Myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) are clonal hematological diseases with immunological, genetic, and epigenetic heterogenicity. MDS and AML both result in ineffective hematopoiesis and dysmorphism of cells, leading to cytopenias and their complications, including infections and bleeding (1). Additionally, the progressive impairment in growth and differentiation through the hypermethylation of the tumor suppressor gene, such as p15INK4B, can promote disease progression from MDS to AML (2–5).

Epigenetic abnormalities, such as hypermethylation of CpG island, are believed to play an essential role in leukemogenesis, which can lead to the silencing of tumor suppressor genes, which, in turn, drives the neoplastic transformation to AML (3, 6–9). MDS and AML tend to occur more often in people over the age of 60 years, with a male predominance (10, 11). Since most patients are of older age, the chances of having more comorbidities and being unfit increase, which, in turn, limits the ability to use curative therapies, such as stem cell transplant (1).

5-azacytidine, an analog of the pyrimidine nucleoside cytidine, has been studied to treat acute leukemia in the United States since the early 1970s. Clinical trials have primarily focused on patients with an illness resistant to conventional chemotherapy. The findings of these studies showed that 5-azacytidine is effective in treating AML. Following that, clinical trials looked at 5-azacytidine’s impact on other hematological diseases including MDS, non-hematological neoplasms, such as solid tumors, hereditary hematologic diseases, and hemoglobinopathies (thalassemia and sickle cell anemia). The Cancer and Leukemia Group B (CALGB) started a series of clinical trials with 5-azacytidine in MDS patients in 1984. These studies, and other supporting evidence, led to the US FDA approval of 5-azacytidine to treat MDS in May 2004. By blocking DNA methyltransferase, 5-azacytidine prevents the methylation of newly produced DNA (DNMT), restoring normal function to genes involved in differentiation and proliferation (12–16). Since then, azacytidine-based combinations have also been tested and developed in MDS, although with limited success, and in AML, where it led to a number of subsequent US-FDA approvals in patients ineligible for high-intensity therapy (10, 17–24).

Azacitidine is considered to have a relatively safe toxicity profile, with the majority being cytopenia and gastrointestinal symptoms. However, there has been growing evidence of lung toxicity complicating treatment with azacitidine. In our review, we aim to shed light on the manifestations and management of this rare but potentially fatal event and further add to the accumulative knowledge published by providing our experience in managing azacitidine-induced pneumonitis in an AML patient.

A 63-year-old male was referred to our institute with a diagnosis of acute leukemia after presenting with fatigue, weight loss, and bruising for two weeks. He was not known to have any past medical illnesses. He had a history of heavy smoking with a pack-year smoking index of 40 but stopped smoking almost five years prior to being diagnosed with leukemia. His initial investigations showed pancytopenia on his complete blood count (CBC) and around 50% blasts on the peripheral blood smear. Bone marrow aspirate and biopsy showed features suggestive of AML, with 56% myeloid blasts detected by flow cytometry. Based on the mentioned investigations along with cytogenetic and molecular testing, he was diagnosed with AML with myelodysplastic-related changes according to the World Health Organization (WHO) classification and stratified as having an adverse-risk disease based on the European Leukemia Network (ELN) risk stratification. The patient refused intensive chemotherapy and stem cell transplant, so he was started on cycle 1 of azacitidine 75 mg/m2 IV for seven days and venetoclax for 21 days. Before starting azacitidine and venetoclax, a high aspergillus galactomannan antigen (14.18) was incidentally found. He was asymptomatic.

Further investigations with imaging of the chest and paranasal sinuses with computed tomography (CT) were done; the chest CT showed multiple bilateral peri-broncho-vascular rounded central consolidative nodules and cavities representing an inflammatory/infectious process. He was started on Posaconazole. Bronchoscopy and bronchoalveolar lavage (BAL) were negative for infections, including aspergillus galactomannan antigen. However, he was continued on antifungal treatment for six weeks as he was labeled to have a probable invasive fungal infection. While on antifungal treatment, he was started on cycle 1 of azacitidine and venetoclax and achieved remission, and a repeat CT chest after completion of antifungal therapy showed resolution of the previous findings. He initially refused to receive further cycles; however, after three months, he had a relapse with 80% circulating blasts, so he was restarted on azacitidine and venetoclax. On day 2 of the second cycle of azacitidine and venetoclax, he developed a fever and tachypnea. Chest examination was normal, and chest x-ray did not reveal any significant abnormality and had neutropenia on CBC. A microbiological work-up was sent, and he was started on empirical antibiotics with meropenem as he was kept on prophylactic levofloxacin, according to the institute’s febrile neutropenia protocol. Despite empirical antibiotics, he continued to have fever and worsening symptoms. Antimicrobial therapy was upgraded with antifungal treatment, vancomycin, and later sulfamethoxazole-trimethoprim, yet there was no significant improvement. CT chest showed bilateral diffuse ground-glass opacities, smooth interlobular septal thickening, severe emphysematous changes, and small bilateral pleural effusions with no evidence of pulmonary embolism. The patient was kept on empirical antibiotics and completed seven days of azacitidine. On day 11, he became more tachypneic and hypoxic and was transferred to the intensive care unit (ICU) and was started on oxygen therapy with a high-flow nasal cannula. He continued empirical antimicrobial therapy despite all infectious work-ups returning negative (see Table 1 ). A bronchoscopy with bronchoalveolar lavage was done and was negative for malignant cells and infectious causes. BAL fluid was bloody with 24,500 x 106/L of RBCs. He was started on Methylprednisolone 40 mg IV BID for five days for possible drug-induced pneumonitis while continuing antibiotics. He improved clinically, was weaned off oxygen therapy, and was discharged from the hospital. He was re-admitted for the third cycle of azacitidine and venetoclax. On day 5 of cycle 3 of azacitidine, he developed a fever, shortness of breath, and hypoxia. Chest x-ray showed bilateral diffuse reticulonodular opacities. An infectious work-up was sent, and he was started on empirical antimicrobial therapy. CT chest showed diffuse bilateral ground glass densities with smooth interlobular septal thickening, centrilobular nodules, and right upper lobe ground glass opacity with focal subpleural infiltrate ( Figure 1 ). Azacitidine-induced pneumonitis was our top differential diagnosis as the symptoms reoccurred with rechallenging with azacitidine, and the Naranjo score was 8, which makes the diagnosis probable ( Table 2 ). Azacitidine was discontinued, and he was started on dexamethasone 4 mg IV BID for three days, then tapered down and changed to prednisone orally, which was stopped after six weeks with clinical and radiological improvement. Treatment of his AML was changed to cladribine, low-dose cytarabine with venetoclax with no similar lung toxicity. The patient achieved complete remission, however he refused further treatment, relapsed 4 months later and died within a month of his relapse.

Azacitidine-induced lung injury is uncommon, occurring in <0.1% of patients (37), but can have terminal outcomes. This makes it an important differential diagnosis when dealing with unexplained fever and respiratory symptoms after exposure to azacitidine. A vigilant evaluation and well-timed management are needed to establish the diagnosis, undo the injury, and prevent atrocious outcomes from happening (30). Several attempts to understand the underlying mechanism have been undertaken, but this remains an area that needs further tackling to establish the predictive factors before starting azacitidine and optimizing the management (38, 49).

RA: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. AA: Writing – review & editing. MAlm: Writing – review & editing. MAly: Writing – review & editing. MAlf: Supervision, Writing – review & editing.