DIAGNOSTIC VALUE OF PERIODIC ACID-SCHIFF POSITIVITY OF BONE MARROW ERYTHROBLASTS IN MYELODYSPLASTIC SYNDROMES
(Abstract release date: 05/19/16)
EHA Library. Invernizzi R. 06/09/16; 134822; PB1922
Prof. Rosangela Invernizzi
Contributions
Contributions
Abstract
Abstract: PB1922
Type: Publication Only
Background
The revised 2008 WHO classification confirmed minimal morphological criteria of myelodysplastic syndrome (MDS) diagnosis: at least 10% of bone marrow (BM) cells of at least one hematopoietic cell lineage must show unequivocal dysplasia to be considered as dysplastic. Morphological abnormalities of erythroid cells include cytoplasmic Periodic acid-Schiff (PAS) positivity, but the PAS positivity diagnostic power is not yet fully clear. Neither was this morphological aspect taken into account by the “Rete Ematologica Lombarda (REL)” clinical network that proposed a structured and reproducible approach for the precise recognition of BM dysplasia (Della Porta et al. Leukemia 2015; 29:66-75).
Aims
Since dysplastic alterations of BM precursors and of peripheral blood cells are still fundamental for diagnostic classifications, the aims of our study were to evaluate the diagnostic value of erythroblast PAS positivity in MDS and to investigate a possible correlation between PAS positivity levels and other morphological and clinical features.
Methods
We retrospectively examined the results of the cytochemical PAS staining for glycogen in BM smears from 120 MDS patients, 105 patients with non-clonal cytopenia and 49 healthy subjects. By counting 100 nucleated cells for the erythroid lineage and classifying them for their degree of PAS reactivity, we developed a PAS score for MDS identification on the basis of a ROC curve analysis.
Results
PAS positive erythroblasts were observed in 74 (62%) MDS patients, 43 (41%) patients with non-clonal cytopenia and 12 (24%) healthy subjects (p<0.0001). In MDS, both positivity rates (mean 3.9%, range 0-31%) and scores (mean 5.8, range 0-49) were significantly higher than those in normal and pathologic controls (p=0.0001 and p<0.0001, respectively). Positive erythroblasts showed diffuse or granular cytoplasmic reactivity. In MDS, no significant relationship was detected between erythroblast PAS positivity rate or score and dyserythropoiesis grading, multilineage dysplasia or excess blasts, whereas there was a significant inverse correlation between PAS score values and percentages of BM erythroblasts (p=0.041) and between PAS score values and percentages of ring sideroblasts (p=0.012). Anemic patients showed higher score values than non-anemic subjects (p=0.016). PAS positivity was unrelated to karyotype abnormalities. A ROC curve analysis allowed us to identify a PAS score value >0 (AUC=0.642, p=0.0001) and a PAS positive erythroblast percentage >0 (AUC=0.629, p=0.0002) as optimal cut-off to discriminate MDS patients from controls with sensitivity and specificity ranging from 62% to 66%. Positive and negative predictive values of PAS positivity were 74% and 58%, respectively. Considering the most discriminant morphological features for dyserythropoiesis, the weight of both PAS positivity rate and score in the recognition of BM dysplasia was lower than that of ring sideroblasts and megaloblastosis, but higher than that of defective hemoglobinisation, nuclear lobulation, multinuclearity, cytoplasmic fraying, pyknosis, and internuclear bridges.
Conclusion
The evaluation of BM erythroblast PAS positivity may be useful in the work-up of patients with suspected MDS, especially if there is only unilineage dysplasia without excess blasts or ring sideroblasts. This parameter should be included in the diagnostic morphological panel to be used for a correct application of the WHO classification.
Session topic: E-poster
Keyword(s): Bone Marrow, Erythroid cells, Myelodysplasia
Type: Publication Only
Background
The revised 2008 WHO classification confirmed minimal morphological criteria of myelodysplastic syndrome (MDS) diagnosis: at least 10% of bone marrow (BM) cells of at least one hematopoietic cell lineage must show unequivocal dysplasia to be considered as dysplastic. Morphological abnormalities of erythroid cells include cytoplasmic Periodic acid-Schiff (PAS) positivity, but the PAS positivity diagnostic power is not yet fully clear. Neither was this morphological aspect taken into account by the “Rete Ematologica Lombarda (REL)” clinical network that proposed a structured and reproducible approach for the precise recognition of BM dysplasia (Della Porta et al. Leukemia 2015; 29:66-75).
Aims
Since dysplastic alterations of BM precursors and of peripheral blood cells are still fundamental for diagnostic classifications, the aims of our study were to evaluate the diagnostic value of erythroblast PAS positivity in MDS and to investigate a possible correlation between PAS positivity levels and other morphological and clinical features.
Methods
We retrospectively examined the results of the cytochemical PAS staining for glycogen in BM smears from 120 MDS patients, 105 patients with non-clonal cytopenia and 49 healthy subjects. By counting 100 nucleated cells for the erythroid lineage and classifying them for their degree of PAS reactivity, we developed a PAS score for MDS identification on the basis of a ROC curve analysis.
Results
PAS positive erythroblasts were observed in 74 (62%) MDS patients, 43 (41%) patients with non-clonal cytopenia and 12 (24%) healthy subjects (p<0.0001). In MDS, both positivity rates (mean 3.9%, range 0-31%) and scores (mean 5.8, range 0-49) were significantly higher than those in normal and pathologic controls (p=0.0001 and p<0.0001, respectively). Positive erythroblasts showed diffuse or granular cytoplasmic reactivity. In MDS, no significant relationship was detected between erythroblast PAS positivity rate or score and dyserythropoiesis grading, multilineage dysplasia or excess blasts, whereas there was a significant inverse correlation between PAS score values and percentages of BM erythroblasts (p=0.041) and between PAS score values and percentages of ring sideroblasts (p=0.012). Anemic patients showed higher score values than non-anemic subjects (p=0.016). PAS positivity was unrelated to karyotype abnormalities. A ROC curve analysis allowed us to identify a PAS score value >0 (AUC=0.642, p=0.0001) and a PAS positive erythroblast percentage >0 (AUC=0.629, p=0.0002) as optimal cut-off to discriminate MDS patients from controls with sensitivity and specificity ranging from 62% to 66%. Positive and negative predictive values of PAS positivity were 74% and 58%, respectively. Considering the most discriminant morphological features for dyserythropoiesis, the weight of both PAS positivity rate and score in the recognition of BM dysplasia was lower than that of ring sideroblasts and megaloblastosis, but higher than that of defective hemoglobinisation, nuclear lobulation, multinuclearity, cytoplasmic fraying, pyknosis, and internuclear bridges.
Conclusion
The evaluation of BM erythroblast PAS positivity may be useful in the work-up of patients with suspected MDS, especially if there is only unilineage dysplasia without excess blasts or ring sideroblasts. This parameter should be included in the diagnostic morphological panel to be used for a correct application of the WHO classification.
Session topic: E-poster
Keyword(s): Bone Marrow, Erythroid cells, Myelodysplasia
Abstract: PB1922
Type: Publication Only
Background
The revised 2008 WHO classification confirmed minimal morphological criteria of myelodysplastic syndrome (MDS) diagnosis: at least 10% of bone marrow (BM) cells of at least one hematopoietic cell lineage must show unequivocal dysplasia to be considered as dysplastic. Morphological abnormalities of erythroid cells include cytoplasmic Periodic acid-Schiff (PAS) positivity, but the PAS positivity diagnostic power is not yet fully clear. Neither was this morphological aspect taken into account by the “Rete Ematologica Lombarda (REL)” clinical network that proposed a structured and reproducible approach for the precise recognition of BM dysplasia (Della Porta et al. Leukemia 2015; 29:66-75).
Aims
Since dysplastic alterations of BM precursors and of peripheral blood cells are still fundamental for diagnostic classifications, the aims of our study were to evaluate the diagnostic value of erythroblast PAS positivity in MDS and to investigate a possible correlation between PAS positivity levels and other morphological and clinical features.
Methods
We retrospectively examined the results of the cytochemical PAS staining for glycogen in BM smears from 120 MDS patients, 105 patients with non-clonal cytopenia and 49 healthy subjects. By counting 100 nucleated cells for the erythroid lineage and classifying them for their degree of PAS reactivity, we developed a PAS score for MDS identification on the basis of a ROC curve analysis.
Results
PAS positive erythroblasts were observed in 74 (62%) MDS patients, 43 (41%) patients with non-clonal cytopenia and 12 (24%) healthy subjects (p<0.0001). In MDS, both positivity rates (mean 3.9%, range 0-31%) and scores (mean 5.8, range 0-49) were significantly higher than those in normal and pathologic controls (p=0.0001 and p<0.0001, respectively). Positive erythroblasts showed diffuse or granular cytoplasmic reactivity. In MDS, no significant relationship was detected between erythroblast PAS positivity rate or score and dyserythropoiesis grading, multilineage dysplasia or excess blasts, whereas there was a significant inverse correlation between PAS score values and percentages of BM erythroblasts (p=0.041) and between PAS score values and percentages of ring sideroblasts (p=0.012). Anemic patients showed higher score values than non-anemic subjects (p=0.016). PAS positivity was unrelated to karyotype abnormalities. A ROC curve analysis allowed us to identify a PAS score value >0 (AUC=0.642, p=0.0001) and a PAS positive erythroblast percentage >0 (AUC=0.629, p=0.0002) as optimal cut-off to discriminate MDS patients from controls with sensitivity and specificity ranging from 62% to 66%. Positive and negative predictive values of PAS positivity were 74% and 58%, respectively. Considering the most discriminant morphological features for dyserythropoiesis, the weight of both PAS positivity rate and score in the recognition of BM dysplasia was lower than that of ring sideroblasts and megaloblastosis, but higher than that of defective hemoglobinisation, nuclear lobulation, multinuclearity, cytoplasmic fraying, pyknosis, and internuclear bridges.
Conclusion
The evaluation of BM erythroblast PAS positivity may be useful in the work-up of patients with suspected MDS, especially if there is only unilineage dysplasia without excess blasts or ring sideroblasts. This parameter should be included in the diagnostic morphological panel to be used for a correct application of the WHO classification.
Session topic: E-poster
Keyword(s): Bone Marrow, Erythroid cells, Myelodysplasia
Type: Publication Only
Background
The revised 2008 WHO classification confirmed minimal morphological criteria of myelodysplastic syndrome (MDS) diagnosis: at least 10% of bone marrow (BM) cells of at least one hematopoietic cell lineage must show unequivocal dysplasia to be considered as dysplastic. Morphological abnormalities of erythroid cells include cytoplasmic Periodic acid-Schiff (PAS) positivity, but the PAS positivity diagnostic power is not yet fully clear. Neither was this morphological aspect taken into account by the “Rete Ematologica Lombarda (REL)” clinical network that proposed a structured and reproducible approach for the precise recognition of BM dysplasia (Della Porta et al. Leukemia 2015; 29:66-75).
Aims
Since dysplastic alterations of BM precursors and of peripheral blood cells are still fundamental for diagnostic classifications, the aims of our study were to evaluate the diagnostic value of erythroblast PAS positivity in MDS and to investigate a possible correlation between PAS positivity levels and other morphological and clinical features.
Methods
We retrospectively examined the results of the cytochemical PAS staining for glycogen in BM smears from 120 MDS patients, 105 patients with non-clonal cytopenia and 49 healthy subjects. By counting 100 nucleated cells for the erythroid lineage and classifying them for their degree of PAS reactivity, we developed a PAS score for MDS identification on the basis of a ROC curve analysis.
Results
PAS positive erythroblasts were observed in 74 (62%) MDS patients, 43 (41%) patients with non-clonal cytopenia and 12 (24%) healthy subjects (p<0.0001). In MDS, both positivity rates (mean 3.9%, range 0-31%) and scores (mean 5.8, range 0-49) were significantly higher than those in normal and pathologic controls (p=0.0001 and p<0.0001, respectively). Positive erythroblasts showed diffuse or granular cytoplasmic reactivity. In MDS, no significant relationship was detected between erythroblast PAS positivity rate or score and dyserythropoiesis grading, multilineage dysplasia or excess blasts, whereas there was a significant inverse correlation between PAS score values and percentages of BM erythroblasts (p=0.041) and between PAS score values and percentages of ring sideroblasts (p=0.012). Anemic patients showed higher score values than non-anemic subjects (p=0.016). PAS positivity was unrelated to karyotype abnormalities. A ROC curve analysis allowed us to identify a PAS score value >0 (AUC=0.642, p=0.0001) and a PAS positive erythroblast percentage >0 (AUC=0.629, p=0.0002) as optimal cut-off to discriminate MDS patients from controls with sensitivity and specificity ranging from 62% to 66%. Positive and negative predictive values of PAS positivity were 74% and 58%, respectively. Considering the most discriminant morphological features for dyserythropoiesis, the weight of both PAS positivity rate and score in the recognition of BM dysplasia was lower than that of ring sideroblasts and megaloblastosis, but higher than that of defective hemoglobinisation, nuclear lobulation, multinuclearity, cytoplasmic fraying, pyknosis, and internuclear bridges.
Conclusion
The evaluation of BM erythroblast PAS positivity may be useful in the work-up of patients with suspected MDS, especially if there is only unilineage dysplasia without excess blasts or ring sideroblasts. This parameter should be included in the diagnostic morphological panel to be used for a correct application of the WHO classification.
Session topic: E-poster
Keyword(s): Bone Marrow, Erythroid cells, Myelodysplasia
{{ help_message }}
{{filter}}