2 and the evolution of O 2 -dependent organisms, from primitive eukaryotic unicellular structures performing metabolism, locomotion and reproduction to present day Homo sapiens.

FROM MOUTH TO MITOCHONDRIA: CONVECTIVE AND DIFFUSIVE DETERMINANTS OF O 2 TRANSPORT
Early measurements describing O 2 uptake (V O 2 ) in humans at the onset of intense movement were conducted by Hill and Lupton (1923) and demonstrated a rapid and exponential response, as skeletal muscle has the capacity to increase rate of metabolism by an astounding 50-to 100-fold above its resting requirements. This challenges a rapid delivery of O 2 to the mitochondrial inner membrane for use as the terminal electron acceptor, whereby oxidative phosphorylation generates ATP. O 2 is transported by convection, which describes its movement within the airways and circulation-driven aero-and hydrostatic pressure gradients, and by diffusion, the passive movement down a concentration gradient such as between the alveolar compartment and pulmonary capillary bed and between the systemic microcirculation and tissue. Figure 2 illustrates the major organs and processes, both convective and diffusive, that describe the 'O 2 cascade' . Following inspiration of air into the lungs, O 2 diffuses down a concentration gradient at the alveolar-capillary membrane, minimally dissolves in plasma and predominantly binds with haemoglobin (Hb), an allosteric protein with affinity for four molecules of O 2 . Deoxygenated venous blood is therefore saturated with O 2 in the pulmonary capillaries, the concentration of which is proportional to the concentration of Hb, its P 50 and the partial pressure exerted by O 2 on the plasma at a given temperature (Henry's law). Oxygenated blood then travels the vascular system driven by the heart. This convective component is referred to as 'O 2 delivery' (Q O 2 ), the product of cardiac output (Q) and arterial O 2 content (Q × C aO 2 ), and is complete when O 2 diffuses across the microcirculatory capillary beds and reaches the mitochondrial matrix where it is used as the terminal electron carrier.V O 2 as described by Fick's principle is equal to the product ofQ and the difference between arterial and mixed venous oxygen content (C aO 2 -C̄v O 2 ).
Notably, in health, the principal 'rate limiting' steps for maximal O 2 uptake (V O 2 max ) are attributed to the perfusive (Q O 2 ) and diffusive components of the cascade (Wagner, 2000).

METRICS AND MEANING: ASSESSMENT OF CARDIORESPIRATORY FITNESS
The advent of breath-by-breath measurement technology has allowed us to measure the capacity of the O 2 transport system and determine metrics describing the magnitude of cardiorespiratory fitness (CRF), which not only describes an individual's ability to perform physical activity, but is linked to cardiovascular health (Ross et al., 2016) and

New Findings
• What is the topic of this review?
The relationships and physiological mechanisms underlying the clinical benefits of cardiorespiratory fitness (CRF) in patients undergoing major intraabdominal surgery.
• What advances does it highlight?
Elevated CRF reduces postoperative morbidity/ mortality, thus highlighting the importance of CRF as an independent risk factor. The vascular protection afforded by exercise prehabilitation can further improve surgical risk stratification and postoperative outcomes. longevity (Blair et al., 1989). Cardiopulmonary exercise testing (CPET) is used to objectively measure the ability of a patient to uptake O 2 and typically involves an incremental exercise test to symptom-limited exhaustion. CPET can also identify underlying pathology and evaluate the impact of chronic comorbidities on O 2 uptake. Recently, the use of CPET has been widely adopted in patients prior to major surgery and approximately 30,000 tests are conducted annually in the UK alone (Reeves et al., 2018). These data are used to support patient care decisions, plan appropriate postoperative critical care, and direct prehabilitation programs aimed at improving CRF .
Three primary metrics describing CRF are typically reported when conducting CPET: 1. Peak oxygen consumption (V O 2 peak ), defined as theV O 2 attained during an incremental test to exhaustion, expressed in absolute terms (ml min −1 ) or relative to body mass (ml kg −1 min −1 ), which can be subject to allometric scaling, and measured as the highest value recorded, often occurring during the final 20 s of a test.
WhilstV O 2 peak is reflective of a patient's 'best effort' , it may not necessarily reflect a true highest value, defined asV O 2 max with an observed plateau present in the O 2 uptake work-rate slope of Hill & Lupton (1923) demonstrated in Figure 3. Controversy exists here, and evidence suggests only a minority of continuous tests, even in young healthy people, yield a measurable plateau (Day et al., 2003;Poole & Jones, 2017). Nevertheless, an exercise test to exhaustion is important since it allows for the site of transport limitation across the O 2 cascade to be identified (Wagner, 2000).
2. Anaerobic threshold (AT), a submaximal index of CRF defined as theV O 2 above which anaerobic metabolism supplements oxidative phosphorylation with additional carbon dioxide (CO 2 ) production, creating a deflection point on a plot of pulmonary CO 2 output versus O 2 uptake ( Figure 4). The AT is also commonly reported as a percentage ofV O 2 peak orV O 2 max . Whilst the AT signifies a transition where increased glycolysis raises the muscles lactate efflux into the blood above its removal rate with associated metabolic acidosis, a multitude of definitions and controversies exist (Poole et al., 2021). Thus in the context of preoperative CPET, AT refers to the gas exchange threshold (GET, sometimes also referred to as the ventilatory threshold), typically measured using the 'gold standard' V-slope (Beaver et al., 1986) method of determination.
GET is expressed in ml kg −1 min −1 or ml min −1 .
3. The ventilatory equivalent for carbon dioxide (V eqCO 2 ), defined as a ratio of minute ventilation to CO 2 production and usually reported at the GET. V eqCO 2 reflects the composite efficiency  The oxygen (O 2 ) transport system characterised by pulmonary ventilation, alveolar-capillary diffusion, circulatory perfusion driven by the cardiovascular system and diffusion across the microcirculatory capillary beds. The volume of O 2 transport, described by Fick's principle, is determined by the product of convective (cardiac output) and diffusive O 2 transport terms (and is the product of cardiac output and the difference between the O 2 content of arterial and venous blood) Workrate (

CRF AND SURGERY: LINK TO SURVIVAL
Mortality following major surgery is a significant risk despite progress being made in surgical technologies, anaesthesia and peri-operative care. In colorectal surgery, mortality is reported at 3.2% within 90 days (NBOCA, 2017) with complication rates above 30% (Lucas & Pawlik, 2014 Schematic representation of the V-slope method (Beaver et al., 1986) for estimation of the gas exchange threshold (GET) during CPET. GET is identified at the intersection of two linear sections of theV CO 2 -V O 2 relationship, represented by the continuous black line. A further deflection point in the relationship may be observed during the latter stages of CPET and represents respiratory compensation (see Older et al., 1993, below) is associated with reduced survival and increased morbidity following major surgery (Moran et al., 2016;Smith et al., 2013).
The seminal work of Older et al. (1993) first described an association between preoperative CRF and postoperative outcome.
They studied 184 elderly patients undergoing elective major intraabdominal surgery and established that patients classified as 'unfit' exhibited markedly higher mortality rates than those deemed 'fit' (18% vs. 0.8%, P < 0.001). Patients were considered unfit by preoperative CPET if O 2 uptake at GET was <11 ml O 2 kg −1 min −1 , a value originally described by Weber and Janicki (1985) that characterised the GET in patients with moderate to severe heart failure. Studies have since used the GET as a measure of CRF, and further supported the inverse association between CRF and postoperative mortality and morbidity in patients undergoing a variety of intra-abdominal surgeries (Table 1).
A theoretical model ( Figure 5) originally developed by Clegg et al.
(2013) helps visualise why elevated CRF is associated with improved postoperative outcome. The model describes potential differences in surgical outcome between a hypothetical patient who is unfit for surgery (for example with a GET < 11 ml O 2 kg −1 min −1 ) compared to a patient deemed fit (GET ≥ 11 ml O 2 kg −1 min −1 ). The unfit patient is more likely to require care in a high dependency unit or intensive care unit with a greater likelihood of complications and risk of mortality, whereas the fit patient may experience a normal and faster recovery on the ward.
Given the importance of assessing CRF in clinical practice, the American Heart Association has published a scientific statement promoting CRF as a clinical vital sign (Ross et al., 2016).

MECHANISTIC LINK BETWEEN CRF AND POSTOPERATIVE OUTCOME
The model presented ( Figure 5) presumes the existence of an obligatory baseline level of CRF (such as the threshold values foṙ V O 2 peak , GET or V eqCO 2 found in Table 1) Clegg et al. (2013). The green plot represents a patient considered (CRF) 'fit' for surgery whereas the red plot represents a patient classified as 'unfit' . The dashed line represents the cut-off between independent patient recovery typically requiring ward-based care, and dependent recovery requiring high dependency unit or intensive care unit admission TA B L E 1 Studies demonstrating an association between CRF and postoperative outcome following non-cardiac intra-abdominal surgery, adapted from (Moran et al., 2016) et al., 2006, 2007). This is particularly prominent during abdominal surgery given the potential for ischemia-reperfusion, leukocyte activation, mitochondrial dysfunction and concurrent depletion of antioxidants in the postoperative period due to increased consumption (Bailey et al., 2006;Musil et al., 2005;Thomas & Balasubramanian, 2004). During laparoscopic surgery, for example, increases in intraabdominal pressure during pneumoperitoneum may cause splanchnic ischemia-reperfusion and subsequent oxidative stress (Leduc & Mitchell, 2006). Furthermore, a reduction inQ contributing to decreased O 2 delivery is observed as systemic venous return is reduced to the right side of the heart and pulmonary venous return

POTENTIAL MECHANISMS THAT ENHANCE SURVIVAL
Whilst mechanistic bases explaining the link between (elevated) CRF and postoperative outcome require further elucidation, evidence demonstrates that patients with low CRF are associated with poor postoperative outcome, likely explained by the prevailing magnitude of perioperative O 2 deficit. Importantly, CRF is a modifiable risk factor and a primary component of prehabilitation strategies (Macmillan, 2019;Tew et al., 2018). Prehabilitation represents an opportunity to improve patient preparation for surgery and is multi-modal in nature comprising exercise training and improving nutritional and psychological status . Prehabilitation aims to improve patient CRF to better tolerate the surgical stress response, leading to a reduced risk of perioperative complications and improved postoperative outcome (Tew et al., 2018). The theoretical potential for this strategy is illustrated in Figure 9.
Few studies have investigated the potential to improve CRF prior to surgery using exercise interventions and those that have mainly comprise small sample sizes demonstrating proof of principle (Rose et al., 2020;Simonsen et al., 2020). West et al. (2015)  From a mechanistic perspective, similarities exist between the physiological insult of surgery and the acute response to an exercise stimulus. Primarily, an increased cellular demand for O 2 , consequent to oxidative phosphorylation required to regenerate ATP, is required to enable continued physical activity. As a chronic adaptive response to exercise, an improved ability to increaseV O 2 is associated with elevated mRNA of peroxisome proliferator-activated receptor γ coactivator 1-α (Gibala et al., 2009), a moderator of skeletal muscle mitochondrial biogenesis. An increase in citrate synthase (a marker of muscle oxidative capacity) has also been reported (Burgomaster et al., 2005), and an increase in oxidative stress (Bailey et al., 2010(Bailey et al., , 2018Davies et al., 1982;Radák et al., 1999), which is attenuated following exercise training (Fatouros et al., 2004).
The mechanisms of this exercise-induced response have been linked to improvements in total antioxidant capacity (Fatouros et al., 2004;Radák et al., 1999), which is considered a marker of the body's defence system to neutralise excessive and deleterious free radical and associated ROS formation (Ghiselli et al., 2000). Total antioxidant capacity has been enhanced following exercise training in both animal (Liu et al., 2000) and human (Fatouros et al., 2004)  Dependent recovery Cardiorespiratory fitness Surgery Time F I G U R E 9 The fundamental principle underlying exercise prehabilitation whereby CRF is improved prior to surgery, thus reducing the risk of postoperative complications, and enhancing recovery as indicted by the green plot. Adapted from Clegg et al. (2013). The dashed line represents the cut-off between independent (ward-based care) and dependent (high dependency unit, intensive care unit) patient recovery been shown to improve following HIIT (Calverley et al., 2020;Molmen-Hansen et al., 2012), the potential consequence of an 'optimised' blood flow-shear phenotype, triggering calcium influx into the hyperpolarised endothelial cells (Cooke et al., 1991) upregulating endothelial nitric oxide synthase (Bolduc et al., 2013).

OPTIMISING RISK QUANTIFICATION AND PATIENT MANAGEMENT
The evidence reviewed suggests that impaired CRF is both an independent and a modifiable risk factor associated with postoperative outcome. Yet the strength of this relationship, used to predict postoperative outcome, is not effectively compared against traditional cardiovascular risk factors such as ischaemic heart disease, lung disease, or diabetes and obesity. This comparison has been addressed epidemiologically for all-cause deaths (outside of the surgical setting) within the Aerobics Centre Longitudinal Study, in which low CRF was found to be a greater risk factor than hypertension, smoking, high cholesterol, diabetes and obesity (Blair, 2009).
Attributable fractions describe the percentage of deaths that would not occur if a risk factor were removed from a population and account for both the risk of mortality associated with that condition and its prevalence in the population, as illustrated in Figure 10. This approach could be conducted in the surgical setting to help optimise risk quantification and further highlight the clinical importance of CRF relative to traditional risk factors.
Like all biomarkers, CRF is a dynamic metric subject to natural variation and thus needs to be interpreted with caution. Such variation encompasses both analytical and biological components which can be described using the concept of critical difference, indicative of the magnitude of variation around a true homeostatic point at any given time. Rose et al. (2018b) introduced the concept of critical difference to preoperative CPET and found differences of ± 19%, 13% and 10%  Table 1 highlights that many studies, including the seminal work of Older et al. (1993), have simply adopted threshold values developed by other studies sometimes using different patient populations and surgical procedures. Furthermore, CRF is commonly described usingV O 2 peak , GET or V eqCO 2 as discussed; however, alternative metrics may provide superior prognostic utility in some settings. For example, if a patient is unable or unwilling to exercise to exhaustion, a submaximal measure of CRF relating O 2 consumption to workload achieved, such as the O 2 uptake efficiency slope (OUES; Hollenberg & Tager, 2000;Bongers et al., 2017), may be more effective.
Female inclusion rate in peer-reviewed publications of perioperative CPET is reported at only 31% and may have a bearing on the interpretation of data (Thomas et al., 2020). Surprisingly, despite evidence that CRF is lower in females across the lifespan, given smaller body size, skeletal muscle mass, peak cardiac output and Hb concentration (Jackson et al., 2009;Fleg et al., 2005), sex is not considered during surgical risk stratification. If a simple doseresponse relationship exists between low CRF and postoperative survival, we would expect females to be at increased risk given these congenital constraints. Furthermore, other risk factors such as cardiovascular disease (CVD), which may vary between the sexes, require investigation to appraise a potential compensatory effect for CRF and consequent changes in its prognostic potential on postoperative outcome.

CONCLUSION
The current review has explored the intimate relationship between O 2 transport and postoperative outcome, emphasising how preoperative CRF is an independent risk factor for postoperative mortality and morbidity, when patients undergo major intra-abdominal surgery.
There is increased O 2 demand during the perioperative period and patients must meet this demand to avoid tissue hypoxia, the presence and magnitude of which dictates postoperative morbidity and mortality. This relationship can be used to assess patient risk, plan perioperative care and optimise patient management using exercise as a modifiable intervention. However, there is a clear need to improve the physiological detection and interpretation of CRF, better quantify risk to specific populations, sex and surgical procedure, and better understand the optimal management of patients including the mode of exercise intervention and its timing. Collectively, a better understanding of CRF used to determine fitness for surgery will enable clinicians and physiologists alike to direct patient care more effectively and ultimately improve survival.