Differential Diagnosis of IPF

Steven D. Nathan, MD, FCCP: IPF [idiopathic pulmonary fibrosis] can be misdiagnosed for many reasons. It’s not very common when 1 compares it with more common causes of shortness of breath such as COPD [chronic obstructive pulmonary disease] and congestive heart failure. Even if the diagnosis of interstitial lung disease is made, then we have 200 different causes that we have to work our way through before we obtain a diagnosis of idiopathic pulmonary fibrosis.

In many ways, IPF is a diagnosis of exclusion. You have to exclude significant exposures—asbestosis, silicosis, chronic hypersensitivity pneumonitis. You have to exclude underlying connective tissue diseases such as rheumatoid arthritis, Sjögren syndrome, and scleroderma as the cause of the interstitial lung disease. You have to exclude drug exposures that they might have received in the past. Infection is usually not an issue to exclude. You know if someone has an infection as a cause of their interstitial lung disease.

Even within the basket of idiopathics, there are many idiopathic ILDs [interstitial lung disease] that need to be sorted out from IPF. If you look at the idiopathic interstitial pneumonias, of which IPF is a prototypical illness, there are some other “cousins” to IPF, such as NSIP, or nonspecific interstitial pneumonia. Then you have DIP, desquamative interstitial pneumonia; RBILD [respiratory bronchiolitis-associated interstitial lung disease]; and COP, or cryptogenic organizing pneumonia, formerly referred to as BOOP or bronchiolitis obliterans organizing pneumonia. Then there’s a group of patients who end up having unclassifiable ILD, which means that despite our best efforts, we don’t have enough to call them IPF, and we don’t have enough to call them anything else. In terms of other idiopathic causes, there are entities such as sarcoidosis, lymphangioleiomyomatosis, and pulmonary Langerhans cell histiocytosis that don’t fall under the classification of IIP [idiopathic interstitial pneumonia], but there are other causes of idiopathic causes of interstitial lung disease.

The central diagnostic modality for IPF and all ILDs is the high-resolution CT [HRCT] scan. That determines which direction 1 is going to go. It’s very unusual for the HRCT by itself to be totally diagnostic. Sometimes you can have changes that are consistent with another form of interstitial lung disease, like sarcoidosis or lymphangioleiomyomatosis. But when we’re dealing with IPF or suspicion of IPF, the way we categorize the HRCT is into 1 of 4 categories in terms of its morphologic appearance. We talk about a UIP [usual interstitial pneumonia] pattern, a probable UIP pattern, and an indeterminate UIP pattern, and then an alternative diagnosis is suggested.

Within the description of a UIP pattern, patients have basilar predominant subpleural reticulation with honeycombing, with or without traction bronchiectasis or traction bronchiolectasis. If you have that and no other atypical features, that’s a UIP pattern. In the appropriate clinical context, that might be enough, together with the patient’s clinical presentation, to make a diagnosis of IPF. Probably UIP pattern is very similar to UIP pattern except there’s no honeycombing. So subpleural reticulation, traction bronchiectasis, traction bronchiectasis, and no atypical features are a probably UIP pattern. In the appropriate clinical context, that also might be enough to make a diagnosis of IPF.

When I talk about an appropriate clinical context—if you have an older man, there’s a slight male predominance vs female predominance. For example, a 75-year-old man, former smoker—smoking is a risk factor—with no other known exposures and no evidence of a connective tissue disease, the likelihood is very high that that patient is going to have IPF. A probable UIP pattern, together with that clinical presentation, might be enough to attain a diagnosis of IPF.

The indeterminate bucket is becoming somewhat of a wastebasket term. If there’s any uncertainty about what it is, or some but not enough atypical features, a lot of HRTCs get lumped into indeterminate. It becomes a little more difficult to make a diagnosis of IPF. But let’s say you have the same gentleman, a 75-year-old smoker with no exposures, no CTD, and the indeterminate HRCT. There’s still a good chance that there could be IPF, and you might want to make a provisional diagnosis with a high level or low level of confidence. What’s the indeterminate pattern? There’s no honeycombing, no traction, maybe subpleural reticulation, and some other plus-or-minus features. So it’s an unsatisfactory wastebasket term, but it’s useful in terms of categorizing many of our CTs.

An alternative diagnosis is suggested when it’s clearly something else going on that’s not IPF: upper lobe distribution, more central infiltrates, more alveolar infiltrates, cysts, mosaicism or air trapping. Those will all fall under alternative diagnosis, such as chronic hypersensitivity pneumonitis or other disorders. But it’s interesting because there have been some series where patients have had alternative diagnoses, suggested HRCTs, and got a lung biopsy, and it’s a UIP histopathologic pattern. What determines a diagnosis is the clinical, the radiographic, and the pathologic that steers us in the right direction for an accurate diagnosis.

Fernando J. Martinez, MD, MS: The role of surgical lung biopsy in the diagnosis of interstitial lung diseases, and idiopathic pulmonary fibrosis, in general is an evolving area. It has changed significantly in the time I’ve been in this field, for the last 30 years, clearly in the last 5 to 10 years. That reflects the role of imaging and the expertise that’s been developed in the clinical community, particularly among pulmonary physicians and thoracic radiologists with regard to the imaging pattern.

Imaging has made a huge difference. We’ve also become much more adept at understanding the clinical features that drive us in 1 diagnostic category or another, and the ability of merging the clinical information with what the thoracic radiologists are providing as the pattern has made it so that we do surgical lung biopsies much less frequently. The availability of antifibrotic therapy for non-IPF progressive fibrosing disorders has also limited the need for surgical biopsy.

All of us, when we’re in these MDDs and think about doing a biopsy, it’s based on if the information you’re going to get is going to change what you’re going to do and recommend that that patient do. If the answer is yes, the features are atypical. And if we make this diagnosis, we’re going to do this vs this. Then it’s a worthwhile approach. That simple binary categorization is no longer as simple as before. We do much fewer surgical biopsies because we’ve developed so much more expertise in the imaging merging with the clinical information and the availability of antifibrotic agents. Many fewer surgical biopsies these days than even 5 years ago.

Transcript Edited for Clarity