Fax: (713) 745-5372
Acute pulmonary failure during remission induction chemotherapy in adults with acute myeloid leukemia or high-risk myelodysplastic syndrome
Article first published online: 27 OCT 2009
Copyright © 2010 American Cancer Society
Volume 116, Issue 1, pages 93–97, 1 January 2010
How to Cite
Al Ameri, A., Koller, C., Kantarjian, H., Ravandi, F., Verstovsek, S., Borthakur, G., Pierce, S. and Mattiuzzi, G. (2010), Acute pulmonary failure during remission induction chemotherapy in adults with acute myeloid leukemia or high-risk myelodysplastic syndrome. Cancer, 116: 93–97. doi: 10.1002/cncr.24711
- Issue published online: 11 JAN 2010
- Article first published online: 27 OCT 2009
- Manuscript Accepted: 27 APR 2009
- Manuscript Revised: 25 MAR 2009
- Manuscript Received: 26 JAN 2009
- acute pulmonary failure;
- acute leukemia;
Acute pulmonary failure during remission induction therapy is a serious complication in patients with acute myeloid leukemia (AML). To the authors' knowledge, the course and prognosis of such patients is not well known.
A total of 1541 patients referred for remission induction chemotherapy of AML or high-risk myelodysplastic syndrome were retrospectively reviewed.
A total of 120 (8%) patients developed acute pulmonary failure within 2 weeks of the initiation of chemotherapy; 87 of these patients (73%) died during remission induction, whereas 17 (14%) achieved a complete response. The median survival among the 120 patients with early acute pulmonary failure was 3 weeks. Predictive factors for the development of early acute pulmonary failure by multivariate analysis were: male sex (P = .00038), acute promyelocytic leukemia (P = .00003), poor performance status (P = .001), lung infiltrates at diagnosis (P = .000001), and increased creatinine (P = .000005). Patients who had 0 to 1, 2, 3, or 4 to 5 adverse factors were found to have estimated predictive incidences of acute pulmonary failure of 3%, 13%, 23%, and 34%, respectively.
Preventive approaches at the start of induction therapy in patients at high risk of pulmonary failure may improve the outcome of these patients. Cancer 2010. © 2010 American Cancer Society.
Modern intensive chemotherapy in patients with acute myeloid leukemia (AML) is associated with complete response (CR) rates of 40% to 80% and cure rates of 10% to 60%, depending on several prognostic factors, including the patient age and performance status, the leukemia karyotype, the molecular abnormalities in patients with a normal karyotype, and the associated comorbid conditions.1-4 Causes of death in patients with AML are most commonly related to leukemia persistence or recurrence, or myelosuppression-associated complications such as infections, bleeding, or multiorgan failure.5 During remission induction therapy, the induction mortality can vary from <5% to 50% as a result of tolerance to chemotherapy, which is primarily related to patient age, performance status, and presenting complications (eg, infection) or organ dysfunctions. Acute pulmonary failure during remission induction therapy is an uncommon but very serious complication. Pulmonary failure may result from 1 or more causes, including pulmonary hemorrhage, capillary leak as a result of tumor lysis and/or fluid overload, infections, and sepsis. Several factors predispose to the development of early pulmonary failure and the need for pulmonary support. Factors predicting for early pulmonary failure during remission induction therapy are not defined, and to our knowledge little is known regarding the course and prognosis of patients who develop early acute pulmonary failure during remission induction therapy.
The purpose of this study was to analyze the course of adults with AML who develop early pulmonary failure during remission induction therapy and to assess the prognostic factors that may predict for the development of such pulmonary failure. This may ultimately help in developing preventive strategies for pulmonary failure in such high-risk patients.
MATERIALS AND METHODS
Adults with newly diagnosed AML or high-risk myelodysplastic syndrome (MDS) referred to the leukemia service at The M. D. Anderson Cancer Center from 1998 to 2005 for remission induction intensive chemotherapy were analyzed. A diagnosis of AML referred to the presence of ≥20% blasts. High-risk MDS referred to the presence of >10% blasts. Only patients with MDS who received intensive chemotherapy were analyzed. Informed consent was obtained according to institutional guidelines. We focused on patients who developed acute pulmonary failure. Because the definition of acute pulmonary failure may vary among investigators and institutions, we defined it in its more serious form, that is, patients who required mechanical ventilatory support in the intensive care unit within 2 weeks of the initiation of remission induction therapy. Pretreatment patient and leukemic cell characteristics as well as follow-up studies and the ultimate course of patients were collected. Complete remission was defined by standard AML criteria.6 Survival was calculated by the Kaplan-Meier method. Survival curves were compared by the log-rank test. A previous prognostic model for early risk of mortality7 within 6 weeks of induction therapy had defined 8 independent poor prognostic factors (age at diagnosis and the following baseline parameters: performance status, bilirubin, creatinine, hemoglobin, albumin, fibrinogen, and absolute neutrophil counts). In the subsequent analyses, we evaluated the composite early risk of mortality for its predictive value for acute pulmonary failure. Univariate and multivariate analyses were conducted to identify potential predictive factors for the development of early acute pulmonary failure. Multivariate analysis used the logistic regression method.8
The characteristics of the study group are shown in Table 1. Among 1541 patients reviewed, 120 (8%) patients developed acute pulmonary failure requiring ventilatory support within 2 weeks of the initiation of remission induction therapy. The characteristics of the latter group are also detailed in Table 1.
|Parameter||Category||No. (%)||Pulmonary Failure (%)||P|
|Age, y||<40||196 (13)||8 (4)||.103|
|40-60||570 (37)||45 (8)|
|>60||775 (50)||67 (9)|
|Sex||Female||646 (42)||61 (9)||.039|
|Male||895 (58)||59 (7)|
|Diagnosis||AML||1390 (90)||113 (8)||.129|
|MDS||151 (10)||7 (5)|
|FAB morphology||M3||109 (7)||20 (18)||<.001|
|M4-M5||314 (20)||32 (10)|
|Others||1118 (73)||68 (6)|
|Karyotype||T(15;17)||97 (6)||17 (18)||<.001|
|T(8;21), inv 16||97 (6)||5 (5)|
|Normal||582 (38)||29 (5)|
|5 or 7 abnormalities||351 (23)||27 (8)|
|Other abnormalities||364 (24)||31 (9)|
|IM/not done||50 (3)||11 (22)|
|Zubrod performance status||0-1||1154 (75)||35 (3)||<.001|
|2-4||387 (25)||85 (22)|
|ERM||<0.1||558 (36)||21 (4)||<.001|
|0.1-0.2||311 (20)||10 (3)|
|Presence of infection||FUO||161 (10)||19 (12)||<.001|
|Pulmonary infiltrates||240 (16)||57 (24)|
|Other infections||102 (7)||4 (4)|
|Initial admission to ICU||Yes||125 (8)||45 (36)||<.001|
|No||1416 (92)||75 (5)|
|Hemoglobin, g/dL||<8||765 (50)||66 (9)||.297|
|8-10||550 (36)||35 (6)|
|>10||226 (15)||19 (8)|
|Platelet count, ×109/L||<20||247 (16)||31 (13)||.001|
|20-49||554 (36)||46 (8)|
|50-99||439 (28)||33 (8)|
|WBC count, K/μL||≤30||1209 (78)||71 (6)||<.001|
|>30||332 (22)||49 (15)|
|Absolute neutrophils, ×109/L||<0.5||612 (40)||29 (5)||<.001|
|0.5 < 1.0||217 (14)||15 (7)|
|Creatinine, mg/dL||<1.3||1266 (82)||75 (6)||<.001|
|1.3-2.0||224 (15)||32 (14)|
|>2.0||51 (3)||13 (25)|
|Bilirubin, mg/dL||<1.3||1393 (90)||91 (7)||<.001|
|1.3-2.0||116 (8)||22 (19)|
|>2.0||32 (2)||7 (22)|
Overall 843 (55%) patients achieved CR, 458 (30%) patients had resistant disease, and 212 (14%) patients died within 6 weeks of remission induction. The CR duration and survival of the total study group are shown in Figure 1.
A total of 120 patients (8%) had early acute pulmonary failure. Among them, 87 (73%) died during remission induction, and 17 (14%) achieved CR.
Their median CR duration was 8 months (Fig. 2). The median survival among patients who developed pulmonary failure was 3 weeks (Fig. 2), and the median survival of patients with pulmonary failure who achieved CR was 15 months (Fig. 2).
Fifty-two of the 120 patients (43%) with acute pulmonary failure also developed renal failure and were referred for dialysis.
We next analyzed the prognostic factors predictive for development of acute pulmonary failure. These are detailed in Table 2. Using univariate analysis, predictive factors for acute pulmonary failure included: male sex, acute promyelocytic leukemia (French-American-British [FAB] M3 morphology), poor performance status, early risk of mortality >0.2, lung infiltrates at presentation, thrombocytopenia, severe leukocytosis, neutropenia, elevated creatinine, and elevated bilirubin (Table 1). Using multivariate analysis the independent poor prognostic factors were: male sex (P = .00038), FAB classification M3 morphology (P = .00003), poor performance status (P < 10−6), lung infiltrates at diagnosis (P = .000001), and increased creatinine (P = .000005) (Table 2).
|Poor Zubrod performance status||−1.6575||0.23432||50.03552||.000000|
|Presence of pulmonary infiltrates||−1.1586||0.23051||25.26273||.000001|
Scoring all patients by the number of these adverse factors, patients who had none or 1, 2, 3, or 4 to 5 adverse factors were found to have estimated predictive incidences of acute pulmonary failure of 3%, 13%, 23%, and 34%, respectively (Table 3).
|No. of Adverse Factors||No. of Patients||No. of Pulmonary Failures (%)|
Acute pulmonary failure is an uncommon but very serious complication of remission induction therapy.9-18 In the current study of 1541 patients with newly diagnosed AML undergoing remission induction therapy, the incidence of acute pulmonary failure requiring ventilatory support was 8%. Using multivariate analysis, the independent prognostic factors for the development of acute pulmonary failure were elevated creatinine, male sex, FAB M3 morphology, poor performance status, and lung infiltrates at the time of diagnosis. Patients with none or 1, 2, 3, or 4 to 5 adverse factors had respective incidences of pulmonary failure of 3%, 13%, 23%, and 34%, respectively. Identifying such high-risk patients early on may allow the implementation of preventive strategies that reduce this risk of acute pulmonary failure and therefore improve the rate of tolerance of such therapies, CR rate, and, eventually, long-term outcome. The causes of pulmonary failure are multiple and often overlapping.9-18 Clinical experience also indicates that it is often difficult to distinguish the single most significant cause of pulmonary failure. Many patients with pulmonary hemorrhage may not develop hemoptysis. Conversely, hemoptysis may be noted in patients with pneumonia, pulmonary capillary leak syndrome, or fluid overload. Therefore, composite supportive measures in these patients may be important. These could include the use of intensive platelet support regardless of evidence of bleeding in such high-risk patients, restriction of fluids or even maneuvers to hypohydrate such patients (eg, diuresis until the creatinine level increases to 1.5× the pretreatment level), continuous venovenous dialysis, the use of tranexamic acid (aminocaproic acid), leukapheresis in patients with severe leukocytosis, or other measures that may reduce the incidence of bleeding.
Some of the independent adverse prognostic factors for early pulmonary failure such as elevated creatinine or bilirubin (indicative of general worse condition), poor performance status, and lung infiltrates were expected. Other factors that might have been suspected to contribute to pulmonary failure (eg, severe leukocytosis, thrombocytopenia) were not selected by multivariate analysis and were found to be strongly correlated with other selected factors (eg, leukocytosis with pulmonary infiltrates, thrombocytopenia with FAB M3 morphology). Although female sex was identified as an independent poor prognostic factor, male sex emerged as a poor prognostic variable because of its strong association with age (median age, 56 years for women vs 64 years for men). Other factors not captured in this analysis may have contributed further to the association of male sex with early pulmonary failure, such as smoking history or the existence of chronic obstructive lung disease.
As expected, survival was significantly worse among patients experiencing early pulmonary failure. Among the 120 such patients, the median survival was 3 weeks. Only 17 (14%) patients achieved CR; their median CR duration was 36 weeks, and their median survival was 66 weeks, with an estimated 1-year survival rate of 63% (Figs. 1 and 2).
In summary, early acute pulmonary failure remains a serious complication of remission induction therapy, which is often complicated by subsequent other organ failure and death. Preventive approaches at the initiation of induction therapy in patients at high risk of pulmonary failure may improve the outcome of these patients.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 8The Analysis of Binary Data Monographs on Statistics and Applied Probability 32. London, UK: Chapman & Hall; 1970..
- 10Fatal haemoptysis in pulmonary filamentous mycosis: an underevaluated cause of death in patients with acute leukaemia in haematological complete remission. A retrospective study and review of the literature. Gimema Infection Program (Gruppo Italiano Malattie Ematologiche dell'Adulto). Br J Haematol. 1995; 89: 500-505., , , et al.