include the clustering of pulmonary brosis and emphysema (e.g. more frequent presence of emphysema on HRCT than expected in several brotic ILDs) [17, 35, 81, 156], and the existence of multiple pathways common to both pulmonarybrosis and emphysema. However, whether CPFE might correspond to a single biologically unique entity in a proportion of cases warrants further study.

emphysema preceding the onset of the pulmonary brosis in the majority of cases [8], and a dismal prognosis. PH may occur in about half of the cases after a median of 18 months following the diagnosis of CPFE [25] (Fig. 33.11). Survival is severely reduced, with a median of 12–18 months from the onset of con rmed precapillary PH. Lung cancer may occur after a median of 2 years after the diagnosis of CPFE (from 0.6 to 11.2 years in our series [11]).

Biology

As in IPF and as observed in the general population, a minority of patients may have antinuclear antibodies, which in the absence of systemic clinical features have limited clinical relevance. As discussed above, anti-neutrophil cytoplasmic antibodies may also be found [14]. The serum levels of KL-6 (Klebs von der Lungen-6) and SP-D (surfactant protein-D) are elevated [187].

The differential cell count of bronchoalveolar lavage fuid is similar in CPFE to that of IPF and does not contribute signi cantly to the diagnosis [8, 22]; bronchoalveolar lavage is useful in the setting of acute worsening or suspicion of lung infection. Chemokines implicated in the recruitment of neutrophils (ENA-78/CXCL5 and IL-8/CXCL8) are elevated in the bronchoalveolar lavage fuid as compared to IPF [22] and may contribute to the development of emphysema.

Complications and Outcome

Making the diagnosis of the CPFE syndrome is highly relevant, because the outcome and risk of complications is distinct from that of either IPF or emphysema alone. The disease course of the CPFE syndrome is often characteristic, with

Mortality

The overall median survival in CPFE is between 2.1 and 8.5 years [8, 13, 15, 20, 21, 129, 188–190]. In CPFE and connective tissue disease, the overall survival was 100% at 1 year, 94% at 2 years, and 73% at 5 years [26]. The main causes of deaths in patients with CPFE are represented by severe PH, intractable hypoxemia, pulmonary infection, and lung cancer. The relatively preserved lung volumes in CPFE may underestimate the severity of the pulmonary brosis.

The main variables associated with an increased risk of death in CPFE include a lower DLco [16, 25, 183, 191–193], FVC < 50% of predicted value [21], composite physiologic index ≥45 [137, 193, 194], oxygen saturation on room air <90% [144], home oxygen use [183], pulmonary hypertension [8, 16, 21, 25, 81, 137, 183, 195], presence of UIP pattern [16, 194] or honeycombing [16, 183, 191], and increased extent of brosis on HRCT [144, 192]. Over the course of disease, decline of ≥10% over 12 months [132], presence of lung cancer [108, 144, 158, 183, 191, 196, 197], or acute exacerbation of brosis [197] are also associated with a poor prognosis.

Comparison of survival between patients with CPFE and IPF has been controversial [189], as results may vary accord-

Fig. 33.11  Diagram of the main complications in tobacco-related CPFE syndrome. The median delay between emphysema andbrosis is 5 years, and pulmonary hypertension may occur after a median of 18 months after the diagnosis of CPFE. Patients may die after a median of about 1 year from the diagnosis of pulmonary hypertension and a median of 7 months from the onset of lung cancer, with great inter-individual variability. PH pulmonary hypertension

Tobacco smoking

Emphysema

Fibrosis

7 months

ing to: (1) diagnostic criteria for CPFE, with survival in CPFE possibly confounded pathologies other than UIP especially NSIP (i.e. non-UIP CPFE); for example, the proportion of patients with NSIP was higher in the CPFE group than in the control group in one study that found better survival in CPFE than in pulmonary brosis without emphysema [188]; (2) dif culties in controlling for the severity ofbrosis; (3) survival time bias, with presumably earlier diagnosis due to more severe symptoms in patients with CPFE as compared to IPF after adjustment for the severity of brosis; patients with CPFE more frequently have chronic bronchitis than those with IPF and may seek medical attention earlier [189], which may explain why brosis and emphysema are inversely correlated at the diagnosis of CPFE [198]; (4) the method used to handle transplantation [189]. A composite physiologic index predicts mortality in isolated IPF and may account for disease severity [5], however, there are uncertainties regarding its use in CPFE [132]. In one series that de ned CPFE as IPF (with a HRCT and/or pathological ­pattern of UIP) associated with emphysema, mortality was similar in patients with CPFE as compared to IPF after adjustment for the severity of brosis at imaging [13]. It should be emphasized that there is great variability of progression between patients with the CPFE syndrome, with individuals with slow or rapid progression of disease especially regarding the brosis (Fig. 33.12).

Comparison between CPFE and isolated IPF with quanti-cation of both pulmonary brosis and emphysema was conducted in two retrospective cohorts of patients with IPF [111, 199], using both visual analysis to the nearest 5% [111, 199] and computer-based analysis with the CALIPER software [111]. The global disease extent on HRCT (i.e. the combined extent of brosis and of emphysema) and the baseline DLco both predicted mortality, refecting the overall severity of parenchymal lung destruction [111, 199]. After correction for baseline severity using DLco, the presence or extent of emphysema did not impact survival [21, 111, 199, 200]. Overall, these data suggest that outcomes are worse for a given extent of brosis, when there is emphysema in addition to brosis, however, the risk of mortality and of developing PH does not differ in patients with both IPF and emphysema compared to those with brosis alone when adjusting for severity using baseline DLco or total disease extent on HRCT (e.g. total extent of brosis and emphysema) [19, 111, 201].

Pulmonary Hypertension

The main determinant of prognosis in CPFE is represented by PH [8, 21], which develops in up to half of patients with CPFE [8]. In various series, PH has been reported in 15–55% of patients with CPFE [8, 21, 26, 81, 151]. When present, PH

may cause dilation of the pulmonary arteries at chest radiograph and HRCT. Magnetic resonance imaging may contribute to the evaluation of PH in CPFE [202], however, right heart catheterization remains the gold standard to diagnose PH.

The additional burden of emphysema, over and above a given extent of brosis, increases the risk of PH. However, the likelihood of PH does not differ for matched extents of disease (combined brosis and emphysema) on HRCT (or when adjusted for DLco) between patients with CPFE and those with brosis alone [19, 111].

PH in CPFE is likely due to the synergistic effect of emphysema and brosis on the pulmonary capillary bed, with further effects of hypoxia, possible sheer stress, and intrinsic pulmonary vascular abnormalities [203, 204]. In addition, there is accumulating evidence that tobacco smoking may directly contribute to PH [205–207], with obvious relevance in the CPFE syndrome.

Precapillary PH is de ned by mean pulmonary artery pressure of 20 mmHg or greater, with pulmonary artery wedge pressure of 15 mmHg or lower, and pulmonary vascular resistance ≥3 Wood units. It is the main determinant of subsequent mortality in CPFE, with an overall probability of survival of only 60% at 1 year once con rmed by right heart catheterization [25]. Low cardiac index and high pulmonary vascular resistance are then the main determinants of shorter survival [25]. According to the World Symposium on Pulmonary Arterial Hypertension, PH is considered severe in patients with mean pulmonary artery pressure of 35 mmHg or greater, or in those with decreased cardiac index (lower than 2 L/min/m2) [208]. In a series of 40 patients with CPFE and PH [25], 68% had a mean pulmonary artery pressure >35 mmHg [25]; 92% of patients needed long-term supplemental oxygen therapy; 15% had developed acute right heart failure after a mean follow-up of 8 ± 8 months; and death was due to hypoxemia and chronic respiratory failure in most cases. In patients with CPFE and connective tissue disease, PH mostly occurs in those with systemic sclerosis [26], a condition in which PH is frequent.

Lung Cancer

Lung cancer is increasingly recognized in patients with the CPFE syndrome (Fig. 33.13), which is hardly surprising given the increased risk of lung cancer in relation to tobacco smoking, chronic obstructive pulmonary disease, and IPF [9]. Lung cancer has been reported in 2–52% of patients with CPFE [108, 111, 112, 129, 130, 151, 181, 187, 190, 191, 209, 210]. A meta-analysis [211] found a higher risk of lung cancer in patients with CPFE (UIP and emphysema) than those with IPF alone, regardless of the extent of emphysema.

Fig. 33.12  Example of slow progression of disease in a woman with smoking-related CPFE syndrome (chest HRCT), who quit smoking at diagnosis. Only mild centrilobular emphysema is present in the upper

lobes. Subpleural reticular changes predominate, which progress to a pattern of probable usual interstitial pneumonia. A thick-walled large cyst is present in the right lower lobe, which increases in size

In a series of 47 patients with lung cancer in the context of CPFE [11], all patients were men, all smokers, with a mean age of 68 years. Importantly, a pathological diagnosis was obtained in only 81% of patients, with the underlying CPFE limiting the diagnostic approach in the remaining cases due to severe functional impairment. Similarly, 43%

of patients could not receive the standard of care treatment for lung cancer following international guidelines, despite a mean FVC of 87 ± 24% of predicted value, and FEV1 of 74 ± 19% of predicted. The mean DLco was 44 ± 16% of predicted (range, 20–80), accounting for a large proportion of diagnostic and treatment limitations together with paren-

Fig. 33.13  Non-small cell lung cancer (squamous cell carcinoma) in a male patient with smoking-related CPFE syndrome (chest HRCT). Upper panel: centrilobular emphysema in the left upper lobe (and post-­ infectious bronchiectasis in the right upper lobe); upper middle panel: lower zones showing usual interstitial pneumonia pattern with honeycombing; lower middle panel: spiculated mass (arrow) in the right lower lobe, adjacent to honeycombing and reticulations; lower panel: thick-walled large cysts and honeycombing in the lung bases

chymal changes at imaging. The most frequent histologic type of lung cancer was squamous cell carcinoma (45% of cases) followed by adenocarcinoma (37%), whereas adenocarcinoma is the most common type of non-small cell carcinoma in Europe; the frequency of squamous cell carcinoma in CPFE may be related to heavy smoking his-

tory. Comparable histologic types were found in series from Japan [110, 130, 187, 212–219]. The majority of the lung cancers are located in the lower lobes [212, 218]. There is greater invasion and the diagnosis is made at a later stage compared to non-small cell lung cancer without CPFE [219, 220].

The molecular mechanisms underlying the carcinogenesis of lung epithelial cells in CPFE is largely unknown. A common susceptibility to emphysema, brosis, and lung cancer, possibly through the telomerase complex, has been hypothesized [11, 221, 222].

Among patients with lung cancer, those with CPFE seem to have a worse outcome as compared to patients with either emphysema or brosis [212, 219]. Acute exacerbation ofbrosis is a major complication in patients with CPFE, and may occur during treatment for lung cancer, especially following radiation therapy [11] or surgery [11, 151, 212, 223]. Acute respiratory distress syndrome may also follow chemotherapy for lung cancer [224]. These complications seem comparable to that occurring in patients with ILD especially IPF and undergoing cancer treatment, however with particularly high severity and risk of death in subjects with the CPFE syndrome and pronounced impairment of gas exchange. Stereotaxic radiotherapy has been suggested in this setting, however acute exacerbation in CPFE is largely unpredictable [11].

Acute Exacerbation of Pulmonary Fibrosis

The natural course of disease in CPFE may encompass episodes of acute exacerbation of pulmonary brosis [26, 151, 158, 182, 197, 212, 225]. In a series of 93 patients with CPFE, 24% developed acute exacerbation [151]. Risk factors have not been well identi ed in the setting of CPFE, however those identi ed in IPF without emphysema, especially low lung function and rapid decline 6 months after baseline, may presumably apply [226]. Interestingly, a study has demonstrated that PH at the time of evaluation for lung transplantation for IPF was associated with a high risk of subsequent acute exacerbation of IPF [227], with no other association with baseline variables being observed. In another study of IPF, presence of emphysema and low DLco were risk factors for acute exacerbation [228]. These observations suggest that patients with the CPFE syndrome may be at particularly high risk of acute exacerbation of pulmonary brosis.

Diffuse ground glass and/or consolidation on chest HRCT help to differentiate exacerbations of brosis from exacerbations of emphysema in CPFE [229]. The prognosis of acute exacerbation in CPFE might be better than that of isolated IPF [20, 225].