serum salicylate level of 30 mg/dL or greater is the threshold above which pulmonary and systemic toxicity is more likely to develop. Blood levels of drug(s) should be measured in the acidotic patient, since dialysis can be life-saving in severe salicylate poisoning. Values above 100 mg/dL are associated with poor outcomes. Management of salicylate NCPE includes serum and urine alkalinization, with hemodialysis being reserved for cases with pulmonary edema, distant organ damage and blood levels >100 mg/dL [81].

Drug-Induced Cardiogenic Pulmonary Edema

Drug-induced and iatrogenic cardiac and/or overload pulmonary edema can complicate chemotherapy [82], treatments with drugs, and/or overzealous administration of fuids [3, 4, 8]. Cardiogenic pulmonary edema may occur early or late [83] as a complication of drugs or radiation therapy causing acute, chronic or delayed leftor bi-ventricular dysfunction. Drugs and abused substances can also cause myocarditis and consequent heart failure [84]. Causal drugs include epinephrine/adrenaline, anthracyclines, chemo agents, fuoropyrimidines, immune checkpoint inhibitors, sunitinib, and drugs of abuse. Acute left ventricular dysfunction may also result from drug-induced coronary vasospasm or myocardial infarction, as seen with amphetamine, cocaine, fuoropyrimidines, or oxaliplatin-based chemo regimens [3, 4, 8]. Cardiac biomarkers can aid in establishing the diagnosis. Drug therapy withdrawal is necessary, underlying disease permitting. Appropriate vigilance is important, because cardiotoxicity impacts the lung in multiple ways [8]. Cardio-oncology is a rapidly emerging eld [85].

The “Chemotherapy Lung”

This complication may occur during or following treatments with many antineoplastic drugs of multiagent chemotherapy regimens, with an average incidence of 1–5% [3, 4]. High drug dosages, rapid as opposed to slow infusion of intravenous drugs, coadministration of bleomycin, gemcitabine, inhaled oxygen, rituximab, radiation therapy, or CSF can be triggering or potentiating. Causal drugs include antibiotics (bleomycin, mitomycin C), alkylating agents (busulfan, chlorambucil, cyclophosphamide, melphalan), antimetabolites (azathioprine, cytosine arabinoside, gemcitabine, fudarabine, 6-mercaptopurine, methotrexate), etoposide, nitrosoureas, oxaliplatin, and oxaliplatin-based regimens, and taxanes. Recent additions to the list include tyrosine kinase inhibitors (TKI) (erlotinib, ge tinib), cetuximab, irinotecan, and pemetrexed. The condition manifests with dyspnea, cough, hypoxemia, diffuse haze or ground-glass that may progress to dense bilateral opacities and volume loss. HRCT discloses interand/or intralobular septal thickening, ground-glass attenuation and in some patients, moderate unilateral or bilateral pleural effusion are also present [86]. Early pulmonary involvement in patients on bleomycin or

chemo agents [87] should be monitored as patients may develop further restrictive lung function, deterioration of diffusing capacity for carbon monoxide, denser pulmonary opacities or full-blown ARDS if the drug is continued. Caution is advised in the asymptomatic patient on bleomycin when the diffusing capacity for CO drops by more than 40% from baseline on serial measurements. Unfortunately, despite awareness of the risks, bleomycin pulmonary toxicity continues to escape early detection of and fatal bleomycin lung still occurs [88]. The potential merit of systematic follow-up of pulmonary physiology in those exposed to other chemo agent remains unclear at the present time. BAL is generally performed in symptomatic patients to exclude Pneumocystis or viral pathogens including COVID19, which drug-induced lung injury may resemble. An increase in BAL neutrophils and hemosiderin-laden alveolar macrophages has been found in BAL. Bizarre type II pneumocytes refecting alkylating agent-induced cellular atypia may be identi ed, particularly in patients exposed to busulfan. Lung biopsy is infrequently performed, because of the risks entailed and its modest diagnostic contributions, as drug-induced changes are largely nonspeci c [60]. Pathology may disclose interstitial edema, alveolar brin, hyaline membranes, resolving or organizing alveolar damage, atypical alveolar lining cells, and/or brosis [60, 61, 71]. The histopathological changes of DAD which are typical in ARDS, can also be observed in the context of an infection, hematopoietic stem cell or solid organ transplantation, or concomitant with an exacerbation of preexisting idiopathic pulmonary brosis of unknown cause. No inciting factor is found in about 20% of DAD cases. For practical reasons, drug-induced ARDS is considered plausible if the workup for an infection or other etiologies is negative and there is a compatible drug history. Antibiotics, corticosteroids, cyclophosphamide, imatinib, or IVIG have been tested in an attempt to improve the chemotherapy lung. Results of these rescue treatments are unpredictable and can be detrimental. About 40% of early chemotherapy lung cases will respond to corticosteroid therapy. More advanced cases may evolve to recalcitrant and progressive ARDS, or transition to irreversible pulmonary brosis. Mortality of drug-­ induced chemotherapy lung can be as high as 45%.

Drug-Induced/Iatrogenic Alveolar Hemorrhage

Drugs

Diffuse alveolar hemorrhage (AH/DAH) occurs when blood enters the alveolar spaces through capillaries rendered permeable by infammation or injury. Bleeding and alveolar lling in the deep lung causes shortness of breath, hypoxemia, anemia, and diffuse haze or ground-glass, mottled opacities, or consolidation. Even though DAH may be suspected on the basis of batwing or diffuse fuffy opacities on imaging, the diagnosis may prove elusive until BAL is performed showing progressively bloodier return on sequential fuid aliquots.

Hemoptysis and an increase in the diffusing capacity for carbon monoxide are inconstant features. A signi cant, abrupt drop in hemoglobin is indicative of severe blood loss. Additional severity may stem from clotting in the distal lung and/or major airways, which may lead to airway obstruction and worsening hypoxia. Prolonged retention of anticoagulants (e.g., the rodenticide brodifacoum) can cause DAH for extended periods of time. Drug-induced alveolar hemorrhage can be isolated (denoted bland DI-DAH), or it can occur in conjunction with such extrapulmonary features as skin necrosis, microscopic hematuria, renal failure, heart, and/or other organ involvement and/or positive serologies for anti-neutrophil cytoplasmic antibodies (ANCAs) of various speci city, or anti-nuclear antibodies (ANAs) [89]. This may have mechanistic relevance. DI-DAH may mimic the clinical manifestations and the laboratory features of systemic ANCA-related vasculitis, lupus, or Goodpasture’s syndrome. Even if the lung biopsy reveals pulmonary capillarities in addition to DAH, whether this nding informs the diagnosis and management of DI-DAH as compared to a more conservative approach is unclear.

History taking in patients with DAH should include exposure to hydrocarbons, crack cocaine, marijuana or tobacco smoke or fumes of snorted crack cocaine or heroin, pesticides, anticoagulants, the rodenticide brodifacoum, paraquat, hyaluronate, and fuid silicone [3, 4]. Early urine drug screen for abused drugs (recognizing that fuid resuscitation can dilute the sample and infuence results), and evaluation of brodifacoum or paraquat in plasma are indicated. Therapy with 152 different drugs can cause AH (up from 103 listed in 2017) [3, 4]. These include compounds which interfere with the coagulation cascade or with platelets such as glycoprotein IIB/IIIA inhibitors of the chemical (clopidogrel, epti - batide, ticlopidine, tiro ban) or biologic type (abciximab), vitamin-K antagonists, heparin, thrombolytic agents, direct (new) oral anticoagulants and superwarfarins, amiodarone, antithyroid drugs (carbimazole, methimazole, propylthiouracil), m-TOR inhibitors (everolimus, sirolimus), cocaine, the adulterant levamisole, all-transretinoic acid (ATRA), arsenic trioxide (As2O3), dextran-70, and penicillamine. Seven drugs have been associated with ANCA-related DAH and ten with Goodpasture-like syndrome [3, 4]. Agranulocytosis and an infection may accompany AH cases resulting from exposure to propylthiouracil or cocaine-levamisole. DAH can occur following percutaneous coronary intervention and is easily mistaken for pulmonary edema unless the BAL is performed. Inhalation of e-cigarette vapor and cryoballoon ablation for atrial brillation have also been associated with the development of AH. Mortality in DI-AH can be as high as 30–50%.

Superwarfarin Rodenticides

4-hydroxycoumarin (a.k.a. brodifacoum) is a rodenticide capable of causing devastating hemorrhage in adults, chil-

dren, and animals including pets and prey birds. Brodifacoum is a vasculotoxic superwarfarin which blocks the actions of vitamin K1. Rodents fed pellets containing the compound die from internal bleeding. Accidental brodifacoum poisoning has been described in nontarget populations (companion animals, humans, and birds) and the clinical characteristics are similar across all species. Brodifacoum poisoning in humans occurs by a number of mechanisms including accidents in factory workers or in children, deliberate ingestions (suicidal, Munchausen syndrome), ingestions by proxy or with criminal intent, by inhalation of laced crack cocaine, marijuana or cannabinoids, inadvertent exposures, and intake that is “impossible-to-track.” The diagnosis of brodifacoum poisoning should be raised in any severe unexplained bleeding (AH, airway, internal bleeding) with persistent and profoundly altered coagulation studies. Pink excreta have been described [90]. Brodifacoum can be detected qualitatively and quantitively in plasma and followed serially. Brodifacoum’s extended biological half-life (1–2 months) accounts for persistent coagulopathy, which requires prolonged (up to 1 year) vitamin-K replacement therapy. Clinical presentation of brodifacoum poisoning can be with epistaxis, hematemesis, AH or neurologic symptoms depending on the predominant site of bleeding. A pink coloration of body fuids from the dye contained in pellets have been reported. Laboratory studies (to be performed before administering Vit-K) reveal prolonged prothrombin and activated thromboplastin times with diminished activity of the vitamin K-dependent coagulation factors II, VII, IX, and X. Patients may improve initially with the administration of fresh frozen plasma (FFP) and vitamin-K [91]. Relapse of bleeding and AH can occur due to the very slow elimination kinetics of brodifacoum requiring prolongation of vitamin K replacement therapy until coagulation returns to normal. A series described three cases intoxicated with inhaled “spice” or “K2” cannabinoid mixed with superwarfarins. All three developed coagulopathy and bleeding. Synthetic opioids and marijuana appear to be now laced with brodifacoum [3, 4]. Inhalation of brodifacoum can cause particularly severe bleeding due to bypass of the enterohepatic circulation andrst pass liver metabolism. Although the majority of brodifacoum poisoning in humans are non-fatal, forensic pathologists, coroners, and veterinarians may be confronted with brodifacoum-related deaths from AH and internal hemorrhage and should be familiar with the clinical presentation.

Fluid silicone injections into the gluteal region or in the buttocks by cosmetic surgeons or by unquali ed illicit “lay” operators during clandestine surgery sessions for the purpose of cosmetic mammaplasty or body augmentation, sometimes in the context of transsexualism, can cause very severe pulmonary and neurologic complications [92]. The “silicone embolism syndrome” (SES) occurs when a fraction of administered silicone gains access to the pulmonary circula-

tion, resulting in acute lung injury, ARDS, and/or AH. Some dermal llers (hyaluronate, polyalkylimide) have also been implicated. The SES shares several clinical and imaging features with the fat embolism syndrome, including petechiae. Right to left shunting enables access of fuid silicone to the systemic circulation, causing brain damage, life-threatening neurological impairment, and/or distant organ failure. The SES may develop within minutes-to-a few hours of the cosmetic procedure. Shorter delay times portend greater severity. Clinical presentation may include any combination of fever, dyspnea, nonproductive cough, chest pain, hypoxemia, hemoptysis, petechiae, and obtundation. Imaging studies may disclose bibasilar or diffuse in ltrates on chest radiography and subpleural basilar or diffuse areas of alveolar shadowing or consolidation on NRCT. Silicone injection in the breast and mammaplasty produce distinctive soft tissue changes on imaging. BAL in SES may reveal silicone vacuoles in macrophages or multinucleated giant cells in the form of large, pleomorphic, cytoplasmic vacuoles, and inclusions on a background of abundant neutrophils and red cells. Pulmonary pathology discloses nonstainable (except with prolonged oil red-0 staining for 72 h) interstitial vacuoles with a peripheral, refractile meniscus of silicone on darkeld microscopy [93]. Silicone droplets may conform to the shape of pulmonary capillaries. Energy dispersive X-ray analysis can con rm the chemical composition of silicone. Outcome of the SES depends on the volume of fuid silicone injected and whether silicone escapes the pulmonary circulation. Early onset of symptoms and neurologic presentations are associated with very high mortality [92].

Inhalation or rarely injection of straight cocaine may be followed in a few hours by an acute episode of DAH. Cocaine may account for up to 12% of DAH cases admitted to the hospital. The true prevalence is likely even higher, as patients may be reluctant to give an accurate drug history. AH is common at autopsy in drug addicts, present in 58% of cases in one study. AH caused by cocaine has a wide spectrum of severity. Indeed, while most crack cocaine users exhibit at least subclinical AH in the form of hemosiderin-laden macrophages in the BAL, both cocaine and crack cocaine can cause clinically-manifest and sometimes massive AH limited to the lung. The etiology and clinical presentation of ­cocaine-­associated AH has changed in the recent past, as most cocaine samples seized in the USA and in Europe are now laced with levamisole, an immunomodulator drug that was once used to treat rheumatoid arthritis in humans and is now only available as a veterinary deworming agent. Pharmaceutical grade levamisole purportedly enhances the euphorisant effects of cocaine, but the compound mainly poses a risk of neutropenia, agranulocytosis, skin vasculitis and extensive necrosis, and end organ involvement. Although this had been described in the 1970s, there is a resurgence of similar such cases recently in levamisole-laced cocaine

users. Cocaine-­levamisole toxicity manifests more often with heated (crack) cocaine inhalation than with cocaine snorting, in the form of malaise, arthralgias involving the larger joints and cutaneous manifestations including a retiform purpura or large painful hemorrhagic bullae or necrosis involving typically, but not invariably, the face, earlobes or other areas of the body, notably the skin of the limbs. This corresponds to focal thrombotic vasculopathy with intravascular brin formation leading to occlusion and ischemic skin necrosis, and less often to true vasculitis with IgM, IgA, IgG, and C3 deposits. Skin involvement may be extensive, requiring reconstructive surgery. Notable laboratory features include neutropenia (<3000) or agranulocytosis, which constitutes a relevant risk factor for superimposed infections, and auto-antibodies (speckled antinuclear antibodies, anticardiolipin antibodies, ANCA). Interestingly, ANCA often occur at a high titer, with a perinuclear (anti-PR3) and/or cytoplasmic (anti-­myeloperoxidase (MPO)) staining pattern or reacting with multiple components of neutrophil granules including, characteristically, human neutrophil elastase (HNE), lactoferrin, cathepsin G in addition to proteinase 3 and MPO. Concomitant dual positivity to both MPO and PR3 targets (100% and 50% in one series, respectively) is suggestive. Rarely, anti-double strand DNA antibodies are found. The panoply of pleiomorphic antibody positivities at high titers should draw attention to drug etiologies. AH develops in a fraction of cocaine-­levamisole poisoned patients, noted in one study of 3 of 30 such cases [3, 4]. Other organ damage can be present in the form of ENT involvement including sinusitis (possibly due also to cocaine snorting) in 44%, kidney injury in eight (severe in two, with evidence of pauci-immune glomerulonephritis in one), and vasculitis in three [3, 4]. Clinicians may arrive at the correct diagnosis of cocaine-levamisole toxicity in the presence of characteristic skin changes, neutropenia, and ANCA antibodies that are now part of the evaluation in the patient with exposure. An algorithm has recently been proposed for patients with cutaneous involvement that would also apply to any AH case in a patient suspect of being a cocaine-user, starting with cocaine urine screen which, if negative, this is followed by GC/MS measurement for levamisole in urine if the clinical suspicion of levamisole toxicity is strong. The next step is measurement of ANCAs which, if positive, will con rm the diagnostic suspicion of cocaine-levamisole toxicity. Work-up for end-organ dysfunction (lung, kidney, liver) is indicated. A toxicology study on urine in addition to paraphernalia in the patient’s residence in legal cases may also suggest the presence of cocaine and levamisole.

Pneumotox lists 38 drugs that have been associated with the development of ANCA-positive vasculitis [3, 4]. The antithyroid drugs propylthiouracil (PTU), benzylthiouracil, and methimazole can also produce a form of vasculopathy that resembles that induced by levamisole. Patients who are