Prognostic value of computed tomography assessment of normal lung volume in upper lobe in idiopathic pleuroparenchymal fibroelastosis

Extract

Idiopathic pleuroparenchymal fibroelastosis (iPPFE) is characterised by fibrosis involving the pleura and subpleural lung parenchyma and it predominantly occurs in the upper lobes [1]. Volume loss in the lungs is considered a hallmark of fibrosing interstitial lung diseases (ILDs), including iPPFE. We recently reported the prognostic importance of assessing the bilateral whole upper-lobe lung volume (UL-volume) by three-dimensional computed tomography (3D-CT) in patients with iPPFE [2]. The UL-volume was less than half the size in iPPFE patients versus controls and was associated with mortality risk, whereas the lower lobe-volume (LL-volume) did not decrease. However, this UL-volume includes dense consolidation, bronchiectasis and pleural thickening, in addition to normal parenchyma. Additionally, certain patients with iPPFE have lower-lobe ILDs. Therefore, we hypothesised that assessing the normal lung volume in the UL (normal UL-volume) would accurately indicate disease severity and mortality risk in patients with iPPFE. Thus, this longitudinal study evaluated the normal UL-volume by 3D-CT and explored its association with mortality.

Shareable abstract

This novel method for quantitatively measuring the normal upper-lobe lung volume using three-dimensional computed tomography is useful for assessing disease severity and mortality risk in idiopathic pleuroparenchymal fibroelastosis https://bit.ly/4fcRCrj

To the Editor:

Idiopathic pleuroparenchymal fibroelastosis (iPPFE) is characterised by fibrosis involving the pleura and subpleural lung parenchyma and it predominantly occurs in the upper lobes [1]. Volume loss in the lungs is considered a hallmark of fibrosing interstitial lung diseases (ILDs), including iPPFE. We recently reported the prognostic importance of assessing the bilateral whole upper-lobe lung volume (UL-volume) by three-dimensional computed tomography (3D-CT) in patients with iPPFE [2]. The UL-volume was less than half the size in iPPFE patients versus controls and was associated with mortality risk, whereas the lower lobe-volume (LL-volume) did not decrease. However, this UL-volume includes dense consolidation, bronchiectasis and pleural thickening, in addition to normal parenchyma. Additionally, certain patients with iPPFE have lower-lobe ILDs. Therefore, we hypothesised that assessing the normal lung volume in the UL (normal UL-volume) would accurately indicate disease severity and mortality risk in patients with iPPFE. Thus, this longitudinal study evaluated the normal UL-volume by 3D-CT and explored its association with mortality.

This multicentre, retrospective, observation study was approved by the Ethical Committees of Hamamatsu University School of Medicine (20–270), Seirei Mikatahara General Hospital (3533) and Seirei Hamamatsu General Hospital (20–65). The requirement for patient approval and/or informed consent was waived because of the retrospective study design.

The medical records of 147 consecutive patients who were diagnosed with iPPFE at participating institutions between March 2004 and December 2021 were retrospectively analysed, and three patients were excluded because of insufficient clinical data. Thus, this study enrolled 144 patients with iPPFE. iPPFE was diagnosed by multi-disciplinary discussion using previously reported criteria [3, 4]. The observation period was calculated from the date of iPPFE diagnosis to the date of the last visit. Patients were censored if they were alive on 30 September 2023 or beyond. Cumulative survival probabilities were estimated above and below the median cut-off using the Kaplan–Meier method and log-rank test. Cox-proportional hazards regression analysis was used to identify factors associated with mortality. All analyses were conducted using R version 4.11.

The lung volume was measured using electronic data from chest CT at iPPFE diagnosis and chest CT scans taken a year later, if available, by SYNAPSE VINCENT (version 3, FUJIFILM Medical Systems, Japan). The standardised lung volume (%) was calculated by dividing the volume measured using 3D-CT by the predicted forced vital capacity (FVC), as previously described [2, 5–7]. The normal lung ratio (NL ratio) was calculated as the ratio of the normal attenuated lung volume (ranged −950 to −750 Hounsfield units (HU)) to the whole-lung volume [8, 9]. Finally, normal lung volume (%) was calculated by multiplying the standardised lung volume (%) by the NL ratio (%); normal lung volume (%) = standardised lung volume (%) × NL ratio (%). The change in the normal lung volume was calculated as a relative change (normal lung volume _baseline − normal lung volume_{1 year}/normal lung volume_baseline).

Among the 144 patients, the median age was 69 years and 95 patients (66%) were male. 30 patients underwent surgical lung biopsy. The median body mass index, FVC and diffusing capacity of the lung for carbon monoxide was 17.0 kg·m⁻², 64.5% and 89.8%, respectively. Lower-lobe ILD was found in 92 (64%) of patients.

The representative segmental lung images of the area between −950 HU (excluding low-attenuation lesions) and −750 HU (excluding consolidation and ground-grass attenuation lesions) in a patient with iPPFE are presented in figure 1a. The normal total lung volume, normal UL-volume and normal LL-volume were 75.4% (range 58.6–93.2%), 18.8% (range 13.9–24.8%) and 45.5% (range 32.8–57.8%), respectively, and were correlated with FVC (r=0.379, 0.542 and 0.258, respectively). The iPPFE patients with lower-lobe ILD had significantly decreased normal UL-volume and normal LL-volume than observed those without lower-lobe ILD.

FIGURE 1.

Three-dimensional lung images and survivals curve of patient with iPPFE. a) Representative segmental lung images of a patient with iPPFE at diagnosis. The patient was a 70-year-old woman of 155.4 cm in height and 31.8 kg in weight (left panel). Computed tomography images for the areas less than −750 and −950 Hounsfield units in the same patient (middle and right panel, respectively). The lungs are colour-coded as follows: right upper lobe: yellow; right middle lobe: blue; right lower lobe: green; left upper lobe: red; left lower lobe: pink. b) Kaplan–Meier curves of patients with iPPFE according to the normal upper-lobe volume. The cut-off was derived from the median normal upper-lobe volume (18.8%). c) Annual changes of the normal upper-lobe volume. d) Kaplan–Meier curves of patients with iPPFE according to age, sex and the normal upper-lobe volume. p-values were determined by the log-rank test. iPPFE: idiopathic pleuroparenchymal fibroelastosis; UL-volume: upper-lobe lung volume.

First, we sought to address whether lower normal lung volume was associated with the risk of mortality. Between groups stratified by the median normal UL-volume (18.8%), patients with normal UL-volume ≤18.8% had significantly shorter survival than those with normal UL-volume >18.8% (34.6 months versus 73.0 months, p<0.001; figure 1b).

Univariate analysis identified age, sex, FVC, diffusing capacity of the lung for carbon monoxide, lower-lobe ILD, normal total lung volume, normal UL-volume and normal LL-volume as significant factors for mortality, whereases NL ratio was not. Multivariate analysis illustrated that lower normal UL-volume <18.8% was significantly associated with mortality after adjusting for age, sex and FVC (hazard ratio 1.873, 95% CI 1.188–2.970, p=0.007).

Next, annual changes in the normal UL-volume were explored. Among 144 patients with iPPFE, 104 patients underwent chest CT 1 year after diagnosis (median 13.4 (range 10.7–19.3) months). Approximately half of the patients exhibited a >30% decline in the normal UL-volume (figure 1c). Notably, the extent of the decline was severe in patients with a lower normal UL-volume at diagnosis. As expected, patients with a relative decline in the normal UL-volume of ≥30% exhibited significantly shorter survival than those with a relative decline of <30% in the combined cohort (47.2 months versus 80.0 months, p=0.005).

Finally, this study attempted to develop a simple composite model using age, gender and the normal UL-volume, because the gender-age-physiology (GAP) index was reported to be less discriminative for disease severity in patients with iPPFE than in those with idiopathic pulmonary fibrosis [2, 10–12]. In line with the GAP index, one point was assigned for age>60 years, male sex and normal UL-volume ≥18.8%. The patients were categorised into three groups based on the total score: mild (0–1), moderate (2) and severe (3). This model successfully classified patients with iPPFE into three groups based on median survival (mild: 95.5 months; moderate: 53.2 months; and severe: 20.8 months; figure 1d). The C-index of this model for mortality was higher than that of the GAP index, but comparable to that of the standardised UL-volume model.

Our study measured normal UL-volume in patients with iPPFE. Although several studies examined NL ratio (%) [13–15] and standardised lung volume (%) [2, 5–7] using 3D-CT images, none of these studies had evaluated normal lung volume (%). Theoretically, normal lung volume enables one to evaluate both fibrotic changes as well as abnormalities such as emphysematous and inflammatory changes.

The benefit of 3D-CT analyses in assessing physiology is that it is a safe and patient-friendly method that is applicable to patients with advanced disease or respiratory failure [16]. In particular, because the distribution of ILD lesions is heterogeneous, 3D-CT permits evaluation of each lobe separately, making it ideal for patients with iPPFE. Indeed, decreased normal UL-volume was associated with shorter survival, suggesting importance of UL-volume rather than those of LL-volume. Further, annual normal UL-volume decline frequently occurred, and the extent of the decline was severe in patients with decreased normal UL-volume, suggesting accelerated disease progression in patients with advanced iPPFE.

To date, several reports have proposed composite models [2, 10–12]. An advantage of this model is that it consists of only three factors. However, its statistical superiority over previously reported models has not been demonstrated, nor has its optimal cut-off for low or high values of normal UL-volume been determined. Thus, the issues of overfitting should be acknowledged in this study. The relevance of the risk factors and cut-offs used in this study should be independently confirmed in the future. Additionally, there were several limitations. First, this was a retrospective study. In addition, quantitative analyses of the normal lung volume by 3D-CT required commercial software and a complicated procedure.

In conclusion, the normal lung volume decreased each year predominantly in the upper lobes, and the normal UL-volume was associated with mortality risk in patients with iPPFE. Collectively, our results suggested that assessing the normal UL-volume is useful for monitoring disease severity and mortality risk in patients with iPPFE.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.