However, the gist of the letter from West China Hospital is valid. INR targets in Chinese patients, and indeed in all patients, need higher-quality evidence than what currently exists. Differing INR targets based on thromboembolic risk is unique to heart valve therapy; for example, the INR target for a patient with atrial fibrillation and a CHA2DS2VASc stroke risk score of 8 is the same as that of a patient with a CHA2DS2VASc score of 3, despite higher thromboembolic risk.² The evidence supporting such differing targets for heart valves is of moderate quality at best according to the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) framework,³ and begs for further studies. One such study, Lowering the Intensity of Oral Anticoagulant Therapy in Patients With Bileaflet Mechanical Aortic Valve Replacement (LOWERING-IT), also supports the approach of lower INR targets in low-risk mechanical aortic valves but needs validation in a larger trial.⁴ The ninth edition of the American College of Chest Physician guidelines presents an objective assessment of the available literature up to October 2009 that is based on the GRADE framework and the resultant recommendations. We look forward to reading publications from the West China Hospital based on its national database and will incorporate any new knowledge into future guidelines.
However, to dismiss our publication as “incomplete, poorly defined and uninterpretable” is unsubstantiated. Performance of spirometry according to American Thoracic Society standards on different spirometers should not diminish the accuracy of the raw values used in our analysis. We maintain that workers should be screened for excessive decline in FEV1. It is our understanding that rotary cutters were still in use during the decade in question. Measurement of respirable, nylon fiber-shaped fragments was undertaken by Black and Veatch Special Projects Corporation using a modified National Institute of Occupational Safety and Health 7400 protocol as part of ongoing work to validate a sampling and analytical technique to estimate exposure. We had access to, but did not publish, these data, given concerns regarding the experimental nature of this sampling method and, hence, uncertainty regarding its validity. While our original postulate of the “potential role of mycotoxins”² proved not to be the case, lack of a definitive known cause should not deter clinicians from publishing case reports or case series. As a point of clarification, desquamative interstitial pneumonitis-like reactions are part of the spectrum of pathologic features of FWL noted by expert pathologists.3,4
We thank Drs BaHammam and Esquinas Rodriguez for their letter about our recent study.¹ As they correctly note, the prevalence of obesity (as reflected by International Classification of Diseases, Ninth Revision, codes) was roughly six times higher among patients with OSA than in those without OSA. Given the high potential for obesity to serve as a confounder, our multivariable analyses included obesity as a covariate, thereby yielding an estimate of the independent association between OSA and the two outcomes we studied: (1) in-hospital mortality and (2) initiation of mechanical ventilation or transfer to the ICU after the second hospital day (a measure of clinical deterioration). In light of the comments by Drs BaHammam and Esquinas Rodriguez, it is also worth noting that in the multivariable analyses (presented in e-Tables 2 and 3 of our article), obesity remained associated with lower in-hospital mortality. However, obesity was not associated with late mechanical ventilation in these analyses. Finally, although we did not include these results in the article, we also investigated the possibility of an interaction between obesity and OSA. The P value for the interaction term was .81, suggesting that patients with OSA and obesity had similar outcomes to patients with OSA who were not obese.
We welcome the opportunity to respond to Dr Noonan’s letter regarding our recent article in CHEST.¹ Statistical power is always a potential limitation in biomedical studies. This is especially true in retrospective observational studies, in which an a priori power analysis cannot be performed. As mentioned in our article, although our reference group of miners with normal high-resolution CT scan (NCTS) studies was relatively small, there was enough power to detect statistically significant differences in mean lung function data between the NCTS reference group and the miner group with pleural plaques and interstitial fibrosis and the miner group with other high-resolution CT (HRCT) scan findings. However, when we compared the miner group with pleural plaques only (PPO) with the NCTS reference group, we were unable to detect a statistical significance in mean lung function data, even though the PPO miner group had many more subjects than did the miner group with pleural plaques and interstitial fibrosis or the miner group with other HRCT scan findings. Information contained within the 95% CI for comparison of the PPO miner group with the NCTS reference group indicates that we should not reject the null hypothesis. Tukey analysis was also used to minimize type 1 error. Thus, our analysis supports the conclusion that pleural plaques alone have no significant effect on lung function in Libby vermiculite miners.
We thank Dr. Lopez-Encuentra et al for their thoughtful and insightful comments. As described in the “Materials and Methods” section, we included only stage I (T1–2N0M0) non-small cell lung cancer patients whose TNM codes were completely recorded in the California Cancer Registry. The T descriptors of all the patients in the study were coded by one of the three extent of disease (EOD) codes: 10, 20, or 40 with no overlap or missing codes indicating excellent internal consistency. The EOD codes and N descriptors were generally abstracted from pathology and/or radiology reports so they can be either pathologically or clinically staged. This limitation was discussed in the article. The Surveillance, Epidemiology, and End Results EOD codes are hierarchically arranged so if a tumor contains both EOD-20 (mainstem bronchus ≥ 2 cm from carina) and EOD-40 (visceral pleura invasion, hilar atelectasis, or obstructive pneumonitis) criteria, only EOD-40 will be coded. We agree with Dr. Lopez-Encuentra et al that we cannot separate the three criteria (visceral pleura invasion, hilar atelectasis, or obstructive pneumonitis) individually, nor can we know how many EOD-40 cases also had EOD-20 criteria, and this is a study limitation. However, we demonstrated that T descriptors coded as T2 due to EOD-20 criteria alone (mainstem bronchus ≥ 2 cm from carina) are infrequent (5.7%).
We also agree that patients with CF have several unique risk factors for the development of renal dysfunction posttransplant, which may increase their risk relative to patients with idiopathic pulmonary fibrosis or COPD. However, we are not aware of any studies that have specifically compared the risk of renal dysfunction in these recipient populations. Our study did not focus on post-lung transplant risk factors, but we appreciate Dr O’Connell and colleagues pointing out that oxalate nephropathy and pigmented tubulopathy are well-recognized histopathologic findings following renal biopsy in the early posttransplant period and are likely related to perioperative stressors such as dehydration, hypoxia, and antibiotics.⁴ Our analysis excluded patients diagnosed with renal dysfunction in the first month post-lung transplant to reduce the chance of including acute cases. Our study found that CF-related diabetes requiring insulin is an important pretransplant risk factor and likely plays an important role in renal function loss in the late posttransplant period. This is in keeping with a published renal biopsy series, which demonstrated that histopathologic findings responsible for late episodes of renal function loss were primarily vascular (ie, diabetic glomerulosclerosis).⁴
We believe that it is important to be clear about the business model. For example, are the costs covered by an institutional grant or a social fund to support disadvantaged patients? Currently, the most common institutional business model is to make up the discrepancy between what is collected for screening and what it actually costs from profits arising from interventions done for screen-detected nodules. It is important to recognize that this represents an inherent conflict: Optimal patient management calls for limited, judicious further intervention, whereas optimal financial management calls for maximizing the number of additional interventions performed. This conflict is particularly poignant because the rate of finding a nodule is high (10%-50% of those screened), the level of anxiety of those with a nodule is high, yet the proportion of inconsequential nodules is high (about 96%). The issue is further magnified when screening is extended beyond the National Lung Screening Trial criteria (currently unsupported by data or modeling studies of efficacy), which increases manyfold those being screened and needing further interventions with a much lower impact on actually preventing deaths from lung cancer.¹ It is certainly possible for this conflict to be managed appropriately, but this business model increases the potential for intended or unintended overuse of subsequent interventions.
We thank Drs Kumar and Chandra for their interest in our work¹ and the opportunity to discuss the use of ultrasound imaging of the diaphragm in the clinical context. Regarding our hypothesis that the diaphragm might show hypertrophy in COPD, this was based on the assumption that the diaphragm may be more active in patients with COPD during quiet breathing than in healthy subjects with normal lung function. This would be consistent with the increased firing rate of diaphragm motor unit potentials reported by Ottenheijm et al² during tidal breathing in patients with COPD. However, our study did not show any such compensatory hypertrophy, and although it is theoretically possible that accelerated protein degradation and associated atrophy could offset any compensatory hypertrophy, Ottenheijm et al² actually did not find a reduction in cross-sectional area of diaphragm fibers in patients with mild to moderate COPD. Although they established a reduction in myosin content (which affects maximum force generation), this did not correlate with a change in cross-sectional area, and the atrophy they refer to appears to be at the molecular level, without affecting the gross anatomic structure. As they pointed out, changes in diaphragm structure and function in COPD are still poorly understood, and there is likely a complex interplay of multiple factors involved, but our goal was to establish normal thickness and thickening ratio in patients with COPD such that we could use these measures to diagnose superimposed phrenic neuropathy or other neuromuscular disease.
Regarding the second point, we concur that diaphragm thickness does not change consistently with lung volume and increases significantly above 50% of vital capacity, but we respectfully disagree with their statement that there is a large increase in thickness between relaxation and 10% of the inspiratory effort.3,4 We studied a group of 150 healthy subjects and found that the degree of diaphragm thickening during tidal breathing was highly variable from one individual to another, and > 30% of diaphragms did not thicken at all during quiet inspiration. Cohn et al³ also found a nonlinear relationship between diaphragm thickness and lung volume, with lower levels of thickening occurring earlier in inspiration and higher levels seen with lung volumes > 50%. Because of the different technique used by Wait et al⁵ (M mode ultrasound in the axial plane, as opposed to B mode ultrasound in the sagittal oblique plane used in our study and several others), direct comparisons cannot be made; however, we would like to point out that Wait et al⁵ only evaluated diaphragm thickening at lung volumes < 50% because of their methodology using M mode, in a small group of 10 normal men, and only reported the relationship of diaphragm thickening fraction to lung volume and not diaphragm thickness to lung volume.⁵ Given that the thickening fraction is derived directly from diaphragm thickness, there should be a similar relationship for both thickness and thickening fraction.
Previous studies showed that > 20% of patients with systemic sarcoidosis had positive LGE, cardiac events occurred in > 10% of patients, and LGE was hypothesized to predict future cardiac events, even in patients with preserved LV systolic function.²⁻⁴ Contrary to these findings, our rates of positive LGE and cardiac events were lower, and, importantly, LGE did not predict adverse events.¹ These differences can be explained by disease severity, including the extent of LGE. Patients in our study had lower frequencies of receiving immunosuppressants and having ECG abnormalities, no cardiac symptoms or impaired LV systolic function, and smaller extent of LGE compared with those in a previous study.³ Ise et al⁵ previously reported that smaller extent of LGE (% LGE mass < 20%) could predict a lower rate of cardiac events. Taking these findings together, the clinical implication of adding LV wall stress analysis would be limited in our population with latent and less severity and preserved LV systolic function. We believe that a larger cohort with a higher event rate (eg, greater disease severity, impaired LV function, and larger extent of LGE) is needed to investigate the association between LV volume and functional variables and long-term prognosis in patients with systemic sarcoidosis.
We thank Dr Guglin for her comments on our recent article¹ regarding the use of pulmonary hypertension (PH)-specific therapies in patients with World Health Organization (WHO) group 2 PH due to left-side heart disease (LHD). As Dr Guglin points out, elevated pulmonary arterial pressures were not required for enrollment in the Flolan International Randomized Survival Trial (FIRST).² However, all patients underwent right-sided heart catheterization, and most subjects (> 75%) did, in fact, have PH as defined by mean pulmonary arterial pressure > 25 mm Hg. This makes it unlikely that epoprostenol is of significant benefit in patients with systolic heart failure and PH. The same is true for the Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in Diastolic Heart Failure (RELAX) trial. The rationale behind the RELAX trial was to target the pleiotropic (cardiac, vascular, and neurohormonal) effects of phosphodiesterase-5 inhibition on cardiovascular function in patients with heart failure with reduced ejection fraction regardless of the pulmonary pressures.³ Based on echocardiographic estimates (which are admittedly imprecise), approximately two-thirds of subjects in this trial also had elevated pulmonary arterial systolic pressure (> 35 mm Hg).
Because of this, we agree that lung transplantation must be considered an option, particularly for the younger miners who are now developing this disorder.⁴ Dr Diaz-Guzman and colleagues encourage increasing awareness of advanced CWP and early referral to a transplant center, in recognition that massive fibrosis in coal miners often progresses even after removal from dust exposure. We concur that the medical community needs to be more effective in enhancing awareness of the continuing human toll from these dust diseases and in assuring optimal medical care and fair compensation for affected miners. Progressive massive fibrosis is entirely preventable, since it is virtually only caused by excessive dust inhalation and does not occur from tobacco use or other causes. Effective dust controls should have eliminated this type of lung disease in a modern mining industry, and the failure of the US industry to tackle this ongoing problem has been highlighted internationally.⁵ In addition to drawing attention to the role of lung transplantation to improve survival and functional status in these patients, we hope our report’s findings can motivate timely implementation of the necessary effective measures to reduce dust exposures and provide a healthful working environment for our country’s coal miners.
The implicit issue in the correspondence appears to be the extent to which change in sweat chloride could be used as a primary end point to establish efficacy for regulatory purposes (ie, approval of a drug for marketing). Such qualification of a pharmacodynamic biomarker as a surrogate end point is a high bar to achieve, as robust scientific evidence would be needed that demonstrates that changes in sweat chloride beyond a specific level in a specific CF population would predict clinical benefit to the same extent that a clinical end point (an improvement in how a patient feels, functions, or survives) would.² Because of such a high level of evidence required, most pharmacodynamic biomarkers are used, as has been the case to date for the ivacaftor program, to guide drug development, whereas clinical end points (or in the case for ivacaftor, the surrogate end point, FEV1) provide the basis for regulatory approval. With the arrival and continued development of a new class or classes of CF therapies that have the potential to address the central defect that results in CF, we, like the CF community, are happy that we have reached such a time that we can have a discussion on the use of change in sweat chloride or other possibly more accurate pharmacodynamic biomarkers that may reflect CF transmembrane conductance regulator function³ as end points in clinical trials.
The acronym PAP in the article depicts positive airway pressure and refers to the use of either continuous positive airway pressure (CPAP) or bi-level positive airway pressure. Since patients may be treated with either modality for sleep-disordered breathing, we intentionally left this open to interpretation for the treating physician, based on the patient’s current treatment. We agree with Dr Gregoretti et al that auto-CPAP should ideally be used for patients with an established diagnosis of OSA, although we also believe that auto-CPAP can be useful in the postoperative setting since opioids can lead to the exacerbation of obstructive respiratory events and higher PAP pressures than previously prescribed may be needed. It is also not uncommon that the patient is unaware of his or her actual PAP settings, and auto-CPAP can be used for in this setting for uncomplicated OSA treated in the hospital wards. It is important to highlight that not all patients with OSA who are being treated with opioids develop central events. However, if hypercapnia and alveolar hypoventilation occur as a result of sedation, noninvasive mechanical ventilation with a backup ventilatory rate should be used.
Quality of life (QOL) is a relatively new clinical end point that is particularly relevant to the typically palliative therapy for non-small cell lung cancer. Patients' assessments of their QOL are shown to differ from their physicians', emphasizing the subjective nature of QOL. A number of relevant instruments and assessment techniques are employed. Results from a study using the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 instrument before and during chemotherapy are presented. Some parameters improved while others did not, preventing a simple interpretation. There are arguments for compiling indexes of QOL while retaining measures for individual parameters and a desire for the consistent international use of an instrument such as the EORTC questionnaire.
Dr Madan and colleagues also recommended that before adopting the FEV3/FVC ratio for common use, it should be validated using other radiographic and clinical assessments. But what if this is the more sensitive measurement? Without concomitant emphysema, a CT scan would appear normal unless air trapping was sufficient to cause a mosaic pattern. Clinically, subjects with milder forms of diseases such as diabetes or hypertension can remain asymptomatic for years. Hansen et al⁴ clearly showed the association of smoking to a reduction in this ratio. Our study supported this relationship. In 1984, Morris et al⁵ and the Intermountain Thoracic Society advocated using this ratio and its lower limit of normal to identify mild/midflow obstruction. Short of pathologic correlation with lung biopsies, large patient databases may be the most sensitive way to identify small physiologic differences between groups. We are not advocating that the FEV3/FVC replace the FEV1/FVC ratio, but rather feel it may have an important complimentary role in identifying milder disease. In response to the opinion that the term lung injury should be restricted to the diagnosis of ARDS, we used these words in the same context as our pathology colleagues who routinely use this phrase to describe damage to the lungs no matter what the cause, including ARDS.
Third, we did not disregard the excellent articles by Reissig and Mathis (actually cited in references 37 and 40, respectively, in our article¹); all the more, because we read with care studies dealing with lung ultrasound. Multicenter studies³ have demonstrated a 74% sensitivity for pleural-based lesions. We actually analyzed the posterior wall in patients with massive embolism; postero-lateral alveolar-pleural syndrome (PLAPS) was found in 52% of cases (see Table 2 in our article¹). What is the meaning of these cases of PLAPS? Infarction is known to be rare in patients who have experienced massive embolism, with proximal obstruction seen early in the course of disease.⁴ An embolism creates a surfactant alteration and local bronchospasm, which generate small atelectases early in the course of the embolism.⁴ Since we are unable to distinguish small atelectases from genuine infarctions using ultrasound, we suppose that the cases of PLAPS seen in our patients were due mainly to atelectases, and not to infarctions. The posterior location of the atelectases was possibly a function of the ventilation/perfusion ratio. Even in patients with nonsevere disease, multicenter studies³ have underlined the rarity of anterior lesions in patients with embolism.