Hoyle C F, Gray R G, Wheatley K, Swirsky D, de Bastos M, Sherrington P, Rees J K, Hayhoe F G
Department of Haematology, Royal Postgraduate Medical School, Hammersmith Hospital, London.
Br J Haematol. 1991 Nov;79(3):398-407. doi: 10.1111/j.1365-2141.1991.tb08047.x.
Analysis of bone marrow slides from 1,386 patients entered into the Medical Research Council's 8th and 9th trials in Acute Myeloid Leukaemia confirmed that features associated with differentiation in blast cells, in particular increasing Sudan Black (SB) positivity, were the most important morphological features for predicting remission achievement (P = 0.002) and hence survival (P less than 0.0001). SB positivity was also weakly predictive of remission duration (P = 0.05). A low complement of maturing granulocytes was associated with early induction death and a high percentage of blasts with shorter remissions. The few patients with acute promyelocytic leukaemia (FAB M3) had a high haemorrhagic death rate during induction and a low relapse rate. Apart from this, lineage involvement was not predictive of outcome. Multiple lineage leukaemias, in particular those with megakaryocytic and/or erythroid involvement, which had been reported previously to have a poor prognosis, did not have any worse remission rates in this series. When more than one cell line was involved, no combination with particularly good or poor prognosis could be identified. Multivariate analysis suggested that percentage SB positivity was adequate on its own to divide granulocytic leukaemias into poorly differentiated (less than 50% SB +ve) and well-differentiated groups (50% or more SB +ve) without the need for further measurements. This simple and reproducible test was strongly predictive of resistant disease but not of induction deaths. It was of considerably greater prognostic value--and was less open to inter-observer disagreement--than the FAB criteria which are usually used to classify granulocytic lineage leukaemias into the M1 and M2 subgroups. It is proposed that greater than or equal to 50% of blasts with SB positivity should replace blasts greater than 10% of maturing myeloid cells for this sub-categorization between M1 and M2.
对参加医学研究委员会第8次和第9次急性髓细胞白血病试验的1386例患者的骨髓涂片分析证实，原始细胞中与分化相关的特征，特别是苏丹黑（SB）阳性增加，是预测缓解实现（P = 0.002）及生存（P＜0.0001）的最重要形态学特征。SB阳性对缓解期也有较弱的预测作用（P = 0.05）。成熟粒细胞数量少与早期诱导死亡相关，原始细胞比例高则缓解期短。少数急性早幼粒细胞白血病（FAB M3）患者在诱导期出血死亡率高，复发率低。除此之外，谱系受累情况不能预测预后。多谱系白血病，特别是那些伴有巨核细胞和/或红系受累的白血病，此前报道预后较差，但在本系列中缓解率并无更差。当涉及多个细胞系时，未发现有预后特别好或特别差的组合。多变量分析表明，仅SB阳性百分比就足以将粒细胞白血病分为低分化组（SB阳性＜50%）和高分化组（SB阳性≥50%），无需进一步测量。这种简单且可重复的检测方法对难治性疾病有很强的预测性，但对诱导死亡无预测性。与通常用于将粒细胞系白血病分为M1和M2亚组的FAB标准相比，它具有更大的预后价值，且观察者间分歧较小。建议用≥50%的原始细胞SB阳性取代＞10%的成熟髓细胞用于M1和M2之间的亚分类。
