Background: In patients with chronic myeloid leukemia (CML), long-term use of tyrosine kinase inhibitors (TKIs) is associated with hematological and non-hematological complications. To investigate the pulmonary complications (pulmonary edema and pulmonary arterial hypertension [PAH]) in CML patients receiving TKIs, we conducted a systematic review and meta-analysis of clinical trials and observational studies.
Methods: We searched MEDLINE, EMBASE and the Cochrane library to May 2015. The studies were included if they enrolled CML patients, used at least one of imatinib, nilotinib, dasatinib, bosutinib and ponatinib, and reported pulmonary effusion or PAH as one of their outcomes. We performed a single proportion meta-analysis for the cumulative incidence of each outcome. Subgroup analysis was carried out based on type of TKI, phase of disease, dosage and line of treatment.
Results: Thirty-nine studies, involving 6562 patients were included in the meta-analysis. The overall cumulative incidence of pleural effusion was 5% (95% confidence interval [CI]; 4.2-5.8, I2=95.9%). Subgroup analysis according to type of TKI revealed that dasatinib was associated with highest incidence of pleural effusion, 26.2% (95%CI; 21.4-31.1) compared to imatinib (0.5%), nilotinib (0.1%), bosutinib (2.1%) and ponatinib (0.3%), Figure 1. Subgroup analysis demonstrated that accelerated/blastic phase, non-first line treatment and higher dose of dasatinib were associated with a higher risk of pleural effusion. There were 7 cohorts reporting PAH. The overall cumulative incidence of pleural effusion was 0.2 per 100 patient year (95% CI; -0.3, I2 =0%) for all TKIs, Figure 2.
Conclusions: Pleural effusion is a frequently found pulmonary complication among CML patients receiving dasatinib. However, PAH is uncommon for TKI treated patients. We recommend a low threshold for performing a chest x-ray in symptomatic patients on dasatinib. Because of the low incidence of PAH, we cannot recommend routine screening for PAH in the absence of highly suspicious symptoms.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.