Myelodysplastic syndromes (MDS) are considered one of the major precursor myeloid neoplasms. It is defined as a group of clonal hematopoietic stem cell neoplasms characterized with bone marrow failure, manifested with peripheral cytopenia, morphologic dysplasia involving ≥ 1 lineage hematopoietic progenitors, variably increased blasts (<20%), and a great risk of leukemic transformation [1,2]. Given its heterogeneous clinical and histologic presentations and the varieties of morphologic mimickers under reactive or autoimmune situations, it clinically encounters diagnostic challenges if there is no clonal cytogenetic abnormality [3,4]. Moreover, it is always conflicting in concern about the degree of morphologic dysplasia or cytopenia.

The 2008 WHO classification has integrated laboratory data, morphology, with cytogenetic findings to sub classify MDS. The updated 2016 revision of World Health Organization (WHO) has modified the sub classification of MDS based on novel molecular data Table 1 [5,6].

Table 1: The Comparison of the Terms Used in sub-classification of MDS in 2008 and 2016 WHO System, No longer shown in the 2016 revision of WHO classification.

In light of the new criteria in AML, the new WHO will also include a subset of patients who were previously called acute erythroid/ myeloid leukemia with the absolute myoblast count <20% of the total cellularity, regardless the percentage of erythroid precursors (Table 1) [5]. The Comparison of the Terms Used in Sub classification of MDS in 2008 and 2016 WHO System [1,5].

Cytogenetic study including conventional karyotyping and fluorescence in situ hybridization (FISH) is one of the common accessary diagnostic tools. Of note, approximately 50% do novo MDS and 75% secondary or therapy related MDS harbor cytogenetic aberrations, frequently associated with del(7q), monosomy 7, del(5q), monosomy 5, and trisomy 8.

Among them, MDS with isolated de(5q) is considered a unique, independent subtype with characteristic megakaryocytic anomaly, macrocytic anemia and erythroid hypoplasia; however, there is no MDS specific cytogenetic abnormality.

Nevertheless, these cytogenetic changes are taken into count in international prognostic scoring system (IPSS) and revised IPSS in prediction of the patients’ clinical outcomes (Table 2a and 2b) [7,8]. IPSS and R-IPSS have been widely accepted in clinical practice for the last decades until a recent multicenter study came out with new results.

Table 2a: Prognostic Score Vales in IPSS and R-IPSS, (≥3 abnormalities) or chromosome 7 anomalies; intermediate=other abnormalities.

Table 2b: Risk Group and Clinical Outcome in IPSS and R-IPSS.

The MDS study collected from 7,212 patients with de novo MDS (untreated) emphasized the risk of transformation and mortality has been changed over time: Hazard scores are reduced in high risk MDS while stable in low risk MDS when analyzed at 3.5 years [9].

The results lead to a new proposal of using cut off of 3.5 points in RIPSS to separate low or high risk groups on the purpose of treatment management. The other risk-stratification systems, e.g., WHO classification–based prognostic scoring system (WPSS), and MD Anderson MDS scoring system have been validated and adopted as needed [10-13].

Emerging next generation sequencing (NGS) technique makes it feasible to identify recurrent somatic mutations in cancer cells and also highlights frequency and importance of these somatic mutations in MDS. Up to 80-90% of MDS patients harbor 1 or more than 1 recurring somatic mutations in epigenetic, signaling, tumor suppressor, or cell cycle pathways, with most common ones including SF3B1, TET2, ASXL1, DNMT3A, EZH2, TP53, SRSF2, RUNX1, ETV6, U2A1 and RUNX1. SF3B1 mutations are found to be associated with ring–sideroblast phenotype in MDS e.g. MDS- unilineage or multilineage dysplasias with RS as well as MDS/MPN with RS and thrombocytosis [5] of prognostic importance, the patients who harbored the 5 key gene mutations including AXSL1, ETV6, TP53, RUNX1 and EZH2 showed short median overall survival when compared with the MDS patients in the same risk group: very low risk, low risk and intermediate, according to R-IPSS [14].

TP53 mutation or overexpression of p53 protein is particularly a negative prognostic predictor [14-17]. The higher variant allele frequency (VAF) of TP53 mutations is associated with the shorter overall survival time [14]. Mutated TP53 status in MDS patients is also associated with a poor response to a long run hypo methylation therapy [18]. MDS phenotyping by flow cytometry is proposed in Europe but have not yet been widely implicated in the USA. Accumulating data and experience might be helpful to make it mature and applicable in daily practice [19-20].

Precursor lesions that might be associated with or lead to MDS included clonal hematopoiesis of indeterminate potential (CHIP), [21,22] idiopathic cytopenias of undetermined significance (ICUS)[23] clonal cytopenias of undetermined significance (CCUS) [24]. Different from the presumable or de novo MDS with clinical presentation or laboratory changes, CHIP is age related hematopoietic clone and is driven by mutations occurred frequently in myeloid neoplasm, commonly DNMT3A, TET2, ASXL1 and less frequently JAK2, SF3B1, SRSF2, and TP53. The incident rate of transformation from CHIP to MDS/AML or other lymphoid neoplasms is 0.5-1.0% per year. Both ICUS and CCUS are possible but not proven to be MDS. The patients with ICUS should have sustained cytopenia for >6 months without explainable etiology and does not meet WHO diagnostic criteria for MDS. Patients with CCUS show persistent unexplainable cytopenia without dysplasia, similar to ICUS, but harbor gene mutations (e.g. DNMT3A, TET2, ASXL1, and TP53), similar to those found in CHIP [25]. Clinical adjustment is needed for the decision whether a longterm follow-up is needed. Of importance, before diagnosis of ICUS and CCUS a complete investigation is highly recommended to exclude other hematologic or non-hematopoietic etiology of cytopenia. The myeloid neoplasms with germline predisposition (MNGP) found in familial MDS or other myeloid neoplasms include 1) AML with germline CEBPA or DDX41 mutations, 2) myeloid neoplasms with germline RUNX-1, ANKRD26 or ETV6 mutations which often have preexisting platelet disorder, and 3) myeloid neoplasms with germline mutations accompanying with organ dysfunction, e.g., Down syndrome, neurofibromatosis, Nooner syndrome, telomere disorder or GATA2 mutation [5]. An accurate diagnosis of MNGP requires a systemic and familial genomic and gene level investigation. There is no discrete treatment plan implicated in the aforementioned situations. However, the potentiality for development of myeloid neoplasm (e.g., MDS or AML) in patients with genetic alterations or mutations, warrant a close clinical monitor and follow-up.

In summary, in the era of molecular diagnosis and personalized medicine, it is important to pay attention to precursor lesions e.g. CHIP, ICUS, CCUS and MNGP that could lead to MDS or AML. Integrating morphology, revision of WHO subclassification, risk stratification according to IPSS and R-IPSS, genetic profiles and immunophenotyping is necessary for clinical diagnosis and management of MDS patients.