Correspondence
Detection of Residual Disease in Childhood Acute Lymphoblastic Leukemia
N Engl J Med 1999; 340:152-154January 14, 1999
- Article
To the Editor:
Cavé et al. (Aug. 27 issue)1 report that the detection of residual disease in patients with childhood acute lymphoblastic leukemia (ALL) by quantitative polymerase-chain-reaction (PCR) analysis within six months after the completion of induction therapy predicts relapse and is an independent prognostic factor. The authors conclude that PCR results should be used in planning treatment. It is important to add some cautionary notes to this conclusion.
It is well documented that PCR can have false positive and false negative results.2 In the study by Cavé et al., residual leukemia was detected in only 42 percent of patients after the completion of induction therapy, which appears to be a low proportion. This result may be due to poor sampling, loss of sensitivity of junctional probes, or degradation of target molecules. Clonal evolution and subclone formation can occur during therapy and change the molecular makeup of leukemic cells sufficiently to render them undetectable by the original PCR primers.2 Furthermore, the high threshold value for residual leukemic blasts (10–2 to 10–3) above which the risk of relapse is significantly increased is rather disappointing. This threshold approaches the sensitivity of light microscopy, and at this level, PCR analysis may not be needed. Independent expert hematopathologists could have been asked to study the slides and address this question adequately.
PCR is not a standardized assay and is subject to a high degree of variability among laboratories. Data obtained from different centers are still too immature and discrepant to be used clinically. Cavé et al. do not provide any data to support the use of alternative treatments on the basis of the results of their PCR assays. Reinduction therapy, intensification therapy, or bone marrow transplantation based on PCR testing may not improve but rather worsen the clinical outcome. We therefore disagree with the authors' conclusion that PCR results should be used clinically. Only controlled clinical trials can determine whether such an approach is warranted.
2 ReferencesStefan Faderl, M.D.
Zeev Estrov, M.D.
M.D. Anderson Cancer Center, Houston, TX 770301
Cave H ,van der Werff ten Bosch J ,Suciu S , et al. Clinical significance of minimal residual disease in childhood acute lymphoblastic leukemia. N Engl J Med 1998;339:591-598
Full Text | Web of Science | Medline2
Faderl S ,Estrov Z . The clinical significance of detection of residual disease in childhood ALL. Crit Rev Oncol Hematol 1998;28:31-55
CrossRef | Web of Science | Medline
To the Editor:
Cavé et al. report that residual disease detected by PCR after induction of a remission is a powerful prognostic factor in patients with childhood leukemia. Their method seems to be useful for patients with childhood leukemia, but we are concerned about some aspects of the study.
Cavé et al. used rearrangements of T-cell–receptor or immunoglobulin heavy-chain genes as clonal markers. However, these genetic changes are not causes of these diseases, but instead are manifestations of clonal expansion of the lymphoid cells. Previous studies have used chimeric messenger RNA such as AML1-MTG81 and BCR/ABL2; in these cases the chimeric products were principal causes of the cancers. In other studies, the usefulness of monitoring for residual leukemia was evaluated in homogenous groups of patients, but the study by Cavé et al. involved a variety of patients with childhood ALL.
If data from patients with different prognoses are analyzed together in an evaluation of the usefulness of monitoring, the efficacy of the monitoring procedure may be overestimated. In high-risk patients, it is difficult to achieve complete remission with conventional induction therapy, and leukemic cells may be detectable at molecular levels even after a hematologically complete remission has been achieved. Because these patients have a high probability of subsequent relapse, positive results of monitoring after induction therapy are not unexpected. In contrast, low-risk patients have a high probability of complete remission and cure with conventional chemotherapy. Their PCR results will probably remain negative after the induction of remission. When these groups are analyzed together, the presence of residual leukemia detected by PCR appears to be correlated with subsequent relapses. However, PCR studies may provide no prognostic information beyond that provided by the initial evaluation of risk factors. It might be useful to know the results of the initial evaluation of risk factors in the patients studied by Cavé et al.
Cavé et al. performed a bivariate analysis after stratification for the classic prognostic factors. However, this would be insufficient to eliminate fully the above-mentioned influences and chromosomal characteristics. These powerful prognostic factors were excluded from the stratified log-rank test because of the high percentage of patients in whom they could not be evaluated. Studies of the efficacy of monitoring for residual leukemia with the use of competitive PCR tests for rearranged T-cell–receptor or immunoglobulin heavy-chain genes should include only patients with uniform prognostic risk factors.
2 ReferencesUtako Machida, M.D.
Masahiro Kami, M.D.
Hisamaru Hirai, M.D.
University of Tokyo, Tokyo 113-8655, Japan1
Tobal K ,Yin JA . Monitoring of minimal residual disease by quantitative reverse transcriptase-polymerase chain reaction for AML1-MTG8 transcripts in AML-M2 with t(8; 21). Blood 1996;88:3704-3709
Web of Science | Medline2
Preudhomme C ,Henic N ,Cazin B , et al. Good correlation between RT-PCR analysis and relapse in Philadelphia (Ph1)-positive acute lymphoblastic leukemia (ALL). Leukemia 1997;11:294-298
CrossRef | Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Faderl and Estrov question the sensitivity of the PCR methods we used. As we reported, the median level of detection is 5×10–5, which is consistent with the number of mononuclear cells analyzed (2×105) and with data in previous reports.1 Such a method usually permits the detection of residual blasts in 30 to 60 percent of patients after the completion of induction therapy that includes anthracyclines.1 Clonal evolution was not a source of false negative results with early samples, since the molecular events associated with clonal evolution occur later.2 We think that reliable detection of 10–2 to 10–3 blasts by cytologic studies is illusive. This would require the systematic examination of more than 103 mononucleated marrow cells. Moreover, a treatment decision that relies on the observation of one or two blasts in a marrow sample would be much more prone to error than a decision based on PCR studies. It is also known that residual leukemic cells can be disguised as differentiated lymphocytes or normal primitive hematopoietic precursor cells and can thus escape cytologic detection.
Monitoring for residual disease in clinical practice is a complex issue. The level of residual disease and the kinetics of the decrease in residual blasts probably reflect the response to the initial chemotherapy, which many investigators consider an important prognostic factor in ALL.3 In one randomized study,3 intensification of postinduction therapy resulted in an improved outcome in a subgroup of patients with a slow response to chemotherapy. For these reasons, we think that patients with levels of residual disease above critical thresholds should receive intensified treatment. All controlled studies aimed at assessing whether monitoring for residual disease improves the outcome have to take into account two pitfalls: only a few patients have high levels of residual disease, which makes randomized studies difficult, and the poor prognosis in these children makes conventional therapy ineffective.
Machida et al. comment on other prognostic factors, especially cytogenetic abnormalities. Of the patients we studied, 12 of 15 with a high level of residual disease after the completion of induction therapy were in the standard-risk group. Of the 13 karyotypes from these patients that could be evaluated, 2 were hyperdiploid (>50 chromosomes), 1 was hypodyploid, 4 showed other abnormalities without known prognostic importance, and 6 were normal, but none showed a chromosomal translocation. For the standard-risk patients (i.e., a more homogeneous group of patients), multivariate analysis showed that at the end of induction chemotherapy and at later points in time, the extent of residual disease remained a strong independent prognostic factor as compared with other risk factors for relapse.
3 ReferencesHélène Cavé, Ph.D.
Hôpital Robert Debré, 75019 Paris, FranceStefan Suciu, M.S.
European Organization for Research and Treatment of Cancer Data Center, B-1200 Brussels, BelgiumEtienne Vilmer, M.D.
Hôpital Robert Debré, 75019 Paris, France1
Jacquy C, Martiat P. Detection of minimal residual disease in lymphoid malignancies. In: Freireich E, Kantarjian H, eds. Therapy of hematopoietic neoplasia. New York: Marcel Dekker, 1997.
2
Marshall GM ,Kwan E ,Haber M , et al. Characterization of clonal immunoglobulin heavy chain and T cell receptor γ gene rearrangements during progression of childhood acute lymphoblastic leukemia. Leukemia 1995;9:1847-1850
Web of Science | Medline3
Nachman JB ,Sather HN ,Sensel MG , et al. Augmented post-induction therapy for children with high-risk acute lymphoblastic leukemia and a slow response to initial therapy. N Engl J Med 1998;338:1663-1671
Full Text | Web of Science | Medline
- Citing Articles (2)
Citing Articles
1
Anna Butturini, John Klein, Robert Peter Gale. (2003) Modeling minimal residual disease (MRD)-testing. Leukemia Research 27:4, 293-300
CrossRef2
Dieter Hoelzer, Nicola Gökbuget. (2000) Recent approaches in acute lymphoblastic leukemia in adults. Critical Reviews in Oncology/Hematology 36:1, 49-58
CrossRef
