Cyclophosphamide treatment expands the circulating hematopoietic stem cell pool in dogs.

Increased numbers of circulating granulocyte-monocyte precursor cells (CFUc) have been observed in the peripheral blood of man after antineoplastic chemotherapy. We have developed a canine model to study the biologic significance of this phenomenon for hematopoietic reconstitution following hematopoietically lethal exposure to total body irradiation (TBI). After cyclophosphamide administration, a 16-fold expansion of circulating CFUc numbers was observed during the period of rapid leukocyte recovery that occurred after the chemotherapy-induced leukocyte nadir. We had previously noted this association between leukocyte recovery and CFUc expansion in our human studies. After 900 rad TBI hematopoietic reconstitution was attempted with autologous, cryopreserved collections of peripheral blood mononuclear cells obtained either at times of post-cyclophosphamide CFUc expansion (group A, 14 dogs) or without CFUc expansion (group B, 12 dogs). Asd compared to group B collections, group A collections contained 11-fold more CFUc and were 12.5-fold more potent in fostering hematopoietic recovery after TBI. These results suggest that the expansion of CFUc numbers we observed was accompanied by a similar expansion of more primitive hematopoietic stem cell numbers. We conclude that chemotherapy-induced expansion of circulating CFUc numbers appears to be of substantial import in effecting hematopoietic reconstitution--an observation that may be of significance for further studies of autologous hematopoietic reconstitution in man.

Interferon-alpha restores the deficient expression of the cytoadhesion molecule lymphocyte function antigen-3 by chronic myelogenous leukemia progenitor cells.

Hematopoietic cells from the malignant clone in chronic myelogenous leukemia (CML) maintain and expand a proliferative advantage over normal hematopoietic cells within the bone marrow. This advantage is often ameliorated or reversed in vivo by IFN alpha. Based upon earlier studies suggesting decreased adhesiveness of CML progenitor cells, we asked whether CML progenitor cells are deficient in their expression of the cytoadhesion molecule lymphocyte function antigen-3 (LFA-3, CD58) which is normally expressed on hematopoietic progenitors. Progenitor cells from untreated CML patients showed greatly reduced or absent LFA-3 expression, whereas progenitors from patients treated with IFN alpha in vivo or in vitro expressed surface LFA-3 at more normal levels. LFA-3-deficient CML progenitor cells were unable to stimulate normal regulatory proliferative responses in autologous T cells. We hypothesize that IFN alpha-sensitive LFA-3 deficiency reflects a cell surface cytoadhesion defect which may help explain adhesive abnormalities of CML progenitor cells in vitro and clonal proliferation in vivo.

International Society for Cell and Gene Therapy of Cancer: 2005 meeting in Shenzhen, China.

The 2005 International Society for Cell and Gene Therapy of Cancer (ISCGT) Congress was held in Shenzhen, China (www.iscgtchina2005.com) from December 9th-11th 2005. Here, we describe a representation of the most seminal presentations providing an overview of the progress in the field of cancer gene therapy including the successful introduction of the first approved gene therapy drug.

A novel interferon-inducible domain: structural and functional analysis of the human interferon regulatory factor 1 gene promoter.

We have cloned and functionally characterized the human interferon regulatory factor 1 (IRF-1) gene promoter. The promoter contains a CpG island, with several GC boxes, a CAAT box, but no TATA box. IRF-1 mRNA is strongly induced by gamma interferon (IFN-gamma) but more weakly and transiently by IFN-alpha. There are several putative kappa B motifs and numerous AA(G/A)G(G/T)A and GAAANN motifs throughout the promoter. The IRF-1 promoter is not autoregulated by the IRF-1 gene product. IFN inducibility of the promoter was studied with 5' deletion mutants linked to a heterologous reporter gene. Gel mobility shift assays were used to show IFN-inducible factor binding to the IRF-1 promoter. These studies showed that IFN inducibility is conferred by a novel imperfect inverted-repeat arrangement of two GAAANN motifs within a domain, 130 nucleotides upstream of transcription initiation. This inverted repeat binds a factor upon induction with IFN and can confer IFN inducibility on a heterologous promoter. Conversely, point mutations of the inverted repeat are not IFN inducible when linked to the same heterologous promoter.

Wild-type human p53 and a temperature-sensitive mutant induce Fas/APO-1 expression.

Fas/APO-1 is a cell surface protein known to trigger apoptosis upon specific antibody engagement. Because wild-type p53 can activate transcription as well as induce apoptosis, we queried whether p53 might upregulate Fas/APO-1. To explore this possibility, we examined human p53-null (H358 non-small-cell lung adenocarcinoma and K562 erythroleukemia) and wild-type p53-containing (H460 non-small-cell lung adenocarcinoma) cell lines. When H358 or H460 cells were transduced with a replication-deficient adenovirus expression construct containing the human wild-type p53 gene but not with vector alone, a marked upregulation (approximately a three-to fourfold increase) of cell surface Fas/APO-1 was observed by flow cytometry. Similarly, K562, cells stably transfected with a plasmid vector containing the temperature-sensitive human p53 mutant Ala-143 demonstrated a four- to sixfold upregulation of Fas/APO-1 by flow-cytometric analysis at the permissive temperature of 32.5 degrees C. Temperature-sensitive upregulation of Fas/APO-1 in K562 Ala-143 cells was verified by immunoprecipitation and demonstrated to result from enhanced mRNA production by nuclear run-on and Northern (RNA) analyses. Stably transfected K562 cells expressing temperature-insensitive, transcriptionally inactive p53 mutants (His-175, Trp-248, His-273, or Gly-281) failed to upregulate Fas/APO-1 at either 32.5 degrees or 37.5 degrees C. The temperature-sensitive transcription of Fas/APO-1 occurred in the presence of cycloheximide, indicating that de novo protein synthesis was not required and suggested a direct involvement of p53. Collectively, these observations argue that Fas/APO-1 is a target gene for transcriptional activation by p53.

Expression of epidermal growth factor receptor and human papillomavirus E6/E7 proteins in cervical carcinoma cells.

BACKGROUND: Epidermal growth factor receptor (EGF-R) proteins are highly expressed in many tumors, including those of the cervix. We have observed previously that the introduction of a transcription unit containing an antisense sequence for the E6/E7 genes of human papillomavirus (HPV) 18, along with a transcription unit containing a sense complementary DNA sequence for the wild-type retinoblastoma (Rb) gene, decreased the growth of human cervical carcinoma HeLa cells (HPV 18 positive) both in vitro and in vivo. To clarify the regulatory mechanisms by which this reduction in cell proliferation occurred, we studied the expression of EGF-R proteins in these cells. METHODS: Western blot and northern blot techniques were used to measure EGF-R expression, and a pulse-chase immunoprecipitation assay was used to measure the stability of EGF-R protein in HeLa cells and HeLa cells that had been transfected with the antisense E6/E7 or sense Rb sequences. Cell proliferation was measured by use of a tetrazolium-based colorimetric assay for numbers of viable cells. RESULTS: The introduction of sense Rb or antisense E6/E7 transcription units or a combination of these two transcription units into HeLa cells dramatically decreased the level of EGF-R proteins in these cells; EGF-R levels were not affected at the transcriptional level but at the post-transcriptional level. Addition of the anti-EGF-R-specific monoclonal antibody 225mAb to HeLa cells caused 53% (95% confidence interval = 44%-62%) growth inhibition. CONCLUSIONS: These results suggest that HeLa cervical carcinoma cells are dependent on EGF-R for proliferation and that changes in functional levels of the E6/E7 HPV proteins and endogenous Rb proteins may alter the growth rate of cervical cancer cell lines by reducing the stability of EGF-R at the post-transcriptional level.

Cisplatin dose intensity in non-small cell lung cancer: phase II results of a day 1 and day 8 high-dose regimen.

Between October 1985 and March 1987, 92 patients were registered on a phase II study of the Northern California Oncology Group investigating the importance of dose intensity in the treatment of advanced non-small cell lung cancer (NSCLC). Treatment consisted of high-dose cisplatin in hypertonic saline (200 mg/m2 on a 28-day cycle) given in a divided day 1 and day 8 schedule. The response rate among 76 assessable patients was 36% (27/76), with complete response (CR) in 8% (6/76) and partial response (PR) in 28% (21/76). If all patients receiving any drug therapy were considered, the overall response rate was 31% (27/87), with CR in 7% (6/87) and PR in 24% (21/87). Median survival times for all assessable patients and all patients receiving any therapy were 37 and 35 weeks, respectively. With the use of a protocol design specifying dose delays rather than dose reduction for toxicity, the mean dose intensity delivered was 47.2 mg/m2 per week, or 94% of projected. Compared with other dose-intensive regimens of cisplatin, this day 1 and day 8 schedule was relatively well tolerated, with peripheral neuropathy as the dose-limiting toxicity. The data on response and median survival times among patients receiving this single-agent therapy are encouraging. They support the potential importance of cisplatin dose intensity in the treatment of NSCLC. Whether these results represent a positive dose-response effect in NSCLC will be tested in a randomized comparative trial of high-dose versus standard-dose cisplatin therapy.

Prognostic indicators of tumor response to Staphylococcus aureus Cowan strain I plasma perfusion.

Ten tumor-bearing dogs were treated with passage of autologous plasma over fixed Staphylococcus aureus Cowan strain I. Five similar dogs were treated identically except for the exposure to S. aureus. These animals have been assessed to identify positive and negative prognostic variables for response. Nonresponder treated animals had significantly larger chest wall tumor bulk than did the responder and control groups (P less than .01). Responder animals had fewer initial circulating immune complexes than did the nonresponders, though each group had similar reductions in immune complexes with therapy. Nonresponder animals had smaller volumes of plasma processed per kilogram of body weight per procedure than did controls (P = .016), whereas responder and control animals had similar volumes processed per kilogram of body weight per procedure (P = .84). These data suggest that the response observed in our original series was significantly related to the larger amount of plasma treated per procedure and suggest that a factor may be eluted from the S. aureus cartridge that mediates this response.

Randomized trial of high-dose chemotherapy and blood cell autografts for high-risk primary breast carcinoma.

BACKGROUND: Uncontrolled studies have reported encouraging outcomes for patients with high-risk primary breast cancer treated with high-dose chemotherapy and autologous hematopoietic stem cell support. We conducted a prospective randomized trial to compare standard-dose chemotherapy with the same therapy followed by high-dose chemotherapy. PATIENTS AND METHODS: Patients with 10 or more positive axillary lymph nodes after primary breast surgery or patients with four or more positive lymph nodes after four cycles of primary (neoadjuvant) chemotherapy were eligible. All patients were to receive eight cycles of 5-fluorouracil, doxorubicin (Adriamycin), and cyclophosphamide (FAC). Patients were stratified by stage and randomly assigned to receive two cycles of high-dose cyclophosphamide, etoposide, and cisplatin with autologous hematopoietic stem cell support or no additional chemotherapy. Tamoxifen was planned for postmenopausal patients with estrogen receptor-positive tumors and chest wall radiotherapy was planned for all. All P values are from two-sided tests. RESULTS: Seventy-eight patients (48 after primary surgery and 30 after primary chemotherapy) were registered. Thirty-nine patients were randomly assigned to FAC and 39 to FAC followed by high-dose chemotherapy. After a median follow-up of 6.5 years, there have been 41 relapses. In intention-to-treat analyses, estimated 3-year relapse-free survival rates were 62% and 48% for FAC and FAC/high-dose chemotherapy, respectively (P =.35), and 3-year survival rates were 77% and 58%, respectively (P =.23). Overall, there was greater and more frequent morbidity associated with high-dose chemotherapy than with FAC; there was one septic death associated with high-dose chemotherapy. CONCLUSIONS: No relapse-free or overall survival advantage was associated with the use of high-dose chemotherapy, and morbidity was increased with its use. Thus, high-dose chemotherapy is not indicated outside a clinical trial.

Changes in the translational activity of polyadenylated messenger RNA of HL60 promyelocytic leukemia cells associated with myeloid or macrophage differentiation.

In order to characterize the events which commit the HL60 human promyelocytic leukemia cell line to differentiate into macrophages or mature myeloid cells, we have analyzed the in vitro [35S]methionine-labeled translational products obtained from polyadenylated messenger RNA of the HL60 cells before and after exposure to: (a) dimethylformamide (DMF), an inducer of myeloid differentiation; (b) 12-O-tetradecanylphorbol-13-acetate (TPA), an inducer of macrophage differentiation; or (c) a combination of the two inducers. Exposure of the HL60 cells to either TPA or DMF results in decreases in the relative abundancy of translational products with molecular weights of 20,000, 17,000, and 15,000. Exposure of the HL60 cells so as to generate macrophage differentiation results in elevations of translational products with molecular weights of 60,000, 47,000, 42,000, 32,000, 27,000, 14,000, and 12,300, while DMF-induced myeloid differentiation is associated with increases in the abundancy of translational products with molecular weights of 60,000, 42,000, 35,000, 32,000, 27,000, 13,000 and 12,300. The addition of the macrophage inducer TPA to HL60 cells previously exposed to the myeloid inducer DMF results in changes in the relative abundance of several translational products, yielding a pattern which differs quantitatively from that obtained from cells treated with DMF or TPA alone. These changes in the relative abundancies of the HL60 translational products suggest that the steady state levels of several different populations of mRNA or the ability of these mRNAs to be translated are being modified during the induction of myeloid or macrophage differentiation in the HL60 promyelocytic leukemia cell line.

Amplification of circulating granulocyte-monocyte stem cell numbers following chemotherapy in patients with extensive small cell carcinoma of the lung.

Circulating numbers of committed granulocyte-monocyte hematopoietic stem cells (CFUc) were measured in the peripheral blood of 20 patients with extensive-stage small cell lung carcinoma during induction chemotherapy. All patients received cyclophosphamide, doxorubicin, VP16-213, and vincristine. CFUc measurements were made either weekly or twice weekly. As leukocytes declined following chemotherapy, circulating CFUc numbers also declined. However, as leukocytes recovered from their nadir levels, circulating CFUc numbers per mononuclear cell and per ml of whole blood became substantially expanded in 19 and 17, respectively, of the 20 patients studied. Per mononuclear cell, the median CFUc expansion was 7.9-fold, and the highest expansion seen was 157-fold. Per mol of blood, the median CFUc expansion was 6.7-fold, and the highest expansion seen was 46-fold. The magnitude of the amplification, its occurrence in 85 to 95% of patients studied, and its association with leukocyte recovery strongly suggest that appropriately timed collections of peripheral blood mononuclear cells obtained during leukocyte recovery from nonablative chemotherapy could be used to provide hematopoietic stem cells in numbers sufficient to effect hematopoietic reconstitution after subsequent marrow-ablative therapy.

Hyperphosphorylation of p53 induced by okadaic acid attenuates its transcriptional activation function.

The tumor suppressor and transcriptional factor p53 is a phosphorylated protein. Its phosphorylation states are regulated by several protein kinases and phosphatases. In this study, the wild-type p53 was transfected and expressed in chronic myelogenous leukemia K-562 cells. Incubation of the transfected cells with okadaic acid, an inhibitor of serine phosphatases 2A and 1, induced hyperphosphorylation of p53 protein. The treatment also increased the steady state level of p53 protein in the cells. However, the hyperphosphorylated p53 protein was less active in promoting transcription mediated by two p53-binding DNA elements, the ribosomal gene cluster and the p53 consensus DNA-binding sequence. Nevertheless, the decreased transcription activation was not due to decreased binding of p53 to these elements, as analyzed by mobility shift DNA-binding assays. In addition, the treatment did not induce a conformational change in p53, as assayed by two conformation-specific anti-p53 monoclonal antibodies, PAb240 and PAb1620. These results suggest that the phosphorylation induced by okadaic acid may selectively modulate the transcription activation function of p53. Consequently, phosphorylation may represent a mechanism of p53 inactivation in tumor cells that harbor the wild-type p53.

Retinoblastoma protein expression is frequently altered in chronic lymphocytic leukemia.

The detection of abnormalities at the retinoblastoma (RB) locus by cytogenetics, Southern blot, and fluorescence in situ hybridization studies suggests that the RB gene has a role in chronic lymphocytic leukemia (CLL). To further study this role, we determined the level of RB protein present in the mononuclear cell fraction derived from peripheral blood or bone marrow samples from 74 patients with CLL, by Western blotting. Compared to similarly prepared samples from the peripheral blood of normal individuals, the level of RB in CLL cells was less than normal in 42% of patients, equal to normal in 22% of patients, and in excess of normal in 36% of patients. Regardless of whether the source of the sample was blood or marrow or if the patients were untreated or previously treated, similar rates of low, normal, and elevated RB levels were observed. RB protein in the CLL patient samples was never phosphorylated. RB levels showed no correlation with the lymphocyte doubling time or with proliferating cell nuclear antigen levels. Low RB levels could arise from genetic alterations of the RB gene or altered regulation of expression. To determine which was occurring, we stimulated the cells from 27 CLL patients in culture with either phytohemagglutinin or pokeweed mitogen in an attempt to induce RB expression and phosphorylation. Among patients with low levels of RB, expression was induced in 46% (6 of 13), and phosphorylation of RB was seen in 31% (4 of 13). Increased expression of phosphorylated RB was induced in 80% (4 of 5) of patients with normal levels of RB and in 78% (7 of 9) of patients with high levels of RB. This study demonstrates that absent RB expression occurs commonly in patients with CLL. Intrinsic abnormalities of the RB gene may be present in those patients with low levels of RB that could not be stimulated by mitogens, while regulatory abnormalities located in trans to the RB gene may occur in the other half. Given the importance that RB levels play in other cancers, the prognostic implication of low RB levels should be studied in CLL.

p53-independent induction of WAF1/CIP1 in human leukemia cells is correlated with growth arrest accompanying monocyte/macrophage differentiation.

The p53 tumor suppressor gene plays a role in controlling a G1 phase checkpoint. The WAF1/CIP1 gene with encodes p21WAF1/CIP1 protein, an inhibitor of cyclin-dependent kinases, is a downstream mediator of p53 function. We examined expression of the WAF1/CIP1 gene and its relationship to growth arrest and differentiation in p53-null human leukemic cell lines. We show that p53-independent induction of WAF1/CIP1 occurs in human leukemia cells treated with 12-O-tetradecanoylphorbol-13-acetate, okadaic acid, or IFN-gamma but not with retinoic acid, vitamin D3, or DMSO. Furthermore, WAF1/CIP1 induction correlates with growth arrest associated with monocyte-macrophage differentiation. The present studies support the idea that WAF1/CIP1 gene expression can be regulated through multiple mechanisms, suggesting that strategies may be designed to restore the G1 checkpoint controls in p53-null cells by targeting these p53-independent mechanisms of WAF1/CIP1 induction.

The requirement of the carboxyl terminus of p53 for DNA binding and transcriptional activation depends on the specific p53 binding DNA element.

The p53 protein specifically binds DNA sequences and activates transcription. In this study, we investigated the requirement of the carboxyl (C)- or amino (N)-terminal domain of p53 for the binding and transactivation of two DNA elements, p53CON and the ribosomal gene cluster (RGC). Three human p53 mutants with deletion in either the C-terminus or N-terminus were used. Mobility-shift assays showed that the oligomerization-defective mutant p53(1-326), from which the final 67 C-terminal amino acids were deleted, retained the wild-type p53's ability to bind the p53CON element in the presence of anti-p53 monoclonal antibody PAb1801. Also, the transient transfection assays showed that the mutant p53(1-326) activated p53CON-mediated transcription. However, this mutant failed to bind the RGC element and was unable to activate RGC-mediated transcription. Thus, the requirement of the C-terminal region of p53 for DNA binding and transcription activation varies with the p53-binding DNA element under study. In contrast, the N-terminus of p53 contains a common transcription activation domain: deletion of the 80 or 159 N-terminal amino acids inactivated both p53CON- and RGC-mediated transactivation. Furthermore, mobility-shift assays did not detect any binding to p53CON- and RGC by either of the two N-terminal-deletion mutants. These results suggest that the N-terminus of p53 affects DNA binding.

Novel DNA binding of p53 mutants and their role in transcriptional activation.

The protein product of the normal p53 gene binds to the DNA p53CON element (GGACATGCCCGGGCATGTCC, Funk et al., 1992), thereby activating transcription from adjacent promoters. Two mutants, 248 (Arg-->Trp) and 281 (Asp-->Gly), failed to bind p53CON and to activate transcription. However, in contrast to previous reports that all p53 mutants fail to bind to the other p53 binding elements, two p53 mutants, 143 (Val-->Ala) and 273 (Arg-->His), retained both p53CON binding and transcriptional activation functions. A third mutant 175 (Arg-->His) bound to the p53CON but did not activate transcription. These data suggest that the DNA binding and transcriptional activation functions of p53 mutants in tumor cells are dependent on the specific missense mutations acquired in the p53 gene and the target sequences of p53 in the genome.

Activation of p21WAF1/Cip1 expression by a temperature-sensitive mutant of human p53 does not lead to apoptosis.

The cyclin-dependent kinase inhibitor p21WAF1/Cip1 (p21) inhibits cellular proliferation and has been implicated in p53-dependent apoptosis. In this study, we examined the regulation of p21 in the K562 erythroleukemia cell line using a human temperature-sensitive mutant of p53, 143Ala. We showed that 143Ala at the permissive temperature (32 degrees C) activated the expression of p21. Stable cell lines expressing 143Ala or other p53 mutants were established. We then examined whether elevation of p21 promotes apoptosis and sensitized cells to radiation-induced apoptosis. Our results showed that p21 elevation did not promote or sensitize K562 cells to apoptosis.

Peptide containing the BCR oligomerization domain (AA 1-160) reverses the transformed phenotype of p210bcr-abl positive 32D myeloid leukemia cells.

We first showed that the introduction of a bcr-abl transcription unit into the 32D murine myeloid cell line (P210bcrabl32D) converts this cell line from an IL3 dependent cell line to an IL3 growth independent cell line. We next cloned a fragment of the bcr-abl cDNA, which codes for the bcr oligomerization domain and neighboring regions. To test for a transformation inhibitory effect of this oligomerization inhibitory peptide transcription unit on the p210bcr-abl mediated IL3 independent growth of the P210bcrabl32D cell line, we transiently co-electroporated into the growth factor dependent 32D cells, mixtures of plasmids which contained varying ratios of the plasmid expression vectors for the bcr oligomerization inhibitory peptide along with a smaller amount of the plasmid expression vector for the full length p210bcr-abl. (The P210bcr-abl protein converts the 32D from a growth factor dependent into a growth factor independent cell line.) We then showed that the oligomerization domain containing fragment from the bcr and bcr-abl proteins, can be used to inhibit the IL3 independent growth of p210bcr-abl positive 32D cells. These studies may be of eventual interest for those investigators whose goal is to design molecular therapeutic approaches to CML based on the use of peptidomimetic chemical functionalities, which mimic the structure and the inhibitory binding properties of the oligomerization domain containing fragment so as to inhibit the transforming function of the P210bcr-abl oncoprotein.

The presence of the Rb c-box peptide in the cytoplasm inhibits p210bcr-abl transforming function.

In order to test if the carboxyl terminal polypeptide of the Retinoblastoma (Rb) tumor suppressor protein, could be used to suppress the growth factor-independent growth phenotype of p210bcr-abl positive myeloid cells, we introduced a truncated form of the 3' end of the Rb cDNA encoding its last 173 amino acid residues (Rb C-box) which localize into the cytoplasm where the p210bcr-abl transforming protein is found, into myeloid cells (32D) which depends on the p210bcr-abl protein for IL3 growth factor-independent growth (32D-p210). The expression of the plasmid vectors carrying the Rb C-box cDNAs was shown to inhibit the abl tyrosine specific protein kinase activity of the p210(bcr-abl) oncoprotein and to suppress the IL3-independent growth phenotype of the 32D-p210 cells. The Rb C-box polypeptides did not suppress the growth of the untransfected 32D parental cell line in methylcellulose in the presence of IL3-conditioned medium. These results suggest that the cytoplasmic localization of the p210(bcr-abl) allows it to escape the effect of intranuclear proteins such as Rb which negatively regulate the p145(c-abl) kinase.

Effect of prior interferon alfa therapy on the outcome of allogeneic bone marrow transplantation for chronic myelogenous leukemia.

PURPOSE: To determine whether prior interferon alfa (IFN-A) treatment affects the outcome of allogeneic bone marrow transplantation. PATIENTS AND METHODS: We analyzed the outcome of 77 patients with chronic myelogenous leukemia (CML) who received transplants from an HLA-identical donor using a total-body irradiation-containing preparative regimen. Engraftment, acute and chronic graft-versus-host disease (GVHD), survival, and disease-free survival were compared between patients who had previously received interferon (IFN+) to those who had not (IFN-). Forty-one patients were transplanted in chronic phase and 36 had more advanced CML. The IFN+ group had received IFN-A in doses of 3 to 5 x 10(6) U/m2 three times a week or more for at least 4 weeks anytime before transplantation. RESULTS: For patients in chronic phase, there were no significant differences between the IFN+ group and the IFN- group in regard to neutrophils recovery more than 1.0 x 10(9)/L (29 v 24), platelet recovery more than 50 x 10(9)/L (33 v 36), incidence of grade II to IV GVHD (23% v 28%), incidence of chronic GVHD (39% v 47%), disease-free survival (46% +/- 11% v 59% +/- 13%), relapse (9% v 11%), or 100-day transplant-related mortality (22% v 16%). In patients with more advanced stage disease, there was also no significant differences between the IFN+ group and the IFN- group in regard to these outcomes. CONCLUSION: Prior treatment with IFN-A did not adversely affect transplant outcome. Further studies are required to better understand the complementary roles of IFN-A and allogeneic bone marrow transplantation for the treatment of CML.