This is the largest histologic analysis of smoking-related parenchymal lung abnormalities since the seminal study of respiratory bronchiolitis by Fraig et al in 2002.10 Compared to that study, this series analyzes a larger number of cases (600 vs. 156), includes not just respiratory bronchiolitis but also emphysema and SRIF, and features a study design in which the histologic analysis was performed by a pathologist blinded to smoking status. Our main finding is that the histologic finding of SRIF is 100% specific for smoking, i.e., it occurs exclusively in current or ex-smokers. This finding validates SRIF as a robust and reliable histologic marker of smoking status. Respiratory bronchiolitis alone, emphysema alone, and the combination of RB and emphysema were also good histologic markers of smoking, but were not as specific as SRIF.
This study was designed with a focus on specificity of histologic findings as markers of smoking status. It is already well known that the sensitivity of these findings is less than perfect since all smokers may not necessarily develop these findings in their lungs, and because some of these findings (such as respiratory bronchiolitis) can regress after cessation of cigarette smoking.10 Why some smokers develop pathologic abnormalities while others do not is not well understood and should be a focus of future inquiry.
In the published literature, all cases of SRIF reported thus far have been from smokers (Table 4). However, since none of these studies have analyzed the histologic findings with the pathologist blinded to smoking status, they do not address the question of specificity of histologic findings for smoking. In the largest series of histologically confirmed SRIF to date, all 65 patients with SRIF were smokers.3 All cases had respiratory bronchiolitis and 61/65 had emphysema. However, these cases were identified by histologic review of resection specimens exclusively from smokers. Since never-smokers were excluded, it is impossible to determine from such a study design whether SRIF could potentially be encountered and diagnosed histologically in patients who have never smoked.
Table 4
Published literature on SRIF and relation to smoking status
First author (Year) | Number of cases of SRIF | Smokers (current or ex) | Never-smokers | Selection criteria |
Katzenstein (2010) | 9 | 9 | 0 | Cases identified by histologic review of resection specimens. Pathologists were not blinded to smoking status. |
Primiani (2014) | 65 | 65 | 0 | Cases identified by histologic review of resection specimens from smokers. Never-smokers were excluded |
Fabre (2017) | 8 | 8 | 0 | Cases were identified by histologic review of resection specimens from smokers. |
SRIF: Smoking-related interstitial fibrosis |
Assessment of histologic findings attributable to smoking is not always a straightforward exercise. Pathologists need to be able to separate smoking-related pathologic findings from their mimics while also dealing with minimal alterations, gray zones, and artefacts. For example, distinguishing respiratory bronchiolitis from hemosiderin-laden macrophages (or a combination of the two pigments) can be challenging, and the significance of very faint pigmentation can be difficult to judge. For example, in Fig. 4, parts C and D were interpreted as respiratory bronchiolitis in this study, but the patient was a never-smoker. In hindsight, on review of these images, macrophages are few in number and pigmentation is subtle. It is possible that respiratory bronchiolitis was over-called in this case, as well as in other cases of respiratory bronchiolitis in which the patients turned out to be never-smokers. Similarly, while severe emphysema is easy to identify on histologic grounds, mild emphysema overlaps with overinflated lung, possibly accounting for cases of emphysema in never-smokers in this study (Fig. 5). Given these problems, the identification of yet another distinctive histologic marker of cigarette smoking (SRIF) is a welcome addition to the armamentarium of pathologists for the identification of smoking-related parenchymal abnormalities.
A limitation of this study - and that of any study that evaluates the accuracy of histology as a marker of smoking status using patients’ self-reported smoking status as a gold standard - is that patients may not always be truthful about their smoking habits. This issue was alluded to by Fraig et al10 and is a potential confounder in studies of this type. One might justifiably ask: are the cases of respiratory bronchiolitis, emphysema or both in never-smokers in this study overcalls, or were some of these patients withholding information about their true smoking status? Having acknowledged this caveat, unless a better, more objective gold standard emerges, the best we can do is to measure pathology against self-reported patient history. One possible way to overcome the limitations of self-reported patient history would be evaluate histologic parameters against more objective parameters such as serum cotinine levels.
Our understanding of the link between smoking and non-neoplastic lung abnormalities has progressed substantially over the past 50 years, but opportunities to improve remain. Although there is agreement among pulmonary pathologists that smoking causes emphysema, accumulation of pigmented macrophages in the lung and interstitial fibrosis, terminology of these findings is confusing, and often uses outdated or inaccurate nomenclature coined in the 1960s and 1970s.12–15 Our study does not attempt to address these complex issues. However, the fact that SRIF is highly specific for smoking does suggest that the prefix “smoking-related” is appropriate for this distinctive histologic finding.
In summary, this study demonstrates that SRIF, when diagnosed solely on the basis of histologic criteria by a pathologist blinded to smoking status, is 100% specific for smoking (current or ex), exceeding the specificity of respiratory bronchiolitis and emphysema, either alone or in combination.