Pulmonary hypertension in combined pulmonary fibrosis and emphysema: A tale of two cities

Emphysema and pulmonary fibrosis, two separate entities, can coexist and give rise to a subgroup of patients who share distinct clinical, physiological and radiological features. The term combined pulmonary fibrosis and emphysema (CPFE) has been used to describe this phenomenon, although a consensus definition of CPFE does not currently exist. To date, this term has been used to include patients with coexistent emphysema and various forms of pulmonary fibrosis, although idiopathic pulmonary fibrosis (IPF) has been the main focus of research. Male gender, a smoking history, an upper lobe predominance of emphysema and lower lobe predominance of fibrosis is typical, together with well-preserved lung volumes and disproportionately low gas transfer measurements.1, 2

Pulmonary hypertension (PHT) in CPFE has been reported to be more common compared with IPF patients without emphysema and is thought to contribute to the increased mortality seen in patients with CPFE.2-5 However, it is not clear whether PHT in CPFE results from a distinct microvascular phenotype or merely the result of an addictive effect of two separate processes that are known to be independently associated with PHT. In this issue of Respirology,6 Jacob et al. evaluated whether patients with IPF and emphysema (CPFE) had an increased likelihood of PHT when compared with IPF patients without emphysema. The likelihood of PHT at baseline was assessed with transthoracic echocardiography using two definitions: (i) increased likelihood of PHT—using a right ventricular systolic pressure (RVSP) >50 mm Hg and (ii) high likelihood of PHT—using consensus guidelines.7 The prevalence of a likelihood of PHT was similar in both CPFE and IPF patients (31% vs 35%, P = 0.52). Disease severity at baseline was quantified as total disease—emphysema and ILD—extent on computed tomography (CT). Two scores were used: (i) sum of visual ILD extent and visual emphysema extent and (ii) sum of CALIPER (Computer-Aided Lung Informatics for Pathology Evaluation and Rating) ILD extent and visual emphysema extent. When patients with a likelihood of PHT were examined, CPFE patients did not have significantly more disease on CT than IPF patients. When patients with CPFE alone were sub-analysed, patients with a likelihood of PHT had significantly more disease on CT than patients without PHT. Total CT disease extent predicted the likelihood of PHT. However, after adjustment for total disease extent, CPFE had no stronger association with PHT likelihood than IPF patients. CPFE patients with the likelihood of PHT had a 60% survival at 12 months, and this was not different compared to patients with IPF with the likelihood of PHT. These findings were robust across two cohorts, using either definition of PHT and either definition of CT disease extent.

Whilst one could be sceptical about the detection of PHT without right heart catheterization (RHC), the term ‘likelihood of PHT’ was used to avoid misleading PHT statements. To reduce the misclassification of PHT, the authors used two definitions of PHT. The crucial findings did not change despite the reduced prevalence when a more stringent definition (high likelihood of PHT) was used. The present cohorts appear to be similar to the cohorts reported in the literature.2, 4 In the study of PHT in CPFE where the presence of PHT was detected by RHC,4 diffusion capacity (DL_CO) characteristics and survival were similar to that of the present cohorts. The prevalence of PHT from this study was 30%, and whilst this is lower than the reported prevalence of 47% in another study,2 this is likely due to the higher threshold used to define PHT in the present study (RVSP > 50 mm Hg vs RVSP > 45 mm Hg). In contrast, Mejia et al.3 reported a significantly higher prevalence of PHT of 90% (defined by an RVSP >50 mm Hg), but this is likely due to the patients having more advanced disease. Therefore, the present CPFE cohort is likely representative of the published CPFE cohorts at large. Whilst it is plausible that PHT in CPFE is not associated with a distinct microvascular phenotype, there might be more to this tale. In a histological analysis of vasculopathy associated with PHT, Awano et al.8 evaluated pulmonary vasculopathy in an autopsy series of patients with CPFE, and compared these findings with those of patients with IPF alone and emphysema alone. Vascular changes were evaluated in the fibrotic areas, emphysematous areas and the relatively unaffected/preserved areas. Moderate-to-severe vasculopathy was seen in the CPFE group, but there were no significant differences in the fibrotic and emphysematous areas among the three groups. However, in the preserved areas, the severity was significantly different among the three groups and vasculopathy in the CPFE groups was the most severe. Therefore, in CPFE, pulmonary vasculopathy occurring in areas of preserved lung, that is assessed as no disease on CT, may play a role in the pathogenesis of PHT.

It is reassuring that the findings were also robust using two definitions of CT disease extent. The use of computer-based CT analysis is being increasingly used in the study of lung disease.9-12 A number of techniques have been described, including the texture classification method9 which allows the differentiation of parenchymal lung disease associated with both ILD and emphysema. The most commonly used texture analysis method is CALIPER. ILD extent quantified by CALIPER has been shown to be superior to visual scores. Compared with visual scores, the CALIPER estimates of ILD extent correlated more strongly with pulmonary function tests12 and prognosis.13 The scoring of emphysema, especially in the presence of pulmonary fibrosis, is more complex. This is largely due to the difficulties in distinguishing between honeycombing and emphysematous changes. Similar to other studies,10, 11 there is less agreement between observers when scoring visually for emphysema extent compared with ILD extent. CALIPER also has difficulties differentiating between small areas of centrilobular emphysema and adjacent normal lung,14 and separately, honeycombing from emphysema when conglomerate destructive emphysema is present within areas of fibrosis.15 Correlations between CT measures of emphysema extent and pulmonary function tests have been weak and did not differ between CALIPER and visual scoring.12 Despite the pitfalls of quantifying emphysema in the presence of pulmonary fibrosis, this is the first study whereby extents of both ILD and emphysema have been used to control for disease severity.

Emphysema is categorized into three major subtypes (paraseptal, centrilobular and panlobular) according to the disease distribution in the secondary pulmonary lobules. Most studies to date in CPFE have not taken into account the different subtypes of emphysema. In a handful of studies where emphysema subtypes were evaluated, paraseptal emphysema was associated with a worse prognosis compared with the other subtypes.5, 16 Interestingly, in the Framingham Heart Study,17 3% (86/2633) of the study population had pure paraseptal emphysema and these changes were predominantly located in the upper zones. Cigarette smoking, ageing and male gender were associated with the presence of paraseptal emphysema. In addition, there was a significant association between the presence of paraseptal emphysema and interstitial lung abnormalities.

For the first time, Jacob et al.6 challenge the dogma that CPFE is associated with a malignant microvascular phenotype. These results clearly need to be validated in further cohorts of IPF patients. Whether the same findings apply to other forms of pulmonary fibrosis, and even different subtypes of emphysema, remain to be seen.