Chan G C, Wang W C, Raimondi S C, Behm F G, Krance R A, Chen G, Freiberg A, Ingram L, Butler D, Head D R
Department of Hematology/Oncology, St Jude Children's Research Hospital, Memphis, TN 38101, USA.
Leukemia. 1997 Feb;11(2):206-11. doi: 10.1038/sj.leu.2400558.
To evaluate diagnostic criteria, disease characteristics, and the clinical course of pediatric myelodysplastic syndrome (MDS), we reviewed 327 consecutive cases diagnosed with de novo acute myeloid leukemia (AML) or MDS at St Jude Children's Research Hospital between February 1980 and January 1993. Among 49 cases with <30% marrow blasts (consistent with FAB criteria and common diagnostic practice for MDS), eight had karyotypes associated with de novo AML (four with t(8;21)(q22;q22) and one each with inv(16)(p13q22), t(11;17)(q23;q21), t(9;11)(p22;q13), and i(1)(ql0)). We termed these cases AML with a low blast count (AML-LBC) and compared their clinical and morphologic features with those of the remaining 41 cases. AML-LBC cases had little or no hematopoietic dysplasia. MDS cases consisted of refractory anemia (RA, n=6), RA with ring sideroblasts (n=2), RA with excess blasts (RAEB, n=4), RAEB in transformation (n=14), and chronic myelomonocytic leukemia (n=15). Most had moderate/severe or multilineage hematopoietic dysplasia, with significantly higher dysplasia scores than AML-LBC cases (P=0.007). Only 30% of patients with MDS achieved complete remission (CR) after two cycles of AML-directed therapy, compared with 88% of patients with AML-LBC (P=0.0001); MDS patients tended to experience prolonged severe cytopenias with chemotherapy. The 4-year survival for MDS patients was 23% +/- 7% (s.e.), vs 50% +/- 18% (s.e.) for AML-LBC (P=0.048). AML-LBC patients frequently had chloromas; none were seen in MDS patients. We conclude that the 30% blast threshold is ineffective for separation of AML and MDS in pediatric patients, and that genetic data should be included in this decision process. AML-LBC, defined by <30% blasts in bone marrow and cyto- (or molecular) genetic abnormalities associated with de novo AML, and characterized by absent or mild marrow dysplasia, is biologically and clinically distinct from MDS and should be treated as de novo AML. Outcome in pediatric MDS remains poor, and new treatment strategies are needed for these patients.
为评估儿童骨髓增生异常综合征（MDS）的诊断标准、疾病特征及临床病程，我们回顾了1980年2月至1993年1月在圣裘德儿童研究医院连续诊断为原发性急性髓系白血病（AML）或MDS的327例病例。在骨髓原始细胞<30%的49例病例中（符合FAB标准及MDS的常见诊断做法），8例具有与原发性AML相关的核型（4例为t(8;21)(q22;q22)，1例分别为inv(16)(p13q22)、t(11;17)(q23;q21)、t(9;11)(p22;q13)和i(1)(q10)）。我们将这些病例称为低原始细胞计数AML（AML-LBC），并将其临床和形态学特征与其余41例病例进行比较。AML-LBC病例几乎没有造血发育异常。MDS病例包括难治性贫血（RA，n = 6）、伴有环形铁粒幼细胞的RA（n = 2）、伴有过多原始细胞的RA（RAEB，n = 4）、转化中的RAEB（n = 14）和慢性粒单核细胞白血病（n = 15）。大多数病例有中度/重度或多系造血发育异常，发育异常评分显著高于AML-LBC病例（P = 0.007）。在接受两个周期AML导向治疗后，只有30%的MDS患者达到完全缓解（CR），而AML-LBC患者为88%（P = 0.0001）；MDS患者化疗后往往会经历长期严重血细胞减少。MDS患者的4年生存率为23%±7%（标准误），而AML-LBC患者为50%±18%（标准误）（P = 0.048）。AML-LBC患者常有绿色瘤；MDS患者未见。我们得出结论，30%的原始细胞阈值在区分儿童患者的AML和MDS方面无效，并且在这一决策过程中应纳入基因数据。AML-LBC定义为骨髓原始细胞<30%且伴有与原发性AML相关的细胞（或分子）遗传学异常，其特征为无或轻度骨髓发育异常，在生物学和临床上与MDS不同，应作为原发性AML治疗。儿童MDS的预后仍然很差，这些患者需要新的治疗策略。
