Thoracic morphological characteristics of primary spontaneous pneumothorax patients requiring surgical intervention

Introduction

In Japan, approximately 14,000 surgeries for spontaneous pneumothorax (SP) are performed annually (1). However, the mechanism underlying primary spontaneous pneumothorax (PSP) remains unclear. In a survey conducted in 2020, we reported that adolescents are more likely to develop SP compared to other age groups (2). It has long been suggested that adolescents, particularly teenagers, are prone to recurrence (2,3). Factors associated with recurrence include physical characteristics such as tall stature (4) and low body mass index (BMI) (5).

Cases of SP in young individuals are generally recognized to exhibit physical features such as tall stature, a flattened thoracic cage, and a lean body type (6-8). Furthermore, studies have also focused on the growth and development of individuals with PSP. According to Fujino et al., patients with PSP are taller than average at the age of six and exhibit a rapid growth rate between the ages of 11 and 14 years. They suggested that imbalances in height and weight changes might be related to pneumothorax (9). Additionally, Mitani et al. reported that a high annual growth rate could lead to asymptomatic pneumothorax (10).

Despite these findings, the physical characteristics of patients with PSP have not been sufficiently compared with those of healthy individuals, leaving room for further investigation. Before conducting a comparison with healthy individuals, we conducted a preliminary study comparing the body types of adolescent pneumothorax patients and those in their twenties (11). This preliminary study revealed that adolescent patients had a significantly lower BMI compared to those in their twenties. Furthermore, there were differences in the relationship between height and thoracic parameters. In patients in their twenties, height was correlated with most thoracic parameters. In contrast, in adolescents, height was correlated only with horizontal and vertical thoracic dimensions, but not with sagittal or anteroposterior dimensions. These findings suggest that a flattened thorax, a recognized characteristic of PSP, may develop during adolescence. Together with low BMI, a flattened thorax might be a key factor in the development of PSP.

In this study, we compared the body types of adolescent PSP patients with those of a control group to identify thoracic dimensions that could serve as risk factors for PSP. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1326/rc).

Methods

Study design and participant selection

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This case-control study was approved by the Institutional Review Board of Tohoku Medical and Pharmaceutical University Hospital (Approval Number: IRB: 2022-2-074). Informed consent was obtained from participants or their parents/legal guardians via an Opt-out methodology.

The observation period was set from January 1, 2018, to December 31, 2022. During this time, adolescent male patients (aged 13–19 years) with PSP who underwent surgery at our hospital and had no history of smoking were included in the PSP group. The control group consisted of age-matched adolescents who underwent chest computed tomography (CT) during the same period for unrelated reasons. Individuals meeting any of the following criteria were excluded from the control group: a history of smoking, prior PSP, the presence of bulla on chest CT, pectus excavatum (defined as a Haller Index ≥3.2), or use of medications affecting bone metabolism (e.g., steroids, immunosuppressants, or anticancer drugs).

Study objectives

This study was designed to address the following two objectives: to compare thoracic dimensions between the adolescent PSP group and the control group; to identify thoracic dimensions associated with increased PSP risk.

Data collection

We compared the physical characteristics of the adolescent PSP group and the control group. Height, weight, and radiological data were retrieved from the electronic medical records database of Tohoku Medical and Pharmaceutical University. BMI was calculated from height and weight.

Radiological data and methods of measuring thoracic parameters

Chest CT scans were utilized to evaluate thoracic dimensions in both groups. The scans were performed as part of routine clinical practice under the following conditions: high-resolution CT (Aquilion 64/One/Prime SP, Toshiba) was used in helical mode with the patients in a supine position during full inspiration. The scans covered the entire thoracic region, from the lung apex to the lung base, without the use of contrast agents. The CT scanning parameters included a matrix size of 512×512 and a slice thickness of 1 mm.

The digital images obtained under these conditions were reconstructed into 3D models using SYNAPSE VINCENT software (FUJIFILM, Tokyo, Japan). Lung segmentation was performed using a threshold of −400 HU to distinguish the lung parenchyma from surrounding soft tissues. Structures such as blood vessels and airways were automatically excluded during the reconstruction.

Measurement methodology

Thoracic dimensions were measured following our prior foundational study (11). Measurements were taken at four sagittal sectional levels: Manubrium, Sternal angle, Midpoint of the sternal body, and Xiphisternal angle. For each level, the following parameters were measured in the horizontal plane: TD, SD (defined as the distance from the posterior surface of the sternum to the anterior edge of the vertebral body), RAPD, and LAPD (Figure 1). Additionally, we measured the vertical length of the thorax using coronal CT images. These measurements included: The vertical distance from the apex of the thorax to just above the diaphragm, and the vertical distance from the thoracic apex to the rib-diaphragm angle (Figure 2).

Figure 1 Measurement of thoracic dimensions. The image illustrates measurements on a transverse CT scan. SD is the anteroposterior diameter of the chest cavity (sternum to vertebra). TD is the widest transverse diameter. R(L)APD measures the anteroposterior dimension of the right (or left) hemithorax. CT, computed tomography; R(L)APD, right (left) anteroposterior diameter; SD, sagittal diameter; TD, transverse diameter.

Figure 2 Measurement of vertical thoracic dimensions. The image illustrates two vertical measurements on a coronal CT scan. The upper arrow shows the distance from the apex of thorax to just above diaphragm (R or L). The lower arrow shows the distance from the apex of thorax to the diaphragmatic angle of ribs (R or L), often referred to as the costodiaphragmatic angle. Both measurements are taken for the R or L hemithorax. CT, computed tomography; L, left; R, right.

Statistical analysis

As this was an exploratory study, no a priori sample size calculation was performed; the sample size was the maximum number of data points obtained during the observation period. Non-normally distributed data are presented as median [interquartile range (IQR)]. All continuous variables were compared between groups using the Mann-Whitney U test, and the IQR is shown, while categorical variables were analyzed using Fisher’s exact test. To examine the relationship between thoracic parameters and height, Spearman’s rank correlation coefficient (γ) was employed. A correlation was deemed statistically significant if |γ|>0.2 and P<0.05.

Next, predictive models were constructed to identify thoracic dimensions associated with pneumothorax risk. Model fit was evaluated using the likelihood ratio test. The baseline model (Model 1) included age, height, and BMI as predictors. Additional variables (e.g., SD/TD ratio, RAPD, LAPD) were incorporated individually, and the changes in residual deviance were assessed to determine whether model fit improved significantly. Reductions in residual deviance and corresponding P values were calculated, with P<0.05 indicating statistically significant improvement. Only variables with significant differences between the PSP and control groups were included in the modeling process. Using the most significant model, we further plotted the receiver operating characteristic (ROC) curve and determined the area under the curve (AUC), sensitivity, and specificity.

All statistical analyses were performed using R (version 4.5.1; https://cran.r-project.org) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan).

Results

PSP group vs. control group: thoracic parameter comparisons

Patient background and thoracic parameter data for all 142 participants (PSP group: 71; control group: 71) are summarized in Table 1. There was no significant age difference between the PSP and control groups, ensuring that participants were appropriately age-matched. While height differences were not significant, the PSP group exhibited significantly lower weight and BMI compared to the control group (Table 1).

Next, we describe the comparisons between the two groups regarding thoracic parameters. Regarding thoracic parameters, the ratios of sagittal diameter to transverse diameter (SD/TD) and right/left anteroposterior diameters to transverse diameter (RAPD/TD, LAPD/TD) were described separately for the PSP group and the control group, and the differences between the two groups were tested (Table 2). At all levels (manubrium, sternal angle, midpoint of sternal body, and Xiphisternal angle), the SD/TD ratio in the PSP group was significantly lower than in the control group. Furthermore, at the level of the Xiphisternal angle, the ratios of the right and left anteroposterior diameters to the transverse diameter were also significantly lower in the PSP group.

Vertical length of thorax

Regarding the vertical length of the thorax, the measurements from the thoracic apex to just above the diaphragm and to the costophrenic angle were both significantly longer in the spontaneous pneumothorax group on both the left and right sides (Table 3).

Relationship with height

The relationship between height and various thoracic parameters is presented separately for each group in Table 4. In the PSP group, TD and vertical thoracic length showed a positive correlation with height. In contrast, the control group demonstrated variability in the upper thoracic parameters but showed a consistent positive correlation between height and all thoracic parameters in the lower thoracic region. These findings suggest distinct patterns in the association between height and thoracic morphology between the two groups.

Development of a predictive model for pneumothorax onset (Table 5)

• Model 1 (age + HT + BMI): the baseline model (age, height, and BMI) had a residual deviance of 123.1 and served as the reference.
• Model 2 (adding SD/TD ratios at measurement sites): incorporating SD/TD ratios at various sites reduced residual deviance and yielded statistically significant P values (P<0.05). Only variables with significant differences between PSP and controls were included:
  • Manubrium: deviance −6.3, P=0.01.
  • Sternal angle: deviance −9.9, P=0.002 (largest improvement).
  • Midpoint of sternal body: deviance −8.2, P=0.004.
  • Xiphisternal angle: deviance −7.5, P=0.006.
  • Interpretation: Model 2-2 (SD/TD ratio at sternal angle) showed the best fit, highlighting thoracic shape at this site as critical for predicting pneumothorax onset.
• Model 3 (adding RAPD or LAPD to SD/TD ratio): based on Model 2-2, additional variables were tested:
  • RAPD/TD ratio (Xiphisternal angle): deviance −7.1, P=0.008 (significant).
  • LAPD/TD ratio (Xiphisternal angle): deviance −1.6, P=0.21 (not significant).
  • Interpretation: RAPD/TD ratio significantly improved model fit, while LAPD did not, suggesting a stronger influence of the right anterior-posterior diameter.

The ROC curve was plotted using Model 3-1. The AUC is 0.882 (95% CI: 0.82–0.95), indicating a strong ability to discriminate between patients and controls. The highlighted optimal cutoff point corresponds to a Youden index of 0.603, which yields a sensitivity of 81.4% and a specificity of 83.0% (Figure 3).

Figure 3 Prediction performance of the thoracic morphological model for PSP. The ROC curve demonstrates the performance of the multivariate logistic regression model in predicting PSP onset. The model incorporated the SD/TD ratio at the sternal angle and the RAPD at the Xiphisternal angle, adjusted for age, height, and BMI. The AUC is 0.882, with a 95% CI of 0.82–0.95, indicating excellent discriminatory power. The highlighted optimal cutoff point corresponds to a Youden index of 0.603, yielding a sensitivity of 81.4% and a specificity of 83.0%. AUC, area under the curve; BMI, body mass index; CI, confidence interval; PSP, primary spontaneous pneumothorax; RAPD, right anterior-posterior diameter; ROC, receiver operating characteristic; SD, sagittal diameter; TD, transverse diameter.

Discussion

It has been widely reported that adolescent patients are more prone to recurrent pneumothorax. Prognostic surveys conducted on pneumothorax patients who underwent surgery at our institution also revealed that adolescents had significantly higher ipsilateral and contralateral pneumothorax recurrence rates compared to other age groups (2). While younger age is often cited as a factor for pneumothorax onset, few studies have examined how age directly contributes to the development of pneumothorax.

Several studies have reported that the unique body habitus of pneumothorax patients may influence its onset. This study focuses on such physical characteristics, specifically tall stature, flattened thorax, and low BMI, commonly referred to in Japan as the “pneumothorax body shape” (11). These three characteristics will be discussed in detail below.

Tall stature

Some previous studies have suggested that pneumothorax patients tend to be tall (12-14). However, in this study, the median height of PSP patients was 173.1 cm, which cannot be considered particularly tall. Additionally, no statistically significant difference in height was observed compared to the control group.

On the other hand, our analysis revealed that the vertical dimension of the thorax was significantly longer in the PSP group compared to the control group. These findings suggest that the focus should shift from overall height to the vertical dimension of the thorax. Supporting this view, studies by Akkas et al. (4) and Peters et al. (14) have also reported that pneumothorax patients had a significantly longer thoracic vertical dimension compared to controls. These results highlight the need to focus on thoracic size rather than height alone.

Flattened thorax

In a previous study, we compared the body shape of teens and 20s PSP cases and suggested that teens PSP might exhibit distinct physical characteristics (11). Notably, teens PSP demonstrated a unique relationship between thoracic parameters and height: while the transverse and vertical dimensions correlated with height, the sagittal (anterior-posterior) dimension did not. This trend was not observed in 20s PSP, suggesting it might be specific to adolescents.

To further investigate, we compared teens PSP with a control group of the same age. In the control group, similar to 20s PSP, most thoracic parameters showed positive correlations with height. In contrast, teens PSP exhibited the same unique pattern as previously reported, indicating that this disproportionate thoracic growth is specific to this group.

To quantify thoracic flattening, we calculated the ratios of TD to SD, as well as RAPD and LAPD. All these ratios were lower in the PSP group, indicating a more flattened thorax. The SD/TD ratio in teens PSP was significantly lower at all heights. Using variables with significant intergroup differences, we analyzed their association with pneumothorax onset and found that the SD/TD ratio at the sternal angle and the RAPD/TD ratio at the Xiphisternal angle were particularly important predictors.

Casha et al. similarly reported that pneumothorax patients had a more flattened and vertically flattened thorax compared to controls (15). They also discussed pleural stress in the lung apex associated with this distinctive thoracic shape, suggesting a potential link to pneumothorax onset. Peters et al. (14) emphasized the role of thoracic shape in pneumothorax development, citing Vawter et al.’s findings that increasing lung height amplifies apex pressures compared to the lung base, potentially affecting lung volume and contributing to bulla formation (16). Further research is warranted to confirm these associations.

Low BMI

Low BMI is another well-documented characteristic of pneumothorax patients (5,13,17). In this study, the median BMI of the pneumothorax group was 17.8, consistent with a thin body shape. Importantly, this low BMI does not indicate malnutrition but rather reflects a specific body composition.

The relationship between low BMI and pneumothorax onset remains controversial. Yu et al. explored the association between BMI and bone metabolism, suggesting that BMI could be utilized for pneumothorax screening (6). Their study found that pneumothorax patients exhibited abnormalities in bone metabolism, with elevated levels of both bone formation and resorption markers. Additionally, it has been reported that BMI is negatively correlated with these bone metabolism markers.

The same study also reported significantly lower levels of 25-hydroxyvitamin D a key indicator of total vitamin D levels in pneumothorax patients. The association between PSP patients and bone metabolism or vitamin D is still speculative, but it may contribute to improving the mechanism of spontaneous pneumothorax onset in the future.

Limitations

This study has the following limitations. The control group in this study was selected as an age-matched control group that underwent chest CT for some disease or trauma. Therefore, it should be noted that they were not purely healthy individuals. Additionally, this study is limited to Japanese people. Differences in living environments and races could result in variations in skeletal formation, growth, and development. Therefore, it is necessary to broaden the range of subjects and conduct prospective studies at multiple facilities in different countries and regions in the future.

Conclusions

This study highlights flattened thorax as a distinctive feature of adolescent PSP patients, differing from normal growth patterns. The SD/TD ratio at the sternal angle and the RAPD/TD ratio at the Xiphisternal angle were identified as potential risk factors for pneumothorax onset. While low BMI is a notable trait of PSP patients, its direct link to pneumothorax onset remains unclear, warranting further investigation. These findings may enhance awareness and understanding of PSP development.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1326/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1326/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1326/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1326/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This case-control study was approved by the Institutional Review Board of Tohoku Medical and Pharmaceutical University Hospital (Approval Number: IRB: 2022-2-074). Informed consent was obtained from participants or their parents/legal guardians via an Opt-out methodology.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Committee for Scientific Affairs, The Japanese Association for Thoracic Surgery. Thoracic and cardiovascular surgeries in Japan during 2019 : Annual report by the Japanese Association for Thoracic Surgery. Gen Thorac Cardiovasc Surg 2023;71:595-628. [Crossref] [PubMed]
2. Nonomura R, Yabe R, Oshima Y, et al. Post-surgery spontaneous pneumothorax: Long-term recurrence rates and follow-up challenges revealed by a written survey. PLoS One 2024;19:e0307910. [Crossref] [PubMed]
3. Yi E, Park JE, Chung JH, et al. Trends in recurrence of primary spontaneous pneumothorax in young population after treatment for first episode based on a nationwide population data. Sci Rep 2023;13:13478. [Crossref] [PubMed]
4. Akkas Y, Peri NG, Kocer B, et al. A novel structural risk index for primary spontaneous pneumothorax: Ankara Numune Risk Index. Asian J Surg 2017;40:249-53. [Crossref] [PubMed]
5. Tan J, Yang Y, Zhong J, et al. Association Between BMI and Recurrence of Primary Spontaneous Pneumothorax. World J Surg 2017;41:1274-80. [Crossref] [PubMed]
6. Yu L, Li H, Hou S, et al. Abnormal bone mineral density and bone turnover marker expression profiles in patients with primary spontaneous pneumothorax. J Thorac Dis 2016;8:1188-96. [Crossref] [PubMed]
7. Nakamura H, Izuchi R, Hagiwara T, et al. Physical constitution and smoking habits of patients with idiopathic spontaneous pneumothorax. Jpn J Med 1983;22:2-8. [Crossref] [PubMed]
8. Hallifax R. Aetiology of Primary Spontaneous Pneumothorax. J Clin Med 2022;11:490. [Crossref] [PubMed]
9. Fujino S, Inoue S, Tezuka N, et al. Physical development of surgically treated patients with primary spontaneous pneumothorax. Chest 1999;116:899-902. [Crossref] [PubMed]
10. Mitani A, Hakamata Y, Hosoi M, et al. The incidence and risk factors of asymptomatic primary spontaneous pneumothorax detected during health check-ups. BMC Pulm Med 2017;17:177. [Crossref] [PubMed]
11. Nonomura R, Sugawara T, Yabe R, et al. Age-Specific Body Shape Characteristics in the Onset of Spontaneous Pneumothorax: A Comparison Between Teens and 20s. Cureus 2024;16:e71922. [Crossref] [PubMed]
12. Lee LP, Lai MH, Chiu WK, et al. Management of primary spontaneous pneumothorax in Chinese children. Hong Kong Med J 2010;16:94-100.
13. Liu G, Yan D, Wang X, et al. Novel insights into causal associations of body mass index or height with pneumothorax: a two-sample Mendelian randomization study. Front Nutr 2024;11:1391017. [Crossref] [PubMed]
14. Peters RM, Peters BA, Benirschke SK, et al. Chest dimensions in young adults with spontaneous pneumothorax. Ann Thorac Surg 1978;25:193-6. [Crossref] [PubMed]
15. Casha AR, Manché A, Gatt R, et al. Is there a biomechanical cause for spontaneous pneumothorax? Eur J Cardiothorac Surg 2014;45:1011-6. [Crossref] [PubMed]
16. Vawter DL, Matthews FL, West JB. Effect of shape and size of lung and chest wall on stresses in the lung. J Appl Physiol 1975;39:9-17. [Crossref] [PubMed]
17. Walker SP, Bibby AC, Halford P, et al. Recurrence rates in primary spontaneous pneumothorax: a systematic review and meta-analysis. Eur Respir J 2018;52:1800864. [Crossref] [PubMed]