What every physician needs to know
Chemotherapeutic agents are widely used to treat solid and hematologic malignancies, and are increasingly used for immunosuppression in the context of autoimmune disease. The first chemotherapy drugs, nitrogen mustards, were developed after the observation in World War I that soldiers who survived inhalation of poison gases often developed cytopenias days later. This observation led to the first class of chemotherapy agents and exemplifies both the desired cytotoxic effects of these drugs as well as the potential for fatal toxicities.
After the success of the nitrogen mustards in the 1950s, several new classes of chemotherapeutic agents were developed including alkylating agents, anti-metabolites, nitrosureas, and antibiotic-derived drugs. Over the last two decades, dozens of novel chemotherapeutic agents have been developed, including targeted inhibitors of tumor growth pathways, immunostimulatory, and immunomodulatory drugs. Despite their diverse mechanisms of action, many of these drugs share the potential to cause pulmonary toxicity.
The differential diagnosis of pulmonary symptoms in oncology patients is broad (Table I), with complications related to the malignancy itself, treatment toxicities, and infections that occur in the context of immunosuppression. Distinguishing between these can be a challenge, and pulmonary symptoms may be multifactorial. Oncology patients are more susceptible to both commonplace and unusual pathogens, particularly respiratory infections. Furthermore, the lung is a common site of neoplastic spread or relapse for many malignancies. Finally, chemotherapeutic agents are often given in combinations, complicating the assessment of which drug is responsible. In nearly all situations, the diagnosis of drug-induced lung disease is one of exclusion, as radiographic and pathologic findings are typically nonspecific. Use of chemotherapy in patients with inflammatory diseases poses similar diagnostic challenges.
Pulmonary complications of chemotherapy may be rapidly progressive, severe, and even fatal. A working knowledge of the spectrum of such complications is essential. Clinicians must also consider downstream consequences; deciding whether a patient can continue a potential culprit drug may have significant impact on the patient’s future cancer treatment.
Individual agents within pharmacologic groups of drugs may be associated with idiosyncratic and distinctive toxicities; however, within a given group, drug toxicity is usually stereotypical. It may be difficult to tease out the toxicity that arises from the use of anti-neoplastic agents from complications related to progression of the malignancy itself or to other treatment side effects, including toxicity of other immunosuppressive medications, radiation, and surgery. The major classes of chemotherapeutic agents and their toxicities are discussed below.
Cytotoxic Antibiotics (Table II)
Pulmonary toxicity due to bleomycin is one of the most common, severe and extensively studied chemotherapy complications. The major toxicities occur in the lung and skin, where the drug is concentrated. Up to 20% of patients treated with bleomycin develop clinical pulmonary disease, and the toxicity is fatal in about 1%. In research settings, bleomycin is used to induce pulmonary fibrosis in animal models. Pathologically, these models demonstrate type I pneumocyte destruction and type II pneumocyte hyperplasia and dysplasia with fibroblast activation, collagen deposition, and eventually fibrosis. Pathologically, lung biopsy specimens demonstrate findings of diffuse alveolar damage.
Risk factors include: (1) cumulative dose > 400 units (although toxicity has been reported at much lower doses); (2) age > 40 years; (3) thoracic radiotherapy (either concurrent, prior, or post-treatment); (4) renal insufficiency (as 80% of drug is renally excreted); (5) combined treatment with other cytotoxic agents; and (6) exposure to high concentrations of inspired oxygen.
In addition, genetic factors may play a role in development of pulmonary injury, as suggested by patient variability to bleomycin-induced injury and strain-specific susceptibility to toxicity in mouse models.
Several syndromes are associated with bleomycin pulmonary toxicity, each characterized by a distinct time course:
Chest pain associated with bleomycin infusion occurs in approximately 3% of patients. This syndrome, which typically resolves with discontinuation of the infusion, is not a contraindication to further treatment with the drug.
Acute hypersensitivity pneumonitis can develop rapidly and is associated with fever, peripheral eosinophilia, and/or eosinophilic alveolitis found on bronchoscopy. The syndrome typically resolves with drug discontinuation and administration of corticosteroids. While re-challenge with bleomycin in patients who demonstrate these type of hypersensitivity reactions is widely described, a reasonable alternative treatment should be strongly considered.
Interstitial lung disease is the most common syndrome; it may progress to end-stage pulmonary fibrosis. The clinical presentation is typically subacute, with symptoms of dry cough and dyspnea accompanied by interstitial changes on imaging occurring a few weeks to six months after treatment. Patients who develop interstitial lung disease due to bleomycin should have the drug withheld; withdrawal of the drug in mild cases may result in improvement. The role of corticosteroids is less clear and they are probably of little benefit in patients with advanced fibrosis. Corticosteroid doses reported in case series typically start at an initial dose of 60 to100 mg/day with slow taper over a period of months, guided by clinical assessment of the patient. Bleomycin-associated interstitial lung disease may also appear months to years after treatment, primarily in patients who receive thoracic radiation and in those exposed to high inspired concentrations of oxygen (e.g., patients who undergo general anesthesia for surgery). Patients should be counseled to avoid exposures to unnecessarily high concentrations of oxygen if possible.
The role of pulmonary function testing during treatment with bleomycin is controversial. As a significant fraction of bleomycin-treated patients develop drug-related lung injury, routine pulmonary function testing is often performed despite a lack of evidence that testing is beneficial. The typical abnormalities seen with bleomycin toxicity include declines in the diffusion capacity for carbon monoxide (DLCO) and a restrictive ventilatory impairment with reduced lung volumes. These findings are not specific for drug toxicity however, and may reflect pulmonary changes related to infection, malignancy, anemia, or even deconditioning. If pulmonary function test abnormalities develop, consideration is usually given to discontinuing the drug, a decision that may have significant clinical consequences. Current National Comprehensive Cancer Network guidelines recommend consideration of baseline pulmonary function tests with repeat testing during treatment if clinically indicated.
Mitomycin-related pulmonary toxicity occurs in 3-12% of patients and is associated with a wide variety of pulmonary syndromes:
Bronchospasm may occur during mitomycin infusion.
Acute respiratory distress syndrome (noncardiogenic pulmonary edema) has been described in the setting of mitomycin given in combination with vinblastine.
Acute interstitial pneumonitis is the most common syndrome and occurs primarily in patients treated with mitomycin in combination with vinca alkaloids. Onset may be abrupt, occurring several hours after administration of the vinca alkaloid, and clinical manifestations may be severe.
Interstitial pneumonitis with pulmonary fibrosis typically occurs 3-4 months after mitomycin treatment and appears to be dose-related; a cumulative dose of 20 mg/m2 or greater is associated with a higher likelihood of toxicity. Case reports suggest that glucocorticoid treatment for interstitial pneumonitis may be of benefit. In one case series of five patients with mitomycin-induced pneumonitis, the authors concluded that withdrawal of glucocorticoids before completing 2-3 weeks of treatment (with a 4 week taper) may result in recrudescence of symptoms and radiographic infiltrates.
Pulmonary hypertension related to pulmonary veno-occlusive disease, typically occurring 4 months after treatment.
Microangiopathic hemolytic anemia and renal failure, with or without pulmonary hemorrhage, have been reported; the syndrome may occur within months after initiation of mitomycin or up to several months after cessation of treatment.
Like bleomycin and mitomycin, Actinomycin D may cause pulmonary toxicity in the form of acute or subacute interstitial lung disease. The drug has been implicated in exacerbations of lung injury related to thoracic radiation due to its radiosensitizing effect.
Alkylating Agents (Table III)
Pulmonary toxicity occurs in less than 5% of patients treated with busulfan and the pathophysiologic mechanism of lung injury remains unknown. Risk factors include cumulative dose, concurrent treatment with other alkylating agents, and concurrent therapeutic radiation. Patients may present with non-productive cough and progressive dyspnea on exertion, with or without systemic symptoms of fever and weight loss. Notably, symptoms often occur years after busulfan exposure. The most common radiographic pattern is one of basilar reticular opacities, though ground glass opacities and centrilobular nodules have also been reported. Pulmonary function testing reveals a reduced DLCO and development of a restrictive ventilatory impairment over time. Busulfan should be withdrawn if possible, if pulmonary toxicity develops; however, many cases will develop long after therapy has been completed. Anecdotal reports of response, along with a lack of other reasonable treatment options, support a trial of corticosteroids (e.g., prednisone 1 mg/kg/day) for patients with progressive symptoms and loss of pulmonary function in whom infection has been excluded.
Pulmonary toxicity related to cyclophosphamide is rare, occurring in less than 1% of patients. Despite the low incidence, it is one of the most common pulmonary toxicities encountered because of its widespread use for both oncologic and non-oncologic diseases. Cyclophosphamide is commonly used in combination cancer therapies and as immunosuppressive therapy in autoimmune diseases. Cyclophosphamide may be used to treat autoimmune interstitial lung disease, and distinction between progression of the underlying lung disease and drug-related pulmonary toxicity may be difficult.
Published reports describing large numbers of patients with cyclophosphamide pulmonary toxicity are lacking. Based on individual case studies and small case series, two syndromes have been described: an acute pneumonitis and a late-onset pneumonitis with progressive pulmonary fibrosis.
Acute pneumonitis occurs within several months of initiation of therapy and presents as cough, dyspnea, and interstitial opacities on chest imaging. The disorder appears to respond to drug discontinuation and administration of corticosteroids, although evidence for the latter is anecdotal. Case reports describe acute, rapidly progressive interstitial disease in patients concurrently receiving cyclophosphamide and other potential pulmonary toxins reinforcing the recommendation that such risky combinations be carefully considered.
Late-onset pulmonary fibrosis is more common in patients treated with low doses of the drug over long periods of time (months to years). In one series reporting late-onset toxicity, symptoms included progressive dyspnea and cough. Pulmonary function testing demonstrated reductions in DLCO, and chest imaging showed parenchymal opacities with bilateral pleural thickening. The course is progressive and may be fatal; no improvement was seen with corticosteroid therapy.
Antimetabolites (Table IV)
Methotrexate is a folate antagonist used broadly as a chemotherapeutic agent and as an immunomodulator in patients with non-neoplastic inflammatory diseases. Up to 30% of patients treated with methotrexate over long periods of time for inflammatory illnesses have toxicity in one or more organs sufficient to warrant drug discontinuation. Pulmonary toxicity from methotrexate occurs in 2-7% of patients with rheumatoid arthritis.
The syndromes, risk factors, and time course for methotrexate-induced lung injury vary, perhaps reflecting the diverse populations in which the drug is used. Patients treated for inflammatory disorders, of which rheumatoid arthritis is the most common, typically receive relatively small doses over long periods of time (e.g., 2.5 to 25 mg per week over years). In contrast, patients treated for malignancy receive much higher doses over relatively shorter time intervals (weeks to months). To mitigate extra-pulmonary toxicities, patients are often prescribed folic acid along with methotrexate, however, this does not prevent or limit pulmonary toxicity.
Several pulmonary toxicity syndromes have been described with methotrexate use:
Hypersensitivity pneumonitis, the most common, is variably accompanied by fever, peripheral eosinophilia, skin rash, or lymphocytic alveolitis noted in bronchoalveolar (BAL) fluid. If a lung biopsy is performed, granulomatous inflammation or mononuclear cell infiltration may be seen. The syndrome may resolve spontaneously, even with continuation or re-challenge with the drug. However, in the absence of a firm indication to expose the patient to ongoing toxicity because of anticipated treatment benefit, continuation of the drug is generally not recommended.
Subacute interstitial pneumonitis has a more insidious presentation and can lead to chronic fibrosis.
Organizing pneumonia (OP) has been reportedin the setting of methotrexate use for rheumatoid arthritis.
Acute respiratory distress syndrome (noncardiogenic pulmonary edema) has been described in the setting of intrathecal administration of drug.
Pleural disease including pleural effusions and pleuritis rarely occur.
A number of risk factors for pulmonary toxicity with methotrexate have been identified:
Renal disease. The drug is primarily excreted by the kidneys and the risk of toxicity is higher in older patients and in those with renal insufficiency.
Hypoalbuminemia. As 50-80% of methotrexate is protein-bound, low protein states may predispose to toxicity.Drug interactions. The risk of toxicity may be increased by concomitant use of drugs that decrease renal excretion of methotrexate (e.g., salicylates, phenylbutazone, penicillin, and sulfonamides) or decrease its protein binding (e.g., salicylates, barbiturates, phenytoin, phenylbutazone, and nonsteroidal anti-inflammatory drugs).
In patients with rheumatoid arthritis, prior use of disease-modifying drugs, underlying rheumatoid pleuropulmonary involvement, and diabetes also appear to increase pulmonary risk.
There are conflicting reports regarding effect of dose intensity, route of drug administration, and dosing interval on pulmonary toxicity.
Withdrawal of the drug is the initial approach in management. In patients with hypersensitivity pneumonitis, withdrawal may be sufficient, with improvement evident within days to weeks. Many anecdotal reports and small case series suggest a role for corticosteroids (e.g., prednisone at a dose of 1 mg/kg/day, followed by slow taper based on clinical response).
Cytosine Arabinoside (Ara-C)
Cytosine arabinoside (Ara-C) is a pyrimidine nucleoside analog that rapidly inhibits DNA synthesis and is used in high-dose chemotherapy regimens, typically to induce remission in acute hematologic malignancies. Pulmonary toxicity appears to parallel the intensity of treatment and can take a number of forms:
Acute respiratory distress syndrome (noncardiogenic pulmonary edema) occurs either during drug infusion or up to several weeks after treatment. Although an incidence of up to 13% has been reported, a large trial of over 1000 patients with acute myeloid leukemia treated with high-dose Ara-C and daunorubicin followed by monthly maintenance with Ara-C, did not result in significant pulmonary toxicity, suggesting the complication is relatively uncommon. Treatment is supportive.
Organizing pneumonia (OP) has been reported when Ara-C is administered with anthracyclines or interferon-alpha. In most patients, clinical and radiographic resolution is seen within days to weeks following drug discontinuation, with or without administration of corticosteroids.
Hypersensitivity pneumonitis is a rare complication, and may present with atypical features such as diffuse centrilobular nodules. This typically resolves spontaneously.
Gemcitabine is a pyrimidine analog that is structurally similar to Ara-C. In a review of pulmonary toxicity in the clinical trial and safety databases performed for its pharmaceutical company, the incidence of serious gemcitabine-related pulmonary toxicity was estimated at less than 1%; however, in Phase II and III clinical trials evaluating patients treated with gemcitabine in combination with other cytotoxic therapy or chest radiation, an incidence of over 10% has been reported. Patterns of gemcitabine-associated lung injury are poorly defined; they include acute respiratory distress syndrome (noncardiogenic pulmonary edema), interstitial pneumonitis, pleural effusion, and pulmonary fibrosis, and potentiation of radiation resulting in radiation pneumonitis.
Risk factors associated with pulmonary toxicity include co-administration of bleomycin and concurrent administration of agents (including the taxanes and vinorelbine), which result in the release of inflammatory cytokines. Gemcitabine is a radiosensitizer, and the possibility of severe radiation-associated pneumonitis should be considered with prior, concurrent, or post-treatment thoracic radiation. Treatment consists of withdrawal of drug and supportive care; anecdotal reports suggest improvement in severe cases with administration of corticosteroids.
Fludarabine, a purine analog, is primarily used for treatment of chronic lymphoproliferative disorders and may be administered over long periods of time. The largest reported series describes pulmonary toxicity in approximately 9% of patients treated at a single institution. Evidence of interstitial pneumonitis developed 3-6 days after treatment. Toxicity did not correlate with age, prior chemotherapy, or underlying lung disease; however, toxicity was more common in patients with chronic lymphocytic leukemia than in other disorders. Other reports describe a more delayed onset of toxicity, occurring within weeks to months of treatment. The presence of interstitial inflammation and granulomas in lung biopsies suggests an underlying hypersensitivity mechanism, a consideration further supported by the observation that improvement is often seen with administration of corticosteroids.
Fludarabine is also associated with profound immunosuppression, which places patients at increased risk of opportunistic infections such as Pneumocystis jirovecii. This effect may persist for months after treatment, warranting a high clinical suspicion for infection.
Nitrosoureas (Table V)
Carmustine is the most widely used of the nitrosureas and is the one most strongly associated with pulmonary toxicity. Like bleomycin, carmustine is used in the laboratory as an animal model of pulmonary fibrosis.
Carmustine pulmonary toxicity is dose-related; a third of patients who receive a cumulative dose of greater than 525 mg/m2 develop severe, progressive interstitial disease, while less than 15% of those who receive a cumulative dose under 475 mg/m2 develop toxicity. Additional risk factors include female sex, concurrent or previous treatment with other chemotherapeutic agents, pre-existing lung disease, chest radiotherapy, concomitant oxygen therapy, and age less than six years at the time of treatment.
Pulmonary toxicity due to carmustine includes interstitial lung disease and pneumothorax.
Interstitial lung disease due to carmustine may be fulminant occurring within days or weeks after treatment, however, the more common presentation is subacute, occurring months to years after exposure. The cardinal pathologic finding is interstitial fibrosis with type II pneumocyte hyperplasia, dysplasia, fibroblast proliferation, and a relative paucity of inflammation. Radiographic evaluation in patients with delayed pulmonary toxicity demonstrates progressive interstitial abnormalities that are typically basilar, although upper lobe predominance has been described. Most patients who develop pulmonary fibrosis do so within 3 years of treatment. However, the risk of developing severe pulmonary fibrosis persists for years. In a 25-year follow-up of seventeen survivors of childhood brain tumors who were treated with high-dose (more than 700 mg/m2) carmustine, 53% died of complications related to pulmonary fibrosis, including two within the first 3 years, four between 6-13 years, and two between 13-25 years after chemotherapy.
Pneumothorax appears to be more commonly associated with carmustine toxicity than with other cytotoxic chemotherapies.
The use of inhaled steroids may reduce the risk of pulmonary toxicity. In a study of patients undergoing high-dose chemotherapy with carmustine in combination with cyclophosphamide and cisplatin as a conditioning regimen prior to autologous stem cell transplantation for breast cancer, high-dose inhaled fluticasone during the initial twelve weeks of treatment was associated with significantly less decline in DLCO at three months than that seen in historical controls. In general, corticosteroid treatment for late-onset pulmonary fibrosis related to carmustine or to any other chemotherapeutic agent is ineffective.
The other members of the nitrosourea family–lomustine (CCNU), semustine (methyl CCNU), and chloroethyl nitrosourea–have not been used or studied as extensively as carmustine has. Reports of pulmonary infiltrates or fibrosis related to these agents are relatively few, but based on the agents’ chemical relationship with carmustine, their potential for causing pulmonary toxicity should be considered.
Biologic Response Modifiers (Table VI)
Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors
Gefitinib and erlotinib are orally active, small-molecule inhibitors of the epidermal growth factor receptor (EGFR) tyrosine kinase. Interstitial lung disease has been well recognized as a complication of both gefitinib and erlotinib.
The incidence of interstitial lung disease related to gefitinib appears to vary by race; rates are higher in Asians (6% in Japan versus 0.3% in the US). In one Japanese series, pre-existing interstitial lung disease, concurrent cardiac disease, older age, poor performance status, and smoking were associated with increased risk for development of interstitial lung disease. Pulmonary toxicity with erlotinib, including fatal cases of progressive interstitial lung disease, has also been described; it appears to occur at a lower frequency than with gefitinib. Other EGFR inhibitors, including cetuximab, may also be associated with pulmonary toxicity, albeit at a lower rate. Patients with non-small cell lung cancer (NSCLC) are more likely to develop interstitial lung disease than patients with other malignancies.
Pulmonary toxicity occurs early, usually within days to one or two months of initiation of treatment. Although the incidence of pulmonary toxicity is fairly low, lung disease may be rapidly progressive and is fatal in up to 45% of cases.
Patients typically present with dyspnea and cough. Radiographs demonstrate progressive ground-glass opacities, consolidation, and bronchiectasis. Distinguishing these findings from worsening pulmonary findings from progression of underlying cancer may be challenging. Furthermore, as the use of these agents is usually based on the identification of patients who have a mutational tumor analysis that suggests potential drug responsiveness, the decision to stop treatment is particularly difficult.
Glucocorticoids are often given, but they are of unproved incremental benefit beyond supportive care and removal of the offending agent.
Tyrosine Kinase Inhibitors Targeting Bcr-Abl
Several small molecule drugs have been developed that inhibit the Bcr-Abl fusion protein formed characteristic of chronic myelogenous leukemia (CML) and other malignancies.
Imatinib can cause interstitial pneumonitis and eosinophilic pneumonia, both of which typically resolve with drug discontinuation with or without adjunctive glucocorticoids.
Dasatinib is associated with lymphocytic, exudative pleural effusions, which occur in as many as 35% of patients, as well as pneumonitis. Rarely, in less than 1% of patients, dasatinib can cause severe pre-capillary pulmonary hypertension.
All-trans Retinoic Acid
All-trans retinoic acid (ATRA) is a vitamin A derivative used in patients with acute promyelocytic leukemia (APL). ATRA induces the maturation of leukemic cells to normal neutrophils. Its use has been associated with the differentiation syndrome (DS, formerly called retinoic acid syndrome). Because these differentiating cells release inflammatory cytokines causing capillary leak syndrome, DS is characterized by fever, edema, interstitial pulmonary opacities, pleural and pericardial effusions, and, in its most severe form, by diffuse alveolar hemorrhage and renal insufficiency. The syndrome occurs in 25% of those treated with the drug, typically within several days to weeks of drug initiation and frequently accompanied by a marked leukocytosis. Risk factors for DS include increased BMI (>35 kg/m2) and WBC count (>5,000/uL).
The incidence appears to be mitigated by the preventive use of corticosteroids (dexamethasone 10 mg twice daily), which are continued until complete resolution of symptoms and then tapered. In contrast to other agents that cause acute pulmonary injury, continued use of ATRA is often justifiable, as the syndrome tends to resolve with corticosteroids; it may not recur with appropriate pre-treatment, and is associated with a good outcome in properly selected patient populations.
Arsenic trioxide (ATO) is used to treat APL by binding to the PML-RARa fusion gene and, like ATRA, causing differentiation of the malignant cells. ATO is also associated with differentiation syndrome, which occurs in up to 30% of patients. Similar to ATRA, DS due to ATO typically resolves with glucocorticoid treatment.
Vascular endothelial growth factor (VEGF) inhibitors
Bevacizumab is a monoclonal antibody that binds circulating VEGF. Since VEGF appears to be a critical modulator of tumor angiogenesis, it is not surprising that VEGF inhibition in patients with lung cancer is associated with complications of pulmonary hemorrhage and hemoptysis. This complication appears to be particularly problematic in patients with squamous cell lung cancer, although hemoptysis and tumor cavitation in the setting of treatment with bevacizumab is described in non-squamous lung cancers as well. Pre-existing hemoptysis and squamous cell lung cancer histology are relative contraindications to treatment with bevacizumab.
Pazopanib is a tyrosine kinase inhibitor targeting both VEGF-1 and 2 receptors. In one series, 14% of patients developed pneumothorax while on pazopanib. Pulmonary metastasis appears to be the major risk factor for development of pneumothorax.
Rituximab is a partially humanized anti-CD 20 monoclonal antibody developed for use in patients with B-cell lymphoma. The drug is now increasingly used in patients with rheumatologic conditions and in solid organ transplantation.
Rituximab causes an immediate infusion reaction in over half of patients receiving the initial dose; findings include headache, fever, rash, nausea, pruritus (with or without urticaria), and possibly angioedema and bronchospasm. Newer anti-CD20 monoclonal antibodies such as obinutuzumab are fully humanized and appear less likely to cause these reactions. Premedication may abrogate symptoms, and subsequent infusions are less likely to cause adverse reactions.
Reports of rituximab-associated interstitial lung disease have increased in frequency. In a series of more than one hundred patients receiving rituximab for non-Hodgkin lymphoma, 8% developed interstitial lung disease presenting with fever, cough, and dyspnea.
Treatment consists of withdrawal of the drug and initiation of corticosteroids; in some cases, progression to fatal pulmonary complications occurs. Since rituximab exerts its effect through B-cell depletion, infection is a common complication. Distinguishing pulmonary infection from drug-related toxicity may be challenging.
Rituximab can cause tumor lysis syndrome (TLS) and cytokine release syndrome (CRS), both of which can have pulmonary effects.
Trastuzumab is an anti-HER2 antibody used in the treatment of breast cancer. Rare but life threatening pulmonary toxicities have been reported including organizing pneumonia, interstitial pneumonitis, and ARDS. In addition to cessation of therapy, response to glucocorticoids has been reported.
Taxanes (Table VII)
The taxane family of drugs exerts chemotherapeutic effects through inhibition of microtubule disassembly and disruption of the G2 and M phases of the cell cycle. The three most commonly used taxanes are docetaxel, paclitaxel, and nabpaclitaxel.
When it was first introduced into clinical practice, paclitaxel was associated with up to a 30% incidence of acute infusion reactions, consisting of bronchospasm, urticaria, and hypotension. These findings are consistent with a hypersensitivity or allergic reaction. Pre-treatment with antihistamines and glucocorticoids is now routine, and the incidence of severe infusion reactions has decreased substantially.
Both paclitaxel and docetaxel are associated with interstitial pneumonitis, which may develop during or following a course of treatment, though most commonly within days to weeks of exposure. Respiratory failure has also been reported. In some cases, the proposed mechanism of injury has been a hypersensitivity reaction, although supporting evidence is limited. Long-term follow-up studies on the clinical course are sparse. The combination of a taxane and other cytotoxic therapies appears to increase the risk of interstitial pneumonitis. In particular, use of paclitaxel or docetaxel with gemcitabine appears to predispose to severe or life-threatening pneumonitis. Taxanes are radiosensitizers and the incidence of pulmonary toxicity is increased with concomitant thoracic radiation.
Docetaxel (but not paclitaxel) has been associated with capillary leak syndrome characterized by pulmonary and peripheral edema, pleural effusions, ascites, and anasarca. Pre-treatment with glucocorticoids appears to decrease the frequency of this complication.
Treatment for taxane-associated pulmonary toxicity consists of withdrawal of the drug and supportive care. Use of glucocorticoids may be reasonable if hypersensitivity is suspected as the underlying mechanism of injury. In addition, a course of corticosteroids may be reasonable if underlying infection and progressive malignancy are excluded.
Mammalian target of rapamycin (mTOR) inhibitors (Table VIII)
Several inhibitors of the mTOR pathways have been developed, including temsirolimus and everolimus. Pneumonitis occurs in 8-14% of patients exposed to everolimus and up to 5% of patients taking temsirolimus. Severity ranges from mild to life threatening, and onset is typically after 2-3 months. Risk factors include NSCLC, pulmonary metastasis, or pre-existing pulmonary disease. Similar findings have been reported with sirolimus, particularly in the solid organ transplant setting, though the occurrence of this toxicity may be rarer, and resolution with withdrawal of the drug may be rapid. In non-oncology populations, case reports demonstrate that pneumonitis can occur after everolimus drug eluting stent placement. Response to corticosteroids has been documented, and recrudesence of pneumonitis after re-challenge is frequent.
Proteosome inhibitors (Table IX)
Proteosome inhibitors, including bortezomib, carfilzomib, and ixazomib are used for treatment of multiple myeloma and often cause minor pulmonary toxicity. Dyspnea occurs in up to 30% of patients receiving proteosome inhibitors and 3% of patients receiving bortezomib develop pulmonary hypertension. Acute pneumonitis, occurring days to weeks after exposure, has also been reported with bortezimib and may resolve with glucocorticoid treatment. Risk factors for acute pneumonitis include male sex, relapsed disease, and IgG subtype.
Immunostimulatory agents (Table X)
Interleukin-2 (IL-2) is a potent immunostimulatory cytokine used for treatment of metastatic renal cell carcinoma and melanoma. IL-2 may be highly toxic, and administration predictably results in capillary leak syndrome with noncardiogenic pulmonary edema, hypotension, cardiac arrhythmias, and renal insufficiency. Drug administration requires a monitored setting such as an ICU, and supportive care needs (including fluids and vasopressors) during administration are high.
Immune checkpoint inhibitors
Many cancers circumvent the host immune response by over-expressing immune checkpoint ligands (PD-L1 and PD-L2), which interact with the immune checkpoint receptors (PD-1 and PD-2) on T-lymphocytes and exert an immune suppressive effect. Immune checkpoint inhibitors (ICIs) are a novel class of immune modulatory chemotherapeutics that block PD-1 and include nivolumab, pembrolizumab, and atezolizumab. Recent trials have demonstrated that ICIs are efficacious against a wide range of cancers, including NSCLC, renal cell cancer, melanoma, and urothelial carcinoma. Unfortunately, due to their potent immune stimulatory effects, ICIs can be associated with autoimmune toxicities, including potentially fatal pneumonitis, as well as other autoimmune phenomena (diabetes, adrenal insufficiency, hypo- and hyperthyroidism). In a large meta-analysis of over 4000 patients treated with ICIs, the incidence of pneumonitis was 2.7% when these were used as monotherapy and 6.6% when used in combination with other immunomodulatory drugs such as ipilimumab. Patients with NSCLC had a higher incidence of pneumonitis (4.1%) compared to patients with other cancer types (1.6%).
The onset of symptoms typically occurs 2-3 months after initiation of ICI therapy, and presenting signs and symptoms are often insidious with dry cough, dyspnea, and crackles that can progress to hypoxemia and respiratory failure. Radiographic findings are protean and include ground glass opacities, reticular opacities, and diffuse consolidations; there may be lower lobe predominance.
Management depends on the severity of the pneumonitis. In patients with significant hypoxemia, the drug should be stopped, bronchoscopy should be performed to exclude infection, and systemic corticosteroid therapy should be promptly initiated. Patients with respiratory failure or those who do not respond to steroids may receive other anti-inflammatory agents such as infliximab. Most patients recover if treated promptly, though pneumonitis has been shown to recur with repeat exposure.
Cell-based immune therapies
Though not yet FDA approved, several promising cell-based therapies are currently in late-stage clinical trials and may pose unique pulmonary challenges. These therapies, including tumor infiltrating lymphocyte (TIL) infusions and chimeric antigen receptor T-cells (CAR-T), involve expansion of patient-derived lymphocytes in vitro, and in the case of CAR-Ts, modification to target specific tumor antigens. These cells are then infused back into the patient where they exert a potent tumor killing effect. Toxicity is characterized by cytokine release syndrome (CRS) often associated with neurotoxicity. Treatment involves administration of glucocorticoids, TNF blockers (e.g., infliximab), and/or IL-6 blockers (e.g., tocilizumab).
Are you sure your patient has chemotherapy-related drug-induced lung injury? What should you expect to find?
Important Considerations in the Evaluation of Patients for Chemotherapy-Related Pulmonary Toxicity
The clinical manifestations of drug-induced pulmonary toxicity are nonspecific. The time course of toxicity is variable, but in most cases, it occurs relatively early within the first weeks or months after initiation of therapy. Notable exceptions include drugs associated with pulmonary fibrosis (bleomycin, busulfan, and nitrosoureas), where disease progression may occur over the course of years.
Symptoms of pulmonary toxicity, such as cough, dyspnea, and chest discomfort, as well as systemic symptoms, including weight loss and fatigue, are nonspecific. They may be difficult to distinguish from symptoms related to infection or the underlying cancer. Failure to consider drug-related pulmonary toxicity as a cause of symptoms may result in progression of adverse effects.
Physical examination in early disease may be completely normal, with progressive disease accompanied by signs related to the type of pulmonary involvement, such as inspiratory crackles in patients with interstitial pneumonitis, wheezing in patients with bronchoconstriction or hypersensitivity syndromes, and dullness to percussion and diminished breath sounds in patients with pleural effusion. These signs are also nonspecific.
The presence of persistent pulmonary symptoms, even mild persistent dyspnea, particularly in the presence of a suggestive lung examination and radiographic abnormalities, should raise the possibility of drug toxicity. Evaluation of a patient with possible chemotherapy-related toxicity is facilitated by asking two questions: Is the drug administered known to cause pulmonary complications when used alone or in combination with other drugs? Is the patient’s pulmonary syndrome consistent with drug-related toxicity? Exclusion of other common causes of pulmonary decompensation, including progression of malignancy and infection, is essential.
Syndromes of Chemotherapy-Related Pulmonary Toxicity (Table XI)
Interstitial Pneumonitis/Pulmonary Fibrosis
Acute pneumonitis is characterized by dry cough, dyspnea, and basilar crackles. A rapid course may mimic noncardiogenic pulmonary edema, while a more subacute course presents insidiously over weeks or months. Progressive pneumonitis may result in pulmonary fibrosis developing over months or years after cessation of drug use; in some cases, respiratory failure may develop. In one study of long-term survivors of childhood and adolescent malignancies, pulmonary fibrosis developed many years after exposure to some chemotherapeutic agents, particularly the nitrosoureas (carmustine and lomustine), bleomycin, cyclophosphamide, and busulfan.
Chest radiographs show reticular infiltrates, which reflect irreversible fibrosis in severe cases; traction bronchiectasis may be seen. Pulmonary function tests demonstrate abnormalities in DLCO and a restrictive ventilatory defect. Lung biopsies characteristically demonstrate proliferation of atypical type II pneumocytes; organizing pneumonia may be seen, a finding which may not enable differentiation between drug-induced pneumonitis and infection.
Noncardiogenic Pulmonary Edema
Noncardiogenic pulmonary edema usually presents abruptly with rapidly progressive respiratory distress developing over hours. Diffuse crackles are noted on exam, and arterial blood gas reveals hypoxemia. Chest radiograph demonstrates diffuse alveolar or reticular infiltrates without cardiomegaly or pleural effusions. BAL findings are nonspecific. In severe cases, diffuse alveolar damage (DAD) may be seen on lung biopsy; however, DAD is a nonspecific finding and may also be seen with infection, shock, transfusion reactions, and tumor lysis syndrome. Withdrawal of the offending drug and supportive care are usually associated with a good prognosis.
Pulmonary eosinophilic syndromes typically occur acutely after the patient is exposed to a drug. Several syndromes that are likely pathophysiologically distinct tend to be included in the hypersensitivity category, all of which result in cough and dyspnea; some are accompanied by fever and fatigue, and less commonly, myalgias, arthralgias, or skin eruption.
Drug toxicity presenting as simple pulmonary eosinophilia (Loeffler’s syndrome) has been described. Symptoms are usually mild, and chest radiographs demonstrate patchy pulmonary infiltrates that may be migratory; peripheral eosinophilia is typically seen. Symptoms usually resolve with withdrawal of the drug, and re-challenge will result in relapse.
Chronic eosinophilic pneumonia associated with drug toxicity follows a subacute course over weeks to months and is often accompanied by systemic symptoms, peripheral eosinophilia, and radiographic infiltrates that may be migratory. Patients usually respond well to corticosteroids, as do patients with idiopathic chronic eosinophilic pneumonia; however, unlike in patients with idiopathic chronic eosinophilic pneumonia, relapse is rare as long as the patient is not re-challenged with the drug. Patients with drug-induced eosinophilic lung disease often demonstrate lymphocytic and eosinophilic alveolitis (more than 25% of nucleated cells) on BAL.
True hypersensitivity pneumonitis due to a cell-mediated (type IV) delayed reaction also appears to occur in some cases of drug toxicity, with symptoms and radiographic changes occurring within hours to days after drug exposure; it may be difficult to distinguish from acute noncardiogenic pulmonary edema.
Pulmonary Vascular Disease
Pulmonary hypertension can occur by several mechanisms including pre-capillary (due to dasatinib), post-capillary (pulmonary veno-occlusive disease due to mitomycin-C), and cardiac toxicities (such as from trastuzumab).
Pulmonary veno-occlusive disease (PVOD), characterized by fibrotic occlusion of pulmonary veins and eventual development of pulmonary hypertension, is an uncommon complication of chemotherapy. The usual symptom is progressive dyspnea. PVOD may be associated with interstitial abnormalities; pleural effusions are commonly seen, which is unusual in patients with pre-capillary pulmonary hypertension. Evaluation for PVOD may be difficult; a definitive diagnosis requires a surgical lung biopsy.
Hepatic sinusoidal obstruction syndrome (previously called veno-occlusive disease) in the context of treatment for hematopoietic malignancies is more common than PVOD and is thought to result from injury to hepatic venules from toxic drug metabolites.
Pleural disease related to drug toxicity usually occurs in conjunction with pulmonary parenchymal toxicity. An isolated pleural effusion or pleuritis without parenchymal lung disease has been described with methotrexate, docetaxel, dasatinib and others. In contrast to non-chemotherapeutic agents, drug-induced lupus has not been described with cancer therapies.
Carmustine and pazopinib are associated with pneumothorax.
Chemotherapeutic agents may produce reactions during their infusion that do not necessarily correlate with the drug’s known toxicities. The reactions usually occur during or within a few hours of termination of the infusion. Pulmonary symptoms may be as mild as dyspnea or chest tightness, or as severe as anaphylaxis or hypoxemia. Systemic symptoms include pruritus, tachycardia, fever, chills, gastrointestinal complaints, and skin rash.
In general, the severity of infusion reactions may be mitigated by premedication with steroids or by slowing the infusion rate. The exception to this rule is anaphylaxis, the most severe and most dangerous form of infusion reaction. In cancer treatment, anaphylaxis is most commonly observed with platinum drugs and taxanes, although it has the potential to occur with any drug.
Many chemotherapeutic agents increase the risk of radiation injury to the lung; particularly noteworthy are bleomycin, actinomycin-D, mitomycin, taxanes, and gemcitabine. Pulmonary toxicity should be of particular concern in patients treated with combined chemotherapy and thoracic radiation, such as in patients with lung cancer, breast cancer, or mediastinal tumors, including lymphoma.
Beware: there are other diseases that can mimic chemotherapy-related drug-induced lung injury
Mimics of chemotherapy-related drug-induced lung injury include complications related to malignancy, radiation exposure, infection, cardiogenic and non-cardiogenic pulmonary edema. A broad differential is warranted and chemotherapy induced toxicity is typically a diagnosis of exclusion.
How and/or why did the patient develop chemotherapy-related drug-induced lung injury?
Which individuals are at greatest risk for developing chemotherapy-related drug-induced lung injury?
Risk factors depend on the particular cancer and chemotherapy.
The diagnosis of pulmonary toxicity related to drug exposure is a diagnosis of exclusion. No laboratory tests are diagnostic, though the presence of peripheral eosinophilia can be suggestive of certain toxicities. Laboratory evaluation should be performed primarily to exclude other causes of progressive pulmonary symptoms and abnormal imaging, including infection, volume overload, or progressive malignancy.
Often, the first evidence of a new pulmonary process may be an abnormal chest radiograph, with or without pulmonary symptoms. In many cases, chest CT will better characterize the pattern and extent of the abnormality. As with routine laboratory studies, radiographic findings are nonspecific. Radiographic patterns that may be useful in determining the basis for clinical findings include that of radiation pneumonitis, which typically follows the distribution of the thoracic radiation portal, or the discovery of mediastinal or hilar adenopathy, findings which would be uncommon with drug toxicity and more suggestive of progressive malignancy.
What imaging studies will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
See discussion of individual agents.
What non-invasive pulmonary diagnostic studies will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
Pulmonary function tests (PFTs) provide an objective measurement of physiologic lung function at the time of initial concern about drug toxicity and may be useful in patient follow-up. The use of PFTs in evaluating patients with various drug toxicities has been widely reported for nearly all drugs known to cause lung injury; however, as with laboratory and radiographic evaluation, PFT abnormalities are nonspecific.
PFTs performed to evaluate patients with suspected or known drug toxicity should include measurements of DLCO and lung volumes; DLCO impairment and a restrictive pattern are the most common abnormalities observed. Spirometry may be influenced by a multitude of factors, including muscular weakness, general fatigue, and pain. Anemia impacts the DLCO. Currently, there are no strong recommendations for routine performance of PFTs before, during, or after chemotherapeutic regimens.
What diagnostic procedures will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
Since symptoms, physical examination findings, laboratory evaluation, PFTs, and imaging studies are nonspecific, consideration should be given to performing bronchoscopy and/or surgical lung biopsy.
Bronchoscopic findings are not diagnostic for drug toxicity. BAL and transbronchial biopsies are performed primarily to identify or exclude recurrent malignancy or infection. Performed in sequential fashion, BAL may be useful in identifying patients with alveolar hemorrhage syndrome, which may occur as an uncommon manifestation of drug-induced pulmonary injury. BAL fluid analysis usually demonstrates a nonspecific elevation in cell counts in patients with drug toxicity, although an elevation in eosinophils may suggest hypersensitivity reactions.
Surgical lung biopsy, including video-assisted thoracoscopic surgery (VATS) and open thoracotomy, are more invasive procedures that allow for larger portions of lung tissue to be sampled compared to transbronchial biopsies. Reliable pathologic recognition of patterns of interstitial pneumonitis, organizing pneumonia, hypersensitivity reactions, diffuse alveolar damage, and other potential manifestations of drug toxicity is much more likely with surgical than bronchoscopic lung biopsy. While the findings may not definitively establish a diagnosis of drug toxicity, the exclusion of malignancy and infection, along with consistent histopathologic findings, supports the diagnosis.
What pathology/cytology/genetic studies will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
Cytologic studies and tissue biopsies obtained from bronchoscopy or surgical lung biopsy may be useful in distinguishing drug toxicity from underlying tumor.
If you decide the patient has chemotherapy-related drug-induced lung injury, how should the patient be managed?
The cardinal rule of management in patients suspected of having pulmonary disease related to drug toxicity is drug withdrawal; few situations (e.g., all-trans retinoic acid) justify continuing an injurious exposure. Since the impact of discontinuing a chemotherapeutic agent beneficial to cancer treatment may be great, the decision should be made only if other causes of the findings–most importantly, infection or cancer progression–are reasonably excluded.
Withdrawal of a drug may be adequate to reverse toxicity, particularly in the case of infusion reactions. Patients with acute or subacute interstitial pneumonitis or hypersensitivity syndromes should be considered for glucocorticoid treatment (e.g., prednisone 1 mg/kg/day tapered over weeks to months). Patients who are unlikely to respond to glucocorticoids include those with chronic pulmonary fibrosis that presents months or years after treatment, noncardiogenic pulmonary edema, or pulmonary vascular disease.
All patients should receive supportive care, including alleviation of dyspnea, supplemental oxygen (if indicated), and adequate pain management.
What is the prognosis for patients managed in the recommended ways?
See discussion under individual agents.
What other considerations exist for patients with chemotherapy-related drug-induced lung injury?
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.
- What every physician needs to know
- Cytosine Arabinoside (Ara-C)
- Carmustine (BCNU)
- Other Nitrosoureas
- Epidermal Growth Factor Receptor Tyrosine Kinase Inhibitors
- Tyrosine Kinase Inhibitors Targeting Bcr-Abl
- All-trans Retinoic Acid
- Arsenic trioxide
- Vascular endothelial growth factor (VEGF) inhibitors
- Monoclonal antibodies
- Immune checkpoint inhibitors
- Cell-based immune therapies
- Are you sure your patient has chemotherapy-related drug-induced lung injury? What should you expect to find?
- Beware: there are other diseases that can mimic chemotherapy-related drug-induced lung injury
- How and/or why did the patient develop chemotherapy-related drug-induced lung injury?
- Which individuals are at greatest risk for developing chemotherapy-related drug-induced lung injury?
- What imaging studies will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
- What non-invasive pulmonary diagnostic studies will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
- What diagnostic procedures will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
- What pathology/cytology/genetic studies will be helpful in making or excluding the diagnosis of chemotherapy-related drug-induced lung injury?
- If you decide the patient has chemotherapy-related drug-induced lung injury, how should the patient be managed?
- What is the prognosis for patients managed in the recommended ways?
- What other considerations exist for patients with chemotherapy-related drug-induced lung injury?