NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Bast RC Jr, Kufe DW, Pollock RE, et al., editors. Holland-Frei Cancer Medicine. 5th edition. Hamilton (ON): BC Decker; 2000.

Cover of Holland-Frei Cancer Medicine

Holland-Frei Cancer Medicine. 5th edition.

Show details

Chapter 150Respiratory Complications

, PhD, MD.

Cancer and its treatment can produce important effects in the lung. These toxic effects are often more easily recognized than toxicity to other visceral organs. Treatment-associated pulmonary injury produces symptoms and abnormalities of functional testing or chest radiographs. While pulmonary disease is often easily detected, the functional abnormalities and chest radiographs often appear quite similar in patients with toxic lung injury, infection, and metastatic cancer. In recent years, the use of bronchoscopy or transthoracic needle biopsy has greatly reduced this diagnostic problem. It is often true, however, that invasive lung studies are required to diagnose cancer-associated pulmonary disease with certainty. Table 150.1 lists common pathophysiologic mechanisms of pulmonary disease in cancer patients. The effects of congestive heart failure, infection, and embolism are discussed elsewhere, so only selected, specific aspects of these problems are discussed here.

Table 150.1. Pulmonary Disease in Cancer Patients.

Table 150.1

Pulmonary Disease in Cancer Patients.

Significant emphasis will be placed on lung disease associated with marrow transplantation. The intensive treatments and the profound immunosuppression associated with this approach produce diseases more acute, and in sharper focus, than those observed during conventional cancer treatment. An understanding of diseases produced in this setting facilitates greater understanding of the smaller subset of complications seen with conventional therapy.

Pulmonary Evaluation

Several factors determine the nature of the evaluation of undiagnosed pulmonary abnormalities in the cancer patient (Table 150.2).

Table 150.2. Factors Influencing Pulmonary Evaluation in Cancer Patients.

Table 150.2

Factors Influencing Pulmonary Evaluation in Cancer Patients.

Focal radiographic abnormalities increase the likelihood that tumor, bacterial or fungal infection, or embolism has caused the pulmonary disease. In patients whose symptoms suggest infection the more common nonpleural abnormalities can often be diagnosed by sputum or blood culture. The differentiation of cancer from infection can usually be made by bronchoscopy with biopsy, when appropriate. Pleural-based abnormalities, particularly those associated with pleuritic pain, are often related to subsegmental pulmonary infarction caused by obstruction of arterial blood flow. Pulmonary embolus is the most common cause of infarction, but the angioinvasive Aspergillus species can produce identical radiographic and clinical findings, particularly in the setting of prolonged or severe immunosuppression. Pulmonary angiography and radionuclide scanning cannot reliably distinguish embolus from Aspergillus, and bronchoscopy with biopsy is often difficult because of the peripheral nature of these lesions. In the setting of profound immunosuppression, surgical lung biopsy is often required to confirm the presence of fungus. The widely disparate and morbid treatments (heparinization versus aggressive antifungal treatment or surgery) demand rapid and accurate diagnosis.

With the exception of the rare bronchioloalveolar carcinoma, primary lung tumors produce focal abnormalities in virtually all patients. A history of tobacco abuse and lack of infectious symptoms are useful guides to diagnosis. Sputum cytology is the initial procedure of choice to identify primary lung tumors. Metastatic cancer most commonly spreads to the lung to form nodules, though adenocarcinomas or hematopoietic tumors, in particular, can spread to the lung by lymphatic extension to form diffuse interstitial lung abnormalities. Some intra-abdominal tumors (particularly ovarian cancer) spread to the lung by direct extension through diaphragmatic lymphatics to form focal basilar lung densities. Bronchoscopy with transbronchial biopsy or transthoracic needle biopsy with computed tomographic (CT) guidance are often the optimal methods for confirming these latter diagnoses.

The patient’s general status is a critical determinant of the nature and rapidity of the evaluation required. A patient who develops a new lung disease while hospitalized and is acutely ill and profoundly immunosuppressed requires rapid and specific diagnosis. These patients should generally proceed quickly to invasive diagnostic testing after radiography with CT scanning, and if possible, blood, sputum, pleural fluid, and other appropriate body fluids have been obtained for analysis. A specific diagnosis within 24 hours is generally required. If interstitial disease is found on radiography, lung biopsy by bronchoscopy or surgery is usually advisable. A coagulation evaluation with platelet or plasma supplementation may be required to perform these procedures. The cancer physician must ensure that specimens obtained by invasive means are processed and delivered appropriately for bacterial, fungal, viral, and cytologic or histologic study. Consultants, pathologists, and microbiologists must be specifically informed of the differential diagnosis and of the need for speedy analysis.

Visually-assisted thoracoscopic surgery (VATS) can frequently be used for lung biopsy in place of thoracotomy. Advantages of VATS include less morbidity, quicker recovery time, and equal or greater ability to inspect the surface of the lung remote from the incision site. In patients with major coagulation defects, bleeding may be harder to control. Additionally, localization of abnormalities deep within the lung parenchyma may be more difficult because this is often done by direct palpation of the lung itself. The selection of the appropriate procedure should be individualized by discussion with the surgical staff and the patient.

Conversely, outpatients with acceptable immune status and no physical signs of acute illness should initially be evaluated by less costly, noninvasive means. In the setting of allogeneic marrow transplantation performed 1 to 3 months previously, the risk of cytomegalovirus (CMV) infection is high, and bronchoscopy is required. Bronchoalveolar lavage (BAL) specimens should be evaluated by viral culture1 and cytology. Blood can be evaluated by antigen detection2 or polymerase chain reaction (PCR) methods,3 to allow rapid initiation of treatment. In recent years, the incidence of CMV infection has been markedly reduced by the use of CMV-negative4 or leukocyte-depleted5 blood products in CMV-negative patients, prophylactic antiviral therapy,6 preemptive BAL 1 month after transplantation, and treatment of patients with CMV in washings.7 Patients most at risk are those previously colonized with CMV (antibody positive) or unexposed patients (CMV antibody-negative) who receive marrow or blood products containing CMV during the immunosuppressed period.

Acute Lung Injury

Patients recently treated with certain chemotherapeutic agents, immunologic agents, or lung irradiation may have experienced a toxic lung injury.8 This syndrome is often diagnosed in the presence of cough, dyspnea, or low-grade fever, in the absence of important immunosuppression. Chest radiographs are initially often normal but, following lung injury from chemotherapy, eventually demonstrate a diffuse interstitial abnormality, more prominent in the basilar lung segments. Radiation-associated lung injury usually occurs within the irradiated lung field but can be seen outside this area.9

There are two extremes of toxic lung injuries, acute and chronic. The acute form is most commonly seen in patients undergoing marrow transplantation or other intensive chemo- or radiotherapy. The injury is usually seen within days to 2 to 3 months after intensive treatment. In general, the more rapid the onset, the more severe is the toxicity and the more acute is the need for rapid diagnosis. The acute toxicity occurring within 1 to 2 weeks of treatment usually coincides with a period of severe immunosuppression and demands rapid, invasive diagnosis. The acute toxicity occurring during the 1- to 3-month interval is usually identified by the development of dyspnea over a period of several days, with minimal or no diffuse interstitial radiographic abnormality.10 The need for invasive diagnosis is proportional to the coinciding degree of immunosuppression. A patient undergoing autotransplantation for a solid tumor who has experienced full hematologic recovery may not require invasive diagnosis, but an allograft recipient with graft-versus-host disease (GVHD) requires invasive diagnosis to distinguish toxic lung injury from viral infection.

The pathologic definition of acute lung injury requires the histologic definition of pneumocyte cellular atypia. Hyperchromatic nuclei, abnormal chromatin clumping and condensation, nuclear swelling and enlargement, and cytoplasmic vacuolation are all observed to varying degrees.11

Importantly, acute toxic lung injury is often effectively treated with systemic, high-dose steroids (Figure 150.1). Improvement of symptoms usually occurs within 72 hours of initiation of treatment. An important determinant of the value of invasive diagnosis of acute toxic lung injury is the risk of high-dose prednisone treatment, tapered from 100 mg daily to cessation over a 1- to 3-month period. If the risk of steroid-exacerbated infection is low, empiric steroid treatment with careful monitoring may be acceptable and can aid in diagnosis. Because the risk of Pneumocystis carinii infection following steroid treatment is high, prophylactic trimethoprim-sulfamethoxazole or inhaled pentamidine is required in conjunction with steroid treatment.

Figure 150.1. Severe drug-induced acute lung injury.

Figure 150.1

Severe drug-induced acute lung injury. A. Day 15: after intensive alkylating agent chemotherapy (cyclotherapy (cyclophosphamide-cisplatin-carmustine (BCNU). B. Day 20: partial resolution on corticosteroids. C. Day 24: residual small septal lilnes; corticosteroids (more...)

Detailed histopathologic evaluation of acute toxic lung injury has been conducted in animals and, to a much lesser extent, in humans. In addition to the pneumocyte abnormalities described above, nonspecific mononuclear infiltrates can be observed (Figure 150.2). Pulmonary veno-occlusive disease, presumably developing as a result of vascular endothelial injury, has also been described.12 This latter finding emphasizes the multiplicity of the cellular targets of toxic cellular injury.

Figure 150.2. Acute lung injury after cyclophosphamide-cisplatin-carmustine (BCNU) treatment obtained at open biopsy.

Figure 150.2

Acute lung injury after cyclophosphamide-cisplatin-carmustine (BCNU) treatment obtained at open biopsy. Type 2 pneumocyte hyperplasia with extensive eosinophilic hyaline membrane remnants are seen at left. Interstitial thickening with marked mononuclear cell (more...)

The chronic form of acute lung injury more frequently occurs months after the initiation of radiation or chemotherapy and is commonly associated with multiple, less intense treatments. Diffuse interstitial lung abnormalities, sometimes associated with nodular densities which can mimic metastatic tumor, are almost always seen radiographically, particularly on CT scanning. Functional respiratory abnormalities, most commonly seen as both a reduced carbon monoxide diffusing capacity (DLCO) and restrictive lung mechanics, are seen with moderate to severe toxicity.13 Lung biopsy reveals diffuse interstitial fibrosis, and both functional abnormality and fibrosis can be progressive, even if drug or radiation therapy is discontinued. Steroid treatment is of no benefit in this setting.

In individual patients, it is often desirable to determine the extent to which an individual toxic injury has an acute component which might be steroid responsive. Greater disruption in respiratory function, as reflected by decreased lung volumes identified by pulmonary function testing, suggests a greater degree of fibrosis. While toxic lung injuries often vary in severity in different areas of the lung, biopsies taken from heavily affected and relatively unaffected areas can be evaluated for extent of fibrosis and cytologic atypia, as described above. Areas for biopsy can be selected radiographically or by the surgeon, if open biopsy is performed. Extensive fibrosis suggests that steroid treatment is unlikely to benefit the patient. Alternatively, severe pneumocyte atypia without fibrosis may benefit from steroid treatment. As a general rule, prednisone doses of 100 mg/d or greater should be initiated and slowly tapered to cessation over 2 to 3 months to test steroid responsiveness adequately and prevent relapse. If a patient’s lung injury is steroid responsive, improvement is usually seen within 3 to 7 days of initiation of treatment.10


The type of infection which might be diagnosed can often be suspected by the setting in which it develops. Increasingly, prophylactic oral antibiotic regimens are used to prevent infection in neutropenic ambulatory patients. In this setting, gram-positive infections from skin organisms assume significant importance, particularly if patients have in-dwelling central venous catheters. Candida infection can occur early in patients experiencing gastrointestinal mucosal injury from chemotherapy. Hospitalized patients with pulmonary infection are much more likely to acquire nosocomially resistant bacteria or fungi. Construction within the hospital produces a greatly increased risk of Aspergillus infection from airborne fungus contained in construction dust.

Pulmonary Hemorrhage

Pulmonary hemorrhage can occur in the setting of severe coagulation defects produced by malnutrition, hepatic insufficiency, or diffuse intravascular coagulation. There is particular risk with severe thrombocytopenia or refractory thrombocytopenia in the setting of marrow transplantation or intensive chemo- or radiotherapy. Repeated donor platelet transfusion can produce both allogeneic and autologous platelet immunity, with failure to augment platelet count with even intensive transfusion support. Radiographs often show patchy upper-lobe densities or migratory densities in multiple lung areas. Acute lung injury from chemo- or radiotherapy in the setting of coagulation defects can also produce hemorrhage in the distal airways. Serial BALs of the same lung segment produce increasing bloody lavage specimens and can assist in the diagnosis of toxic lung injury.14

Specific Complications

Hypersensitivity Pneumonitis

The clinical and radiographic presentations of hypersensitivity pneumonitis (HP) mimic entirely those of toxic acute lung injury, as described above. On pathologic evaluation, however, eosinophilic infiltrates, noncaseating granulomas, and macrophage proliferation dominate. Hypereosinophilia in the circulation is variably present. As with acute lung injury, treatment with high-dose steroids and withdrawal of any possible offending drug result in rapid improvement. Methotrexate is the most common agent associated with HP, but bleomycin and procarbazine have also been associated with this condition.8

Noncardiogenic Pulmonary Edema

Noncardiogenic pulmonary edema (NCPE), defined as the accumulation of fluid in the distal airways and absent cardiac dysfunction, has become an important cause of cancer-associated pulmonary disease. Increased intensity of cancer treatment is the likely explanation. Operationally, NCPE is often defined by demonstrating adequate cardiac function, absence of infection, and a diffuse interstitial lung abnormality responsive to diuretics. The infiltrates are frequently found in the lower lung segments, where pulmonary blood flow is the greatest. It is critical to recognize that diuretic responsiveness is not helpful in determining the etiology of the edema.

Adult respiratory distress syndrome associated with sepsis or shock from other causes can produce NCPE. Radiation- or chemotherapy-induced toxic lung injury can also produce NCPE, and such patients often benefit significantly from intravascular volume depletion. The improvement which can be seen with diuretic therapy should never deter evaluation of noncardiac causes of pulmonary edema in the immediate period following intensive chemo- or radiotherapy or marrow transplantation, unless diuresis completely reverses the edema rapidly.

Many intensive chemotherapy regimens must be administered with vigorous hydration in order to minimize visceral organ injury. Daily urine output in excess of 15 to 20 L is occasionally required. Drugs commonly utilized with hyperhydration include cyclophosphamide/ ifosfamide, nitrosoureas, platinum analogues, and methotrexate. In spite of normal cardiac function, NCPE often occurs in these patients, and diuresis is required to maintain adequate pulmonary function.15

Hemolytic-Uremic Syndrome

The syndrome of microangiopathic hemolytic anemia, thrombocytopenia, renal insufficiency, severe hypertension, and pulmonary infiltrates is increasingly described following intensive alkylating-agent chemotherapy and marrow transplantation. The most common conventionally dosed drug associated with this syndrome is mitomycin C,16 especially when given in cumulative doses in excess of 40 mg/m2. Biopsies of both lungs and kidneys reveal vascular endothelial injury with intimal proliferation and microvascular thrombi. NCPE and congestive heart failure, both responsive to intravascular volume depletion, can complicate this syndrome. The fully developed syndrome is highly lethal but may be treated with plasmapheresis, steroids, or immune complex removal in selected patients. The mechanism of these treatments’ effectiveness is unknown.

Generalized Capillary Leak Syndrome

Biologic cancer treatments are increasingly being associated with a syndrome of diffuse capillary leak, which is rapidly responsive to discontinuation of the agent. Interleukin-2 (IL-2), particularly in high doses, can produce both NCPE and pleural effusion as part of a systemic process.18 High doses of granulocyte-macrophage colony–stimulating factor (GM-CSF) can produce the same effect.19 IL-11 has also been reported to cause this effect. Other investigational cytokines may produce similar toxicity.20

Docetaxel (Taxotere), a tubulin-binding antitumor agent, heavily used in Europe and increasingly in the United States, also produces this syndrome. The probability of edema is increased with increasing doses of the drug, but pretreatment with steroid, mast cell–stabilizing agents, and antihistamines can markedly reduce this complication.

Bronchiolitis Obliterans

Bronchiolitis obliterans (BO) is an obstructive pulmonary disease characterized by a chronic inflammatory cell infiltrate surrounding the terminal airways and alveoli.21 This chronic inflammation progresses to fibrosis and granulation, producing a radiographic and functional picture compatible with chronic obstructive pulmonary disease.22

BO is clearly associated with allogeneic marrow and solid organ transplantation.23 Within the group of allogeneic marrow transplantations, development of chronic GVHD is correlated with risk of BO. BO rarely occurs in patients receiving autologous transplants, strongly implying that immune mechanisms underlie this disorder. No clear-cut treatment, other than lung transplantation, has been developed, but a lessened incidence of chronic GVHD may lower the frequency of this syndrome.24 Increases in immunosuppressive treatment may also reverse this disorder if it occurs in association with chronic GVHD.

Clinically, BO generally develops within months of allogeneic transplantation and presents as a small-airway obstructive disease on pulmonary function testing. The lungs appear hyperinflated on radiographs. Rapid development of this syndrome is occasionally associated with patchy interstitial infiltrates which may reflect the lymphocytic or mixed inflammatory cell bronchiolitis seen pathologically. A significant mortality rate is associated with this condition.

Toxic Lung Injury

Most cancer drugs have been associated with toxic lung injury. Table 150.3 lists a number of agents associated with this condition. Several representative agents from this list will be discussed in detail.

Table 150.3. Agents That Cause Toxic or Allergic Lung Injury.

Table 150.3

Agents That Cause Toxic or Allergic Lung Injury.

Alkylating Agents

Alkylating agents (AAs) are defined as compounds which form small-molecular-weight reactive intermediates capable of covalent binding to tissue nucleophiles. These nucleophiles include nucleic acids and proteins with sulfydryl, amino, or hydroxyl groups available for binding. The reactive intermediates are highly unstable and often bind to structures in proximity to the site of their formation. While many AAs form reactive intermediates spontaneously, others, such as cyclophosphamide, ifosfamide, procarbazine, and dacarbazine require activation by microsomal enzymes (or other oxidative systems, such as xanthine oxidase or prostaglandin synthase) before reactive intermediates can be formed. Like the liver and kidney, the lung contains cells with drug-activating capability. The nonciliated bronchiolar epithelial cell (Clara cell) has been shown to have microsomal enzymes, thus permitting drug activation within the lung itself. When combined with the fact that the lungs are perfused by the entire cardiac output—and thus are maximally exposed to circulating reactive intermediates—it is not surprising that the lung is a major target for nonspecific reactive intermediate binding and toxicity.

Nitrosoureas are the most common class of AAs producing pulmonary injury. Carmustine (BCNU) is the most commonly used agent in this class and illustrates the spectrum of drug-induced lung injury that nitrosoureas can produce. Of note, most nitrosoureas activate spontaneously to produce alkylating and carbamoylating portions of the molecule as shown in Figure 150.3.

Figure 150.3. Metabolic products of carmustine.

Figure 150.3

Metabolic products of carmustine.

In the case of BCNU, a chloroethyl carbonium equivalent (the alkylation portion) and chloroethyl isocyanate (the carbamoylation portion) are produced.25 Interestingly, a heavily studied chemical congener, methyl isocyanate, is a known pulmonary toxicant and was a major cause of inhalation pulmonary injury following an industrial accident in Bhopal, India. Whether chloroethyl isocyanate is fully responsible for BCNU-associated pulmonary injury is unknown. It is of interest that nitrosoureas designed to have either no (streptozotocin) or less (chlorozotocin) carbamoylating activity rarely cause pulmonary injury. Further study of this association is also important because investigators have suggested that the antitumor effects of nitrosoureas are associated with the alkylating portion of nitrosoureas and not with their carbamoylating activity.26

After initial case reports, Aronin and colleagues described the consistent development of diffuse interstitial fibrosis and respiratory insufficiency in patients with brain tumors treated as outpatients with repetitive doses of BCNU.27 They identified a direct relationship between total cumulative BCNU dose and toxic risk and also identified a history of smoking as an independent risk factor for toxicity. It is attractive to hypothesize that BCNU-associated pulmonary fibrosis in the setting of repetitive outpatient doses is produced by multiple subclinical toxic events producing fibrosis, but studies to support this hypothesis are lacking. The development of late pulmonary fibrosis in patients treated years previously, as children, has also been reported.28 No similar reports of delayed lung injury have been reported in adults, but this subject deserves careful study in long-term survivors.

Over the last 15 years, BCNU has been used increasingly as an important component of high-dose chemotherapy regimens given with autologous hematopoietic cell support (AHCS) to treat patients with breast cancer29 or lymphomas.30 In these settings, acute lung injury has been reported in 10 to 40% of patients and usually occurs 1 to 3 months following treatment. Within this range, there is a dose-related increase in the risk of acute toxicity between 350 and 600 mg/m2, used in most intensive-dose regimens. Pharmacokinetic-pharmacodynamic studies of the cyclophosphamide-cisplatin/BCNU regimen demonstrated that individual patient blood exposure to BCNU (as measured by area under the curve [AUC]) was highly correlated with the risk of developing acute lung injury.10

When used in single high doses, BCNU produces an acute lung injury, as opposed to fibrosis, and this acute injury is reversible with steroid treatment, as described previously. The clinical course, radiographic findings, and physiologic testing results from this injury are typical of those described for acute toxic lung injury in the introductory portion of this section. To date, no late fibrosis or clinical respiratory decompensation has been reported with this or similar regimens but this subject requires further and more prolonged study before firm conclusions can be drawn. Whether steroid treatment after repetitive outpatient doses would reduce the risk of lung fibrosis is unknown.

Lomustine (CCNU) has also been associated with acute lung injury, but its manifestations, frequency, and clinical course are less well described than those of BCNU, perhaps because of its infrequent use.31

Busulfan was for many years the standard treatment for chronic myeloid leukemia (CML). It was administered orally in periodic doses, and this schedule produced a notable incidence of chronic pulmonary injury.32 The risk of pulmonary injury increases markedly when cumulative doses above 500 mg are administered but is still generally less than 5% for treated patients.33 As with nitrosoureas, chronic, prolonged lower-dose administration leading to fibrosis produces irreversible injury and risk of death. In occasional patients, a more acute lung injury is produced which responds to steroid therapy. Limited data suggest that drug interruption or steroid treatment is more likely to improve the lung toxicity than is the case for nitrosoureas. In part because of this toxicity, hydroxyurea has now supplanted busulfan as the chemotherapeutic agent of choice for suppression of CML.

The busulfan/cyclophosphamide (Bu/Cy) regimen is now frequently employed as an alternative to cyclophosphamide plus total body irradiation for patients undergoing allogeneic marrow transplantation and is used for selected autologous transplantation procedures as well.34 The incidence of pulmonary toxicity with this regimen is less than 5%, as is the case with chronic lower-dose administration. When acute lung injury is seen and found by biopsy to be consistent with drug injury, steroid treatment is often effective in reducing its severity.

Other Agents Forming Reactive Intermediates

Bleomycin is used in multiple outpatient doses to treat patients with testicular cancer and lymphoma. Like the AAs, higher cumulative doses delivered in this manner increase the risk of chronic, fibrosing pulmonary injury. Cumulative doses in excess of 450 units appear to greatly increase the risk of pulmonary fibrosis. Administration of bleomycin with certain combination chemotherapy regimens, such as BACOP (bleomycin, doxorubicin, cyclophosphamide, vincristine, and prednisone),35 appears to reduce markedly the cumulative dose required for toxicity.

Unlike the AAs, there is evidence that bleomycin can chelate iron; in the process, the complex can form an intermediate capable of interacting with oxygen to produce reactive oxygen species, particularly superoxide.36 These toxic species can injure multiple cellular components, including lipid membranes. This indirect effect appears to persist for prolonged periods. Administering high inspired oxygen concentrations during anesthesia or acute illness many months after bleomycin treatment has been reported to exacerbate pulmonary injury through unknown mechanisms.37 This and other potential toxic interactions are discussed below. It is advisable to monitor alveolar function by DLCO measurements at intervals in any patient who is to receive more than 200 units bleomycin or who is receiving bleomycin as part of a nonstandard chemotherapy regimen. It is suggested that bleomycin be discontinued if a greater than 60% fall of DLCO from baseline is noted, in the absence of an alternative explanation for the change.38 Mortality from important bleomycin-induced fibrosis has been reported to be in excess of 50% in some series.39

Paclitaxel has gained wide acceptance as an effective treatment for breast, ovarian, and lung cancers as well as other less common tumors. To date, there are no reports of chronic lung toxicity from Taxol, but acute lung injury responding to steroid treatment has been reported following a single ambulatory dose of the drug.40 Additionally, acute pulmonary toxicity with associated NCPE has been reported during a high-dose trial of paclitaxel-cyclophosphamide-cisplatin and is likely due to paclitaxel.41 Docetaxel is associated with NCPE and pleural effusions related to increasing cumulative doses of drug. The vascular leak responsible for this responds to discontinuation of docetaxel.

Radiation-Induced Toxic Lung Injury

Similar to drugs, radiation to the lung can produce either an acute toxic injury or chronic fibrosis. Which injury is produced is determined by the dose rate of radiation delivery but also by other factors, such as pre-existing lung disease or concomitant treatments, or timing and extent of steroid treatment for acute lung injury.

As with many chemotherapeutic agents, acute radiation injury usually occurs from 2 weeks to 3 months after treatment. In most patients, the acute injury is limited to the irradiated field, but nonadjacent and even contralateral lung injury has been reported.9 While mild injury often resolves without treatment, more serious toxicity results in fibrosis 6 to 12 months following treatment. It is not necessary that this fibrosis be preceded by a definable syndrome of acute radiation pneumonitis.

The extent of both acute injury and eventual fibrosis can be related to the dose rate of radiation delivery, and perhaps to the volume of lung irradiated. Acute lung injury is frequently produced by total lung radiation in excess of 1,200 cGy delivered over a 2- to 4-day period, as is often done with total body irradiation used in preparation for bone marrow transplantation.42 Alternately, cumulative doses of 4,000 to 5,000 cGy delivered to only part of one lung over 5 to 6 weeks infrequently produce acute lung injury requiring steroid treatment. These higher cumulative doses, however, often result in late radiation-induced fibrosis.

Histopathologic evaluation of acute lung injury from radiation demonstrates injury to both vascular endothelial cells and alveolar lining cells, particularly type 2 pneumocytes. These changes are usually indistinguishable from drug-induced acute lung injury, suggesting the possibility that similar pathogenic mechanisms may underlie these toxicities. Radiation is generally believed to produce cytotoxicity via reactive oxygen species, capable of covalently binding to target molecules within cells and altering cellular detoxification mechanisms in a manner identical to reactive intermediates produced by chemotherapeutic agents. Reports of radiation-associated acute lung injury occurring outside the irradiated lung field, while not common, suggest the possibility that circulating factors can contribute to radiation injury. Definitive studies are lacking.

Radiation-induced acute lung injury produces clinical and radiographic findings similar to those described for acute chemotherapy-induced lung injury, except that the inflammation is usually sharply demarcated by the radiation field. The more acute-onset and severe toxicities should be treated with steroids in the manner described above for chemotherapy. Lesser injuries will often resolve completely and do not require steroid treatment. There is no consensus about objective criteria to define when steroids are required for acute lung injury from radiation, and clinical judgment is required.

Diffuse Interstitial Pneumonitis After Allogeneic Marrow Transplantation

Following total-body irradiation and chemotherapy in preparation for allogeneic marrow transplantation, it is important to consider the range of potential etiologies for acute lung injury. Numerous infections, diffuse alveolar hemorrhage from coagulopathy, and GVHD-associated immunologic injury are difficult to differentiate from treatment-related acute lung injury. BAL or lung biopsy is often required for diagnosis. Diffuse interstitial pneumonitis (DIP) syndrome occurring during the first 75 to 100 days after allogeneic marrow transplantation is usually considered not to be steroid responsive, perhaps because of the variety of mechanisms and pathogens which can produce similar radiographic and clinical findings. During this period of intense immunosuppression, it is critical that invasive diagnostic methods be employed quickly, to allow rapid institution of specific therapy. As the incidence of CMV pneumonitis declines with the greater use of CMV-negative or leukocyte-depleted blood products, improved drug treatment, and preemptive diagnosis by BAL, it may be useful to re-address the steroid responsiveness of DIP, which demonstrates substantial type 2 pneumocyte injury without inflammatory infiltrate on biopsy. Further research is required to identify the pathologic mechanisms and causes of this evolving syndrome.

Drug and Modality Interactions Associated with Lung Injury

Interactions at the pharmacokinetic or cellular level can alter the risk for and intensity of acute lung injury and its sequelae. When radiation to the lung is administered simultaneously with drugs capable of producing lung injury when used alone, the resulting lung injury can be severe and additive. Other agents which produce reactive intermediates but which are not independently associated with lung injury can augment radiation-associated lung injury. Simultaneous administration of bleomycin and pulmonary radiation illustrates this potential.43 Caution should always be used when any chemotherapeutic agent is administered simultaneously with lung radiation.

Pharmacokinetic drug interactions can increase the risk of chemotherapy-induced acute lung injury when combination regimens are used. The cyclophosphamide-cisplatin-BCNU regimen has frequently been employed with autologous marrow or blood stem cell transplantation to treat breast cancer. Administering cyclophosphamide and cisplatin immediately before BCNU produces a decreased rate of BCNU elimination and higher rates of acute lung injury than would be anticipated if an identical dose of BCNU were used alone.44

The BACOP regimen designed to treat lymphomas, frequently produces bleomycin lung injury at cumulative doses of less than 100 units.35 The mechanism of this rapid development of lung injury is unclear; but this illustrates the importance of always exercising caution when new combination chemotherapy regimens containing lung-toxic agents are tested.

Conclusion and Perspective

The mechanisms by which antineoplastic agents produce acute and chronic lung injury are likely shared by toxic syndromes in other organs, particularly the liver. The lung offers a unique setting to study these toxicities because they are easily observed radiographically and evaluated bronchoscopically and produce easily recognized symptoms.

Recognition of acute lung injury following conventional-dose treatments may allow effective treatment of it or discontinuation of a regimen which might otherwise produce morbid complications or death.

Extensive study of the mechanisms of toxic lung injury is important, since dose-intense radiation and chemotherapy regimens are frequently used in the treatment of cancer. Once mechanistic differences between antitumor effects and normal tissue injury are defined, great improvements in the therapeutic index of these regimens should be possible. Hematologic transplantation has removed marrow toxicity as a limitation to dose increases of anticancer agents. Lung injury looms next as an important barrier to further increases in dose. Clinical data suggest that further increases in tumor eradication may accompany relatively modest further dose increases in antineoplastic agents. Amelioration of lung injury may accompany a greater understanding of its mechanisms and greatly improve cancer treatment in the future.


Gleaves C A, Smith T F, Shuster R A, Pearson G R. Rapid detection of cytomegalovirus in MRC-5 cells inoculated with urine specimens by using low-speed centrifugation and monoclonal antibody to an early antigen. J Clin Microbiol. 1984;19:917–919. [PMC free article: PMC271213] [PubMed: 6088574]
van der Bij W, Torensma R, van Son W J. et al. Rapid immunodiagnosis of active cytomegalovirus infection by monoclonal antibody staining of blood leukocytes. J Med Virol. 1988;25:179–188. [PubMed: 2839611]
Gema G, Zipeto D, Parea M. et al. Monitoring of human cytomegalovirus infections and gancyclovir treatment in heart transplant recipients by determination of viremia, antigenemia, and DNAemia. J Infect Dis. 1991;164:488–498. [PubMed: 1651361]
Bowden R A, Sayers M, Floumoy N. et al. Cytomegalovirus immune globulin and seronegative blood products to prevent primary cytomegalovirus infection after marrow transplantation. N Engl J Med. 1986;314:1006–1010. [PubMed: 3007984]
Bowden R A, Slichter S J, Sayers M H. et al. Use of leukocyte-depleted platelets and leukocyte-poor red blood cells for prevention of primary cytomegalovirus infection after marrow transplant. Blood. 1991;78:227–234. [PubMed: 1648976]
Goodrich J M, Bowden R A, Fisher L. et al. Prevention of cytomegalovirus disease after allogeneic marrow transplant by gancyclovir prophylaxis. Ann Intern Med. 1993;118:173–178. [PubMed: 8380242]
Schmidt G M, Horak D A, Niland J C. et al. A randomized controlled trial of prophylactic gancyclovir for cytomegalovirus infection in recipients of allogeneic bone marrow transplants. N Engl J Med. 1991;324:1005–1011. [PubMed: 1848679]
Cooper J A D Jr, White D A, Matthay R A. Drug-induced pulmonary disease. Part I: Cytotoxic drugs. Am Rev Respir Dis. 1986;133:321. [PubMed: 3511808]
Bennett D E, Million R R, Ackerman L V. Bilateral radiation pneumonitis, a complication of the radiotherapy of bronchogenic carcinoma. Cancer. 1969;23:1001. [PubMed: 4976030]
Jones R B, Matthes S, Shpall E J. et al. Acute lung injury following treatment with high-dose cyclophosphamide, cisplatin, and carmustine: pharmacodynamic evaluation of carmustine. J Natl Cancer Inst. 1993;85:640–647. [PubMed: 8468721]
Smith G J W. The histopathology of pulmonary reactions to drugs. Clin Chest Med. 1990;11:95. [PubMed: 1691070]
Lombard C M, Churg A, Winokur S. Pulmonary veno-occlusive disease following therapy for malignant neoplasms. Chest. 1987;92:871. [PubMed: 3665603]
Sutegda T G, Apperley J F, Hughes J M B. et al. Pulmonary function after bone marrow transplantation for chronic myeloid leukemia. Thorax. 1988;43:163–169. [PMC free article: PMC461154] [PubMed: 3043753]
Robbins R A, Linder J, Stahl M G. et al. Diffuse alveolar hemorrhage in autologous bone marrow transplant recipients. Am J Med. 1989;87:511. [PubMed: 2816966]
Dickout W J, Chan C K, Hyland R H. et al. Prevention of acute pulmonary edema after bone marrow transplantation. Chest. 1987;92:303–309. [PubMed: 3301221]
Verwejj J, van der Burg M E L, Pinedo H M. Mitomycin C–induced hemolytic uremic syndrome: six case reports and review of the literature on renal, pulmonary, and cardiac side effects of the drug. Radiother Oncol. 1987;8:33–41. [PubMed: 3101142]
Korec S, Schein P S, Smith F P. et al. Treatment of cancer-associated hemolytic-uremic syndrome with staphylococcal protein A immunoperfusion. J Clin Oncol. 1986;4:210–215. [PubMed: 3944604]
Margolin K A, Rayner A A, Hawkins M J. et al. Interleukin-2 and lymphokine-activated killer cell therapy of solid tumors: analysis of toxicity and management guidelines. J Clin Oncol. 1989;7:486–498. [PubMed: 2647914]
Brandt S J, Peters W P, Atwater S K. et al. Effect of recombinant human granulocyte-macrophage colony–stimulating factor on hematopoietic reconstitution after high-dose chemotherapy and autologous bone marrow transplantation. N Engl J Med. 1988;318:869–876. [PubMed: 3281007]
Conkling P R, Chua C C, Nadler P. et al. Clinical trials with human tumor necrosis factor: in vivo and in vitro effects on human mononuclear phagocyte function. Cancer Res. 1988;48:5604. [PubMed: 3046744]
Epler G R. Bronchiolitis obliterans and airways obstruction associated with graft-versus-host disease. Clin Chest Med. 1988;9:551. [PubMed: 3069287]
Wyatt S E, Nunn P, Hows J M. et al. Airways obstruction associated with graft versus host disease after bone marrow transplantation. Thorax. 1984;39:887. [PMC free article: PMC459948] [PubMed: 6393415]
Ralph D D, Springmeyer S C, Sullivan K M. Rapidly progressive airflow obstruction in marrow transplant recipients: possible association between obliterative bronchiolitis and chronic graft-versus-host disease. Am Rev Respir Dis. 1984;129:641. [PubMed: 6370061]
Holland H K, Wingard J R, Beschorner W E. et al. Bronchiolitis obliterans in bone marrow transplantation and its relationship to chronic graft-vs-host disease and low serum IgG. Blood. 1988;72:621. [PubMed: 3042044]
Chatterji U C, Greene R F, Gallelli J F. Mechanism of hydrolysis of halogenated nitrosoureas. J Pharm Sci. 1978;67:1527–1532. [PubMed: 30837]
Wheeler G P, Bowdon B J, Grimsley J A. et al. Interrelationships of some chemical, physicochemical, and biological activities of several 1-(2-haloethyl)-1-nitrosoureas. Cancer Res. 1974;34:194–200. [PubMed: 4809457]
Aronin P A, Mahaley M S J, Rudnick S A. et al. Prediction of BCNU pulmonary toxicity in patients with malignant gliomas: an assessment of risk factors. N Engl J Med. 1980;303:183–188. [PubMed: 7383088]
O’Driscoll B R, Haselton P S, Taylor P M. et al. Active lung fibrosis up to 17 years after chemotherapy with carmustine (BCNU) in childhood. N Engl J Med. 1990;323:378–382. [PubMed: 2370889]
Peters W P, Shpall E J, Jones R B. et al. High-dose alkylating agents with bone marrow support as initial treatment for metastatic breast cancer. J Clin Oncol. 1988;6:1368–1376. [PubMed: 3047332]
Wheeler C, Strawderman M, Ayash L. et al. Prognostic factors for treatment outcome in autotransplantation of intermediate grade and high grade non-Hodgkin’s lymphoma with cyclophosphamide, carmustine and etoposide. J Clin Oncol. 1993;11:1085–1091. [PubMed: 8099120]
Vats T S, Langston C M. Pulmonary fibrosis associated with lomustine (CCNU): a case report. Cancer Treat Rep. 1982;66:1881–1882. [PubMed: 7127329]
Oliner H, Schwartz R, Rubio R Jr, Dameshek W. Interstitial pulmonary fibrosis following busulfan therapy. Am J Med. 1961;31:134. [PubMed: 13730733]
Ginsberg S J, Comis R L. The pulmonary toxicity of antineoplastic agents. Semin Oncol. 1982;9:34. [PubMed: 6176027]
Santos G W, Tutschka P J, Brookmeyer R. et al. Marrow transplantation for acute nonlymphocytic leukemia after treatment with busulfan and cyclophosphamide. N Engl J Med. 1983;309:1347–1353. [PubMed: 6355849]
Schein P S, DeVita V T, Hubbard S. et al. Bleomycin, Adriamycin, cyclophosphamide, vincristine and prednisone (BACOP) combination chemotherapy in the treatment of advanced diffuse histiocytic lymphoma. Ann Intern Med. 1976;85:417. [PubMed: 61732]
Lown J W, Sim S K. The mechanism of bleomycin-induced cleavage of DNA. Biochem Biophys Res Commun. 1977;77:1150–1157. [PubMed: 71149]
Goldiner P L, Carlon G C, Cvitkovic E. et al. Factors influencing postoperative morbidity and mortality in patients treated with bleomycin. BMJ. 1978;1:1664. [PMC free article: PMC1605498] [PubMed: 77697]
Comis R L, Kuppinger M S, Ginsberg S J. et al. The role of single-breath carbon monoxide diffusing capacity in monitoring pulmonary effects of bleomycin in germ cell tumor patients. Cancer Res. 1979;39:50–76. [PubMed: 91430]
White D A, Stover D E. Severe bleomycin-induced pneumonitis: clinical features and response to corticosteroids. Chest. 1984;86:723. [PubMed: 6207992]
Goldberg H L, Vannice S B. Pneumonitis related to treatment with paclitaxel. J Clin Oncol. 1995;13:534. [PubMed: 7844618]
Cagnoni P J, Stemmer S M, Matthes S. et al. Phase I trial of high-dose paclitaxel (Taxol), cyclophosphamide (CPA), and cisplatin (cDDP) with autologous hematopoietic progenitor cell (AHPC) rescue. Proc Am Soc Clin Oncol. 1995;14:477.
Deeg H J, Sullivan K M, Buckner C D. Marrow transplantation for acute nonlymphocytic leukemia in first remission: toxicity and long-term follow up of patients conditioned with single-dose or fractionated total body irradiation. Bone Marrow Transplant. 1986;1:151–157. [PubMed: 3332129]
Samuels M L, Johnson D E, Holoye P Y, Lanzotti V J. Large-dose bleomycin therapy and pulmonary toxicity: a possible role of prior radiotherapy. JAMA. 1976;235:1117. [PubMed: 55502]
Jones R B, Matthes S, Kemme D. et al. Cyclophosphamide, cisplatin, and carmustine: pharmacokinetics of carmustine following multiple alkylating agent interactions. Cancer Chemother Pharmacol. 1994;35:59–63. [PubMed: 7987978]
© 2000, BC Decker Inc.
Bookshelf ID: NBK20772


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...