Update in Diffuse Parenchymal Lung Diseases 2005

CLINICAL ADVANCES

A number of publications have shed light on important clinical concepts surrounding the clinical evaluation and management of diffuse parenchymal lung diseases (DPLDs). These include approaches to idiopathic interstitial pneumonias (IIPs), including nonspecific interstitial pneumonia (NSIP) and idiopathic pulmonary fibrosis (IPF; idiopathic usual interstitial pneumonia [UIP]), connective tissue-associated DPLD, hypersensitivity pneumonitis (HP), and lymphangioleiomyomatosis (LAM).

IIPs

A series of sentinel observations have addressed diagnostic issues, staging approaches, unique phenotypes, and management strategies for IIPs.

Diagnostic approach. The approach to diagnosis in patients with DPLDs has continued to evolve over the past decade. A recent centennial review described the major impact of high-resolution computed tomography (HRCT) in guiding the diagnostic approach to patients with suspected IIP (1). It is evident that stereotypical imaging features have been described for many DPLDs, including some of the IIPs (2). The diagnostic accuracy of the radiological features of UIP has been demonstrated in well-designed clinical studies (3-5). Despite these data, there remains diagnostic confusion regarding the separation of UIP from other IIPs, particularly NSIP (6). On the other hand, increasing data have suggested that predominance of ground-glass opacity with no honeycombing is more common in NSIP. In a recent series, such a pattern was seen in 96% of patients with NSIP and 59% of patients with UIP (7). Unfortunately, the overall diagnostic accuracy of HRCT for separating these two disorders was 70% in this series.

The contrast between expert and less experienced radiologists in applying HRCT criteria to the diagnosis of IPF has recently been reported in data from a landmark clinical therapeutic trial (8). As part of this trial, HRCT scans were interpreted by clinical site radiologists and an expert radiology core using predefined criteria. The core radiologists agreed with a diagnosis of IPF in 99% of those diagnosed by site radiologists; little difference was found if the clinical site was an "academic" center or a community-based practice. A survey of American College of Chest Physician members (52.6% responses to the survey) suggested that 67% of respondents accepted HRCT diagnosis for IPF (9). Interestingly, one investigative group has demonstrated that interaction between expert clinicians and radiologists improves the interobserver agreement of the former in evaluating patients with IIPs (10). It is evident that a consensus is developing that typical features of UIP in the appropriate clinical setting can be used to accurately diagnose IPF. In other IIPs, particularly NSIP, the data may be more variable and a surgical lung biopsy remains an important part of the diagnostic approach.

Concurrent emphysema and IPF on HRCT may identify a varying phenotype in IPF. Two groups have recently described such a picture. In one cohort, 47% of patients exhibited pulmonary hypertension at presentation and 55% during a mean of 2.1 yr of follow-up (11). Survival was 87.5% at 2 yr and 54.6% at 5 yr, with a median survival of 6.1 yr. In a smaller cohort of eight patients with upper lobe predominant emphysema and basilar fibrotic changes consistent with UIP, seven of eight patients exhibited echocardiographic pulmonary hypertension (12).

The role of histopathologic evaluation has similarly evolved over the past several decades (1). The histologic criteria for DPLDs, including the IIPs, have become better accepted. Unfortunately, significant interobserver disagreement remains among expert pathologists (13). The majority of the variability relates to distinguishing NSIP from UIP. This has been highlighted by a retrospective study comparing local, general pathologist interpretation with that of an expert pathologist (14). The histopathologic interpretation by the specialist differed from the general pathologist in 52% of cases. It is evident that additional work is required to standardize pathologic interpretation.

Staging. Several investigators have continued to highlight the prognostic value of pulmonary function studies in both IFF and NSIP (8, 15). Two recent studies have confirmed that a baseline decrease in DL^sub CO^ is highly predictive of mortality in IPF and NSIP (8, 15). The work of legal and colleagues is particularly germane because they examined a large cohort of patients with IFF (n = 131) and NSIP (n = 48). Multivariate modeling confirmed that a very low baseline DL^sub CO^ seems to predict impaired survival, independent of histologic diagnosis, confirming the previous findings of another group (16). Based partly on these results, one group has proposed classifying patients with IPF and NSIP as exhibiting advanced disease if the baseline DL^sub CO^ is less than 39% predicted and limited if the DL^sub CO^ is greater than 40% predicted (17). The prognostic value of serial changes in physiologic parameters has been further supported by the finding that a 6-mo decrease in FVC of greater than 10% was predictive of survival (15), confirming the findings of other groups (18, 19). On the other hand, in the placebo-treated arm of a large therapeutic trial in IPF, a decrease in FVC of more than 10% predicted exhibited a sensitivity of 60%, specificity of 75%, positive predictive value of 31%, and negative predictive value of 91% in predicting survival (20).

Several groups have confirmed the prognostic value of a simple 6-min walk study in IPF (21-23). Two of these groups confirmed a remarkably similar mortality in patients with IPF who demonstrated a trough saturation of less than 88% during a room-air 6-min walk study (21, 23). One of these groups examined a cohort of patients with fibrotic IIP serially and confirmed an excellent reproducibility in 6-min walk distance but poor reproducibility in the amplitude of oxygen desaturation (23). On the other hand, they confirmed that a trough saturation of less than 88% was highly reproducible in the short term. It is evident that baseline and longitudinal changes in pulmonary physiologic parameters have a clear role in the clinical assessment and management of patients with IIP.

Unique phenotypes. The importance of a family history of fibrosing lung disease was demonstrated by a report characterizing 111 families with two or more cases of IIPs among first-degree family members (24). Within these families were 309 affected and 360 unaffected individuals (237 members, 25%, refused to participate). A thorough review confirmed that older age, male sex, and a history of smoking cigarettes were associated with the development of fibrosing interstitial pneumonia (IP). Importantly, 45% of the pedigrees demonstrated phenotypic variability with various subtypes of IPs present within the families. A smaller, retrospective analysis of 15 families with familial IPF suggested that the clinical, radiologic, and pathologic findings were more similar in familial IPF cases than those with nonfamilial disease (25). Importantly, outcome was similar between familial and nonfamilial IPF.

The natural history of IPF has been further defined by the identification that acute exacerbations of disease occur with a greater frequency than previously suspected. Examination of the natural history and pace of death in a carefully evaluated cohort of patients with IPF enrolled in the placebo arm of a multicenter therapeutic trial suggested that 47% of patients who died during a median follow-up of 72 wk died with an abrupt (occurring within minutes to hours) or acute (= 4 wk) pace of deterioration before death (26). Overall, approximately 8.9% of deaths in this cohort died after abrupt or acute deterioration. A small retrospective study presented details in seven patients who presented with an acute deterioration in dyspnea (< 28 d; median, 14 d) (27). CT scans in all patients demonstrated ground-glass opacities and airspace consolidation, whereas surgical lung biopsies (in five patients) or autopsies (in two patients) demonstrated underlying UIP and superimposed diffuse alveolar damage. Further prospective data are required to define the incidence, natural history, and the optimal diagnostic and therapeutic approach to acute exacerbations of IPF.

Management strategies. Therapeutic options for the IIPs have evolved over the past decade (1). In the case of IPF, this has been particularly evident given the lack of definitive data supporting efficacy for currently available therapies. The standard paradigm of disease pathogenesis, which suggested a predominantly inflammatory process, has been challenged by an aberrant host response to epithelial cell injury with resulting fibroproliferation.

Several groups have provided additional controlled data that have shed further light, and raised new questions, regarding therapeutic options in IPF. Azuma and colleagues reported the results of a prospective, double-blind, placebo-controlled trial of pirfenidone, a novel compound with pleiotropic antifibrotic effects (28). One hundred and seven patients with IPF were randomized to pirfenidone or placebo in a study with a novel primary endpoint, the change in lowest Sp^sub O2^ during a steady-state 6-min exercise test. The primary endpoint did not reach significance in the overall group. However, in the 80 of 107 patients who were able to complete the exercise test at baseline, improvement favoring pirfenidone was noted. Change in VC at 9 mo favored pirfenidone but not TLC, DL^sub CO^, or resting Pa^sub O2^. The study was terminated early by the data safety monitoring board after 14% of the patients in the placebo group met predetermined criteria for an acute exacerbation compared with none in the pirfenidone group. Additional therapeutic trials are ongoing to better define the therapeutic effect, optimal dose, and risk-benefit of pirfenidone in IPF.

A multicenter European group applied concepts of increased oxidative stress in IPF (described below) by performing an elegant controlled trial of azathioprine/prednisone/placebo versus azathioprine/prednisone/N-acetyl cysteine (NAC) in 182 patients with IPF (29). The study was powered to detect a treatment difference of 15% for VC and 20% for DL^sub CO^ after 1 yr. Using last-observation-carried-forward methodology, there was a statistically significant difference between the two treatment groups favoring the NAC arm (relative difference of 9% in VC and 24% in DL^sub CO^). In addition, the NAC-treated group experienced significantly lesser bone marrow toxicity. These data clearly support that NAC provides beneficial clinical and physiologic effects when added to azathioprine and prednisone in IPF. A thoughtful editorial highlighted the inherent difficulty in extrapolating that such "triple combination" therapy should be considered standard therapy in patients with IPF given the absence of placebo control for the azathioprine/prednisone combination (30). It is interesting to note that the longitudinal change in FVC and DL^sub CO^ with the triple combination was remarkably similar to that noted in the placebo arm of two recent trials (26, 28).

An additional interesting study examined the role of modulating the prothrombotic state in patients with IPF by completing an open-labeled, placebo-controlled trial of anticoagulation in addition to prednisone (31). Unfortunately, there was differential dropout, with 8 of 31 patients in the anticoagulation not participating compared with none of 33 patients in the anticoagulant arm. There was no dramatic difference in survival time or hospitalization-free survival, the primary endpoints, although overall mortality was worse without anticoagulation (hazard ratio, 2.9; 95% confidence interval, 1.0-8.0). The major difference appeared to relate to the mortality from acute exacerbations of IPF. The dropouts and open-label nature of the study limit the interpretation of its results, but certainly suggests that additional investigation into this novel therapeutic approach is warranted.

Miscellaneous DPLDs

Several investigative groups provided important insight into the characteristics and complications of LAM. Ryu and colleagues presented preliminary results of a 3-yr, multicenter registry of patients with LAM (32). With aggressive recruitment, the six centers recruited and characterized 243 patients (both incident and prevalent cases) with both sporadic and tuberous sclerosis complex (TSC)-associated LAM. From this well-characterized cohort, numerous important observations flowed, including that spontaneous pneumothorax was the sentinel event leading to diagnosis in about one-third of cases. In contrast to previous reports, the age range was relatively broad (18-76 yr), with more than one-third of patients being postmenopausal. Comparison between sporadic (n = 196) and TSC-associated disease (n = 34) yielded important information, including similar symptoms but a greater prevalence of renal angiomyolipomas and nephrectomy in patients with TSC. Overall for the cohort there was significant impairment in health-related quality of life. Additional clinically relevant information was provided by TaveiraDaSilva and colleagues who examined bone mineral density (BMD) in a large cohort (n = 211) of patients with LAM (33). Importantly, 70% of patients had abnormal BMD, with greater abnormality associated with greater severity of lung disease, menopause, and oophorectomy. After adjusting for baseline lung function and baseline BMD, biphosphonate therapy but not progesterone therapy was associated with lower rate of bone loss. These data highlight an important comorbid condition in patients with LAM.

HP remains a challenging diagnostic and therapeutic dilemma. Fink and colleagues reported the results of a National Heart, Lung, and Blood Institute and the Office of Rare Diseases workshop exploring needs and opportunities for HP research (34). The authors of this workshop summary highlighted recent advances, and continuing limitations of diagnosis in HP, suggesting the need to consider revising diagnostic criteria to include a presumptive diagnosis in the absence of a clear-cut source of exposure. The wide variety of exposures resulting in HP was briefly reviewed as were concepts in radiologic, physiologic, and histopathologic assessment. In conclusion, a series of recommendations were made to advance future research in this area, including the need for the following: (1) establishing a multicenter, collaborative network; (2) defining risk factors affecting occurrence and natural history of disease; (3) establishing reasonable and validated case definitions; (4) exploring the use of biomarkers; (5) developing and supporting population-based studies; (6) better defining the natural history of disease; (7) developing a standardized battery of antigens known to cause HP; and (8) using quantitative CT and HRCT in prospective studies of well-defined HP cases.