Does functional evaluation before lung cancer surgery need reappraisal?

There is nothing wrong with change, if it is in the right direction.

–Winston Churchill

The most frequently cited algorithms for functional assessment before lung resection are those published by the European Respiratory Society/European Society of Thoracic Surgeons (ERS/ESTS, 2009), the British Thoracic Society (BTS, 2010), and the American College of Chest Physicians (ACCP, 2013) [1–3]. All of them include a preliminary cardiac evaluation, lung function tests, and according to the results, additional exercise tests. The aim is to categorize patients as low, moderate or high risk for complications before thoracotomy and major lung resection, i.e. lobectomy or pneumonectomy. According to the most recent recommendations [2], patients classified as high risk ‘should be counselled about alternative surgical (minor resections or minimally invasive surgery) or non-surgical options’.

Although published, none of these algorithms have been validated by prospective studies. However, the recommended lung function tests and the associated cut-off values have been broadly accepted and implemented in the selection process of lung resection candidates. On the other hand, video-assisted thoracoscopic surgery (VATS) has already replaced open thoracotomy (open) for early-stage lung cancers and is also currently used by many surgeons for locally advanced lung cancers. As a result, updating the functional assessment before lung resection would be a means of bringing this evaluative tool up to speed with today’s most cutting-edge thoracoscopic surgery techniques. Indeed, the algorithms have not been designed to select patients before sublobar resection and/or mini-invasive surgery, and no specific recommendations are made regarding minor resections or minimally invasive surgery counselled to high-risk patients. In addition, besides the significant increase in use of these surgical techniques, advances in anaesthesiology and perioperative management have likely contributed to the decrease in postoperative mortality and complications observed over the last decades. Furthermore, the fact that most of the studies on which the 2009–2013 recommendations are based were published before the 2000s [1–3] is an additional point that illustrates the need to re-evaluate the functional assessment before lung resection, including the relevance of the tests and parameters included or to be included in the algorithms, the sequence of the steps and the cut-off values.

The first step of the current algorithms is a cardiac evaluation, based on the guidelines of the American College of Cardiology and the American Heart Association. The ERS/ESTS and the ACCP recommend the use of a score system, either the revised cardiac risk index (RCRI) or the thoracic revised cardiac risk index (Th)RCRI [1, 2], based on medical history, physical examination, baseline ECG and plasma creatinine measurement. In summary, patients with ThRCRI beyond the threshold, or any cardiac condition requiring medication, or a newly suspected cardiac condition or limited exercise tolerance (inability to climb 2 flights of stairs), should be referred for a cardiac consultation. Interestingly, Falcoz et al. [4] and Kneuertz et al. [5] observed similar cardiac complication rates in open and VATS groups. Moreover, Law et al. [6] found that cardiovascular complications were more strongly associated with mortality in the VATS cohort than in the open cohort (odds ratio [OR], 2.19). Consequently, these data clearly point to the fact that cardiac assessment before mini-invasive surgery is necessary and cannot be foregone.

Lung function assessment requires the measurement of forced expiratory volume in 1 s (FEV1) and diffusing capacity of the lung carbon monoxide (DLCO). The use of calculated residual lung function after surgery [predicted postoperative (ppo) values] should be preferred to that of preoperative FEV1 and DLCO values, since both the patient’s lung function and the extent of resection are considered. The formula used to calculate ppo FEV1 or DLCO are based on the number of segments. As a result, one may speculate that ppo values may be included in the algorithms regardless of the intended resection, from segmentectomy to pneumonectomy. However, we should keep in mind what the concept of ppo FEV1 covers. FEV1 stabilizes several months after lung resection. The lowest FEV1 is measured the first day after surgery and is lower than the ppo FEV1 [7]. After 12 months, the loss of FEV1 is grossly −5% (−2% to −7%) after segmentectomy and −11% (−8% to −13%) after lobectomy, and FEV1 is higher than the ppo FEV1 [8]. Additional factors affect the residual lung function. Indeed, ppo and measured FEV1 at 3 months differ according to the degree of preoperative lung function impairment [9]. Furthermore, the functional impact of the resection differs according to the anatomical situation of the removed segments [10]. Eventually, a better compensatory adaption of the remaining lung after lobectomy, rather than after segmentectomy, could reduce the functional benefit of segmentectomy [11]. Consequently, ppo values are not used to predict residual FEV1 or DLCO, but to evaluate the risk of perioperative complications. In a study of >1400 subjects undergoing lung resection, Alam et al. found that the OR for postoperative respiratory complications increases as ppo FEV1 and ppo DLCO decrease. Specifically, a 10% increase in the risk of complications is associated with every 5% decrease in ppo lung function [12]. Preoperative risk stratification based on lung function tests before VATS has been called into question [13]. However, an increase in risk of complications in patients with low ppo DLCO has also been described after VATS [14, 15]. In a series of 1088 patients who underwent robotic lobectomy, pulmonary complications were still associated with ppo FEV1 or ppo DLCO ≤50% on multivariable analysis [16].

If assessing ppo FEV1 and ppo DLCO before mini-invasive surgery is likely still of use, the cut-off values applied to classify patients as low, moderate, or high risk for complications deserve to be reconsidered. A few studies provide lung function evaluation after VATS. In a small series of patients, 7 days after surgery, FEV1 tended to be lower after thoracotomy than after thoracoscopy (open: 88.6 ± 13.1%; VATS: 78.2 ± 12.9%) [17]. However, 3 months after surgery, a propensity analysis of 83 well-matched pairs of open and VATS patients found no further difference [18]. The most striking results supporting the re-assessment of current recommendations are the low mortality and complication rates recorded after VATS in patients with compromised lung function. In a large series of patients from the Society of Thoracic Surgeons General Thoracic Database, 4215 propensity-matched patients per group were analysed. For patients with ppo FEV1 <40%, mortality was greater in the open (4.8%) than in the matched VATS group (0.7%). Similar results were observed for ppo DLCO <40% (5.2% in the open, 2.0% in the VATS group). The rate of complications was significantly greater at ppo FEV1 <40% in the open (21.9%) than in the matched VATS (12.8%) group, with similar results with ppo DLCO% <40% (14.9% in the open, 10.4% in the VATS group) [19]. In a previous retrospective series of 12 970 patients from the Society of Thoracic Surgeons Database, complications were significantly lower in the VATS compared with the open group for patients with FEV1 <60% of predicted; however, no matching of patients was done [20]. Once again, these results suggest that lowering the ppo FEV1 and ppo DLCO cut-off values used to define patients as high risk for complications before VATS should be examined.

Exercise tests are the most frequent additional tests performed to assess fitness before lung surgery and are included in the 3 main published guidelines. Three questions deserve to be examined: are cardiopulmonary exercise tests (CPET) or other exercise tests able to identify patients at high risk for complications after VATS? Should the maximal oxygen consumption (VO₂max) threshold, below which patients are considered as high risk for complications, be revised? Do we need to revise the FEV1 and DLCO cut-off values below which exercise tests are prescribed, or offer exercise tests, especially low technology ones, to all patients? The prognostic value of exercise tests before lung resection is still questioned [3, 21]. In the general population, including patients and healthy subjects, better cardiorespiratory fitness is associated with lower risk of all-cause mortality [22]. In lung cancer patients, a meta-analysis showed that the lower the VO₂max, the higher the risk of lung resection complications [23]. The relevance of using exercise tests to select patients before surgery is that the cardiorespiratory response to exercise mimics the capacity of the body to cope with surgical stress and possible surgical complications. VO₂max has demonstrated its ability to identify patients at increased risk for postoperative complications after major non-cardiac surgery [24]. Another rationale for using CPET is the detection of the limiting factor of O₂ supply to the body.

With regard to the second question, the VO₂max level below which patients are at risk for complications in the perioperative period remains to be determined. Most experts agree that a VO₂max >20 ml/min/kg is associated with a low risk, and patients with VO₂max >15 ml/min/kg are considered at moderate risk for complications [2]. A VO₂max <10 ml/min/kg is associated with a high risk for complications. Should a lower cut-off in the case of VATS procedures be adopted? To our knowledge, no study specifically addressed the question of VO₂max measurement before VATS. However, if complications are less frequent when using mini-invasive procedures, severe complications can still occur. A 10 ml/min/kg VO₂max reflects poor aerobic capacity, as illustrated by the following values: a 4-km/h walk requires around 10 ml/min/kg VO₂, shopping requires 11 ml/min/kg VO₂, and dressing and bathing require 8 ml/min/kg VO₂. Besides VO₂max, other parameters such as desaturation >4% [25, 26] and respiratory equivalent VE/VCO₂ measured during CPET may also help identify patients at risk.

The lack of CPET availability has prompted studies on low technology and field tests. Some of them, such as stair-climbing tests and shuttle tests, are part of the published algorithms. Others that are easier to integrate into an established routine (e.g. 6-minute stepper test or sit to stand test) are currently being evaluated. Their ability to predict complications [27], as well as their role and place in the process of patient selection and rehabilitation, deserves to be examined.

Over the last decade, patient outcomes after thoracic surgery have changed—mainly because of the adoption of the concept of enhanced recovery after surgery (ERAS) [28]—but not the way in which these patients are assessed preoperatively. To provide a more comprehensive preoperative evaluation approach, the next 2 points definitely deserve mention. First, using a physical activity questionnaire such as the Duke Activity Status Index [24] to make a more refined selection of patients could be interesting to explore, not only for cardiac risk assessment but also for the further development of patient-reported outcome measures, which could be added to the existing algorithms. Second, the preoperative investigation of nutritional status is also of utmost importance, with sarcopenia playing a key role. Indeed, a clear association between sarcopenia (or pre-sarcopenia) and physical function among preoperative patients with lung cancer has been demonstrated [29]. Sarcopenia also impacts survival after resected lung cancer, independently of tumour stage [30, 31].

In the current era of thoracic surgery, functional evaluation before lung cancer surgery needs reappraisal. With the emergence of minimally invasive surgery, the cut-off values used to classify patients as low, moderate or high risk for complications have changed and definitely deserve to be reconsidered. Some other tools directly involving the patient (patient-reported outcome measures) or related to the preoperative nutritional status could be added to more precisely screen high-risk patients and refer them to specific tailored prehabilitation programs.

In addition, the development of prehabilitation raises the question of whether a completely different approach to preoperative evaluation—maybe other than a step-by-step design—is necessary. We must also bear in mind that almost all published studies evaluate the ability of a given test, or a few tests, to predict postoperative complications within a very specific setting and context. Studies aiming to validate a global preoperative strategy, as done by C. Bolliger’s team [32], are very much lacking. As a result, the forthcoming recommendations and algorithms will need to be prospectively validated. In addition, mortality, complications, long-term disability and quality of life should be measured in operated as well as in non-operated lung cancer patients for functional reasons, to evaluate the benefits and risks of each treatment.

Acknowledgement

The authors thank E.F. for her expert editorial review of the manuscript.

Conflict of interest: none declared.

REFERENCES