Discussion
In young adults referred to our center for evaluation of pectus, we found most have normal resting lung volumes, flows and normal aerobic fitness. Among those with abnormal PFT patterns, most were mild and were not associated with increased symptomatology, reduced aerobic fitness, or magnitude of the pectus defect compared to patients with normal patterns. Analysis of ventilatory responses during maximal exercise revealed a pulmonary limitation in less than 3% of patients. Taken together, we interpret these results to indicate that the pectus chest wall deformity found in most pectus patients does not directly lead to dyspnea or pulmonary limitations with exercise.
Obstruction was the most common abnormal pattern on PFTs identified in 10.7% of all referrals tested. This prevalence is higher than Lawson’s prior study of 218 pectus patients where obstruction was reported in less than 2%25. Lawson used a more stringent criteria for obstruction of a FEV1/FVC of <67% which likely explains the discrepancy between our studies. In a study of over 3,000 young adults in the general US population using a similar definition of obstruction to ours the prevalence was 10.1%26. Overall, these findings suggest an obstructive pattern in pectus is similar to what would be expected from the general population.
Using both spirometry and plethysmography criteria we identified a restrictive pattern in 7.5% of referrals. Lawson reported 14% of pectus patients exhibited a restrictive pattern using only spirometry criteria25. However, using spirometry without plethysmography has been shown to have a low predictive value for restriction and overestimates the incidence. Studies of spirometry with matching plethysmography in adult PFT labs reveal approximately 50% of restrictive patterns on spirometry have normal TLC on plethysmography22,27. Applying this correction to Lawson’s data would reduce the prevalence of restriction to 7%, mirroring our findings. Both Lawson’s and our data suggest that the prevalence of restrictive patterns in pectus is higher than would be expected from the general population. A longitudinal study in Austria of nearly 10,000 people in the general population using both spirometry and plethysmography found a prevalence of restriction of 0.9% with no restrictive patterns identified among those under 18 years of age28.
The NSVL is an often-unrecognized distinct PFT pattern, which has not been previously reported in prior pectus studies. The NSVL has not been shown to associate with any specific clinical phenotype and can be found in patients with a variety of underlying respiratory disorders including asthma, chronic obstructive pulmonary disease, interstitial lung disease and obesity29. Adult PFT laboratories report the prevalence of NSVL pattern in 6.6% to 15% suggesting our finding of NSVL in 7.5% of referrals may not represent a unique underlying pathology21,29,30.
While the majority of CPET studies were classified as having normal fitness, the average VO2 peak was 89% of predicted. Our findings are consistent with several prior reports CPET studies in pectus patients which demonstrate mild but significant reductions in mean VO2 and O2 pulse values when compared with either normative standards or age-matched controls5,7,31,32. Findings from these studies and ours thus support a physiologic impairment with exercise is likely in some pectus patients.
Using scatterplots, we demonstrate a minor inverse association between the correction index and both the TLC and FEV1. The finding of lower lung volumes with a more extensive chest wall deformity was also reported by Lawson who noted the likelihood of a restrictive pattern was four times higher for pectus patients with a Haller index of seven25. Multivariate analysis also demonstrated a higher correction index and lower FEV1 were associated with reductions in VO2 peak. Thus, the association of both a lower VO2 and lower resting lung volumes with more severe pectus may suggest a pulmonary limitation from a restrictive defect contributes to the high rate of dyspnea or lower mean VO2 peak in the pectus population. However, despite a high number of patients with severe pectus, we identified a restrictive ventilatory limitation during exercise in only six subjects. There was no relationship between either the BR or Vt/IC with the correction index (Figure 3) suggesting there are no limitations in the expansion of tidal volume breathing during maximal exercise with more severe pectus defects. Multivariate analysis also demonstrated peak VO2 was positively associated with a higher Vt/IC and inversely associated with the BR (Figure 4). These findings demonstrate patients with pectus can appropriately increase their tidal volume to achieve higher VO2. And there were no differences in the mean BR or Vt/IC between pectus referrals with normal or abnormal PFT patterns (Table 2) suggesting a resting ventilatory limitation was not associated with limitations in tidal volume expansion with exercise.
Our finding in this study of a stronger correlation in the Vt/IC with VO2 peak in females over males suggests females may have greater effective ventilatory expansion of their tidal volume during maximal exercise. We previously demonstrated that females with pectus have significantly deeper chest wall deformity as represented by higher pectus indices and higher frequency of chest pain and dyspnea on exertion than males33. However, females have significantly increased VO2, O2 pulse, a higher breathing reserve than males, and right and left ventricular ejection fraction greater than males. The mechanism for these findings is unclear. One theory postulates females may have a more compliant chest wall which thereby allows less fixed cardiac compression and a more minor degree of cardiac impairment33.
We found 64% of pectus patients reported dyspnea on exertion, a finding almost identical to Kelly’s 63% in a prospective observational study of 327 pectus patients34. Prior studies in adolescents report an overall prevalence of dyspnea with exertion ranging from 7% to 14% suggesting individuals with pectus experience greater dyspnea symptoms than would be expected in the general population35,36. In our study, among the those reporting dyspnea on exertion nearly three-quarters had normal CPET, and dyspnea was not associated with a reduced VO2 by multivariable linear regression modeling. These results suggest dyspnea in many pectus patients is likely caused by mechanisms other than reduced oxygen delivery during exercise. Prior studies in adolescents presenting with dyspnea collectively report the most common cause is normal physiological limitation, a condition believed to be caused by an excessive sensation of the perceived work of breathing with increased ventilation during exercise37,38. Other common causes include vocal cord dysfunction (VCD), exercise-induced bronchospasm (EIB), exercise-induced hyperventilation and deconditioning. We did not assess directly for EIB as lung function was not performed after exercise nor were patients exercised on an EIB protocol. VCD was also not directly assessed as we did not perform laryngoscopy during exercise, however the incidence is likely low as stridor was not auscultated during exercise and flattening of inspiratory loops was not observed during exercise.
A little over one-quarter of pectus referrals reporting dyspnea demonstrated reduced fitness. Deconditioning is reported in 10-23% of adolescents with dyspnea and may be a significant contributing factor to dyspnea in pectus38,39. However several reports demonstrated cardiac filling in pectus is reduced in some patients as a result of compression by the displaced sternum on the right heart chambers or torqueing of the great vessels, thereby constraining increases in cardiac output during vigorous exercise3-5,40. Limitations of CPET in identifying precise mechanisms for exercise intolerance include specific response patterns are rarely pathognomonic and indirectly informative regarding the central hemodynamic responses to exertion. Further studies including assessment of cardiac MRI endpoints will be important to differentiate if dyspnea with reduced VO2 peak in pectus is from direct anatomic compression of cardiovascular structures versus deconditioning or other mechanisms.
We conclude that in most pectus referrals to our center, respiratory limitations during maximal exercise are rare and an unlikely cause for the observed high the rates of dyspnea on exertion or reduced aerobic fitness. Resting lung volume measurements on PFTs were associated with the anatomic degree of pectus severity but not symptoms or physiologic outcomes measured during maximal exercise testing. Future prospective studies enlisting age and sex-matched controls are needed to better understand the apparent male-female differences, the underlying mechanisms for the high rates of symptomatology and inform under which conditions there are meaningful physiologic benefits for surgical repair of the defect.