Ann Thorac Surg 2003;76:362
© 2003 The Society of Thoracic Surgeons
Original article: general thoracic
Invited commentary
Eddie L. Hoover, MD (112)a
a Buffalo VAMC3495 Bailey Avenue, Buffalo, NY 14215, USA
e-mail: eddie.hoover{at}med.va.gov
In this paper, Sekine and colleagues focused on a prior observation that some patients with moderate emphysema had a better post-operative pulmonary function than expected when compared to a control group without this disease. At first pass, this would appear to be contrary to logic and pulmonary physiology as most of us have understood and practiced it over the years. They then hypothesized that resection for cancer in these patients may have some beneficial effect similar to that of lung volume reduction surgery, but noted that they were dealing with much less severe disease. They also attempted to validate their observations through a formula that would be useful in a predictive mode to pulmonologists and thoracic surgeons who must make crucial decisions as to whether these patients can tolerate resection for cure. Although the sample size from which these observations and conclusions were drawn was small, 48/521 patients, this was apparently felt to be adequate by The Annals of Thoracic Surgery Editorial Board and the statisticians involved to empower the study. Nevertheless, one must seriously consider whether this would hold up in a much larger cohort of such patients. Of significance, standard nomenclature involving pulmonary function tests and arterial blood gases was used throughout the manuscript; therefore, interested readers do not have to subscribe to any new lexicon or technology. Their observation that the PaO2, FEV1, and FVC were better maintained in the COPD group than the non-COPD group is provocative, especially when this was found to be more pronounced in lower lobe resections, since most of the early changes in emphysema occur the upper lobes. They make some inferences about the upper lobe versus lower lobe differences, which may or may not hold up in the final analysis, but it makes for interesting speculation with perhaps some potential therapeutic applications.
The ratio of the apoFEV1 to ppoFEV1 is critical in this paper and was calculated by the formula: FEV1 (1 - S x 0.0526), which was appropriately referenced in the bibliography (17). However, after analyzing the data in Table 4, where unexpecteddifferences were observed, they "multiplied a specific coefficient to ppoFEV1 to improve predictive accuracy." No explanation was given as to how these new "specific coefficients" were derived nor were they referenced. This may be problematic for other investigators who chose to incorporate this model in their work. The authors were able to test their formula in a prospective manner in 142 patients over a 2-year period, and the modified ppoFEV1 was significantly related with the apoFEV1 at a R2 of 0.619. The acknowledged single concern with this retrospective study is that post-operative PFTs were obtained at one month while recovery of function continues out to perhaps 6 months. Therefore, assuming the validity of their observations, the results can only be enhanced had the PFTs been done at a later date.
In summary, the authors make some speculations that may or may not be valid over time and, in deed, some of their observations may ultimately prove to be real, but not in a cause-effect dynamic. However, the paper does pose interesting questions for future investigations that may answer questions about this difficult group of patients. Finally, the formula presented may indeed be helpful to clinicians in deciding operability in these patients.
