【作者】 李英梅；

【导师】 王建祥； 郑以州； 秘营昌； 王敏；

【作者基本信息】 中国协和医科大学 ， 血液内科学， 2010， 博士

【摘要】 目的探讨再生障碍性贫血(AA)演变为血液学克隆性疾病的发生率及其危险因素。方法应用Cumulative incidence curves分析1991年1月到2007年12月802例住院AA患者转化为阵发性睡眠性血红蛋白尿症(PNH)、骨髓增生异常综合征／急性髓系白血病(MDS/AML)的累积发生率；并应用Cox回归模型分析其可能的危险因素。结果①802例患者中超重型AA (VSAA)174例(21.7%),重型AA (SAA) 335例(41.8%),非重型AA (NSAA)293例(36.5%)。中位年龄23岁,中位随访时间71月,5年总体生存率77.2%。共有4.7%(38／802)转化为克隆性疾病,估计5年克隆性演变累积发生率为3.7%(95%可信区间2.6%—5.3%)；MDS/AML累积发生率1.7%(1.0%—3.0%),PNH 2.1%(1.3%—3.4%)。②19例转化为MDS/AML的AA,中位发病年龄37(9—59)岁,转化为MDS/AML的中位时间为确诊AA后33(12—240)月。rhuG-CSF疗程大于300天者中位转化时间20.5(12—37)月,明显早于rhuG-CSF疗程0—100天者[88(29—240)月,P=.006]。转化为MDS/AML时核型分析为11例—7,1例+8,1例20q-,5例正常,1例未查。转化为MDS/AML后中位随访时间13(2-95)月,12／19例(63.2%)死亡,预后差。③Cox回归多因素分析：疾病严重程度(VSAA/SAA/NSAA, P=.001)、年龄(P=.007,RR=1.043)、rhuG-CSF疗程(P=.000,RR=1.005)是AA演变为MDS/AML的危险因素。VSAA危险性是SAA/NSAA的6—7倍(P=.001),而SAA与NSAA相比无统计学差异(P=.743)。rhuG—CSF疗程大于300天组危险性明显增高(P=.000,RR=21.919),而疗程为0天／1—89天／90-300天者,3组间无统计学差异(P>.10)。④另有9例患者治疗后出现异常核型但形态学不符合MDS,包括3例+8,2例—Y,2例13P+,—14及t(3,12)各一例。中位随访时间114(51—210)月,均生存。⑤21例转化为PNH者(2例合并MDS),中位年龄25(11—43)岁,中位转化时间为确诊AA后24(6—216)月。转化为PNH后中位随访时间52(14—145)月,4／21(19.1%)例死亡。单因素分析,转化为PNH不增加死亡危险性(P=.216)。⑥多因素分析AA转化为PNH的危险因素,仅与AA初诊时白细胞计数有关(P=.007,RR=1.453),与疾病严重程度、治疗方案、初诊时是否有PNH克隆等无关。⑦确诊AA时有405例患者检测CD55/CD59,117例(28.9%)PNH克隆阳性。系列监测237例AA治疗前后PNH克隆的演变过程：PNH克隆持续阴性者140例(59.0%),间断阳性者62例(26.2%),持续阳性者35例(14.8%),后者10例(4.2%)转化为PNH。结论①年龄、VSAA及长期应用rhuG-CSF是AA转化为MDS/AML的危险因素；②应尽量避免长期应用rhuG-CSF,对长期应用者须加强骨髓细胞形态及细胞遗传学监测。③AA转化为PNH更倾向于自然病程,与初诊时是否有PNH克隆无关,治疗后应定期监测患者PNH克隆水平。目的探讨再生障碍性贫血(AA)单体7克隆的演变过程及初诊时潜在低比例—7克隆的临床意义。方法应用间期荧光原位杂交(FISH)回顾性分析81例核型正常的初诊AA及46例免疫抑制(IST)联合重组人粒细胞集落刺激因子(rhuG-CSF,疗程大于6个月)治疗后AA的—7克隆。结果81例初诊AA中,11例(13.6%)—7克隆阳性,阳性细胞比例5.4%—7.6%,—7阳性患者疗效及生存率与—7阴性者相比无统计学差异(P=.481,.865),11例阳性患者(包括5例rhuG-CSF疗程大于6个月者)治疗后—7比例均下降至正常,中位随访时间44个月,未发现转化为骨髓增生异常综合征／急性髓系白血病(MDS/AML)的证据；系列随访46例AA的—7克隆,治疗后3—6个月均为阴性,治疗后12—15月5例阳性,中位随访时间48月,FISH共检测到6例—7阳性,均进展为MDS/AML,5例为—7核型。FISH检出阳性克隆的时间较常规核型分析提前3—18个月；FISH回顾性分析了4／6例转化为MDS/AML的AA初诊时标本,—7均为阴性。结论部分初诊AA潜在低比例的—7克隆,但与治疗反应及最终进展为克隆性疾病无关；rhuG-CSF可能促进—7扩增,长期应用rhuG-CSF治疗的AA需用间期FISH密切监测以尽早发现。更多还原

【Abstract】 Objective:To elucidate the incidence and risk factors for clonal evolution in aplastic anemia (AA) patients.Methods:We used the Cumulative incidence curves to estimate the probability of evolution to paroxysmal nocturnal haemoglobinuria (PNH) or myelodysplastic syndrome/acute myeloid leukemia (MDS/AML) in 802 hospitalized AA patients from January,1991 through December,2007. The Cox proportional hazards model was used to assess the risk factors for evolution to MDS/AML or PNH.Results:①A total of 802 patients met the eligibility criteria including 174 very severe AA (VSAA,21.7%),335 severe AA (SAA,41.8%) and 293 non-severe AA (NSAA, 36.5%)。Median follow-up period for all patients is 71 months and the median age was 23.0 years. The overall survival at 5 years after diagnosis was 77.2%. Thirty-eight patients (38/802,4.7%) were considered to have evolved to a new hematologic disorder. The cumulative incidence of clonal evolution was estimated as 3.7%(95 percent confidence interval,2.6%—5.3%) and the incidence of MDS/AML was 1.7%(1.0%—3.0%), PNH 2.1%(1.3%-3.4%) at 5 years.②We found 19 patients evolved to MDS/AML between 12 and 240 months (median,33 months) following initial AA diagnosis. Median age at diagnosis of AA was 37 years (9-59 years). The median evolution time of patients treated with rhuG-CSF more than 300 days was 20.5 months (12-37 months) compared with 88 months (29-240 months) for patients with 0-100 days rhuG-CSF therapy (P=0.001). Cytogenetic analysis of bone marrow at the time of diagnosis of MDS revealed monosomy 7 (11 patients), trisomy 8 (1 patient),20q-(1 patient) and normal karyotype (5 patient), while at initial diagnosis they all had a normal karyotype. The prognosis was very poor and 12/19(63.2%) died at the median follow-up time of 13 months (2-95 months) after evolution.③By multivariate analysis, age (P=.007, RR=1.043), severity of disease (P=.001), and the number of days rhuG-CSF therapy (P=.000, RR=1.005) are associated with translation to MDS/AML. The RR for VSAA is 6-7 times higher than that for SAA and NSAA (P=.001), but the last two do not differ significantly (P=.743). Age and the number of days rhuG-CSF therapy were analyzed as a continuous variable and each year or day increases the risk of evolution by about 4.3% or 0.5%, respectively. RR for patients who received rhuG-CSF longer than 300 days was 21 times higher than that for patients who did not receive rhuG-CSF (P< 0.001, RR=21.919), but there was no statistical difference between 0 day and 1-89 days or 90-300 days of rhuG-CSF therapy (P>.10).④In addition, there are 9 patients developing clonal abnormalities without the morphologic features of MDS. Trisomy 8 (3 patients) and monosomy y (2patients) and add 13 (2 patient) were noted in these patients. All of them responsed to IST and were alive at a median follow up 114 months (51-210 months).⑤Twenty-one patients developed to overt PNH, and 2 of them combined with MDS. Median age was 25 years (11-43 years) and the median time of evolution was 24 months (6-216 month).17/21(81%) patients are alive at 92 months (29-313 moths) after dianosis of AA or 52 months after evolution. By univariate analysis, transformation to PNH was not statistically significantly associated with death (P=0.216).⑥As for evolving to PNH, white blood cell count at initial diagnosis is the only significant risk factor (P=.007, RR=1.453). It was worth mentioning that Hemoglobin, Reticulocyte count, with or without PNH clone at initial diagnosis, treatment and response to IST were all not linked with evolving to PNH.⑦PNH clone was monitored sequentially in 237 patients and transient or instable PNH clone were detected in 26.2%patients compared with persistent PNH clone in 14.8%patients including 4.2% developed to overt PNH.Conclusions:①Patients with older age, VSAA and long-term rhuG-CSF therapy are at higher risk for evolving to MDS/AML.②long-time administration of rhuG-CSF should be avoided and it is necessary to monitor morphology and cytogenetics in AA with long-term rhuG-CSF. Furthermore, normalizing hematopoiesis of AA may represent a viable approach to prevent clone evolutions, especially to MDS/AML.③The appearance of abnormal clone does not necessarily mean the subsequent expansion of those clones. Transformation to PNH for subpopulations of AA patients may be natural evolution and PNH clone should be detected serially after therapy. Objective To explore the clonal evolution of monosomy 7 in patients with aplastic anemia (AA) and the clinical implications of small monosomy 7 clones at initial diagnosis.Methods We retrospectively detected-7 clones by interphase-fluorescence in situ hybridization (FISH) in 81 AA patients with normal karyotype at initial diagnosis. We also performed serial analyses of -7 clones in 46 AA treated with immunosuppressive therapy (IST) and more than 6 months of recombinant human granulocyte colony-stimulating factor (rhuG-CSF).Results We disclosed 5.4%—7.6% monosomy 7 cells in 11 (13.6%) of 81 patients at initial diagnosis of AA, the survival and response rate to IST of the 11 patients did not differ significantly from that of the other 70 patients (P=.481,.865); Monosomy 7 cells disappeared after IST in all of 11 patients including 5 received long-term rhuG-CSF administration, and none of them evolved to myelodysplastic syndromes/acute myeloid leukemia (MDS/AML) at a median follow-up of 44 months. Serial assessments of-7 clones were performed in 46 patients, none of them detected-7 clones 3-6 months after IST, but we detected-7 in 5 patients 12-15 months after IST, at a median follow up time of 48 months, FISH identified 6 patients with-7 clones while the conventional cytogenetic analysis (CCA) recognized in 5. Moreover, the time when-7 was first evident by FISH was 3-18 months earlier than that by CCA. All of the 6 patients evolved to MDS/AML with -7 and four of them were retrospectively analysed -7 at initial diagnosis of AA, but none of them waspositive.Conclusions Monosomy 7 existed in a part of AA onset, but the preexisting -7 cells seemed neither associated with fatality nor evolving to MDS/AML. rhuG-CSF might facilitate the expansion of -7 clones; thus it should be necessary to monitor -7 in AA, especially received long-term rhuG-CSF administration.更多还原