Drugs without benefits? Confronting the challenges of drug-induced interstitial lung disease

Drug-induced interstitial lung disease (DILD) with an estimated incidence of approximately 4.1–12.4 cases/million/year is implicated in ~5% of ILD cases.1 Data vary by country, with DILD more commonly diagnosed in Japan, perhaps due to higher reporting.2 The Common Terminology Criteria for Adverse Events (CTCAE) scale3 (table 1) helps quantify severity and treatment includes cessation of the drug and, in more severe cases, corticosteroid use. At least 350 drugs have been implicated in causing lung toxicity, or pneumonitis, across a spectrum from mild radiological infiltrates to life-threatening respiratory failure.4 This heterogeneity of presentation and lack of diagnostic standards, with rechallenge to confirm toxicity rarely justified, makes DILD difficult to identify, even at individual patient level.4 The result is a heavy reliance on databases such as pneumotox5 that collate evidence from the literature, often case reports or small series, but with no large-scale assessment. View this table: Table 1 Common Terminology Criteria for Adverse Events (CTCAE), V.5.0 Jo et al 6 take a different approach using a large, nationally representative dataset of hospitalised patients in Japan to retrospectively identify patients that had developed DILD, severe enough to warrant corticosteroid therapy, after receiving a ‘risk drug’ from one of 42 categories associated with lung toxicity. We applaud the methodology used in this study which identified 2342 cases of DILD out of ~42 million hospital admissions (0.0056%). For each case the authors selected four controls matched for known DILD risk factors (primary diagnosis, …