Epidemiology

The incidence of pulmonary toxicity due to most drugs is low-to-very low. Given that DIRDs generally have a prevalence of under 1/2000 (and many much lower), a large number of these conditions qualify as orphan diseases. Among the 1650 current drugs and agents known to cause respira-

tory injury, 1011 do so rarely with fewer than ten reports in the world literature [3, 4]. The rigor of published reports varies widely [25] and some date back many years. Guidelines and good publication practice for classifying and characterizing adverse drug reactions are not always followed. In addition, patients may have received corticosteroid or other disease-modifying therapy, making it dif cult to evaluate the

effect of drug withdrawal and to assign causality; an assessment which (unfortunately) is often inadequately performed [5, 6, 26].

Relatively “common” pulmonary toxins such as nitrofurantoin or amiodarone have a DIRD incidence rate of 0%, 2%, and 1.9%, respectively. Incidence is possibly higher in Japanese compared to caucasians patients. Chemotherapeutic agents with relatively well characterized incidence rates of DIRD including chemo agents: 1–3%, crizotinib: 1.8%, brigatinib: 7%, ceritinib: 1.1%, alectinib 2.6%, and lorlatinib: 1.8%. Even higher incidence rates (3.5–20%) can be observed with the use the novel immune checkpoint inhibitors (ICI) [27], or when more than one ICI is given concomitantly. DIRD due to tyrosine kinase inhibitors is also relatively common (1–5% depending on severity grade). These estimates should be viewed with caution, as both the de nition of DILD and diagnostic criteria are variable between studies and countries. Reported DIRD incidences as high as 50% in early phase 1 or 2 combination chemotherapy trials with bleomycin, CCNU, gemcitabine, or dasatinib and irradiation may have hampered the development of these agents if methods of assessment had not advanced. Recently, histopathology descriptors have been used to name roentgenographic patterns of involvement [28–30]. The accuracy, reliability, and reproducibility of these methods are unknown. Studies blindly comparing imaging and pathology have shown suboptimal agreement [31, 32]. The Fleischner Society glossary of terms remains an important reference which can be used to describe and classify imaging characteristics [33], and as was employed to construct the imaging section in Pneumotox [3, 4]. Estimates of the fraction of ILD due to drugs in ILD populations ranged between 3% and 5% [34], 9.5% in ARDS [35], 11–18% in alveolar hemorrhage [36], 12–17% in eosinophilic pneumonia [37], 3% for organizing pneumonia due to amiodarone [38], or drugs 28% [39], and 1–3% for breast radiation-induced organizing pneumonia or cryptogenic organizing pneumonia (COP) [40].

Overall, it is estimated that 8% of DIRD are preventable if risk factors, adjusted drug dosage and guidelines are implemented and followed [41, 42]. Approximately 4% of all DIRDs are lethal, partly because patients are deliberately or inadvertently rechallenged with the culprit drug [43]. Atopy and asthma are considered risk factors for allergic reactions and anaphylaxis. Pre-existing ILD may predispose to the risk of developing drugor radiation-induced pneumonitis [44, 45].

Timing, Chronology, Delay Time

About a fth of all DIRDs exhibit a rapid or precipitous onset immediately after the rst administration of the drug or

agent. Such reactions include drug-induced fash pulmonary edema [46], catastrophic bronchospasm [47], hypersensitivity, anaphylaxis [48], acute ventilatory depression and respiratory failure [49, 50], coronary vasospasm [51], suffocation:asphyxia from inhaled substances, gases or chemicals [52, 53]. However, the majority of drug-induced or iatrogenic pulmonary, pleural, vascular, mediastinal, cardiovascular or lymphatic reactions develop more slowly months or years into treatment. Rarely, the adverse reaction develops after termination of treatment with the drug, typically­ a chemo agent such as nitrosoureas or amiodarone, or chest radiation therapy.

Route of Administration

Drugs can cause respiratory disease regardless of the route of administration, although the oral and parenteral routes are more commonly implicated. Drugs that are inhaled, given topically in the form of application or instillations into the pleural space, epidural or intrathecal space, vertebral body, eyes, urinary bladder, female genital tract, or delivered subcutaneously or transdermally, can all cause respiratory injury.

Patterns of Involvement [3, 4]

DIRD may manifest in the form of parenchymal lung disease, pulmonary in ltrates, pulmonary hemorrhage, storage lung diseases, or involvement of the central, large or distal airways, pleural and/or pericardial surface, heart, pulmonary circulation, coronary arteries, mediastinum, respiratory muscles, central nervous system, nerve and nerve endings, or hemoglobin [3, 4]. Drugs or families [e.g., beta-blockers, nonsteroidal anti-infammatory drugs (NSAIDs)] or salicylate may cause a stereotypical pattern of injury [e.g., acute bronchospasm, angioedema, pulmonary edema, or ILD] [3, 4] (Table 42.2). In general, drugs produce more than one pattern of respiratory involvement, as exempli ed by amiodarone and immune checkpoint inhibitors (ICI) which outnumber almost any other drugs in terms of patterns of injury (with 83 and 90 patterns, respectively). It is equally important to be vigilant for adverse effects (sometimes severe) of drugs in extrapulmonary organs or organ systems, especially during treatments with amiodarone, TKI or ICI.

Illicit/abused drugs and agents including ethanol can cause severe respiratory issues and are therefore also harbored in Pneumotox. In the past 40 years, the death toll from opioids due either to ventilatory depression, falls or aspiration increased several-fold in the US [54], now as high as 3.3% of the exposed population. Subversion of drug use, cinnamon, pepper, herbal therapy, dietary supplements, energy drinks, “incense” and concoctions purchased via the Internet

[55] emerged as a novel cause of life-threatening injury, burns, fainting, choking, bronchospasm, or acute respiratory failure or ARDS [56]. Room air in “Meth-Labs” can harbor toxic chemicals capable of causing life-threatening or fatal injury to the owner, manufacturer, family, kids, law enforcement of cers, coroner, re ghters, forensic personnel and/or healthcare workers [57, 58]. Other presentations of DIRD include drug-induced occupational asthma in the pharmaceutical industry [59], and lung injury in pets or other animals that can be similar to those in humans [3, 4].

Diagnosing Drug-Induced Respiratory

Disease (DIRD)

Early consideration that a drug or a combination of drugs, abused substances, electronic cigarettes (-e-cigarettes), electronic nicotine delivery systems (ENDS), chemicals, household chemicals, or radiation therapy may be the cause for a respiratory problem is important for optimal patient care, because: (1) Further damage can be avoided by withdrawing the drug or exposure early, (2) Invasive procedures such as lung biopsy and/or empiric corticosteroid therapy can be avoided pending the outcome of drug holiday or exposure withdrawal (which can be diagnostic if followed by resolution of all presenting signs and symptoms), (3) Inadvertent re-administration of the causal drug can be fatal. A high index of suspicion is needed at all times in adults, children, and in the newborn, where vertical transmission of drug-­ induced respiratory adverse effects from the mother to the infant is possible. Classic papers on causality assessment and scoring remain valuable resources [6].

A few drugs are reserved for treatments in females or males, and DIRD occur preferentially or exclusively in one gender (e.g., nilutamide is a therapy used to treat prostate carcinoma, and i.v. tocolytics can produce pulmonary edema in women). Altered sensorium and inability to communicate with the patient in the emergency setting may complicate history taking at the bedside, leading to reliance on information from relatives, friends, family physician or pharmacists. Urine drug screens and measurements of blood levels of drugs (e.g., salicylate), illicit drugs, chemicals (e.g., paraquat, brodifacoum, organophosphates) in blood are helpful but not all results are immediately available. It is important to obtain samples early in the course of events, before the drug or metabolites are cleared from the blood or urine. Drugs differ widely in terms of time to onset, symptoms, clinical radiographic, laboratory, bronchoalveolar lavage, and pathological pattern and outcome (see Pneumotox and [3, 4]). Determination of latency period and time course of symptoms vs. exposure and matching the suspected drug with the observed pattern of involvement is key to establishing the correct diagnosis. Diligent consideration of other

plausible causes and etiologies of the presenting illness is required, as drug-induced pneumonitis is a diagnosis of exclusion and infectious lung diseases and pulmonary involvement from systemic diseases may present similarly. The expansive differential diagnosis depends on drug, underlying background, comorbidities, clinical and epidemiologic context, presentation, pattern of involvement and whether the patient was being exposed to one or more than one drug, immunosuppressive agents, dietary supplements and/or drug of abuse. Clinical improvement or resolution of symptoms following drug discontinuance favors a drug etiology. However, acute severe drug-induced lung disease, including pulmonary edema, alveolar hemorrhage, ARDS or systemic reactions may not improve in an immediate and linear fashion upon simple drug discontinuance. In such situations, corticosteroid therapy or immunosuppressants may be indicated. Lack of recurrence within an appropriate observational period off the drug also supports the drug etiology, while relapse after rechallenge with the drug, be it inadvertent or deliberate, is indisputable evidence for causality of the drug or its excipients. Of course, rechallenging patients with the putative offending drug can be hazardous and sometimes fatal, so appropriate consideration for that approach is imperative. Criteria for rechallenging patients include: (1) the drug is vital, (2) there is no adequate substitute, (3) rechallenge is not always followed by relapse (based on literature), (4) a protocol for induction of tolerance with or without corticosteroid therapy is available, (5) rechallenge is validated by a multidisciplinary discussion,

(6) rechallenge can be conducted in a safe environment (i.e., in a setting with adequate resuscitative equipment), (7) detailed informed consent is obtained. Overall, 1.4% of all DIRD cases are rechallenged, leading to death in as many as 7% of those so tested (extracted from references in [3, 4]). Immediately negative rechallenge does not exclude the diagnosis of DIRD.

A lung biopsy is obtained in approximately 6.2% of all DILD cases reported in the literature. Except in rare circumstances (e.g., with amiodarone pulmonary toxicity or when distinctive foreign drug material is present in lung tissue), changes are at most “compatible” or “consistent” with, and less often, “suggestive” of the drug etiology [60, 61]. This limits the potential contribution of lung tissue sampling, inasmuch as complications may occur following lung biopsy be it open-, transbronchial-, or cryo-. Still, lung biopsy may prove useful to rule out a non-iatrogenic etiology. Table 42.5 lists the pathology patterns of iatrogenic lung disease and whether the bronchoalveolar lavage can be used as a surrogate test [62]. Diagnostic criteria are summarized in Table 42.3 and on the Pneumotox frontpage. Though in a number of cases respiratory reactions were de nitely ascribed to a drug, more often the drug etiology is represented as likely, probable or possible, with all the uncertainty

that such wording may carry, and nuances that are dependent on who is formulating the opinion and the language of reporting person [63, 64]. In the transportation accident reporting world, virtually certain or almost certain corresponds to >99% probability of occurrence, very or highly likely to >90%, likely or probable to >66%, about as likely as not or more or less likely to 33–66%, unlikely or improbable to <33%, very or highly unlikely to <10% and exceptionally unlikely to <1% probability [65].

Uncommon and Rare Drug-Induced/

Iatrogenic Respiratory Disease

It is beyond the reasonable scope of this chapter to cover all drugs known to cause respiratory problems. A comprehensive list is available since 1997 in Pneumotox along with references, the reader is referred to [3, 4]. We shall concentrate here on severe, rare, poorly known, problematic and preventable DIRDs.