Feasibility of hyperthermia as a purging modality in autologous bone marrow transplantation Wierenga, Pieter Klaas

Transcription

1 University of Groningen Feasibility of hyperthermia as a purging modality in autologous bone marrow transplantation Wierenga, Pieter Klaas IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 1997 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Wierenga, P. K. (1997). Feasibility of hyperthermia as a purging modality in autologous bone marrow transplantation Groningen: s.n. Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date:

2 CHAPTER SEVEN Reduction of Heat Induced Hematoxicity in a Hyperthermic Purging Protocol of Murine Acute Myeloid leukemic Stem Cells by AcSDKP (Goralatide) 1 2 Pieter K. Wierenga, Jan H. Dillingh and Antonius W.T. Konings 1 1) Department of Radiobiology and 2) Department of Physiological Chemistry University of Groningen Groningen, The Netherlands From: Brit. J. Haematol., in press.

3 Chapter 7 ABSTRACT The tetrapeptide AcSDKP (Goralatide) is known as a cytokine with inhibitory effects on cell proliferation. Clinically, the use of this cytokine is challenging. Many purging agents used in autologous bone marrow transplantation protocols, including hyperthermia, preferentially kill cycling cells. A pretreatment with this negative hematopoietic regulator offers a possibility to reduce the hematotoxicity in purging settings. The impact of Goralatide on the hyperthermic purging protocol was investigated on normal and myeloid leukemic (SA8) murine cells. The median survival time after transplantation (ie, leukemia incidence) was used as an in vivo parameter to learn the effects on leukemic cells. The hyperthermic effect on normal and leukemic cells was also investigated in vitro using the cobblestone area-forming cell (CAFC) assay. It could be demonstrated that a heat treatment of 90 minutes at 43EC results in a 4-log depletion of leukemic stem cells. For normal progenitor cells (CFU-GM) a 2-log cell kill is shown. The reduction in proliferative activity of the CFU-GM after an 8-hour -9 incubation with 10 M Goralatide results in a decrease in the heat sensitivity of the progenitor subset to approximately 1-log cell kill. The leukemic precursor cells seem insensitive for the inhibiting feature of Goralatide implicating an increase in the therapeutic window of the hyperthermic purging protocol. Finally, simulated remission bone marrow (5% leukemic blasts) was incubated with Goralatide followed by a heat treatment of 90 minutes at 43EC. Lethally irradiated (10 Gy) mice transplanted with 6 4 heat-treated remission bone marrow (10 normal bone marrow cells versus 5.10 leukemic cells) died of aplasia while Goralatide-pretreated remission bone marrow could rescue the irradiated mice without revealing leukemic engraftment. These findings confirm the enhanced protection against hyperthermia of the normal hematopoietic subsets and the elimination of leukemic cells by the hyperthermic purging protocol. 126

4 Effect of Goralatide/Hyperthermia on remission bone marrow INTRODUCTION In patients with acute myeloid leukemia in first remission and without a matching sibling donor, autologous bone marrow transplantation would offer a potentially curative strategy. However, it is well recognized that disease recurrence is one major problem in autologous bone marrow transplantation. A relapse may originate from residual disease in the patient or it may result from malignant cells contaminating the transplant, or both. Recurrence from residual leukemic cells would require intensification of the cytotoxic conditioning regimen while contaminating cells in the transplant can be diminished by purging techniques. Although the efficacy of attempts to eliminate residual leukemic cells from the transplant is still controversial (Rill et al., 1992; Brenner et al., 1993) some encouraging data have been reported in clinical trials (Keating, 1991; Lazarus et al., 1993). Moreover, evidence has emerged showing the contribution of the transplant to relapse (Gorin et al., 1991; Selvaggi et al., 1994). These findings strengthen the necessity to purge the autologous remission bone marrow before transplantation. The feasibility of hyperthermia as a purging modality was investigated before (Moriyama et al., 1986; Da et al., 1989; Murphy and Richman, 1989; Gidali et al., 1990; Iwasawa et al., 1991; Moriyama et al., 1991, 1992). Most of these studies were directed toward the determination of the hyperthermic effect on normal hematopoietic progenitor subsets and their leukemic counterparts. Until then, studies on the heat sensitivity of human primitive precursors were hampered by the lack of suitable assays. In the murine system, however, we were able to investigate the heat sensitivity of both the hematopoietic progenitors and primitive stem cells capable of long-term reconstitution. It appeared that the primitive stem cells are less sensitive to hyperthermia compared with the committed subsets (Wierenga and Konings, 1993; Wierenga et al., 1995). The apparent relationship between the stem cell hierarchy and heat sensitivity is predominantly based on the differences in proliferative activity. Moreover, within the same subset a decrease in proliferative activity confers a decrease in heat sensitivity (Baeza et al., 1987; Wierenga and Konings, 1990, 1991). It will be appreciated that the possibility of arresting active proliferating hematopoietic cells can be used to protect these cells in a hyperthermic purging protocol. The tetrapeptide Acetyl-N-Ser-Asp-Lys-Pro (Goralatide) is known as a selective inhibitor of normal hematopoietic progenitor cells (Monpezat and Frindel, 1989; Guigon et al., 1990; Bonnet et al., 1992; Cashman et al., 1994; Wierenga and Konings, 1996) to enter the S-phase and can therefore be considered as a potent protector of these cells in purging protocols. The introduction of a protected stem cell compartment will enhance hematopoietic recovery by conserving a larger population of progenitors. This 127

5 Chapter 7 will help to alleviate the delay in engraftment as a dose-limiting factor. On the other hand, it permits an increase in treatment temperature augmenting the elimination of leukemic cells. In the current study an in vivo murine myeloid leukemia model was used. This model is believed to be more useful and clinically relevant to investigate the therapeutic responses of myeloid leukemia (Hepburn et al., 1987) than routinely passaged cell lines used thus far (Gidali et al., 1990; Iwasawa et al., 1991; Wierenga and Konings, 1996). In the clinical studies published on the outcome of autologous bone marrow transplantation purged by hyperthermia, 43EC was considered the optimal temperature (Higuchi et al., 1991; Herrmann et al., 1992). The rapid decline in progenitor viability was the dose-limiting factor in the hyperthermic treatment. Therefore, the hyperthermic sensitivity at 43EC of the myeloid leukemic stem cells and the effect of Goralatide was investigated in vivo by determining the median survival time after transplantation. A comparison with the effects on different subsets in the hematopoietic stem cell compartment was made in vitro using the cobblestone areaforming cell (CAFC) assay. This assay permits the calculation of frequencies of colonyforming cells within the stem cell compartment. The early appearing cobblestone areas (CAs) represent the progenitor subsets while the late appearing (CAs) are considered primitive stem cells. Furthermore, a simulated remission bone marrow was incubated with Goralatide preceding the hyperthermic treatment and transplanted in lethally irradiated mice. The in vivo and in vitro results demonstrate the use of Goralatide toward improvement of a hyperthermic purging protocol. MATERIALS AND METHODS Mice. Female CBA/H mice were used as the source of the hematopoietic and leukemic cells. The mice were bred at the Netherlands Energy Research Foundation, Petten, The Netherlands and maintained under clean conventional conditions in the animal facilities of the Department of Radiobiology, University of Groningen, The Netherlands. Mice were fed ad libitum with food pellets and acidified tap water (ph=2.8). Hematopoietic cells. Normal bone marrow cells were obtained by flushing the femoral cell content with alpha-medium buffered with 20 mm morpholinopropane sulphonic acid and in the presence of 10% fetal calf serum. The cell suspension was dispersed through a 21-gauge needle. Nucleated cell count was done on a Coulter Counter Model Z1 (Coulter Electronics, Hialeah, FL, USA). The cell suspensions were kept on ice until use. 128

6 Effect of Goralatide/Hyperthermia on remission bone marrow Leukemic cells. Murine myeloid leukemic cells (obtained from Dr. A.C. Riches, Univ. St. Andrews, Scotland) were used throughout this study. This myeloid leukemia was originated from irradiated CBA/H mice (Mole et al., 1983) and maintained by serial passage of spleen cell suspensions (Riches et al., 1987). Cells from passage eight 5 (SA8) were transplanted at a cell dose of 10 in CBA/H mice. Mice were observed daily and white blood cell counts (WBC) were performed on weekly basis. At the time of 6 leukemia incidence (WBC=> 50x10 /ml) single cell suspensions were prepared from the spleen of these leukemic mice. TB-exclusion was always >80%. Nucleated cell count was done on a Coulter Counter Model Z1 (Coulter Electronics, Hialeah, FL, USA). The cell suspensions were kept on ice until use. Goralatide treatment. Goralatide (kindly provided by IPSEN-BIOTECH, Paris, -7 France) was dissolved in alpha-medium as a stock solution of 10 M. Leukemic cells 6 or remission bone marrow cells were plated out at a concentration of 2-3x10 /ml in culture flasks (Falcon, Becton and Dickinson, Lincoln Park, NJ, USA) and incubated -9 for eight hours ± 10 M Goralatide in alpha-medium supplemented with 10% fetal calf serum. After the incubation the cells were washed twice and immediately heat treated. Proliferative activity. The proliferative activity of the leukemic cells and hematopoietic progenitor cells was determined by the hydroxyurea-kill assay as described before (Wierenga and Konings, 1990). Hydroxyurea (200 Fg/ml) was added to aliquots of cell suspensions ± Goralatide after zero and seven hours, respectively and incubated for another 60 minutes at 37EC. Then the cells were washed and plated out for the colony-forming assays. Hyperthermic treatments. The cell suspensions were placed in culture tubes at 7 a concentration of 2-3x10 /ml in alpha-medium supplemented with 10% fetal calf serum. The hyperthermic treatments were performed at 43 ± 0.1EC and interrupted by chilling. Cells were diluted to the appropriate concentration and immediately used for the colony-forming assays. Colony-forming assays. The progenitor cell CFU-GM was assayed as described earlier (Wierenga and Konings, 1993). Briefly, cells were plated out in alpha-medium -4 containing 1.2% methylcellulose, 30% fetal calf serum and 10 M 2-mercaptoethanol. The medium was buffered with sodium bicarbonate and HEPES. Cultures were plated in 35-mm polystyrene culture dishes (Falcon, Becton and Dickinson, Lincoln Park, NJ, USA) and grown at 37EC in a 5% CO 2 humidified atmosphere. Colony growth was stimulated by the addition of pokeweed mitogen-stimulated spleen-conditioned medium (10 µl/ml). CFU-GM colonies (>50 cells) were scored after eight days of culture. 129

7 Chapter 7 The cobblestone-area forming cell (CAFC) assay was performed as described by Ploemacher and colleagues (1993). Confluent stromal layers of FBMD-1 cells in 96- wells plates were overlaid with normal bone marrow or leukemic spleen cells in a limiting dilution setup. Eight dilutions twofold apart were used for each sample, with 15 replicate wells per dilution. The percentage of wells with at least one phase-dark hematopoietic clone of at least five cells beneath the stromal layer was determined about every week and CAFC frequencies were calculated using Poisson statistics, as described (1994). Leukemia incidences. The leukemia engraftment rate was determined at different cell doses of control and heat-treated leukemic cell suspensions in groups of five normal CBA/H mice. Control and heat-treated leukemic cells were also 5 transplanted in lethally irradiated mice concomitantly with 2x10 normal bone marrow cells. Remission bone marrow was transplanted in lethally irradiated mice. Mice were 137 irradiated with 10.0 Gy (-rays (0.89 Gy/min) in a Cs-source. The irradiated recipients were put on acidified water supplemented with neomycin sulphate (3.5 g/l). Mice were observed daily and white blood cell counts were performed on weekly basis. At cell 6 counts >50x10 /ml the moribund mice were sacrificed. Splenomegaly was observed in all mice. The median survival time after transplantation was used as parameter of a leukemia incidence. RESULTS Hyperthermic sensitivity of leukemic cells. Control and heat-treated cells were transplanted at different cell doses in normal or lethally irradiated mice. In the latter case a fixed number of normal bone marrow cells was concomitantly transplanted (see Materials and Methods). Figure 7.1. Median survival time (MST) at different cell doses as described in Materials and Methods. Control or heat-treated leukemic cells were transplanted in untreated recipients: control cells =, slope= ± 0.015, r2= , heat-treated cells =, slope= ± 0.011, r2= or in irradiated recipients: control cells = :, slope= ± 0.010, r2= , heat-treated cells =, slope= ± 0.012, r2=

8 Effect of Goralatide/Hyperthermia on remission bone marrow In Figure 7.1, the median survival time (ie, leukemia incidence) of the different groups is plotted against the cell doses. It is clear from this figure that in both normal and lethally irradiated recipients, the dose-response curves of control and heat-treated cells share a similar slope. The difference in median survival time that separates these curves appears independent of the cell dose. Although in lethally irradiated recipients the slope of the dose-response curves for both control and heat-treated leukemic cells differ from that in normal mice, the cell dose ratio does not differ from that detected in normal recipients. At a given iso-effect level, the number of surviving leukemic stem cells should theoretically be the same and therefore a cell dose ratio measures the surviving fraction (Figure 7.2). For comparative reasons, the survival curves for the progenitor (CFU-GM) and the primitive stem cell with long-term repopulating ability (LTRA) from a previous study (Wierenga and Konings, 1996) are incorporated in this figure. These data clearly suggest the feasibility of hyperthermia as purging modality. Figure 7.2. Survival curve of leukemic cells at 43EC. Data points are derived from three separate experiments. Bars = SEM. LTRA and CFU-GM curves are adapted from a previous study (Wierenga and Konings, 1993). Effect Goralatide on normal and leukemic cells. In Table 7.1, the effect of a pretreatment with Goralatide on the proliferative activity and heat sensitivity of the CFU-GM is shown. An incubation of eight hours -9 with 10 M Goralatide results in a dramatic decrease in the percentage of S-phase cells and concomitantly in a decrease in the hyperthermic sensitivity. The effect of Goralatide on leukemic stem cells was investigated in vivo. Leukemic cells were pretreated as normal bone marrow and subsequently transplanted in normal mice. Figure 7.3 shows that in both control and heat-treated samples the same median survival times are observed indicative for the absence of a Goralatide effect on leukemic cells. 131

9 Chapter 7 Table 7.1. Effect of Goralatide on the proliferative activity (% HU-kill) and hyperthermic sensitivity (% survival) of the CFU-GM in normal bone marrow. See materials and Methods for details. Data point (± 1 SEM) are derived from three separate experiments. % HU-kill % survival Control 50 ± ± Goralatide 14 ± ± Figure 7.3. Survival of normal recipients transplanted with 10 leukemic cells. Cells where either control ( = Control, b = + Goralatide) or heat-treated for 90 minutes at 43EC (: = Control, = + Goralatide). CAFC assay. The CAFC-assay enables us to examine the hierarchical structure of the leukemic stem cell compartment and to compare that with the hierarchy in normal bone marrow. In Table 7.2 it is shown that in normal bone marrow the surviving fraction of the CAFC-day 7 clearly confirms that of the CFU-GM (Table 7.1) and the data on CAFC-day 28 corresponds very well with those on LTRA (Figure 7.2). 132

10 Effect of Goralatide/Hyperthermia on remission bone marrow Table 7.2. CAFC frequencies (mean ± 1 SEM), proliferative activity (% S-phase cells determined by hydroxyurea-kill) and hyperthermic sensitivity (% survival) of normal and leukemic cells in a representative experiment. See Materials and Methods for details. (n.d. = not detectable) Normal bone marrow CAFC-d7 d14 d21 d28 Frequency/105 n.c. 81 ± ± ± 7 12 ± 5 % S-phase cells Control n.d. n.d. + Gor. <10 <10 n.d. n.d. % survival Control (HT: 90' 43EC) + Gor Leukemic spleen cells Frequency/105 n.c. 485 ± ± ± ± 56 % S-phase cells Control 41 n.d. n.d. n.d. + Gor. 44 n.d. n.d. n.d. % survival Control (HT: 30' 43EC) + Gor The frequency of the subsets in normal bone marrow is decreasing toward the primitive precursor. Beyond day 28 the frequency and heat-sensitivity of the CAs are comparable with that of CAFC-day 28 (data not shown). In the leukemic stem cell compartment early and late appearing CAs can be detected as well. Comparable with the normal subsets, the leukemic CAFC-day 7 is proliferating while the late appearing leukemic CAs are quiescent. The differences in hyperthermic sensitivity of the leukemic subsets, as shown in this table, may be related to the proliferative activity as it is for normal hematopoietic subsets. The surviving fraction of the late appearing leukemic CAs nicely corresponds with the in vivo data (Figure 7.2). In contrast to normal bone marrow, the subsets in the leukemic stem cell compartment do not decrease toward the primitive subset. Furthermore, the results from the CAFC-assay confirm the protective effect of Goralatide on the progenitor subsets in normal bone marrow by displaying a decrease in the proliferative activity and hyperthermic sensitivity of the CAFC-day 7. No inhibitory effect on the leukemic CAFC-day 7 could be demonstrated. The ultimate effect of this pretreatment implies an increase in the therapeutic window of the purging modality. 133

11 Chapter 7 In vivo effect of a Goralatide-pretreated remission bone marrow transplant after hyperthermia. To test this hypothesis, heat-treated simulated remission bone marrow (5% leukemic cells) was transplanted in lethally irradiated (10.0 Gy) mice. Figure 7.4 shows the in vivo results of the remission bone marrow transplants purged by hyperthermia. Mice transplanted with heat-treated remission bone marrow died before day 15 due to aplasia. However, mice transplanted with Goralatide pretreated remission bone marrow were rescued without revealing leukemic engraftment. These data confirm the protective effect of Goralatide against hyperthermic damage of the normal hematopoietic stem cell compartment and implicate the elimination of leukemic cells by the hyperthermic purging protocol. Figure 7.4. Survival of lethally irradiated recipients transplanted with heat 6 4 treated simulated remission bone marrow (10 n.b.m./5.10 leukemic cells). Hyperthermic treatment was 90 minutes at 43EC. = untreated L = heat treated = Goralatide + heat treated DISCUSSION In most studies attempting to prove the feasibility of hyperthermia as purging modality, the hyperthermic sensitivity of normal hematopoietic progenitors was compared with that of leukemic cell lines (Gidali et al., 1990; Wierenga and Konings, 1990; Iwasawa et al., 1991; Moriyama et al., 1992; Wierenga and Konings, 1996). The disadvantage of using cell lines is that they consist of a homogeneous cell population and might not be representative for the clinical situation (Trott, 1994). The leukemic 134

12 Effect of Goralatide/Hyperthermia on remission bone marrow model used in this study is a radiation-induced myeloid leukemia. Serial passages at low cell dose conserve the primary kinetic and morphological properties of the leukemia (Hepburn et al., 1987). Data from the CAFC-assay reveal both short and long-term growth capacity of leukemic cells and also differences in proliferative activity between the subsets, confirming the heterogeneity in the clonogenic cell population (Table 7.2). This heterogeneity is also demonstrated in human AML bone marrow using the CAFC-assay (Terpstra et al., 1996a, 1996b). The malignant cell population of AML is maintained by a small fraction of clonogenic cells (Griffin and Löwenberg, 1986). Although it is shown that the expression of CD34 does not necessarily correlates with long-term leukemia-initiating capacity (Terpstra et al., 1996a) it is generally regarded to be associated with an immature phenotype (Lapidot et al., 1994). Furthermore, it could be shown that CFU-AML subsets are eliminated after a 5-FU treatment without affecting the late appearing CAs (Lenfant et al., 1989). The persistent cell population is still capable to initiate leukemia in SCID mice indicative for a correlation between long-term leukemia-initiating capacity and late appearing CAs. In the murine system the CAFC-day 28 resembles such cell type. The dormant cell cycle state of the leukemic stem cell detected in the present study (Table 7.2) corresponds with the immature phenotype of the leukemia-initiating cell. The in vivo determination of the surviving fractions of leukemic cells after hyperthermia is based on the median survival time after transplantation. It appears that the slopes of the cell dose-response curves differ in normal and lethally irradiated recipients. The differences in median survival time at the low cell doses are responsible for these changes in the slope. The reasons for these differences are yet unknown. One explanation might be that in lethally irradiated recipients the created space for hematopoietic cells to reside will preferentially be occupied by normal bone marrow cells instead of the leukemic spleen cells. Because of the co-transplantation of a fixed number of normal bone marrow cells, the cell ratio favors the normal bone marrow cells at these low leukemic cell doses and will restrain leukemic engraftment. Nevertheless, both experimental approaches result in the same surviving fraction after hyperthermia. The surviving fraction of leukemic cells obtained in the CAFC-assay corresponds with the in vivo data. After 30 minutes at 43EC, a 2-log leukemic cell kill is demonstrated (Figure 7.2, Table 7.2). These data are consistent with those published on the heat sensitivity of leukemic progenitors in human AML bone marrow (Da et al., 1989; Moriyama et al., 1991). After 90 minutes at 43EC, the hyperthermic-induced cell kill of the leukemic cells is increased to 4-log (Figure 7.2) while the surviving fraction of normal subsets varies between 1-30% (Figure 7.2, Table 7.2). These data clearly show a significant difference in heat sensitivity between normal and leukemic stem cells and demonstrate the feasibility of hyperthermia as purging modality. It is important to notice 135

13 Chapter 7 that the primitive precursor cell (CAFC-day 28) responsible for permanent long-term engraftment, is less sensitive to heat than the committed progenitor (CAFC-day 7) as is shown in Table 7.2. Because of the observed differences in heat sensitivity between the early and late-appearing CAs, a hyperthermic purging regimen could be determined from the effects on the primitive stem cell. However, the hyperthermic effect on progenitors is still relevant because a depletion of the progenitor pool prohibits the short-term reconstitution of the recipient and will cause aplasia. This demands the protection of the progenitor subsets without reducing the anti-tumor efficacy of the purging modality. As previously demonstrated, the heat sensitivity within the stem cell hierarchy is predominantly based on the differences in proliferative activity (Baeza et al., 1987; Wierenga and Konings, 1990, 1993). A strategy to increase the therapeutic window of a hyperthermic purging protocol is decreasing the proliferative activity of the normal hematopoietic progenitors. An inhibiting factor for CFU-S proliferation could be isolated from fetal calf bone marrow (Frindel and Guigon, 1977) which was characterized as the tetrapeptide Acetyl-N-Ser-Asp-Lys-Pro (Lenfant et al., 1989). This tetrapeptide is now chemically synthesized and called Goralatide. The protective effect of Goralatide is explained by the inhibition of entry of the cells into the S-phase (Monpezat and Frindel, 1989). In our experiments the number of CFU-GM in S-phase decreases from - 30% in the -9 control situation to below 10% after eight hours incubation with 10 M Goralatide (Table 7.1). In other studies (Bonnet et al., 1992; Guigon et al., 1990) using the same range of Goralatide concentration, a comparable decrease in the percentage progenitors in S-phase was reported. A decrease in proliferative activity of the progenitors results in a decrease in hyperthermic sensitivity as is shown in Tables 7.1 and 7.2. The hyperthermic sensitivity of the murine myeloid leukemic cells is not affected by a pretreatment with Goralatide (Figure 7.3) corroborating the insensitivity of the leukemia-initiating cell for Goralatide (Bonnet et al., 1992; Cashman et al., 1994; Wierenga and Konings, 1996). The insensitivity can be explained by the detected quiescent cell cycle state of the leukemia-initiating cell (Table 7.2). However, the proliferative activity of the leukemic CAFC-day 7 is not decreased after the pretreatment with Goralatide. This finding is consistent with the observation that active proliferating leukemic cell lines are insensitive for Goralatide (Bonnet et al., 1992; Wierenga and Konings, 1996). The selective inhibitory effect of Goralatide might be caused by the inhibition of the synthesis of D-type cyclins. These cyclins regulate the progression from G 1 to S-phase and might be abundant in malignant cells (Gimenez- Conti et al., 1996). In both normal and leukemic cells an inhibition in synthesis of D- type cyclins might occur after a Goralatide treatment but the nadir of cyclin levels in leukemic cells does not fall below the theoretical threshold needed for the inhibition of 136

14 Effect of Goralatide/Hyperthermia on remission bone marrow cell cycle progression. This in contrast to the levels in normal hematopoietic progenitors. This hypothesis is currently under investigation. It is still unclear why normal and leukemic cells differ in their sensitivity toward hyperthermia. In normal bone marrow the proliferative activity of the subsets is a major determinant for heat induced cell kill (Baeza et al., 1987; Wierenga and Konings, 1990, 1993, 1995, 1996). In this study a comparable relationship is shown for leukemic cells (Table 7.2). However, although the normal and leukemic CAFC-day 28 are both quiescent, they display a significant difference in heat sensitivity. This implicates that other cell characteristics than proliferative activity are responsible for the observed differences between normal and leukemic stem cells. Denaturation of membrane proteins probably resulting in protein aggregation is observed after hyperthermia (Burgman and Konings, 1992, Stege et al., 1994). Protein aggregation may impair several membrane proteins associated functions that could lead to cell death. Differences in heat-induced damage to membrane proteins might be the cause for the observed effect. Another intriguing possibility is that differences in heat shock protein expression might be the reason for differences in heat sensitivity (Stege et al., 1994). As shown in Figure 7.4, mice transplanted with heat-treated remission bone marrow after a Goralatide pretreatment were rescued from aplasia related death. Apparently the number of transplanted progenitors (- 150 CAFC-day 7) and stem cells responsible for long-term engraftment (- 30 CAFC-day 28) were sufficient to empower short and long-term reconstitution. If the CAFC-day 28 can be considered as the leukemia-initiating cell, calculating the effect of the hyperthermic treatment on the actual number of cells is interesting. The frequency of the leukemic CAFC-day 28 is per 10 cells (Table 7.2). Although in Figure 7.1 the threshold for a leukemia 3 incidence is approximately 10 cells, a leukemic take could be observed after transplanting 200 cells (data not shown). According to the data in Table 7.2, this corresponds with - 1 CAFC-day 28. This is consistent with the perception that the leukemic CAFC-day 28 is the genuine leukemia-initiating stem cell. After the 4 hyperthermic dose the remission bone marrow transplant containing 5.10 heat-treated leukemic cells includes <1 CAFC-day 28, obviously to low to reveal leukemia incidences. Taken together, this study indicates that hyperthermia is a potent purging modality for AML cells and that the therapeutic window of this modality can be improved by a pretreatment with the negative hematopoietic regulator Goralatide. 137