Respiratory Muscle Training Before Surgery in Preventing Lung Complications in Patients With Stage I-IIIB Lung Cancer

Will be treated as a continuous variable and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The effectiveness of the respiratory muscle training (RMT) program on each respiratory outcome will be assessed by comparing the preoperative change between groups using an analysis of covariance (ANCOVA) model, with an adjustment for the pretreatment levels. For each outcome, the preoperative change (T1-T0) will be modeled as a function of treatment group (RMT versus usual care) and pre-treatment levels. A one-sided Wald type-test about coefficient for treatment group will evaluate whether the RMT program had a beneficial impact on the given respiratory outcome. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be treated as a continuous variable and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The effectiveness of the RMT program on each respiratory outcome will be assessed by comparing the preoperative change between groups using an ANCOVA model, with an adjustment for the pretreatment levels. For each outcome, the preoperative change (T1-T0) will be modeled as a function of treatment group (RMT versus usual care) and pre-treatment levels. A one-sided Wald type-test about coefficient for treatment group will evaluate whether the RMT program had a beneficial impact on the given respiratory outcome. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be treated as a continuous variable and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The effectiveness of the RMT program on each respiratory outcome will be assessed by comparing the preoperative change between groups using an ANCOVA model, with an adjustment for the pretreatment levels. For each outcome, the preoperative change (T1-T0) will be modeled as a function of treatment group (RMT versus usual care) and pre-treatment levels. A one-sided Wald type-test about coefficient for treatment group will evaluate whether the RMT program had a beneficial impact on the given respiratory outcome. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Assays of muscle biopsies will be performed for metabolic and muscle physiology markers. The correlative markers will be compared between RMT responders, RMT non-responders, and usual care (control) in a pairwise fashion using Holm-Bonferroni adjusted t-tests. Responders will be those who present with a > 15% increase in inspiratory and expiratory muscle strength. The gene-level raw count values of micro ribonucleic acid (mRNA)s will be analyzed with the edgeR Bioconductor package in R, first for normalization with the trimmed mean of M-values method, and then for comparison of expression between treatments using generalized linear models with negative binomial distribution and a likelihood ratio test to generate p values. False discovery rates (FDR) will be estimated from p-values with the Benjamini-Hochberg method, and mRNAs/genes with FDR < 0.05 and fold-change values of >= 1 log2 unit will be considered as differentially expressed.

Assays of muscle biopsies will be performed for gene expression of RNA extraction, reverse transcription and real-time PCR. The correlative markers will be compared between RMT responders, RMT non-responders, and usual care (control) in a pairwise fashion using Holm-Bonferroni adjusted t-tests. Responders will be those who present with a > 15% increase in inspiratory and expiratory muscle strength. The gene-level raw count values of mRNAs will be analyzed with the edgeR Bioconductor package in R, first for normalization with the trimmed mean of M-values method, and then for comparison of expression between treatments using generalized linear models with negative binomial distribution and a likelihood ratio test to generate p values. FDR will be estimated from p-values with the Benjamini-Hochberg method, and mRNAs/genes with FDR < 0.05 and fold-change values of >= 1 log2 unit will be considered as differentially expressed.

Will be measured by European Organization for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire (QLQ) - Core (C)30. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by EORTC QLQ - Lung Cancer 13. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by Functional Assessment of Chronic Illness Therapy Fatigue. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by the Epworth Sleepiness Scale. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by the Stop-Bang Questionnaire. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by Pittsburgh Sleep Quality Index. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by Hospital Anxiety and Depression Scale. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Will be measured by the Borg Dyspnea Scale. The QoL measures are treated as continuous and will be summarized by treatment group and time-point using the mean, median, standard deviation, and the appropriate percentiles. The change in QoL measures (from baseline) will be modeled as a function of treatment group, time-point, their two-way interaction, and baseline levels using a general linear model. Comparisons of QoL at each time-point will utilize Holm-Bonferroni adjusted tests about the appropriate contrasts of model estimates. All model assumptions will be verified graphically using quantile-quantile and residual plots. Transformations will be applied as appropriate.

Pneumonia status is treated as dichotomous data and will be summarized by treatment group using frequencies and relative frequencies. The pneumonia rates will be compared between treatment groups using a one-sided Fisher exact test, as we expect the RMT program to reduce rates.

Will be treated as continuous data and will be summarized by treatment group using the mean, median, standard deviation, and the appropriate percentiles. Comparisons between treatment groups will be made using one-sided t-tests or Mann-Whitney U tests (as appropriate).

Will be treated as continuous data and will be summarized by treatment group using the mean, median, standard deviation, and the appropriate percentiles. Comparisons between treatment groups will be made using one-sided t-tests or Mann-Whitney U tests (as appropriate).

Will be treated as continuous data and will be summarized by treatment group using the mean, median, standard deviation, and the appropriate percentiles. Comparisons between treatment groups will be made using one-sided t-tests or Mann-Whitney U tests (as appropriate).

Lung infection will be treated as dichotomous data and will be summarized by treatment group using frequencies and relative frequencies. Comparison of infection rates between treatment groups will be made using Fisher?s exact test.

Analysis of identified molecular markers will be completed to correlate with patient outcome and potentially differentiate responders from non-responders.