Agreement between observed and predicted postoperative forced expiratory volume in one second, forced vital capacity, and diffusing capacity for carbon monoxide after anatomic lung resection

Introduction

The recently published results of JCOG 0802 and CALBG 140503 demonstrated no clinically meaningful difference in lung function loss between patients randomised to sub-lobar resection or lobectomy for lung cancer (1,2). The findings were both surprising and counterintuitive given the inclusion criteria of small lesions less than 2 cm and led many surgeons to question if current assessment of postoperative lung function is accurate and pertinent.

Prediction of postoperative lung function is most commonly achieved with segment counting with the assumption that each segment contributes 5% of lung function. It is the principal method recommended in surgical guidelines worldwide (3,4) and currently how most surgeons predict anticipated losses in lung function for patients who are considering surgery for lung cancer.

Despite its importance, little work has been undertaken on the agreement between expected lung function loss and actual observed values. There is currently no information with regards to the validation of segment counting between effort-dependent measures such as forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) (as reported by trials) and effort independent measures such as transfer factor or diffusing capacity for carbon monoxide (DL_CO). Previous studies have reported conflicting evidence in relation to the residual lung function preservation benefits of segmentectomy in comparison to lobectomy (5,6). Possible hypothesis includes variation of lung function at different test time intervals and differing degrees of compensation by amount of resection (7).

In this study, we collated lung function study results before (closest to surgery) and after lung resection surgery (furthest from index surgery) with the aim to examine the agreement between predicted postoperative lung function through segment counting on both effort-dependent and -independent measures of lung function. Our aim is to determine the agreement of postoperative lung function prediction for patients through segment counting to inform surgeon and patient decision-making processes and future guidelines. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1390/rc).

Methods

Participants

We performed a retrospective analysis of prospectively collected data from patients undergoing lung surgery at a tertiary academic hospital. We included all patients undergoing lung resection at the Royal Brompton and Harefield hospitals and excluded those not receiving anatomic lung resection (e.g., wedge resections) or if there was no paired lung function testing both before and after surgery. Anatomical lung resections were performed using stapler-based (fissure and segmental plane) approaches or open blunt dissection of the segmental planes specific to the operating surgeon.

All formal lung function test was performed according to the United Kingdom Association for Respiratory Technology & Physiology guidelines. Patient results were reported as percentage predicted values adjusted for gender, age, height and weight and obtained from existing hospital electronic records (T.C.A.).

We used a 20-segment model (ten segments for each lung) with each segment contributing to 5% of lung function. Segment counts were assigned as follows: right upper lobe [3], middle lobe [2], left or right lower lobe [5], left upper division/tri-segmentectomy [3], left lingula [2] and left upper lobe [5]. The reason for a 20-segment model was because of lack of consensus on how nine segments are allocated in the left lung (either left upper division having two segments or left lower lobe having four segments).

The predicted postoperative values for FVC, FEV1 and DL_CO were calculated using the following formula: predicted postoperative value = preoperative value × (20 − the number of resected segments)/20.

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Individual patient consent and the National Health Service (NHS) Research Ethics Committee (REC) review for this retrospective analysis were waived.

Statistical analysis

We plotted longitudinal plots for individual patients’ lung function results of FEV1, FVC and DL_CO to determine trajectory with time (Figures S1-S3) to determine if lung function parameters were sufficiently stable (to ascertain the need to standardise the follow up time if there were any important variations in lung function) in the presence of multiple lung function tests.

Performance was assessed visually using scatter plots and formal testing of agreement undertaken using Bland and Altman methods (8). We defined a deviation of more than 5% as a clinically minimally important difference (corresponding to the contribution of one segment) and statistical analyses was undertaken (Eric Lim) using R version 4.3.1 (9).

Results

Between January 2013 to July 2023, we obtained lung function results from 1,174 patients and removed records from 1,016 patients who did not fit our inclusion/exclusion criteria leaving 158 patients for analysis. The baseline characteristics are presented in Table 1.

Post operative longitudinal lung function plots (Figures S1-S3) demonstrated sufficient stability with time and therefore preoperative lung function results were selected the time closest before surgery and (based on longitudinal plots) postoperative lung function was defined as the longest time from surgery.

Scatter plots for effort-dependent measures suggested the tendency for underprediction (observed values were higher than predicted) for FEV1 and FVC but good agreement for DL_CO (Figure 1). Formal agreement confirmed mean difference for FEV1 was −9.84% [95% confidence interval (CI): −39.33% to 19.65%], FVC −11.39% (95% CI: −50.14% to 27.36%) and DL_CO −4.83% (95% CI: −25.59% to 15.92%) as demonstrated in Figure 2.

Figure 1 Scatter plots graphically representing the agreement between observed and predicted FEV1, FVC and DL_CO. FEV1, forced expiratory volume in one second; FVC, forced vital capacity; DL_CO, diffusing capacity for carbon monoxide.

Figure 2 The Bland and Altman plots for FEV1 (A), FVC (B) and DL_CO (C). Formal agreement confirmed mean difference for FEV1 was −9.84% (95% CI: −39.33% to 19.65%), FVC −11.39% (95% CI: −50.14% to 27.36%) and DL_CO −4.83% (95% CI: −25.59% to 15.92%). FEV1, forced expiratory volume in one second; FVC, forced vital capacity; DL_CO, diffusing capacity for carbon monoxide; CI, confidence interval.

Discussion

The results of our study suggested that current methods for prediction of postoperative lung function underestimates the observed values after surgery, reflecting the results reported by other groups (10). For effort-dependent parameters (FEV1 and FVC) the underestimation was considerable of approximately 10% (equivalent of two segments) but there was good agreement with effort independent measures of lung function (DL_CO) which was within the 5% of our predefined clinically important threshold. It also validates that a 20-segment model has good agreement with actual amount of (predicted) functional loss from lung resection.

Effort-dependent measures of lung function

There are several plausible considerations as to why postoperative effort-dependent measures of FEV1 and FVC (the two most commonly used estimates in clinical trials as spirometry is easily accessible) are not reliable. We do know however it is not due to lung “recovery” as postulated by some surgeons to explain the higher-than-expected lung function results for JCOG 0802 (1), because the DL_CO estimates were very close to that predicted/expected from the amount of lung tissue lost. More plausible explanation might include better technique or efforts after surgery in patients who recovered from surgery (with additional experience of lung function testing) or the possibility that a smaller intrathoracic volume can lead to better diaphragmatic excursion and hence FEV1 and FVC as observed in lung volume reduction surgery (11).

The differences in FEV1 and FVC were clinically significantly higher (better) than expected, and this has important clinical implications. Surgeons often estimate that the amount of lung function loss is directly proportional to the amount of lung tissue resected and the result of our work suggests that this is not accurate. The variation between predicted and observed was considerable of approximately 10% with very wide confidence intervals.

Effort independent measures of lung function

When postoperative predicted lung function was estimated with diffusion capacity for carbon monoxide, the estimated and observed were very close (within 5%, or one segment) leading us to infer that (I) it is a better measure of lung function loss as it is not nearly as dependent on technique and (II) that the 20-segment model for lung function holds true.

Clinical implications

It is unlikely that variation between expected and observed differences in FEV1 and FVC could account for the lack of a difference between segmentectomy and lobectomy in the two large clinical trials [JCOG 0802 (1) and CALGB 140503 (2)]. This is because the degree of error would be expected to be the same on both sides of the randomisation arms. The more likely explanation would be that the amount of lung tissue resected in the sublobar arms would be similar to “lobectomy” to achieve cancer clearance as the 2 cm lesions do not usually sit nicely within the centre of one segment or peripherally to allow a wedge with minimal tissue clearance (as expected when the trials were designed). Also, the concept of a lobectomy is often misinterpreted as a larger resection given left upper division segmentectomy is the same amount of tissue as right upper lobectomy and left lingula segmentectomy is the same amount of tissue as a right middle lobectomy.

Our results suggest that when consenting patients in clinic prior to lung resection, it is no longer accurate to simply estimate the amount of lung tissue loss if FEV1 is used as a yardstick measure, as it overestimates postoperative losses, and prohibits patients proceeding to surgery or alters the extent of resection recommended on inaccurate measures (12). Future clinical guidelines also need to be aware of our findings to allow for better recommendations for both surgeons and patients. For example, the use of 40% postoperative predicted value is often used as a yardstick for patient selection for lung resection (12) and our results suggest that on average the actual values are expected to be approximately 10% higher.

Limitations

The widespread use of postoperative estimation of lung function seems academic, as it overestimates lung function losses, an outcome that needs to be interpreted by patients rather than doctors or guidelines experts as stated in the 2010 British guidelines (3). For example, patients may accept much lower values of “predicted” lung function losses and choose surgery over non-surgical options. Ultimately, what we need is better measures of functional loss rather than abstract concepts of predicted postoperative lung function, and we hope that this will be a focus for future research efforts. Our patient cohort had principally received lobectomy, so we were unable to ascertain any differential effect on segmentectomy alone, but this was not the focus of our study and may be an aspiration for future work (especially in the setting of “complex” and unusual segment combinations). We did not perform comparisons using other alternatives to segment counting, such as functional lung volume measurement nor were we able to analyse the effect of staples versus blunt dissection on intersegmental planes on postoperative lung function test. Furthermore, we did not distinguish between surgical access (video-assisted thoracoscopic surgery or open) based on the assumption that extent of resection would not differ.

Conclusions

Effort-dependent parameters of lung function (spirometry) including FEV1 and FVC tends to overestimate the amount of lung function loss after anatomic lung resection, clinicians should be cautious in using these measures to determine suitability of surgery based on current established guidelines. However, independent measures such as DL_CO demonstrate good agreement suggesting that predicted lung tissue loss is consistent with a 20-segment lung model.

Acknowledgments

Funding: None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1390/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1390/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1390/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1390/coif). Eric Lim is the chief investigator of VIOLET (NIHR 13/04/03), MARS 2 (NIHR 15/188/31) and RAMON (NIHR 131306). He is also the founder of My Cancer Companion, Healthcare Companion Ltd. His other financial disclosures include: consulting payment from Beigene, Roche and BMS and the recipient of grants from AstraZeneca, Medela, Johnson and Johnson/Ethicon, Covidien/Medtronic, Takeda, Lily Oncology. P.D.S. is a recipient of personal fees from Vitae Professionals. J.J.L. received the 2023/2024 Society for Cardiothoracic Surgery and AstraZeneca Thoracic Oncology Educational Fellowship Award. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Individual patient consent and the National Health Service (NHS) Research Ethics Committee (REC) review for this retrospective analysis were waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Saji H, Okada M, Tsuboi M, et al. Segmentectomy versus lobectomy in small-sized peripheral non-small-cell lung cancer (JCOG0802/WJOG4607L): a multicentre, open-label, phase 3, randomised, controlled, non-inferiority trial. Lancet 2022;399:1607-17. [Crossref] [PubMed]
2. Altorki N, Wang X, Kozono D, et al. Lobar or Sublobar Resection for Peripheral Stage IA Non-Small-Cell Lung Cancer. N Engl J Med 2023;388:489-98. [Crossref] [PubMed]
3. Lim E, Baldwin D, Beckles M, et al. Guidelines on the radical management of patients with lung cancer. Thorax 2010;65:iii1-27. [Crossref] [PubMed]
4. Brunelli A, Charloux A, Bolliger CT, et al. ERS/ESTS clinical guidelines on fitness for radical therapy in lung cancer patients (surgery and chemo-radiotherapy). Eur Respir J 2009;34:17-41. [Crossref] [PubMed]
5. Tane S, Nishio W, Nishioka Y, et al. Evaluation of the Residual Lung Function After Thoracoscopic Segmentectomy Compared With Lobectomy. Ann Thorac Surg 2019;108:1543-50. [Crossref] [PubMed]
6. Ueda K, Tanaka T, Hayashi M, et al. Computed tomography-defined functional lung volume after segmentectomy versus lobectomy. Eur J Cardiothorac Surg 2010;37:1433-7. [Crossref] [PubMed]
7. Nomori H, Shiraishi A, Cong Y, et al. Differences in postoperative changes in pulmonary functions following segmentectomy compared with lobectomy. Eur J Cardiothorac Surg 2018;53:640-7. [Crossref] [PubMed]
8. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10.
9. Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. 2023.
10. Chen L, Gu Z, Lin B, et al. Pulmonary function changes after thoracoscopic lobectomy versus intentional thoracoscopic segmentectomy for early-stage non-small cell lung cancer. Transl Lung Cancer Res 2021;10:4141-51. [Crossref] [PubMed]
11. Lim E, Sousa I, Shah PL, et al. Lung Volume Reduction Surgery: Reinterpreted With Longitudinal Data Analyses Methodology. Ann Thorac Surg 2020;109:1496-501. [Crossref] [PubMed]
12. Batchelor TJP, Rasburn NJ, Abdelnour-Berchtold E, et al. Guidelines for enhanced recovery after lung surgery: recommendations of the Enhanced Recovery After Surgery (ERAS®) Society and the European Society of Thoracic Surgeons (ESTS). Eur J Cardiothorac Surg 2019;55:91-115. [Crossref] [PubMed]