Administration Of BCNU Research Articles

Effects of hypoxia on drug resistance phenotype and genotype in human glioma cell lines.

Recurrent gliomas are most often treated by chemotherapy. However, these tumors typically acquire resistance to most drugs administered, and patients will usually die of recurrent tumor. Factors which may play a role include overexpression of putative multidrug resistance genes, such as the multidrug resistance gene 1 (MDR1), multidrug resistance associated protein gene (MRP), 06-alkylguanine, DNA alkyltransferase gene (06MT) and excision repair cross complementing gene 1 (ERCC1). Tumor hypoxia has also been shown to be associated with drug resistance in other soft tissue tumors. Since gliomas have regions of diminished oxygenation, and have clinical resistance to chemotherapy, the relationship between phenotypic resistance to chemotherapy after hypoxic exposure and expression of drug resistance genes was investigated in glioma cell lines (U373 MG, PFAT-MT). After a 24 hour exposure to hypoxia, drugs 1, 3-bis, 2-chloroethyl-1-nitrosurea (BCNU) and cis-diammine, dichloroplatinum II (CDDP) were administered, and cell survival was determined. Hypoxic exposure was associated with increased survival of the cell lines after administration of BCNU and CDDP, with resistance to BCNU 15 to 30-fold when compared to cells which did not undergo hypoxic exposure. Both tumor cell lines also showed some degree of resistance to CDDP, although not to the extent of BCNU (2 to 3-fold increased resistance). The expression of the drug resistance genes was found to be unchanged when comparing cells which had undergone hypoxic exposure and those which had not. Thus, hypoxic exposure is associated with substantial drug resistance in brain tumor cell lines. The lack of correlation between the induced phenotype and known drug resistance genes suggests other mechanisms may be acting in these tumors in hypoxic conditions.

Treatment of subcutaneous and intracranial brain tumor xenografts with O6-benzylguanine and 1,3-bis(2-chloroethyl)-1-nitrosourea.

O6-Alkylguanine-DNA alkyltransferase (AT) is a cellular protein that protects cells from the cytotoxic effects of nitrosoureas by repairing alkyl lesions at the O6 position of guanine. We have studied the ability of O6-benzylguanine to deplete AT activity in brain tumor xenografts and thereby increase the sensitivity of these tumors to 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). In toxicity studies, pretreatment of athymic mice with O6-benzylguanine increased the toxicity of BCNU significantly. After i.p. injection of O6-benzylguanine into athymic mice carrying subcutaneous (s.c.) D341MED, a human medulloblastoma xenograft with a high AT activity, the AT activity of the tumors became undetectable within 1 h and remained depleted until 36 h. In s.c. xenografts to D341MED, treatment with O6-benzylguanine followed 1 h later by BCNU produced a significantly greater growth delay (14.8 days) than was seen with BCNU alone (2.3 days). A lower pretreatment dose of O6-benzylguanine produced a significantly smaller therapeutic effect. Delaying the administration of BCNU until 36 h after O6-benzylguanine resulted in a growth delay (1.2 days) that was not significantly different from that produced by the control or BCNU alone. In athymic mice with intracranial (i.c.) xenografts of D341MED, pretreatment with O6-benzylguanine followed 1 h later by BCNU produced a significantly increased survival as compared with that of the control, BCNU alone, O6-benzylguanine alone, and O6-benzylguanine followed 36 h later by BCNU. In experiments with s.c. xenografts of D245MG, a human glioma xenograft with undetectable AT activity, pretreatment with O6-benzylguanine 1 h prior to BCNU produced a significantly greater effect than was seen with BCNU treatment alone. The combination regimen, however, was not as effective as an equitoxic dose of BCNU alone. These studies suggest that O6-benzylguanine may be a useful adjuvant to nitrosourea therapy in human malignancies that exhibit a range of AT activities and that dose and timing are important variables in achieving therapeutic success. These data also indicate that therapeutic potentiation of BCNU by O6-benzylguanine can be achieved in i.c. tumors. As a result, this approach may be useful in the treatment of neoplasms of the central nervous system.

Site‐specific drug delivery to the lung

AbstractInvestigations were undertaken to determine whether anti‐cancer drugs introduced locally into the lung using bioerodible polyanhydride microspheres as carriers would demonstrate both efficacy and reduced toxicity.BCNU (carmustine) and CDDP (cisplatin), loaded in polyanhydride microspheres, were administered to mice bearing either of two tumors selected for their affinity for the lung: the B16F10 melanoma and the recently established GL26F4 glioma. Following intravenous inoculation of these tumor cells, the number of metastatic foci formed in the lung follows a predictable time course and can readily be determined. Comparisons were made between the efficacy of microspheres introduced into the lung by intratracheal intubation (IT) and of those administered by intraperitoneal injection (IP). Administration of microspheres loaded either with BCNU or cisplatin reduced the detectable metastatic foci by 25–90% depending on the tumor load, both when administered IP and IT.Toxicity was assessed by flow cytometric analysis of bone marrow cellularity as well as by determination of mortality rates. Intratracheal administration of either cisplatin or BCNU reduced the deleterious systemic effects observed when the drug was administered by IP injection. This was seen at high drug levels, where significant mortality occurred only in animals given drug injections IP; and at lower levels where IP injection led to a reduction of bone marrow blast cells, while IT administration caused no detectable effect on marrow cellularity.Since local delivery of BCNU or cisplatin through bioerodible polyanhydride microspheres induces significantly less systemic damage while demonstrating efficacy equal to or exceeding that of IP injection, this mode of drug delivery warrants further and more detailed exploration. Moreover, the method may be applicable to the treatment of other chronic pulmonary pathologies.

Cyclophosphamide decreases O6-alkylguanine-DNA alkyltransferase activity in peripheral lymphocytes of patients undergoing bone marrow transplantation.

O6-alkylguanine-DNA-alkyltransferase (ATase) levels were measured in extracts of peripheral blood lymphocytes taken at various times during chemotherapy from 19 patients with various haematological malignancies. Seven patients with advanced Hodgkin's disease received preparative treatment consisting of cyclophosphamide (1.5 g m-2, daily) administered on days 1 to 4 and BCNU (600 mg m-2) on day 5 prior to autologous bone marrow rescue (ABMR) delivered on day 7. Treatment in the remaining 12 patients consisted of cyclophosphamide (1.8 g m-2, daily) given on days 1 and 2 followed at day 4 with total body irradiation (TBI) administered in six fractions over the subsequent 3 days to a total dose of 1200 cGy prior to bone marrow transplantation. In the Hodgkin's group, significant decreases in ATase activity were seen during the cyclophosphamide treatment, and the median ATase nadir was 32% (range 0% to 57%) of pretreatment levels following 4 days of cyclophosphamide. In one patient, no ATase activity was detectable following the 4th cyclophosphamide treatment. ATase activities decreased further after BCNU administration to a median of 19% (range 0% to 32%) of pretreatment levels. Extensive cyclophosphamide-induced reduction of lymphocyte ATase levels was also seen in the other group of 12 patients treated with cyclophosphamide/TBI: postcyclophosphamide median ATase nadir was 35% (range 12% to 78%) of the pretreatment levels. No ATase depletion was seen when cyclophosphamide (up to 10 mM) was incubated for 2 h with pure recombinant human ATase in vitro whereas ATase activity was reduced by 90% on preincubation with 100 microns acrolein or with greater than 1 mM phosphoramide mustard. This suggests that a cyclophosphamide-induced decrease in ATase levels in human peripheral lymphocytes in vivo may be due to depletion mediated by the production of intracellular acrolein. Since ATase appears to be a principal mechanism in cellular resistance to the cytotoxic effects of BCNU and related alkylating agents, these observations suggest that a cyclophosphamide-induced reduction in ATase activity may be an additional factor in the effectiveness of the combined sequential therapy.

Neurotoxicity after intracarotid 1,3-bis(2-chloroethyl)-1-nitrosourea administration in the rat

Changes in blood-brain (BBB) permeability and local cerebral blood flow after intracarotid administration of 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) were examined quantitatively in rats with double-tracer autoradiography using [14C]alpha-amino-isobutyric acid and [18F]fluoroantipyrine. Forty-eight female Wistar rats were divided into four groups. The control group (Group 1) received 1 ml of 5% dextrose. The other three groups received three different doses of BCNU dissolved in 5% dextrose: Group II rats received 1 mg, Group III 3 mg, and Group IV 10 mg. The tracer study was performed on Day 1 or Days 4 to 12 after intracarotid administration of BCNU. In 11 rats in Group II, there were no changes of BBB permeability. Transient BBB permeability changes were seen in the striatum or hippocampus in 3 of the 5 rats (60%) in Group III within 24 hours. In 8 of 9 rats (89%) in the same group, late BBB permeability changes were observed in the hypothalamus with or without histological changes. BBB permeability changes were seen in all rats of Group IV. Focal increase of local cerebral blood flow on the infused side compared with the non-infused side of the brain was observed, although not at a significant level, in 5 of 25 rats examined with [18F]fluoroantipyrine. The results of BBB permeability and histological examinations and study of heterogenous distribution by [18F]fluorodeoxyglucose indicated that the ipsilateral subcortical structures such as the hypothalamus, amygdala, internal capsule, and caudate putamen have the highest incidence of neurotoxicity, which are closely related to histopathological damage seen in human BCNU leucoencephalopathy.(ABSTRACT TRUNCATED AT 250 WORDS)

Chromosome numbers and DNA-content in intra-cerebrally transplanted experimental gliomas

Serially transplanted experimental tumors of the central and peripheral nervous system can be used as models to investigate open questions in human neurooncology. Altered susceptibility of higher passages to chemotherapy might be correlated with chromosome number and DNA-content variations which would be partly expressed as changes in proliferation behaviour. Karyotypes therefore were analysed in the 73rd to 90th generations of transplanted experimental gliomas. Wide variation of chromosome number was observed; 2 major types of distribution occurred, the one presenting with, the other without stemlines. Large chromosomes # 1 and # 4 were often monosomic, while small chromosomes of ## 8 to 20 were increased up to the fivefold. Lines with prominent and few markers were observed. On the whole, cells of the proliferating pool of the tumor had to be considered as hypotetraploid. Comparison of chromosome numbers and DNA content gave good correlation; differences between the 2 were explained by the fact that only the number of chromosomes was taken into account, regardless of whether small or large chromosomes were lacking or in excess. When intracerebrally transplanted tumors had been previously treated by administration of BCNU, the DNA content was altered, indicating an increased share of diploid cells in the proliferation pool. Results are at variance with earlier findings in tissue cultures of directly induced malignant gliomas and neurinomas in rats. The findings in transplanted tumors can be interpreted as a result of increased malignancy in transplantation tumors, documented by rapid growth in the animal and dedifferentiated histologic morphology.

Perioperative cytostatic treatment of malignant supratentorial brain tumors with BCNU with special reference to tumor site

During the period from 1979 until 1986, sixty male and female patients were treated perioperatively in the Neurosurgical Department of the Surgical Clinic and OPD of the Munich Technical University with the alkylating agent BCNU. Up to five treatment cycles were employed, the first treatment cycle being initiated pre-operatively. In none of the cases were disorders of wound healing seen. Among these sixty patients there was one case of suspected BCNU lung and one case of massive leukopenia. Considering the rate of complications described in the international literature, this figure is extremely small. As to the study of the mean survival time, only 30 patients presenting supratentorial malignant tumours of comparable localisation were included. In light of these strict criteria the perioperative BCNU treatment at our clinic resulted in a mean survival time of 67.1 weeks which in comparison with the literature can be regarded as a good result. We would therefore only recommend cytostatic treatment with peroperative administration of BCNU.

Blood vessel endothelia in rat brain following experimental administration of BCND (Carmustine): Karyometric studies and cytophotometric studies on DNA

Studies on the effect of BCNU (Carmustine), administered in 4 intraperitoneal injections of 7.5 mg each, on cell nuclei of blood vessel endothelia were performed on the parietal cortex and the thalamus of 20 adult rats of Wistar strain. Karyometric studies demonstrated increased circumference and cross-section area of cell nuclei in vascular endothelia following administration of the cytostatic drug. The nuclei acquired a more spheric shape and a decreased surface irregularity (decreased folding) of their cross-sections. Extinction of endothelial cell nuclei estimated following administration of BCNU showed also an increased content of DNA as well as a more compact and concentrated nuclear chromatin. The alterations noted in endothelial nuclei of thalamic blood vessels were more pronounced than those seen in the parietal cortex.

Histoenzymatic, karyometric, and DNA cytophotometric studies on rat neuroglia following administration of BCNU (Carmustine)

BCNU is a nitrosourea derivative applied in anti-neoplastic therapy. As a result of conversion under in vivo-conditions, products of alkylating properties are formed which derrange DNA metabolism, and products of carbamylating activity which disturb protein metabolism. Studies on BCNU neurotoxicity were performed on 20 adult rats of Wistar strain which were given the compound intraperitoneally, in doses corresponding to those used for tumour treatment in humans. Material for estimating unfavorable effects of the drug on central nervous system involved brains, the sections of which were subjected to histoenzymatic reactions and stained according to FEULGEN for karyometric studies and DNA cytophotometry. Size and shape estimates for nuclei of the cells and their DNA content were obtained using Morphoquant automatic microscopic image analyzer (VEB Carl Zeiss, Jena) and compared to those of neuroglia of control animals, given no drug. Multiple intraperitoneal injections of BCNU induced distinct histoenzymatic, karyometric, and cytophotometric changes in oligodendroglia and astroglia. In oligodendroglia, lowered NsE and AcP activities and augmented TPPase activity were noted while in astroglia decreased activity of the former 2 enzymes was associated with an augmented ATPase activity. In karyometric studies, significant decrease in the size of astroglia and in DNA extinction was noted. No karyometric or cytophotometric alterations were noted in oligodendroglia nuclei. Results of the presented studies may indicate primary injury to vascular walls and secondary damage to neuroglia.

Combined effects of chemotherapy and host antitumor response in a murine histocompatible lymphoma model

Combined effects of 1,3,-bis-2-chloroethyl)-1-nitrosourea (BCNU) and host antitumor immune response were studied in mice inoculated intraperitoneally with histocompatible LSTRA leukemia cells carrying virus-induced transplantation antigens. Marked chemo-immune collaborative activity was found to occur when selected schedules of BCNU administration were employed. Moreover, synergist effects were also detected between chemotherapy and both specific and non-specific immunotherapy.

Relationships Among Tumor Load, Route of Tumor Inoculation, and Response to Immunochemotherapy in a Murine Lymphoma Model<xref ref-type="fn" rid="FN2">2</xref><xref ref-type="fn" rid="FN3">3</xref>

The combined effects of nonspecific immunostimulation with Candida albicans (CA) and chemotherapy were studied in (BALB/cCr X DBA/2Cr)F1 and (C57BL/6Cr X DBA/2Cr)F1 mice bearing virus-induced LSTRA lymphomas. Paradoxically, animals treated with a relatively high number of tumor cells responded better to therapy with 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) than those challenged with a low number of tumor cells. However, the majority of mice subjected to low initial tumor load were cured when they were treated with chemotherapy or chemotherapy plus booster injection of CA at a relatively "late" stage of the disease, i.e., when high tumor load was present in tumor-bearing hosts. It has been shown that this phenomenon, provisionally called high tumor load protection, occurs when the animals are challenged ip but not when they are challenged iv with the tumor and is abolished by total-body gamma-irradiation. Moreover, marked host protection can be attained when immunostimulated mice, inoculated iv with lymphoma cells, are subjected to simultaneous challenge with high inocula of the same tumor ip, followed by BCNU administration. These data stress the importance of the peritoneal cavity for successful CA plus drug treatment and suggest that optimal tumor "antigen load" should be present at the time of CA and/or BCNU administration.

High dose BCNU with autologous bone marrow rescue in the treatment of recurrent malignant gliomas.

Eleven patients with malignant gliomas recurring after surgery and radiation therapy, were treated with high dose BCNU 1 050-1 200 mg/M2 with autologous bone marrow rescue. Four patients also received concomitant 5-fluorouracil 1 000 mg/M2/24 hr daily for three days. Eight of ten evaluable patients demonstrated improvement on CAT scan as well as a decrease in steroid requirement. All patients surviving longer than two weeks after BCNU administration experienced full hematologic recovery. No delayed myelosuppression was seen after a single course of high dose therapy. Two patients died as a result of therapy, one following a second induction of BCNU for a total cumulative BCNU dose of 2 400 mg/M2 and one of infection while cytopenic. Additional toxicity includes one steroid-responsive interstitial pneumonitis, one centrilobular necrosis of the liver which spontaneously resolved and one episode of deep vein thrombosis. With limitation on the maximum BCNU dose and distribution of the total dose over three days, high dose BCNU can be administered with acceptable toxicity. This approach may offer a higher response rate than that expected for standard dose BCNU.

Vinylchloride: decomposition product of BCNU in vivo and in vitro.

Vinylchloride (VC) was identified by gas chromatography-mass spectrometry as a decomposition product of 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU). VC was observed in blood and exhaled air from patients after administration of BCNU. Production of VC is evident even after incubation of blood from normal donors with BCNU.

Distribution of 13N in rat tissues following intravenous administration of nitroso-labeled BCNU.

The concentrations of label in 16 major organs and tissues of pentobarbital-sedated normal male rats were measured at six time points ranging from 0.2 to 50 min after IV injection of the antitumor drug BCNU labeled in the nitroso position with cyclotron-produced nitrogen-13. Initial (12-s) concentrations in the lungs, kidneys, and heart were 41, 13, and 11 times the whole-body average, respectively. Time for clearance of the first 50% of the injected dose from the circulation was of the order of several seconds. Estimated first-pass extractions of 70% or more were noted in the heart, kidneys, brain, stomach, small intestine, muscle, fat, and bone. Washout of label from the heart and lungs was quite rapid, removing most of the initially extracted 13N from these organs by 2 min after injection. Label concentrations in the kidneys exceeded those in all other tissues studied between 2 and 50 min. Secondary accumulations of 13N were observed in muscle, skin, liver, small intestine, and fat. Label concentrations in a number of tissues closely paralleled the steadily decreasing concentration in blood for various intervals between 5 and 50 min. The results suggest that the toxic insult to lung tissue from IV administered BCNU is effected in a period of several minutes. They also suggest that intra-arterial administration of the drug would significantly raise the target/non-target dose ratio and lower the incidence of pulmonary toxicity.

Hemorrhagic encephalitis produced by selective non-occlusive intracarotid BCNU injection in dogs.

A selective non-occlusive technique was developed for administration of BCNU (1,3-bis(2-chloroethyl)-1-nitrosourea) into the internal carotid artery of the dog, and the neuropathological effects in the brain were studied. One out of three dogs showed ipsilateral hemorrhagic necrotizing encephalitis at doses of 102 mg/sq m, and all of three dogs showed similar but more severe pathology at doses of 215 to 232 mg/sq m. This study and previous studies in the dog and monkey suggest definite thresholds above which cerebral toxicity occurs when BCNU is administered via the intracarotid route. Greater dilution of drug in the larger territory of supply of the human internal carotid artery allows somewhat higher doses in man. The pathology of the lesion induced by BCNU suggests a primary vascular injury as a pathogenic mechanism, consonant with similar findings following high-dose systemic BCNU administration in man. Investigators conducting ongoing and future trials of intracarotid BCNU in the human for the treatment of intracranial neoplasms should be especially vigilant for a similar toxic effect.

In vivo response of KHT sarcomas to combination chemotherapy with radiosensitizers and BCNU.

Female C3H/HeJ mice bearing intramuscularly transplanted KHT sarcomas were treated with a single dose of 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU, 30 mg/kg, i.p.) alone or in combination with a single dose of misonidazole (MISO, 1.0 mg/g, i.p.) or its desmethylated metabolite Ro-05-9963 (2.0 mg/g, i.p.). The effectiveness of drug therapy was assessed by a tumour growth-delay assay (i.e. measuring the median time required for tumours to grow to treatment size x 4). The relative efficacy of administering the nitroimidazoles in various schedules ranging from 12 h before to 12 h after BCNU administration also was evaluated. Untreated control KHT tumours grew to the initial size x 4 in a median time of 4 days. No significant growth delay was seen in mice treated with either nitroimidazole alone, whilst treatment with BCNU alone produced a median growth delay of 7 days. Combination chemotherapy with 9963 administration 3 h after BCNU significantly increased the median tumour growth delay to 9 days. However, no significant growth delay was produced in any of the other combinations of these agents. The median growth delay was significantly reduced to 5 days when MISO was administered 3 h before BCNU, whereas MISO administered simultaneously 3,6, or 12 h after BCNU significantly enhanced delays ( 9 days). These results indicate that both MISO and 0063 may be combined with conventional therapeutic agents, in this particular case a nitrosourea, to produce an enhanced tumour response. The production of such a response appears to be nitroimidazole as well as schedule dependent.

Peritoneal exudate mononuclear phagocyte colony-forming cells: their proliferative state in vivo and sensitivity to cytotoxic agents.

We investigated the proliferative state of peritoneal exudate macrophage colony-forming cells (PE-CFC) by exposing them to [3H]thymidine with a high specific activity in vitro and by administering cytosine arabinoside and methotrexate in vivo. Since [3H]TdR had no appreciable killing effect on PE-CFC and since the drug treatment caused no reduction in the appearance of PE-CFC, we concluded that the majority of PE-CFC, if not all, are not in active cell cycle. The production of PE-CFC was, however, suppressed by the administration of nitrogen mustard, cyclophosphamide, BCNU and vinlastine.

The effect of time between X-irradiation and chemotherapy on the growth of three solid mouse tumors. VI. BCNU

Experiments have been carried out to determine the effect of different time intervals between the administration of X-irradiation (1200 rad) and BCNU (15 mg/kg) on the growth delay produced in three mouse tumors. The tumors used were the EMT6 tumor in BALB/c mice and the KHT and RIF-1 sarcomas in C3H mice. All tumors were grown intramuscularly in the gastrocnemius muscle and treatment was carried out at a mean tumor weight of 450 mg. Time to reach 2X (for KHT) or 4X (for EMT6 and RIF-1) treatment volume was used as the endpoint of response. The drug was administered by the intraperitoneal route either 24, 6, or 2 hr before radiation. All irradiations were carried out in unanesthetized mice. The growth delays due to the drug alone were 2, 6, and 11 days for the RIF-1, EMT6, and KHT tumors, respectively. No consistent general pattern emerged from the results of combination treatments. For the RIF-1 tumor, the growth delays following combination treatments were generally less than predicted by the simple addition of the growth delays for the single modalities. For EMT6 this was true when BCNU was administered immediately before X-rays, but not for other timings. In the KHT tumor an unexpectedly high incidence of long-term tumor controls was seen in the group which received BCNU at 2 hr before X-rays. In addition to the single dose studies (above), fractionated regimens in which radiation and BCNU were combined in several different ways were tested with the RIF-1 tumor. None of the combination schedules tested showed a greater-than-additive effect.

In Vivo Erythropoietin Requirements of Regenerating Erythroid Progenitors (BFU-e, CFU-e) in Bone Marrow of Mice

Erythroid progenitors (B-8, B-4, CFU-e) in the femoral marrow of polycythemic mice were measured by in vitro culture assays after a single administration of BCNU or Myleran. BCNU reduced pluripotent stem cells to 40% and erythroid progenitors to less than 5% of normal. B-8, the earliest erythroid progenitors, regenerated without erythropoietin (Epo) completely within 5 days. At 14 days after BCNU, intermediate progenitors (B-4) attained 60% of their normal numbers and CFU-e attained approximately 30%. Daily injections of Epo promptly restored normal B-4 numbers and near-normal CFU-e numbers in BCNU-treated mice. After Myleran, CFU-s remained below 2% of normal for 14 days, and no regeneration of the B-8 occurred with or without daily Epo injections. The findings suggest that regneration of B-8 was dependent on cell inflow from the pluripotent stem cell compartment but was independent of the presence of Epo. Intermediate progenitors (B-4) required Epo and the presence of B-8 for complete and permanent regeneration. CFU-e were the most Epo-dependent of the three progenitors. B-4, recruited by Epo, required after their formation a second exposure to the hormone in order to progress into the CFU-e stage.

In vivo erythropoietin requirements of regenerating erythroid progenitors (BFU-e, CFU-e) in bone marrow of mice

Erythroid progenitors (B-8, B-4, CFU-e) in the femoral marrow of polycythemic mice were measured by in vitro culture assays after a single administration of BCNU or Myleran. BCNU reduced pluripotent stem cells to 40% and erythroid progenitors to less than 5% of normal. B-8, the earliest erythroid progenitors, regenerated without erythropoietin (Epo) completely within 5 days. At 14 days after BCNU, intermediate progenitors (B-4) attained 60% of their normal numbers and CFU-e attained approximately 30%. Daily injections of Epo promptly restored normal B-4 numbers and near-normal CFU-e numbers in BCNU-treated mice. After Myleran, CFU-s remained below 2% of normal for 14 days, and no regeneration of the B-8 occurred with or without daily Epo injections. The findings suggest that regneration of B-8 was dependent on cell inflow from the pluripotent stem cell compartment but was independent of the presence of Epo. Intermediate progenitors (B-4) required Epo and the presence of B-8 for complete and permanent regeneration. CFU-e were the most Epo-dependent of the three progenitors. B-4, recruited by Epo, required after their formation a second exposure to the hormone in order to progress into the CFU-e stage.