Amiodarone, methotrexate-, nitrofurantoin-, chemotherapy agents and radiation cause lung injury with a significant incidence rate (nitrofurantoin: 0.2 %; amiodarone 2–4 % per year). Incidence with most other drugs is low to very low, fitting the numerical definition for orphan diseases of <1/2,000 patients affected. The base of evidence for drug causality is heterogenous as many reports date back several years and few specific tests were or are available. In a few phase I or II combination chemotherapy regimens where bleomycin, CCNU, gemcitabine, or dasatinib were given concomitantly with irradiation, the incidence of DIRD could be as high as 50 % or more [16].

Drugs are a significant contributor for pulmonary diseases accounting for 3–5 % of all ILD cases [17], 10 % of ARDS [18], 11–18 % of DAH [19] or eosinophilic pneumonia and 30 % of organizing pneumonia cases. Four percent of all DIRD are fatal and about 8 % are preventable although risk evaluation and prevention are still in their infancy.

Drugs cause DIRD after variable times on the medication, from a few seconds for flash pulmonary edema [20], catastrophic bronchospasm or anaphylaxis [21], to months or years into treatment for many ILD, or even after termination of treatment for iatrogenic fibrotic conditions [22]. DIRD may develop regardless of route of administration of drugs, although the oral and parenteral routes are foremost. Drugs given via topical instillation (transdermal, ophtalmic, gingival, intra-uterine, endomyometrial, subcutaneously, or in the vertebral body, pleura, urinary bladder or intrathecally), via inhalation or aspiration may also cause substantial lung injury [2].

Clinically, DIRD may manifest with involvement of the lung, central or peripheral airways, pleural and/or pericardial surface, circulation, mediastinum, neuromuscular system, heart or hemoglobin [2]. Although drugs or drug families may cause stereotyped patterns of adverse effects (Table 34.2), drugs generally may cause more than one pattern of involvement depending on the patient, with amiodarone outdoing any other drug with over 20 possible patterns [2].

Illicit/abused drugs including alcohol (also harbored in Pneumotox) are relevant DIRD providers [2]. In the past 40 years, the death toll from opioids in the US increased nine-fold, due to ventilatory depression, falls, and aspiration [23]. Subversion of drug use, herbals, dietary supplements, incense and chemical concoctions obtained through the Internet emerge as a cause of life-threatening respiratory disease [24, 25]. The new trend of ‘Met-Labs’ may cause life-threatening injuries to the manufacturer, family, law enforcement officers, firefighters and forensic personnel [26]. Other forms of DIRD include occupational asthma in pharmaceutical industry workers, and lung injury in pets and other animals, which closely resemble those seen in humans [27–29].

To raise the possibility that a new respiratory problem is caused by the administration of a drug, a combination of drugs and/or radiation therapy can be important for the patient: further/irreversible damage can be prevented if use of the harmful drug is stopped (although the underlying illness needs be managed using other medications so that the patient is not exposed to a flare of the underlying disease), patients may be spared unnecessary invasive evaluation (e.g. a lung biopsy and/or empiric corticosteroid therapy) pending the result of a drug holiday which can be diagnostic, and unnecessary readministration of the causal drug can be avoided. A high index of suspicion for drugs or chemicals as the possible cause for any new respiratory problem is warranted regardless of age as DIRD can occur in children or in newborns, gender though a few DIRD are female- or male-specific or are unique to the pregnancy state and underlying condition. Altered sensorium and inability of the patient to communicate in the emergency setting may complexify history taking, which may then rest on relatives, family physician or pharmacist. A urine drug screen and blood levels of the suspect drug or chemical (e.g. paraquat, brodifacoum) are indicated as early as possible after admission, so as the compound has not cleared to insignificant levels [30]. Analysis of the time course of symptoms vs. exposure is indicated as both drugs and patterns differ in this respect. Literature should confirm that the observed pattern (symptoms, imaging, laboratory findings, BAL, pathology, outcome) is appropriate for the specific drug. Pneumotox was designed to expedite this search [2]. Diligent exclusion of other possible causes is required as drug-induced pneumonitis and infectious pneumonia may present similarly. The expanse of differential diagnosis varies according to drug, basic disease, clinical context, presentation, pattern of injury and whether the patient was being exposed to one or several drugs or immunosuppressive agents. Abatement or resolution of symptoms should follow drug discontinuance, preferably without adjunctive corticosteroid therapy otherwise drug dechallenge is more difficult to interpret. Hyperacute drug-induced cellular ILD, pulmonary edema, alveolar hemorrhage, ARDS and systemic reactions may not improve upon simple drug withdrawal and corticosteroid therapy is indicated in such case and can be life-saving. Absence of recurrence of symptoms within an appropriate observational period off the drug will support the drug etiology. Relapse after inadvertent or deliberate rechallenge with the drug is undisputable evidence for drug causality, but this test often is not available as rechallenge can be hazardous or cause patient’s demise. Rechallenging the patient can be considered only if the drug is vital, there is no substitute for it, safe rechallenge has been described in the literature (e.g. imatinib, dasatinib, m-TOR inhibitors), and/or a prococol for rechallenge or induction of tolerance is available. Overall, 1.4 % of all DIRD cases have been rechallenged, leading to death of 7 % of those so managed. Lung pathology is rarely available in DIRD, being described in 3.8 % of all DIRD reports. Tissue abnormalities in drug-induced lung disease may be ‘consistent with’, ‘suggestive of’ or rarely ‘diagnostic of’ the drug etiology, which limits the contribution of this test. Table 34.3 lists the pathology patterns and whether BAL can be used as a surrogate test. The lung biopsy may have an important exclusionary role, whereby other illnesses or an infection can be ruled out with greater certainty. The recently described cryobiopsy technique is being currently evaluated [31]. Diagnostic criteria are summarized on Table 34.4. Overall, only a few cases can be definitely ascribed to drugs, namely those with a distinctive pathology or positive rechallenge. In a fraction, drugs can be ruled out, and the vast majority is labled possible or probable, with all uncertainty left to such wording [32]. Classic papers by Hill [33] and the Naranjo scoring system [34] still are valuable resources. The topic of counterfeit drugs [35] and the recent issue of fake journals [36] is a further challenge in evaluation the materiality of DIRD; continued vigilance is necessary.

The orphan patient is the one presenting with DIRD in whom the diagnosis is not raised and drug dechallenge is not attempted. A list of common inducers and the corresponding clinical-imaging-pathologic patterns of involvement is given in Table 34.5. Being supported by a wealthy and persistently fueled literature, involvement from these drugs is not reviewed here in great detail although atypical respiratory presentations from some of these compounds is reviewed later in this text.

About half the 950 drugs and agents censored in Pneumotox at this time cause ILD with a low to very low incidence rate (<10 reports overall despite many years on the market), indicating circumstantial evidence for causality. Reports are heterogenous [37], often date back many years, details were not given, patients were managed empirically with drug withdrawal and corticosteroid therapy making retrospective assessment difficult. The issue of causation vs. chance association is an unresolved one at this time and may remain so in the foreseable future. These drugs appear in Pneumotox with a 0–1 frequency [2].

Mecamylamine, hexamethonium, aminorex, tryptophan, fenfluramine, dex fenfluramine and benfluorex have all been recalled. We keep them in Pneumotox though, as any drug may regain popularity in light of a new indication as is the case with thalidomide, or because related congeners may reproduce DIRD similar to those of the parent compound.

The antneoplastic antibiotic mitomycin-C is now less commonly used than it once was. Mitomycin pneumonitis developed in about 5–10 % of the patients in the form of acute lung injury and DAD followed, in some patients by irreversible pulmonary fibrosis [38]. Corticosteroids have been effective in reversing early cases [39]. Mitomycin-associated hemolytic uremic syndrome is a rare (incidence 0.015 %) dreadful dose-related complication of mitomycin, gemcitabine and other chemotherapy regimens [40–42]. The syndrome is characterized by any association of hemolytic anemia, circulating schizocytes, thrombocytopenia, reticulocytosis, hyper- or hypotension, renal failure central nervous system symptoms, renal failure, pulmonary edema, alveolar hemorrhage, ARDS and/or pulmonary hypertension. Blood transfusions can trigger the onset of or aggravate the syndrome. Pathologic appearance is with intravascular pulmonary fibrin thrombi, pulmonary edema, hemorrhage and/or DAD. Management includes plasmapheresis, renal replacement therapy and high-dose corticosteroids. Serial monitoring of renal function and urinalysis may indicate those patients with early impending toxicity [43].

The once popular antirheumatoid drugs gold and penicillamine have mostly fallen into disuse since the advent of methotrexate, leflunomide and TNF-alfa antagonists. Severe gold-induced NSIP-like disease, obliterative bronchiolitis and Goodpasture like disease from penicillamine are for now diseases of the past [2].

Amphetamine-like anorectics (aminorex in the 1960s, fenfluramine-dexfenfluramine in the 1990s, benfluorex in the 2000s,) were recalled on the basis of acquired pulmonary hypertension (PHT) or valvular heart disease. As early as 1977, Widgren et al. reported on plexiform arterial lesions in the pulmonary circulation of young ladies who had used the anorectic aminorex [44]. In the series by Gurtner [45] prior to the advent of new therapies for PHT, death was mainly from right ventricular failure after an average of 3.5 years. Patients with higher grades of pulmonary artery pressures had worse outcomes. However, the prognosis was not as severe as that of primary PHT. Pathologically, aminorex vasculopathy is similar to primary PHT and similar findings have been reported with the use of other anorexigens [46, 47]. The aminorex epidemic peaked in the 1960s and declined in the early 1970s concomitant with aminorex discontinuation. Although household manufacture of aminorex has been responsible for sporadic PHT cases [48], most of the concern with aminorex now resides in its precursor levamisole, a veterinary antihelminthic agent that is widely used since the early 2000s to cut street cocaine [49]. Levamisole is bioconverted to aminorex in mammals [50]. One PHT case has been documented following cocaine-levamisole abuse [51], and the topic is closely monitored. Accordingly, history of exposure to cocaine and other PHT inducers is indicated in any PHT case [52]. On the other hand, levamisole can produce a distinctive p-ANCA (with dual anti-PR3 and -MPO specificity) cutaneous vasculopathy mainly involving the face, earlobes and limbs, which is covered in more detail below [12].

The potent 5-HT2B receptor antagonist anorexigen fenfluramine and its dextro-isomer dexfenfluramine have also caused a PHT epidemic with a 6.3-fold increase in risk compared to nonusers [53]. Current data from the UK and Ireland Registry indicates a 1.7 % prevalence of antecedent anorexigen use among PHT cases [54]. In addition, fenfluramine caused valvular heart disease in the form of mainly left-sided valvular retraction with regurgitation or stenosis [55]. Fenfluramine and its dextro-isomer were recalled for that reason in the late 1990s. Benfluorex was promoted as an antidiabetic drug while in fact this is a fenfluramine prodrug. Vast numbers of valvular heart disease [56] and a few PHT cases developed in France following longterm indiscriminate exposure to the drug [57]. The drug was recalled in 2009.

Illicit amphetamines remain of concern as regards ‘primary’ PHT. An undeniably strong association between stimulant use (amphetamine, methamphetamine, cocaine) and ‘idiopathic’ PHT cases has been noted in western USA [58]. Chin et al. found a history of stimulant use in 28.9 % of their ‘idiopathic’ pulmonary hypertension cases, ten-fold that in nonidiopathic PHT [58].

Several members of the thiazolinedione (glitazones) family were discontinued in some countries owing to their propensity to cause capillary leak, pleural effusions and at times irreversible heart failure [2].

L-tryptophan once was a popular neutraceutical. The compound has produced an epidemic of ‘eosinophilia-myalgia syndrome’ [59]. Blood eosinophils were increased, about half of affected patients presented with small irregular or dense radiographic opacities and pleural effusion. An ARDS picture was noted in a few [2]. Minute amounts of contaminants (denoted peak and a suffix) formed during the bioengineered-driven synthesis of tryptophan at the Showa Denko plant were blamed as the cause for the syndrome. Incident cases diminished sharply after l-tryptophan was recalled. Only rare sporadic cases are now being reported [60].

Pulmonologists, anesthesiologists, emergency physicians, intensive care and ENT specialists, dental and other surgeons can be confronted with acute life-threatening drug-induced respiratory and/or systemic emergencies. These reactions can affect (by decreasing order of frequency) the lung, upper or lower airways, pulmonary circulation, pleura or they are systemic in nature with pulmonary involvement in association. Clinical presentations are in the form of diffuse white-out of the lung corresponding to several forms or acute lung injury or airway emergencies. The causal relationship with administration of a drug is deduced from the brief time-to-onset of the reaction upon exposure, which can be of seconds. Some reactions develop electively during the intra- or perioperative period, causing difficult diagnostic and management issues. Recognition of the drug etiology is paramount and can be life-saving.