marrow transplant for chronic myelogenous leukemia ...

From bloodjournal.hematologylibrary.org by guest on July 21, 2011. For personal use only.

1992 79: 276-282

Prognostic significance of Philadelphia chromosome-positive cells detected by the polymerase chain reaction after allogeneic bone marrow transplant for chronic myelogenous leukemia MS Roth, JH Antin, R Ash, VH Terry, M Gotlieb, SM Silver and D Ginsburg

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.


Prognostic Significance of Philadelphia Chromosome-Positive Cells Detected by the Polymerase Chain Reaction After Allogeneic Bone Marrow Transplant for Chronic Myelogenous Leukemia By Mark S. Roth, Joseph H. Antin, Robert Ash, Valeri H. Terry, Mark Gotlieb, Sam M. Silver, and David Ginsburg Although rare cells expressing the bcrlabl fusion transcript can be detected by the polymerase chain reaction (PCR) in patient blood or marrow after allogeneic bone marrow transplant (BMT) for Philadelphia chromosome (Ph+)-positive chronic myelogenous leukemia (CML), the prognostic significance of this finding is unknown. This paper reports clinical, cytogenetic, and molecular data derived from 64 CML patients following allogeneic BMT. Nested primer PCR was performed on patient blood and bone marrow samples t o detect the presence of residual bcrlabl (+) cells in CML patients considered t o be in clinical remission at the time of study. Bcrlabl transcripts were detected in 37 of 64 patients for at least one timepoint post-BMT. Thirteen of these 37 bcrlabl (+) patients have subsequently relapsed, as defined by clinical and/or persistent cytogenetic findings, in contrast t o 0 relapses among the 27 bcrlabl (-) patients (P = .0025).

The median time from first (+) bcrlabl PCR signal t o relapse was 150 days (range 90 t o 832). Fifty-four patients were studied at two or more timepoints post-BMT: five of eight patients persistently bcrlabl (+) have relapsed; 5 of 23 patients with both bcrlabl (+) and (-) assays during follow-up have relapsed; and none of 23 patients persistently (-) have relapsed (cumulative actuarial relapse rates 77%. 20%. and 0%, respectively, P = .0017). These data indicate that among CML patients in apparent clinical remission after BMT, nested primer bcrlabl PCR can define subgroups with low, intermediate, and high risk of relapse. The pattern of bcrlabl PCR detection after transplant may aid in the development of trials designed t o reduce the risk of relapse, or allow for early intervention in patients who fail t o clear the malignant clone. 0 1992by The American Society of Hematology.

A

consistent correlation with clinical outcome has been identified. We now report clinical, cytogenetic, and molecular data on 64 Ph’ CML patients after allogeneic BMT. Our data indicate that CML patients with bcrlabl transcripts detected by PCR following allogeneic BMT are at a significantly increased risk of subsequent relapse, while those patients persistently bcrlabl (-) are less likely to relapse. The pattern of bcr/abl detection after transplant may aid in the development of trials designed to reduce the risk of relapse, or allow for early intervention in patients who fail to clear the malignant clone.

LLOGENEIC BONE marrow transplantation (BMT) is the only known curative treatment for patients with Philadelphia chromosome-positive (Ph’) chronic myelogenous leukemia (CML).’” Disease recurrence remains a major obstacle, however, with the overall 4-year probability of relapse ranging from 20% to 60%, depending on the stage of disease at the time of transplant.’” Relapse from clinically undetectable residual malignant cells is thought to be the most likely source of r e c ~ r r e n c eRecently, .~ we and others have adapted the polymerase chain reaction (PCR) to detect the chimeric bcrlabl RNA transcript unique to Ph’ CML cells.’~’‘This assay is significantly more sensitive than cytogenetics or Southern b l ~ t t i n gThe . ~ application of this approach to detect minimal residual bcrlabl (+) cells after allogeneic BMT might identify subsets of patients who are more likely to relapse after transplant and thus might benefit from additional therapy or alternative conditioning regimens. However, initial reports of bcrlabl PCR analysis after BMT have yielded conflicting results.’.’’ To date, no

From the Departments of Medicine and Human Genetics, University of Michigan Medical School; the Howard Hughes Medical Institute, Ann Arbor, MI; the Department of Medicine, Brigham and Women’sHospital Boston, MA; and the Departments of Medicine and Pediatrics, the Medical College of Wisconsin, Milwaukee. Submitted July 10, 1991; accepted September 5, 1991. Supported in part by grants from the American Cancer Society, Children’s Leukemia Foundation of Michigan, the Koehring Foundation, the Freedman Memorial Fund, and National Institutes of Health Grant CA39542. Presented in part at the American Society of Hematology annual meeting, Boston, MA, December 1990. Address reprint requests to Mark S. Roth, MD, C570D MSRB II, 1150 WMedical Center Dr, Ann Arbor, MI 48109-0668. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section I734 solely to indicate this fact. 0 1992 by The American Society of Hematology. 0006-49711921 7901-0020$3.00/0 276

MATERIALS AND METHODS Patient material. Sixty-four patients with Ph’ CML in clinical remission were studied between April 1985 and November 1990 after allogeneic BMT at either the Brigham and Women’s Hospital in Boston, MA (33 patients) or the Medical College of Wisconsin in Milwaukee (31 patients). At the time of transplantation, 37 patients were in chronic phase (CP), 25 in accelerated phase (AP), and two in blast crisis (BC). Sixty-one patients received high-dose cytosine-arabinoside, cyclophosphamide, and total body irradiation as preparative regimen and three received busulfan and cyclophosphamide.“” Twenty patients received unmanipulated marrow and 44 received grafts depleted of CD5’ or CD3’ T lymphocytes. Forty-seven patients received related donor grafts and 17 received closely HLA-matched unrelated donor grafts. Informed consent was obtained and treatments administered according to protocols and procedures approved by the appropriate institutional review committees on human investigations. Details of the chemo-irradiation regimens, supportive care, and graft-versus-host disease (GVHD) prophylaxis at the respective transplant centers have been previously published.’“’’ RNApreparation and PCR. Peripheral blood bu@ coats, mononuclear cells, or bone marrow samples were obtained at various timepoints post-BMT ranging from 7 days to over 4 years. Cells were either viably frozen or directly added to 4 mol1L guanidinium isothiocyanate and RNA prepared with minor modifications of previously described methods.’.’’ RNA PCR was performed using an ab1 sequence-specific reverse transcriptase primer and bcrlablspecific PCR primers as previously reported.’ Second-round PCR with nested internal bcrlabl primers was performed in all cases and the PCRs were evaluated for the bcrlabl transcript by direct Blood, Vol79, No 1 (January l), 1992: pp 276-282

From bloodjournal.hematologylibrary.org by guest on July 21, 2011. For personal use only. PROGNOSTIC SIGNIFICANCE OF PCR IN CML

visualization on ethidium bromide stained agarose gels after electrophoresis? In reconstitution experiments, one CML cell per million normal mononuclear cells could be detected by this approach.’ The presence of intact RNA and adequate cDNA preparation was evaluated by a single round of PCR using ab1 sequence-specificPCR primers as previously reported? Controls for PCR canyover. To avoid sample contamination, RNA extraction, cDNA synthesis, and PCR preparation were performed in an area physically isolated from the gel electrophoresis analysis. Mock RNA preparations were also run in parallel as negative controls. Sample RNA were scored as bcrlabl (+) only if the corresponding set of controls was negative. In addition, when adequate RNA was available, nested primer PCR was repeated on a separate RNA aliquot from the bcrlabl (+) samples to confirm the presence of the bcrlabl transcript. If a separate assay gave discordant results, the data point was discarded. For 53 of 82 repeat PCR analyses two new bcr oligonucleotide primers were used (bcr-3 [AGCAAGGCTACGGAGAGGCTGAAGAAGAAG bp (-)365 - (-)336] and bcr-4 [TCGGAGCAGGAGTCACTGCTGCTGCTTATG bp (-)333 - (-)304]).*” These primers designed 5‘ to the earlier reported bcr primers were used in conjunction with the previously described abl-1 and abl-2 primers for nested primer PCR. Although PCR using the new primers resulted in sensitivity and specificity similar to the previous primer sets (data not shown), carryover from earlier PCRs would not amplify and thus could be eliminated as a source of contamination. Analysis for the last 10 patients studied used the new primers exclusively. Cytogenetics. Bone marrow cells from 31 patients were analyzed at the Mayo Clinic CytogeneticsLaboratory (Rochester, MN) for the presence of the Ph’ chromosome. Analyses used direct preparation of aspirates with slides stained by quinacrine mustard with fluorescence (QFQ banding).” For each analysis, at least 20 metaphases were evaluated by light microscopy. The median number of separate cytogenetic evaluations after BMT for these 31 patients was four (range 1to 7). Statistical analysis and dejinition of relapse. Actuarial relapsefree survival and cumulative relapse probabilities were calculated using the Kaplan-Meier method.*’ The Cox-Mantel statistic and Chi-squared analysis were used to compare bcrlabl status with clinical parameter^.^^.'^ Relapse was defined by the occurrence of clinical hematologic features (abnormal blood counts and/or marrow consistent with CML), or progressive cytogenetic changes (the appearance of Ph’ metaphases in increasing percentages on consecutive marrow examinations). RESULTS

Detection of bcrlabl by PCR after allogeneic BMT identifies patients at significant risk of relapse. Samples from all 64 patients were adequate for evaluation based on the presence of intact ab1 mRNA by PCR or the detection of bcrlabl after two rounds of PCR. All 40 patients on whom pretransplant samples were available had detectable bcrlabl transcripts before BMT. After BMT, a total of 221 R N A samples were analyzed, ranging from one to eight per patient, with a median number of three samples per patient. The median follow-up time post-BMT for the entire group was 365 days with a range of 14 days to over 5 years. In all, PCR-detectable bcrlabl transcripts were present in 37 of 64 patients. The bcr exon IIIlabl exon 2 fusion was detected in 22 patients, the bcr exon IIlabl exon 2 product in 12 patients, and both in three patients. Twenty-six of 40 patients with available pretransplant samples had a (+)

277

assay after BMT. In two patients both bands were seen pretransplant, while only the larger transcript (bcr exon IIIlabZ exon 2) remained post-BMT. In the remaining 24 patients, pre- and post-BMT transcripts matched in size. Repeat nested primer PCR analysis was performed o n 82 of the 85 (+) R N A samples as described in Materials and Methods. Sixty-eight of the 82 (+) assays remained (+). The 14 discordant data points were discarded from the statistical analysis presented below. Inclusion of these additional points did not significantly alter the results. Thirteen of the 37 bcrlabl (+) patients (5CP and 8AP) have relapsed, as documented by clinical parameters and cytogenetics in 11o r by a progressive increase in the percentage of Ph‘ chromosomes detected by cytogenetics in two. The median time from first (+) bcrlnbl signal to relapse was 150 days (range 90 to 832). The larger bcrlabl transcript was detected in eight of the relapsing patients while the smaller transcript was detected in four and both transcripts in one. Ten of the 37 patients with a (+) bcrlabl assay had a detectable signal with one round of PCR. Seven of these patients have relapsed with a median time from (+) signal to relapse of 120 days (range 30 to 270). None of the 27 bcrlabl (-) patients have relapsed with a median follow-up time of 372 days (range 14 to 1,289). The difference in the incidence of disease relapse between patients with 0 versus 1 or more (+) assays was highly significant (P = .0025, Fig 1). Distinct bcrlabl detection pattems identi& patients at high, intermediate, and low risk of relapse after BMT. The 54 patients studied at two or more timepoints post-BMT could be classified into three groups (Table 1). Group I consisted of eight patients with a persistent bcrlabl (+) assay. Median follow-up for all group I patients was 268 days (range 63 to 1,140). In this group, five patients have relapsed, one died in clinical remission at day 60, and two are alive and clinically free of disease at days 63 and 1,140. The latter patient was first evaluated at day 660 after BMT. Group I1 consisted of 23 patients with both bcrlabl (+) and (-) timepoints. The median follow-up time posttransplantation was 480 days (range 50 to 1,460). The patterns of bcrlabl detection and disease outcome for group I1 patients 100,

2

I

+

1 bcrlabl assays (t)

(56 t 12.5%)

I

r‘

p = 0.0025

I

0 bcrlabl assays (t) r‘ (0%) 7

6

12

18

24

30

36

42

40

54

Time (months) Post Transplant

Fig 1. Kaplan-Meier plot of the actuarial probability of relapse for all 64 patients comparing those with no (+) assays to those with one or more (+) assays. The difference in the probability of relapse is statistically significant (P = ,0025).

From bloodjournal.hematologylibrary.org by guest on July 21, 2011. For personal use only. 278

ROTH ET AL Table 1. Patterns of bcrlabl Detection Post-BMT Group*

No. of patients No. relapsing Actuarial relapse probability (%) Median follow-up days (range)

8 23 5 5 77 2 18 20 2 18 268 480 (634,140) (50-1,460)

23 0 0 372 (14-1,289)

Derived from 54 patients with two or more assays. *Group I patients were persistently bcr/ab/(+), group II patients had both (+) and (-1 bcr/ab/assays, and group Ill patients were persistently bcr/ab/(-) after BMT.

are shown graphically in Fig 2. Among the 23 patients in this group, five patients (nos. 18 through 22, Fig 2) have relapsed. Ten patients (nos. 1 through 10, Fig 2) had one or two initial (+) assays followed by one or more subsequent negative assays. Eleven patients (nos. 11 through 21, Fig 2) had one or more initial (-) assays within the first 180 days post-BMT followed by at least one (+) assay at a later timepoint. Two patients (nos. 22 and 23, Fig 2) had an initial (+) assay followed by (-) and then (+) assays. Group 111 consisted of 23 patients with persistent bcrlabl negative assays after BMT. None of these patients have relapsed, with a median follow-up time of 372 days (range 14 to 1,289). Analysis of bcrlabl in patients from all three groups are shown in Fig 3. To further determine if the number of (+) or (-) bcrlabl PCR assays was associated with disease relapse, relapse rates were compared among patients with 0, 1, or 2 2 (+) assays (Fig 4). The actuarial risk of relapse increased with the number of (+) assays (0% v 41% v 75%, respectively; P = .0082) (Fig 4). An increased risk of relapse with increasing number of positive bcrlabl assays was also observed when patients were stratified into CP v AP/BC (data not shown). Early detection ( .7). Seven patients (nos. 1 through 6 and 8, Fig 2) who were initially bcrlabl (+) in this time period became bcrlabl (-) at later timepoints. In contrast to the finding in the first 30 days, patients who relapsed were more likely to be bcrlabl (+) between days 30 and 365 when compared with patients who remained in remission (Table 2). Detection of bcrlabl transcripts after BMT is not associated with T-cell depletion, donor source, or GVHD. There was no significant association between marrow manipulation or donor source with overall detection of bcrlabl transcripts. A PCR bcrlabl (+) assay was present in 28 of 44 T-celldepleted and in 9 of the 20 unmanipulated transplants (P > .25). Eight of 17 patients who received unrelated but closely HLA-matched transplants had detectable bcrlabl signals compared with 29 of 47 related donors (P > .4). A detectable bcrlabl signal was not statistically associated with the presence or absence of acute (grades 111 or IV) or chronic (moderate or severe) GVHD. In 13 patients, simultaneous blood and marrow samples were available for study. Ten of these samples had similar PCR assay results, while three had a (-) blood assay with a (+) marrow. The bcrlabl PCR assay is more sensitive than cytogenetics. Concurrent cytogenetic and PCR bcrlabl analysis was available for 30 patients. No patient had a positive cytogenetic analysis with a negative bcrlabl PCR assay. Thirteen patients had exclusivelynormal donor karyotype analysis by cytogenetics and a negative bcrlabl assay. None of these patients have relapsed, with a median follow-up of 420 days (range 60 to 1,460). Twelve patients had detectable bcrlabl transcripts with normal cytogenetics. Three of these patients have relapsed. One of these patients developed concurrent clinical and cytogenetic relapse, while in the other two detectable Ph' chromosomes appeared before clinical relapse. Five patients had both a bcrlabl (+) assay and detection by cytogenetics of a Ph' chromosome when first evaluated. Three of these five patients progressed to clinical relapse. DISCUSSION

This report defines the prognostic significance for patterns of RNA PCR detection of residual bcrlabl containing

Fig 2. Graphic representation of bcrlabl detection for group II patients (both [+] and [-] assays after BMT). Solid circles represent (+) assays and open circles (-1 assays. Clinical follow-up in days after BMT is shown for each patient. rem, remission; rel, relapse; D, deceased. The positive signals seen on day 1,095 for patient 10 and day 730 for patient 16 were obtained from bone marrow cells. Peripheral blood mononuclear cells evaluated simultaneously were bcr/abl(-).


PROGNOSTIC SIGNIFICANCE OF PCR IN CML

279

Fig 3. Patterns of bcr/abl expression detected after BMT. (Athrough D) Ethidium bromide-stained agarose gel electrophoresis analysis of PCR products from four patients. The bcl/ab/ analysis represents two rounds of PCR with nested primers, while the control ab1 PCR (for intact RNA and cDNA) is only after one round. pBR 322 plasmid digested with restriction enzyme Mspl was used as a size marker. (A) Patient with persistent detection of the b c r l a b l transcript. Despite the detection of bcr/ab/-containing cells at days 108,210, and 324. this patient remained in clinical and cytogenetic remission. By day 357, 8/20 marrow metaphases were Ph' by cytogenetics. Clinical disease relapse occurred at day 480 post-BMT. (B) Complete elimination of detectable b c r l a b l transcripts. The presenceof intact RNA for each sample is demonstrated by the ( + ) ab/ control. The variable intensities of the ab/ signal seen may reflect RNA quality or RNA PCR efficiency. These factors should be less significant after two rounds of PCR as performed for all b c r l a b l detection. This patient remains in clinical and cytogenetic continuous remission at day 570 after transplantation. (C) Intermittent bcrlabl-positive timepoints. A ( + ) b c r l a b l assay was obtained from a bone marrow sample at day 30 but ( - 1 assays from the blood at day 60 and marrow at days 120 and 180. Reappearance of a b c r l a b l signal occurred in the blood at day 480 after transplant. This patient (no. 23, Fig 2) remained in clinical and cytogenetic remission until his death at day 618. (D) Transient detection of the b c r l a b l transcript. Both sizes of the bcr/ab/ transcript were present pre-BMT and at day 90 after transplant with subsequent disappearanceat all later timepoints. Simultaneous blood (B) and marrow (M) were analyzed at day 730. This patient (no. 9, Fig 2) remains in clinical and cytogenetic remission at day 730 after transplant.

cells after allogeneic BMT for CML. In all, bcrlahl transcripts were detected by PCR in 37 of 64 patients. Previous studies have reported widely varying frequencies of detection of the bcrlahl transcript after allogeneic BMT for CML."" Consistent with our data, Hughes et all' recently

reported a high frequency of (+) hcrlabl PCR analysis in high-risk patients, including eight patients analyzed at more than one timepoint. Although not statistically significant, their analysis suggested an association with disease relapse and a (+) bcrlabl PCR assay, particularly between 6 and 24


ROTH ET AL

lool

h._ a

1 bcriabl assay

401

(t)

(41 f 13%)

6

12

18 24 30 36 42 Time (months) Post Transplant

48

54

Fig 4. Kaplan-Meier plot of actuarial probability of relapse for all patients having 0, 1, or 2 or more (+) bcr/ab/ assays (N = 64).

months post-BMT. However, some studies have found that the majority of patients remain bcrlabl (+) for years after BMT regardless of the stage of disease or the techniques used for GVHD prophylaxis, while still others have reported a very low frequency of detection and occasional transient The basis for these discordant results is unclear. While major differences may exist among the patient populations studied, these results could also reflect differences in assay sensitivity or, alternatively, the significant problem of PCR carryo~er.~' To minimize possible cross-contamination in our study, extensive technical precautions were implemented.z In addition, repeat RNA PCR analysis verified a (+) signal in 68 of 82 (+) samples. The discordant results from 14 samples (in 10 patients) could reflect contamination of the original sample or, alternatively, the presence of a small number of Ph' cells near the threshold for detection by the PCR assay. However, the overall conclusions from the statistical analysis were not altered by exclusion of these data. In the present study the presence of a (+) bcrlabl assay after BMT was associated with a significantly increased risk of disease relapse (Figs 1 and 4). Among relapsers, the median time from first (+) bcrlabl assay to relapse was 150 days (range 90 to 832). Evaluation of patients at multiple timepoints after BMT identified three distinct patterns of bcrlabl mRNA detection strongly associated with risk of relapse (Table 1). The highest incidence of relapse occurred in group I patients (persistently bcrlabl (+), 5 / 8 [actuarial relapse risk, 77%]), the lowest in group 111 patients (persistently bcrlabl (-), 0/23 [O%]), and intermediate in group I1 patients (both (+) and (-) assays, 5/23 [actuarial relapse risk, 20%]). Within this latter group, Table 2. Distribution of (+) bcrlabl Signals Relative to Time After BMT and Disease Outcome Time of Study (dl

< 30 31-120 121-365 > 365

Relapse*

6 (50) 9 (89) 8 (87) 2 (100)

Remission"

26 (46) 31 (6.5) 33 (12) 23 (39)

P

> .70 < ,001 < ,001 > .30

'Refers to the total number of patients studied within the indicated time period for each group. The numbers in parentheses represent the percent of patients with a (+) bcrlablassay.

some patients were transiently bcrlabl (+) in the early posttransplant period, while others were initially (-) or even episodically (+) (Figs 2 and 3). The finding of (+) and (-) assays in the same patient over time may reflect the presence in the RNA sample of an absolute number of Ph' cells near the threshold for detection by the PCR assay. Patients in this group may either not have had sufficient time to develop clinical relapse or to completely eliminate the remaining (+) cells. Only one patient was persistently (+) and alive beyond 100 days without relapse. This patient was first evaluated at day 660 after BMT with no data available from earlier timepoints. As shown in Fig 4, disease relapse was significantly more frequent as the number of bcrlabl (+) assays increased. Multiple (+) assays may reflect a higher number of persisting bcrlabl (+) tumor cells. Mixed (+) and (-) assays, as in group 11, might then be indicative of a number of Ph' cells close to the threshold of the RNA PCR assay. The relative number of Ph' cells may be directly related to the differing risk of relapse in these groups. Although not a quantitative assay, the observation that 7 of 10 patients with a (+) assay after only one round of PCR have subsequently relapsed is consistent with this hypothesis. Although classification as group 111 [bcrlabl (-)I was associated with an excellent prognosis in this study, median follow-up time is still relatively short (372 days). Late relapses of CML greater than 4 years after BMT have been reported as has a slow evolution to clinical relapse among patients known to be Ph' chromosome (+) after BMT,26 particularly among T-cell-depleted transplant recipienkzb Eight patients (nos. 11,13,15 through 18,22, and 23, Fig 2) with bcrlabl (-) assays within the first year post-BMT have become bcrlabl (+) after 1year. Two of these patients (nos. 18 and 22, Fig 2) have relapsed. Interestingly, this finding parallels the reports of late first detection ( > 1 year) of Ph' marrow metaphases in patients after BMT.27The initial RNA PCR (-) assays may again reflect numbers of Ph' cells below the threshold for detection, or alternatively the existence of Philadelphia chromosome (-) precursor leukemic cells. However, this latter hypothesis is not supported by recent gene transfer experiments demonstrating that expression of the bcrlabl oncogene and the p210 protein in mouse marrow is sufficient to produce a myeloproliferative disorder closely resembling human chronic-phase CML.*' A late perturbation in immunoregulatory mechanisms responsible for early leukemia cell suppression may account for delayed expansion of the leukemic clone from a previously low residual or dormant tumor The finding in some patients of a (+) bcrlabl assay from bone marrow cells with a simultaneous (-) assay in peripheral blood (patients 10 and 16, Fig 2) might also reflect early sequestration of tumor cells. Thirty patients in our study had concurrent cytogenetic and PCR analysis. The PCR assay was significantly more sensitive than cytogenetics, with 12 patients demonstrating detectable bcrlabl transcripts by PCR despite normal cytogenetics. No patient had a positive cytogenetic analysis with a negative bcrlabl PCR assay. In two patients, initial (+) bcrlabl analysis followed by subsequent (+) cytogenetics predated clinical relapse. Taken together, these data sug-


PROGNOSTIC SIGNIFICANCE OF PCR IN CML

gest a gradual increase in tumor-cell burden in those patients destined to relapse after allogeneic BMT. Leukemia is first detectable by PCR, followed by cytogenetic relapse, and finally clinical disease. We conclude that the bcrlabl PCR assay is a useful prognostic tool in the management of CML patients postBMT. The persistent detection of the bcrlabl transcript by PCR after allogeneic BMT for CML is associated with a significantly increased risk of relapse. In contrast, persistently bcrlabl (-) patients have a very low risk of relapse. As a significant number of patients may be transiently (+) in the early post-BMT period ( < 3 0 days), it is important that any assessment for prognosis should include multiple evaluations. The presence of multiple (+) assays between 1 month and 1year post-BMT is associated with a significant risk of early disease relapse. Cells containing bcrlabl may also persist for prolonged periods in some patients who

remain in clinical remission. The long-term outlook in these patients remains to be determined. NOTE ADDED IN PROOF

Since submission of this manuscript, two additional bcrlabl (+) patients (nos. 11 and 17, Fig 2) have progressed to clinical relapse. Both patients were first bcrlabl (+) at > 365 days post-BMT. The crossover of these patients into the relapse group now makes the association of a (+) bcrlabl assay beyond day 365 with disease relapse statistically significant (P < .06, Table 2).

ACKNOWLEDGMENT

We thank Sheila Norton and Danie Osbome for preparation of oligonucleotide primers and Mara Van Dusen for her excellent secretarial assistance in the preparation of this manuscript.

REFERENCES

1. Thomas ED, Clift RA, Fefer A, Appelbaum FR, Beatty P, Bensinger WI, Buckner D, Cheever MA, Deeg HJ, Doney K, Floumoy N, Greenberg P, Hansen JA, Martin P, McGuffin R, Ramberg R, Sanders JE, Singer J, Stewart P, Storb R, Sullivan K, Weiden PL, Witherspoon R: Marrow transplantation for the treatment of chronic myelogenous leukemia. Ann Intern Med 104155,1986 2. Champlin RE, Golde D W Chronic myelogenous leukemia: Recent advances. Blood 651039,1985 3. Goldman JM, Gale RP, Horowitz MM, Biggs JC, Champlin RE, Gluckman E, H o h a n n RG, Jacobsen SJ, Marmont AM, McGlave PB, Messner HA, Rimm AA, Rozman C, Speck B, Tura S, Weiner RS, Bortin MM: Bone marrow transplantation for chronic myelogenous leukemia in chronic phase. Ann Intern Med 108:806,1988 4. Baruchel A, Schaison G: Detection of minimal residual disease in leukemias: A critical approach. Nouv Rev Fr Hematol 32:21,1990 5. Roth MS, Antin JH, Bingham EL, Ginsburg D: Detection of Philadelphia chromosome positive cells by the polymerase chain reaction following bone marrow transplant for chronic myelogenous leukemia. Blood 742382,1989 6. Morgan GJ, Hughes T, Janssen JW, Cow J, Guo AP, Goldman JM, Wiedemann LM, Bartram CR: Polymerase chain reaction for detection of residual leukaemia. Lancet 1:928,1989 7. Lange W, Synder DS, Castro R, Rossi JJ, Blume K G Detection by enzymatic amplification of bcr-ab1 mRNA in peripheral blood and bone marrow cells of patients with chronic myelogenous leukemia. Blood 73:1735,1989 8. Gabert J, Lafage M, Maraninchi D, Thuret I, Carcassone Y, Mannoni P: Detection of residual bcrlabl translocation by polymerase chain reaction in chronic myeloid leukaemia patients after bone-marrow transplantation. Lancet 2:1125,1989 9. Bartram CR, Janssen JW, Schmidberger M, Lyons J, Arnold R Minimal residual leukaemia in chronic myeloid leukaemia patients after T-cell depleted bone-marrow transplantation (letter). Lancet 1:1260, 1989 10. Sawyers CL, Timson L, Kawasaki ES, Clark SS, Witte ON, Champlin R: Molecular relapse in chronic myelogenous leukemia patients after bone marrow transplantation detected by polymerase chain reaction. Proc Natl Acad Sci USA 87563,1990 11. Delfau M-H, Kerckaert J-P, d‘Hooghe MC, Fenaux P, Lai J-L, Jouet J-P, Grandchamp B: Detection of minimal residual disease in chronic myeloid leukemia patients after bone marrow transplantation by polymerase chain reaction. Leukemia 4:1, 1990

12. Hughes T, Martiat P, Morgan G, Sawyers C, Witte JM: Significance of residual leukaemia transcripts after bone marrow transplant for CML. Lancet 33550,1990 13. Hughes T, Morgan GJ, Martiat P, Goldman JM: Detection of residual leukemia after bone marrow transplant for chronic myeloid leukemia: Role of polymerase chain reaction in predicting relapse. Blood 772374,1991 14. Martiat P, Maisin D, Philippe M, Ferrant A, Michaux JL, Cassiman JJ, Van den Berghe H, Sokal G: Detection of residual BCRlABL transcripts in chronic myeloid leukaemia patients in complete remission using the polymerase chain reaction and nested primers. Br J Haematol75:355,1990 15. Pignon JM, Henni T, Amselem S, Vidaud M, Duquesnoy P, Vemant JP, Kuentz M, Cordonnier C, Rochant H, Goossens M: Frequent detection of minimal residual disease by use of the polymerase chain reaction in long-term survivors after bone marrow transplantation for chronic myeloid leukemia. Leukemia 4:83, 1990 16. Ash RC, Casper JT, Chitambar CR, Hansen R, Bunin N, Truitt RL, Lawton C, Murray K, Hunter J, Baxter-Lowe LA, Gottschall JL, Oldham K, Anderson T, Camitta B, Menitove J: Successful allogeneic transplantation of T-cell-depleted bone marrow from closely HLA-matched unrelated donors. N Engl J Med 322485,1990 17. Ash RC, Casper J, Lawton C, Hansen R, Chitambar C, Murray K, Camitta B, Anderson T, Truitt R: “Third generation” T depletion GVHD prophylaxis in genotypically HLA-matched sibling marrow transplantation: Durable engraftment with apparent preservation of the graft-versus leukemia (GVL) effect. Blood 74:163a, 1989 (abstr, suppl) 18. Antin JH, Bierer BE, Smith BR, Ferrara J, Guinen EC, Sieff C, Macklis RM, Tarbell NJ, Golan D, Lynch E, Reichert T, Blythman H, Bouloux C, Rappaport JM, Burakoff SJ, Weinstein HJ: Selective depletion of bone marrow T-lymphocytes with anti-CD5 monoclonal antibodies: Effective prophylaxis for graftversus-host disease in patients with hematologic malignancies. Blood (in press) 19. Hanson CA, Holbrook EA, Sheldon S, Schnitzer B, Roth MS: Detection of Philadelphia chromosome-positive cells from glass slide smears using the polymerase chain reaction. Am J Pathol 137:1,1990 20. Canaani E, Shtivelman E, Lifshitz B, Gale RP: An abl-bcr hybrid RNA in chronic myelogenous leukemia, in Gale RP, Golde DW (eds): Leukemia: Recent Advances in Biology and Treatment. New York, NY, Liss, 1985, p 29


21. Drobyski W, Thibodeau S, Truitt RL, Baxter-Lowe LA, Gorski J, Jenkins R, Gottschall J, Ash RC: Third-party-mediated graft rejection and graft-versus-host disease after T-cell-depleted bone marrow transplantation, as demonstrated by hypervariable DNA probes and HLA-DR polymorphism. Blood 74:2285,1989 22. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53:457, 1958 23. Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50:163,1966 24. Cox DR: Regression models and life-tables. J R Stat Soc [B] 34:187, 1972 25. Kwok S, Higuchi R: Avoiding false positives with PCR. Nature 339:237,1989 26. Hughes TP, Economou K, Mackinnon S, Vlitos M, Arthur CK, Guo AP, Rassool F, Apperley JF, Hows J, Goldman JM: Slow

ROTH ET AL

evolution of chronic myeloid leukaemia relapsing after BMT with T-cell depleted donor marrow. Br J Haematol73:462, 1989 27. Arthur CK, Apperley JF, Guo AP, Rassool F, Gao LM, Goldman JM: Cytogenetic events after bone marrow transplantation for chronic myeloid leukemia in chronic phase. Blood 71:1179, 1988 28. Daley GQ, Van Etten RA, Baltimore D: Induction of chronic myelogenous leukemia in mice by the P21(Y"'IRM gene of the Philadelphia chromosome. Science 247:824, 1990 29. Slavin S, Ackerstein A, Naparstek E, Or R, Weiss L The graft-versus-leukemia (GVL) phenomenon: Is GVL separable from GVHD? Bone Marrow Transplant 6:155, 1990 30. Truitt RL, Atasoylu AA: Impact of pretransplant conditioning and donor T cells on chimerism, graft-versus-host disease, graft-versus-leukemia reactivity, and tolerance after bone marrow transplantation. Blood (in press)