Bone marrow (BM) aspiration plays an important role in hematologic malignancies diagnosis. Access and cost of diagnostic flow cytometry remains a problem in low and middle-income countries. In this context, morphological diagnosis by BM smear often represents the only means to rapidly diagnose our patients. Therefore, in this context obtaining the highest quality sample possible during the procedure is paramount. Despite being a well-known problem, evidence-based recommendations to improve BM aspirate quality are few, with studies evaluating factors associated with poor quality samples lacking.
Objectives
To determine factors associated with poor quality BM aspirates defined by an aspicular or hemodiluted sample in a hematology referral center.
Materials and methods
We conducted a retrospective study in our University Hospital and analyzed the BM smear samples stored in our center performed from October 2014 to December 2018. We collected and analyzed data based on diagnosis, age, gender, recent chemotherapy, and the variables of a complete blood count performed just before each BM aspiration. The quality of the BM smear was defined in any of the following: aspicular (without spicules), pauciaspicular (1-3 spicules), spicular (> 3 spicules), defining aspicular BM smear as non-diagnostic samples. Univariate analysis was performed looking for diferences between operators (in a 3-year residency program). In the other hand, in the multivariate analysis we seek to reveal the factors associated with obtaining hemodiluted (aspicular) bone marrow aspirate-smears.
Results
A total of 1,073 BM aspirates were evaluated. Hematology fellows performed 97% of BM aspirates; the remaining 3% were performed by attending physicians. In our analysis, 301 aspirates were aspicular, constituting 28.1% of the total number of aspirate smears. Most BM aspirates were performed for a diagnostic evaluation (66.3%) with the rest of the procedures for subsequent hematologic malignancy response assessments. In the univariate analysis, no differences were observed between operators. In a multivariate analysis the presence of an age >65 years (OR 3.1, 95% CI 2.3 to 4.1) and hemoglobin <6.0 g/dL (OR 2.7, 95% CI 1.4 to 4.5) at the time of the procedure were significantly associated to obtaining a non-diagnostic sample. Diagnosis, WBC count, platelet count, operator experience or other variables did not show statistical relevance. In our center, 18.81% of diagnostic patient samples that had acute leukemia were diagnosed without flow cytometry and through BM aspirate morphology alone. A second procedure to reach a diagnosis was necessary in 7.97% of the patients due to aspicular samples, obtaining a mean of diagnostic-treatment delay of 18.3 days (±5.7 days).
Conclusions
We found no differences between operators, emphasizing that there are other factors to consider in addition to a correct BM aspiration technique or operator experience. We believe this is crucial to recognize, especially in developing countries where morphological diagnosis remains the only means for the diagnosis or response evaluation of our patients. BM aspirate sample quality is multifactorial, being age and hemoglobin important factors. In addition, obtaining aspicular or hemodiluted samples represents not only a diagnostic challenge, but also delays the treatment of our patients.
Disclosures
Gomez-Almaguer: Celgene: Consultancy, Speakers Bureau; Takeda: Consultancy, Speakers Bureau; Teva: Consultancy, Speakers Bureau; Amgen: Consultancy, Speakers Bureau; Janssen: Consultancy, Speakers Bureau.
Anidulafungin compared with fluconazole for treatment of candidemia and other forms of invasive candidiasis caused by Candida albicans: a multivariate analysis of factors associated with improved outcome
