Methods
A retrospective chart review was done on patients with pectus referred to CCHMC who completed a symptom questionnaire recorded at the time of initial assessment and performed a cardiopulmonary exercise test (CPET), pulmonary function tests (PFT), and chest magnetic resonance imaging (cMRI). Patients with incomplete records and evidence of submaximal effort on CPET were excluded from the analysis. Testing on all patients was completed using the same equipment and methodology thereby allowing uniform comparison of data.
All CPET were performed on a cycle ergometer (Ergoline ViaSprint 150p; Germany) using a 15-25 Watt/ min ramping protocol under the supervision of clinical exercise physiologists. Continuous breath-by-breath measurements (Vmax Encore metabolic cart, Vyaire Medical; Yorba Linda, CA), pulse oximetry (Masimo Corp; Irvine, CA), and a 12-lead electrocardiogram (Cardiosoft, GE Healthcare; Milwaukee, WI) were recorded throughout exercise., Tests were determined maximal effort if meeting 2 of the 3 following criteria:1) peak heart rate ≥85% age-predicted maximum (220-age); 2) respiratory exchange ratio ≥ 1.1; or peak minute ventilation was >50% of FEV1x40. Endpoints were compared to published normative standards14,15.
Tests were classified as normal aerobic fitness if the percent predicted peak oxygen uptake (VO2) was 80% or greater of predicted. For tests with reduced aerobic fitness, gradation of severity was based on the VO2 percent of predicted and was defined as mild: (70-79%), moderate (60-69%), and moderate-severe (50-59%). CPET were classified as demonstrating pulmonary limitation if the VO2 was < 80% predicted and the ratio of tidal volume (Vt) during exercise divided by resting inspiratory capacity (IC) was greater than 85% or if the breathing reserve (BR) during exercise was either less than 12 liters or 12% of predicted. Hyperventilation was defined by a VO2 > 80% predicted with the ventilatory equivalent for oxygen (VE/VO2) greater than 40, a respiratory rate (RR) greater than 60 breaths/minute, and end-tidal carbon dioxide (PETCO2) less than 34 mmHg. Tachypnea was defined as a RR >60 breaths/minute in the absence of a low VO2 or criteria for hyperventilation.
Spirometry and plethysmography were performed on one of three flow-sensing spirometers and whole-body boxes (Vyaire, Yorba Linda, CA) with calibration performed prior to testing daily per American Thoracic Society (ATS) recommendations16,17. All measurements were body temperature, pressure and saturation corrected. Only pre-bronchodilator maneuvers were evaluated. The respiratory therapists supervising tests evaluated if each maneuver met all ATS criteria for acceptability and testing was discontinued once the patient produced efforts which met acceptable criteria that were also repeatable. These tests were classified as acceptable and repeatable. Spirometry which did not reach end of test criteria but had satisfactory start and no artifacts with repeatable FEV1 were classified as usable. Testing was discontinued if the subject was unable to produce any acceptable, usable, or repeatable efforts after 8 attempts. PFTs were classified as uninterpretable if both spirometry and plethysmography were unacceptable or not repeatable, or if spirometry was usable but plethysmography was not acceptable or repeatable. Spirometry endpoints included the FVC, FEV1, FEV1/FVC and the forced expiratory flow at 25-75% of the FVC (FEF25-75). Each flow-volume curve was inspected for evidence of coving, defined as concavity on expiratory limb of the flow-volume curve. Each volume time curve assessed for volume plateau, defined as volume change of < 25 ml/s16. Plethysmography endpoints included the total lung capacity (TLC), the residual volume (RV) and the RV/TLC.
PFTs were interpreted as normal if both spirometry and plethysmography endpoints fell within the 95% confidence intervals (95% CI) of published normative standards18-20. Tests were interpreted as obstructive if the FEV1/FVC was reduced and one of the following: reduced FEF25-75, coving, failure to plateau, or an elevated RV/TLC16. Tests were interpreted as restrictive if both the TLC and FVC were reduced and the FEV1/FVC normal. Tests were interpreted as a non-specific ventilatory limitation (NSVL) if the FVC was reduced but the TLC fell within the 95% CI21. Tests with both obstructive and restrictive patterns were categorized as mixed. Gradation of the severity for obstructive defects was based on the FEV1 percent of predicted and gradation for the severity of restrictive and NSVL defects was based on the FVC as follows: mild > 70% up to 95% ci; moderate 60-69%; moderate-severe 50-59%22.
The cMRI was performed on a 1.5 T scanner (Philips Ingenia; Best, Netherlands) for assessment of the Haller and correction index. The Haller index is a ratio of the transverse diameter of the chest to the anterior-posterior diameter, measured from the inner aspect of the sternum to the anterior aspect of the vertebral body at the level of greatest sternal depression23. The correction index was evaluated according to previously described methods24.
Categorical data are presented as frequencies and percentages, while continuous variables are reported as mean (standard deviation) or median (first and third quartiles). ANOVA, chi-square, and Kruskal-Wallis tests were used to evaluate demographic and clinical characteristics by PFT classification. Pearson correlation coefficients were calculated for the relationships between correction index and PFT/CPET metrics. Multivariable linear regression was used to assess the association of VO2 with demographic and clinical characteristics. Results from the final model are presented as beta estimates with 95% CI. Data were analyzed using the SAS v9.4 (Cary, NC). All reported p-values are two-sided and considered statistically significant when < 0.05.
Results
From September 2016 through February 2019, 259 patients with pectus completed maximal effort CPET, PFTs and cMRI. Patients were predominantly male (86%), white race (98%), with an average age of 15.8 years. Most were healthy with no known underlying co-morbidities. Twenty-one patients (8%) had an underlying connective tissue syndrome including hypermobility (12 patients), Ehlers-Danlos syndrome (8 patients) and Marfan syndrome (1 patient). Demographic and clinical characteristics data including PFT and CPET primary endpoints are presented in Table 1. A high percentage of pectus patients reported dyspnea on exertion (64%) and chest pain (41%). Mean lung volumes on spirometry and plethysmography were normal with the FVC 96.6% and TLC 96.9%. Measures of aerobic fitness on CPET were low-normal with the mean percent predicted VO2 peak 89.0% and oxygen pulse 90.8%.
Aerobic fitness was normal in 181 patients (70%) and reduced in 78 (30%) (Figure 1A). Reduced aerobic fitness was classified as mild in 51, moderate in 21 and moderate-severe in 6. Six patients (2.3%) demonstrated a pulmonary limitation with reduced fitness. Among patients with normal aerobic fitness 18 (7%) demonstrated tachypnea and 7 (2.7%) hyperventilation. Among the 164 patients reporting dyspnea on exertion 119 (72.6%) exhibited normal fitness and 17 (10.3%) demonstrated tachypnea, hyperventilation, or a pulmonary limitation (Figure 1B).
Among those completing maximal CPET, six could not perform spirometry and plethysmography and were excluded from further analysis (Figure 2). For the remaining 253, spirometry was classified as acceptable and repeatable for 133 (53%) and usable for 120 (47%). Pulmonary function testing were normal in 188 (74%) with a resting ventilatory limitation identified in 65 (26%). Obstruction was the most common limitation pattern in 27 (11%) followed by 7% with a NSVL and 7% with a restrictive pattern. Two had mixed obstructive and restrictive defects. Of the 65 patients with any resting ventilatory limitation pattern, 55 (85%) fell into the mild category.
Pectus patients with normal PFTs were compared against those with obstructive, NSVL or restrictive patterns (Table 2). There were no differences between groups for the degree of pectus malformation or CPET endpoints. Patients with abnormal patterns did not have increased dyspnea on exertion or chest pain than those with normal patterns.
Scatter plots depicting correction index with TLC% and peak VO2% show significant inverse relationships between degree of the pectus deformity and lower lung volumes (r=-0.26; p<0.001) and reduced aerobic fitness (r=-0.20; p<0.001) (Figure 3 A, B). A similar pattern also exists with FEV1% (data not shown). In contrast, there was no significant correlation between an increase in the degree of the pectus deformity and BR (r=-0.004; p=0.95) or Vt/IC (r=-0.11; p=0.09) (Figure 3 C, D).
Multivariable linear regression modeling was used to evaluate pectus severity indices and PFT/CPET measurements as predictors of peak VO2 (Figure 4, Panel A). After adjustment, increasing correction index values were associated with decreasing VO2 (p<0.001). A similar inverse relationship was also noted for BR (p<0.001). Peak VO2 was found to increase as body mass index (BMI) and FEV1increased (each p<0.001). Also, peak VO2 was significantly higher for those with tachypnea (no hyperinflation) compared to patients without the condition (p<0.001). A significant interaction was noted between Vt/IC values and sex (p for interaction <0.001). The Vt/IC was positively associated with VO2 peak for both males and females, with a greater increase in VO2 noted in females (Figure 4, Panel B). No associations were noted with TLC, FVC, Haller Index, symptoms of dyspnea or chest pain.