Autologous Transplantation with Philadelphia-Negative Progenitor Cells for Patients with Chronic Myeloid Leukaemia (CML) Failing

Home , Interferon alfa

Bone Marrow Transplantation (2000) 26, 1165–1172  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt Autologous transplantation with Philadelphia-negative progenitor cells for patients with chronic myeloid leukaemia (CML) failing to attain a cytogenetic response to alpha interferon

NC McBride1, JD Cavenagh1, AC Newland1, DM Lillington2, C Murrell1 and SM Kelsey1

1Department of Haematology and 2ICRF Department of Medical Oncology, St Bartholomew’s and Royal London Hospital School of Medicine and Dentistry, London, UK

Summary: rest period is used. Bone Marrow Transplantation (2000) 26, 1165–1172. Between October 1993 and March 1999, 29 patients Keywords: CML; peripheral blood stem cell transplan- with CML who were ineligible for allogeneic BMT tation; ␣IFN underwent PBSC harvest using idarubicin, cytarabine and G-CSF. In 9/29 (31%) patients all collected stem cells were Ph-negative, and 15/29 patients’ (52%) collections were substantially (Ͼ95%) Ph-negative. The pro- Allogeneic BMT is the only currently available curative portion of patients from whom Ph-negative stem cells option for CML,1,2 but the majority of patients will be were obtained was similar between patients who had, ineligible due to chronological age or absence of an HLA- or had not, received prior ␣IFN. Fifteen patients in identical donor. In addition, procedural-related mortality is chronic phase (median age 45) proceeded to PBSCT fol- relatively high. lowing busulphan 16 mg/m2 and cyclophosphamide There is evidence that patients with CML have residual 120 mg/m2. Nine of the 13 patients who had failed to Ph-negative haemopoiesis, at least in the early stages of the respond to prior ␣IFN proceeded to stem cell transplan- disease, which is presumed to be non-leukaemic.3 As a tation as soon as was feasible and six of the newly diag- result, high-dose chemotherapy followed by unmanipulated nosed patients were transplanted after failing to achieve or ex vivo cultured autologous stem cell transplantation in a cytogenetic response after a minimum of 12 months patients with CML may result in transient Ph-negative on ␣IFN following progenitor cell harvest. The median haemopoiesis and encouraging survival durations, parti- number of days to neutrophils Ͼ0.5 and platelet Ͼ50 cularly if followed by ␣IFN therapy.4–6 was 18 (range 13–69) and 28 (range 13–234), respect- Following administration of myelosuppressive chemo- ively. There was no procedure-related mortality. At therapy to patients with CML, progenitor cells appear in median follow-up of 2.3 years post autograft 10 of 15 the peripheral blood. In the early stages of regeneration patients remain alive and in chronic phase. Overall sur- these cells may be substantially Ph-negative while still cap- vival for all 27 patients at 5 years after initial diagnosis able of supporting multi-lineage haemopoiesis in long-term is 70% and median survival from diagnosis 7.3 years. culture and supporting haemopoietic engraftment in Survival for ␣IFN non-responders who were trans- patients conditioned with high-dose chemotherapy.7,8 With planted is 74% at 5 years from diagnosis and 75% at 3 the assistance of G-CSF these cells may be harvested from years from transplant. Cytogenetic analysis performed the peripheral blood by leukapheresis and used for 3 months post transplant demonstrated one patient with subsequent transplantation.8–10 a complete cytogenetic response, seven with a partial ␣IFN prolongs chronic phase and overall survival in response and three with no response. Six patients patients with CML not undergoing high-dose therapy.11–14 remain partially Ph-negative, with one major CR. Sur- Patients achieving a cytogenetic response to ␣IFN survive vival for all patients in the protocol is favourable com- longer than those who do not and ␣IFN is the treatment pared with conventional therapy and is particularly of choice for patients who achieve a major or complete encouraging following PBSCT for ␣IFN non-responsive cytogenetic response due to potentially long survival (9 patients. Patients not responding to ␣IFN can be years in low-risk patients).15 Median survival for patients induced into Ph-negativity with PBSCT but this may who fail to achieve a good cytogenetic response to ␣IFN not always be sustainable. There seems to be no obvious is around 40 months.12,13,16 Therefore, failure to achieve disadvantage in harvesting stem cells after prior a haematological or cytogenetic response to ␣IFN places exposure to ␣IFN, providing an adequate ␣IFN-free patients with CML in a prognostic group with a relatively poor outcome. Improved therapies are therefore required for subjects who are fit enough to tolerate a bone marrow transplant, but lack a suitable donor and fail to obtain a ␣ Correspondence: Dr SM Kelsey, Pharmacia, Via Robert Koch 1.2, 20152 favourable cytogenetic response to IFN. Milano, Italy The purpose of this study was to prospectively evaluate Received 9 June 2000; accepted 1 September 2000 the feasibility, efficacy and toxicity of high-dose chemo- Autologous transplantation in CML NC McBride et al 1166 therapy with peripheral blood progenitor cell rescue, lowing collection of progenitor cells. These patients pro- specifically for patients with CML failing to achieve a ceeded to high-dose therapy and autologous progenitor cell major cytogenetic response to ␣IFN. Patients were trans- rescue if they were intolerant of ␣IFN, or failed to achieve a planted with substantially Ph-negative progenitor cells after complete or major cytogenetic response (Ͻ35% Ph-positive disease cytoreduction. All patients were then placed back metaphases in bone marrow) after at least 12 months ther- on ␣IFN in an attempt to achieve, or maintain, Ph-negative apy with ␣IFN. Patients who, after 12 months on ␣IFN, haemopoiesis. As prior exposure to ␣IFN may damage stem demonstrated a minor cytogenetic response, continued with cells and delay engraftment following subsequent trans- ␣IFN for a further 6 months prior to evaluation. Patients plant, collection of the stem cells was performed before previously exposed to ␣IFN (protocol 2) who had failed to exposure to ␣IFN where possible. achieve or maintain a major cytogenetic response proceeded to autologous progenitor cell transplant as soon as was feasible. Patients and methods Transplant methodology Patients and mobilisation protocol The conditioning therapy consisted of standard busulphan Patients with a diagnosis of CML in first chronic phase 16 mg/kg over 4 days and cyclophosphamide 120 mg/kg who were not eligible for, or refused, allogeneic BMT were over 2 days. Stem cell rescue was performed on day 0, 1 recruited to the study. Patients previously untreated with day after the last day of cyclophosphamide. ␣IFN was interferon underwent a buffy coat harvest by leukapheresis restarted when neutrophils were Ͼ2.0 ϫ 109/l and platelets Ͼ ϫ 9 if their white blood cell count (WCC) was 50 10 /l were Ͼ100 ϫ 109/l. A bone marrow aspirate and a trephine and were then treated with oral hydroxyurea. A progenitor were performed 3 months post PBSCT and samples sent for cell harvest was performed once the WCC had been under cytogenetic analysis. Marrow cytogenetics were re-checked control for 4 to 6 weeks. Patients who had been previously every 6 months thereafter. treated with ␣IFN, but had failed to obtain a major cytogenetic response (р35% Ph-positive marrow) or had a cytogenetic relapse following an initial response, proceeded Statistical analysis ␣ to harvest after IFN therapy had been stopped for a mini- Survival curves were estimated by the method of Kaplan mum of 8 weeks. Risk score at diagnosis was calculated and Meier using Statistica software (StatSoft Inc (1997) using the new prognostic score (modified Sokal score) for Statistica for windows, Tulsa, OK, USA). patients with CML treated with ␣IFN.15 All patients gave written informed consent and the protocol was approved by the local Research Ethics Committee. Results For collection of Ph-negative progenitor cells patients received idarubicin 8 mg/m2 i.v. on days 1 and 2, plus cytarabine 200 mg/m2 i.v. on days 1–5. Administration of gra- Study population nulocyte colony-stimulating factor 263 ␮g s.c. daily From October 1993 to March 1999, 29 patients entered the (Lenograstim, Chugai Pharma, London, UK) was com- study (19 males; 10 females). Thirteen (45%) patients had menced when the total WCC fell below 0.5 ϫ 109/l after been previously treated with ␣IFN and 16 (55%) had not. chemotherapy and was continued until mobilisation was The median age of the patients was 46 years (range 20– complete. Stem cell collections started when the total WCC 64). The median time from diagnosis to entry into the study was Ͼ1 ϫ 109/l. Adequate mobilisation of progenitor cells for all patients was 7 months (range 2–78). The median ϩ was defined as у2 ϫ 108 MNCs and у1 ϫ 106 CD34 time from diagnosis to study entry for patients in protocol cells per kg body weight, with no greater than 10% Ph- 1 (no previous exposure to ␣IFN) was 5 months (range 2– positivity. If patients failed to fulfil these criteria then re- 22) and for patients in protocol 2 was 24 (range 4–78) induction and repeat harvest was considered. Samples of months, respectively. Twenty-five (86%) patients were in collected stem cells were evaluated by conventional cyto- chronic phase at the time of harvest and four (14%) patients genetics and RT-PCR analysis for bcr-abl mRNA were in semi-controlled accelerating phase. Patient charac- (sensitivity 1:105 cells). teristics are summarised in Table 1. Harvest data and early outcome for four patients has been previously reported.7,9 Cytogenetic analysis Harvest data Cytogenetic analysis was performed using conventional G banding, with analysis of 100 metaphases from both periph- A total of 37 harvests were performed on 29 patients. One eral blood derived, or bone marrow cells. Nested RT-PCR hundred and thirteen leukaphereses were performed overall, was performed as per method of Hughes et al.17 starting at a median of 16 days (range 12–35) from the last day of chemotherapy. The median WCC at the time of first ϫ 9 Progression to transplant harvest was 1.60 10 /l (range 0.35–55.27). The median number of PBSC collections for each harvest was three Patients who had not previously been exposed to ␣IFN (range 1–6) from which the median numbers of MNCs and (protocol 1) were all given a therapeutic trial of ␣IFN fol- CD34ϩ cells collected were 3.45 ϫ 108/kg (range 1.2–

Bone Marrow Transplantation Autologous transplantation in CML NC McBride et al 1167 Table 1 Patient characteristics at baseline removal. GI toxicity (grade 1–3) was experienced in 12 (41%) patients. Skin rashes (grade 1–2) associated with Protocol 1 Protocol 2 All patients drug therapy occurred in five (17%) patients. Other non- (n ϭ 16) (n ϭ 13) haematological adverse events included headache (7%) and grade 1–3 mucositis (19%). Four (10%) patients experi- Median age 48 (20–64) 46 (28–56) 46 (20–64) enced reactions considered to be related to treatment with (years) (range) G-CSF (fever, bone pain and rash). There was no pro- Sex Male 10 9 19 cedure-related mortality or toxic death related to chemo- Female 6 4 10 therapy. Improved Sokal score High risk 1 1 2 Transplant outcome Intermediate 3 2 5 risk Low risk 12 9 21 To date, 15 patients (10 male, five female), with a median Not known 0 1 1 age of 45 years (range 25–59) have proceeded to PBSCT. Time (months) 5 (2–22) 24 (4–78) Nine of the 13 patients who had failed to respond to prior since CML ␣IFN proceeded to stem cell transplantation as soon as was diagnosis (range) feasible. Three patients were not transplanted because their collections were substantially Ph-positive and one patient refused to undergo the procedure and was lost to follow-up. Six of the newly diagnosed patients were transplanted 15.49) and 3.37 ϫ 106/kg (range 0.43–167), respectively. after failing to achieve a cytogenetic response after a mini- 28/29 patients had у2.0 ϫ 108/kg MNCs collected (min mum of 6 months on ␣IFN, post harvest. All patients were 1.20) and 26/29 patients had у1.0 ϫ 106/kg CD34ϩ cells in chronic phase at the time of transplant. Ten patients have collected (min 0.43). No significant differences in yield not been transplanted at the time of writing. Five had sub- were observed between patients who had or had not stantially Ph-positive collections; two were controlled with received prior ␣IFN (P ϭ 0.6). IFN post harvest (one has since undergone a MUD) and three remain on a trial of IFN post harvest. Cytogenetic analysis The median Ph-negativity of stem cells transplanted was 98% (range 61–100) with 47% of patients receiving 100% The degree of Ph-negativity in PBPC collections varied Ph-negative transplants. The median numbers of MNCs and considerably. In 9/29 (31%) patients all collected stem cells CD34ϩ cells transplanted were 3.24 ϫ 108/kg (range 0.69– were Ph-negative and 15/29 (52%) all collections were sub- 7.60) and 2.97 ϫ 106/kg (range 0.624–33), respectively. stantially (Ͼ95%) Ph-negative. The likelihood of Ph-nega- The median number of days required for neutrophils to tive phereses decreased with increasing number of phereses reach Ͼ0.5 ϫ 109/l was 18 (range 13–69). The median performed, but did not correlate with total WCC at the time number of days required for platelets to reach Ͼ20 ϫ 109/l of pheresis (P ϭ 0.22). The proportion of patients from and platelets Ͼ50 ϫ 109/l was 20 (range 10–42) and 28 whom sufficient numbers of Ph-negative stem cells were (range 13–234) respectively. Patients were hospitalised for obtained was similar between the groups who had, or had a median number of 22 days (range 17–71) in total and not received prior IFN (protocol 1 ϭ 9/16 patients with there was no procedure-related mortality. у95% Ph-negativity and protocol 2 ϭ 9/13 patients with Cytogenetic analysis of bone marrow was informative in у95% Ph-negativity). All phereses were positive for bcr- 11/15 patients when performed 3 months following trans- abl by RT-PCR. plant. This demonstrated one patient with 100% Ph-nega- Eight patients required a second harvest either because tivity; seven patients with Ph-negativity of 31, 40, 50, 58, the collections were substantially Ph-positive (n ϭ 5), or 65, 70 and 80% and three patients with 100% Ph-positivity. because of failure to collect adequate numbers of cells (n Six of 15 patients remain partially Ph-negative and one ϭ 4). Of these eight patients, five went on to have success- patient has had a major cytogenetic response (Ͼ90% Ph- ful second harvests. negativity) to ␣IFN, persisting at 23 months following PBSCT (see Table 2). Harvest tolerability At median follow-up of 2.3 years post transplant, 10 of the 15 autografted patients remain alive and most are in Patients spent a median of 25 days (range 13–59) in hospi- chronic phase. Four patients died of disease progression and tal for the harvest period. All 29 patients experienced one died from incidental trauma. This death has not been haematological toxicity (grade у3). Twenty-four (83%) censored from the survival curves, although the patient patients developed a fever following chemotherapy and 12 remained in chronic phase at the time of his death. The (41%) patients developed an infection that required treat- majority of patients are maintained on ␣IFN therapy with ment with intravenous antibiotics. Four (14%) patients or without hydroxyurea. Overall survival for all 29 patients developed proven or suspected fungal infections and were at 5 years from diagnosis is 70% and median survival 7.3 treated with AmBisome (NeXstar Pharmaceuticals, Cam- years (Figure 1). Survival for IFN non-responders who bridge, UK). One developed a Staphylococcus aureus sep- were transplanted is 74% at 5 years from diagnosis (Figure sis with an infected right atrial thrombus requiring surgical 2) and 75% at 3 years from transplant (Figure 3). Although

Bone Marrow Transplantation Autologous transplantation in CML NC McBride et al 1168 matched unrelated transplant; Cyt ϭ harvest. hydroxyurea; MUD ϭ interferon; HU ϭ traumatic death; IFN ϭ blast crisis; TR ϭ diagnosis transplant PBSCT % Ph-ve % Ph-ve % Ph-ve accelerated phase; BC ϭ chronic phase; AP ϭ transplant status from post prior to 3 months 9 months last FU interferon therapy Results of PBSCT in CML not available; CP ϭ cytarabine. 123456789 CP CP CP CP CP Dead (BC) CP Dead (BC) CP CML CP CP Dead (AP) CP CML 3.5 CP Dead (TR) 7.3 CML CP 6.7 CML 5.0 CP CML CP 8.9 6.0 2.2 4.9 9.2 5.3 5.0 3.9 6.8 4.8 2.1 0% N/A 4.8 3.3 N/A 0% 4.0 N/A N/A 0% Failed 15% Failed 0% 80% N/A 65% Failed 0% 0% 40% 0% 50% 15% 70% 25% 0% 0% 0% 20% 0% 5% 13% 0% 0% 0% Yes 0% Yes Yes 10% 8% Yes Yes Yes — Yes — IFN Yes Yes — IFN — IFN IFN HU Table 2 Patient Status at Current Survival SurvivalTime % in Ph-ve years. All patients with cytogenetic data not available prior to PBSCT Cyt were R patients at who, according to protocol, proceeded immediately to transplant after Cyt R at Cyt R at Post BMT Current N/A ϭ 101112131415 CP CP CP CP CP CML CP CP Dead (AP) CML CP CML CP CML CP 5.2 CML 3.7 CP 6.8 3.8 3.1 3.2 2.7 2.0 2.7 1.9 1.1 0.3 54% 54% 96% N/A 9% 0% 0% 0% N/A 100% 58% 31% 6% 0% 22% 80% N/A N/A 0% 0% 14% 95% 58% 31% Yes Yes Yes Yes Yes Yes IFN/HU IFN — IFN IFN IFN

Bone Marrow Transplantation Autologous transplantation in CML NC McBride et al

Dead Alive 1169 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2

Cumulative proportion surviving 0.1 0.0 0246810 Survival in years

Figure 1 Overall survival from diagnosis for all patients entered on to the study (n ϭ 29).

Dead Alive

1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2

Cumulative proportion surviving 0.1 0.0 012345678910 Survival in years

Figure 2 Overall survival from diagnosis for 15 interferon non-responders who proceeded to autologous transplantation. the numbers are small, revised Sokal risk score15 had no problems and had a rapid resolution of her encephalopathy. obvious impact on survival in this protocol (Figure 4). She was discharged on day 38 post transplant.

Transplant toxicity Discussion There were no transplant-related deaths. GI toxicity (grade у3) was experienced in 13 (87%) patients and seven (47%) Administration of G-CSF following cytoreductive chemo- patients developed moderate to severe mucositis. One therapy can result in the mobilisation of residual Ph-nega- patient developed an oesophageal ulcer which signiﬁcantly tive haemopoietic progenitor cells. Recent studies have prolonged his stay in hospital following PBSCT. Eleven shown that autologous progenitor cell rescue with Ph-nega- (73%) patients developed a bacterial infection in the post- tive cells after myeloablative chemotherapy may induce transplant period and all required treatment with intra- prolonged periods of haematologic and cytogenetic venous antibiotics. One patient developed fungal pneu- remission, and may be associated with prolonged sur- monia. One patient had reversible myocardial ischemia and vival.8,10,14,18 rapid atrial ﬁbrillation and another developed veno-occlus- In this study we report the outcome of several years fol- ive disease (VOD) of the liver on day 19 having regener- low-up of patients with CML undergoing high-dose chemo- ated her WCC. Despite being thrombocytopenic and plate- therapy and autologous progenitor cell rescue using pro- let refractory at this stage, she did not develop bleeding genitor cells which were completely, or substantially, Ph

Bone Marrow Transplantation Autologous transplantation in CML NC McBride et al

1170 Dead Alive 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2

Cumulative proportion surviving 0.1 0.0 0123456 Survival in years

Figure 3 Overall survival from transplant (n = 15).

Dead Alive 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 Low risk (n = 21) 0.4 0.3 0.2

Cumulative proportion surviving 0.1 Intermediate/high risk (n = 7) 0.0 0246810 Survival in years

Figure 4 Overall survival from diagnosis for all 29 patients entering the protocol according to the modiﬁed Sokal risk score.

negative by conventional cytogenetics. All subjects cessful second harvests following re-induction with chemo- undergoing high-dose therapy had failed to achieve a major therapy. cytogenetic response on ␣IFN and would have a relatively The actual timing of the harvest was influenced by the poor outcome independently of Sokal risk score. All white cell count and the yield of CD34ϩ cells increased as patients were placed back on ␣IFN following stem cell res- the white cell count regenerated after chemotherapy. How- cue in an attempt to achieve or maintain Ph-negative ever, we found no correlation between white cell count at haemopoiesis. the time of harvest and the likelihood of collecting Ph-nega- We have confirmed that Ph-negative cells can be suc- tive cells. Earlier phereses were more likely to be Ph-nega- cessfully collected from the peripheral blood of patients tive, consistent with the hypothesis that negative cells either with CML and can be used for subsequent autologous trans- mobilize earlier, or Ph-positive progenitors are more sensi- plant after myeloablation. Sufficient numbers of CD34ϩ tive to chemotherapy and as a result, take longer to regener- cells and MNCs to proceed safely to transplantation were ate. Continuing phereses in order to obtain more cells collected in 90% and 97% of patients respectively, which was therefore self-limiting because of the increase in Ph- compares favourably with other reported studies.7–10,19 Ph- positivity of the pheresis product. negative stem cells were also successfully mobilised in the Previous reports have shown a significant difference in majority of patients. 9/29 (31%) collections were all Ph- the PBSC harvest between patients who had not been negative and 15/29 (52%) collections were substantially treated with ␣IFN and were mobilised in the first few (Ͼ95%) Ph-negative. Although the first harvest failed in weeks after diagnosis (with WBC controlled by eight patients, five of these patients went on to have suc- hydroxyurea) and patients who had received ␣IFN for Ͼ1

Bone Marrow Transplantation Autologous transplantation in CML NC McBride et al 1171 year.8,19 We found no obvious disadvantage in harvesting ing autografting with cultured marrow.5 Carella’s group cells after prior exposure to ␣IFN, providing an adequate was able to restore and maintain a major or complete cyto- IFN-free rest period was used – however, our numbers are genetic response in greater than 50% of their patient comparatively small. There was also no evidence that the population.10 percentage of Ph-negative stem cells harvested was influ- In conclusion, PBSCT could be an option for patients enced by time from diagnosis, or by revised Sokal score.15 who are ineligible for allogeneic transplantation and may We have confirmed that autologous transplantation with prolong chronic phase. The procedure is reasonably well Ph-negative cells, and subsequent ␣IFN maintenance, in tolerated and patients not responding to ␣IFN may be patients failing to achieve an initial response to ␣IFN can induced into Ph-negativity with PBSCT, although this is induce a cytogenetic response post PBSCT. Clearly inten- not sustainable in the majority of patients. Further studies sive chemotherapy alone, without progenitor cell rescue, are needed to investigate more effective purging techniques may produce transient Ph-negative haemopoiesis, but it is in order to reduce contamination with Ph-positive cells dur- rarely durable. Moreover, it can result in encouraging sur- ing transplant. Another option is to improve post-transplant vival times in an otherwise relatively poor risk group of maintenance by using pegylated interferon, cytarabine or patients as survival for patients not responding to ␣IFN is even STI571. consistently lower than in interferon responders.11–13,20 Median survival for patients failing to achieve a major cytogenetic response to ␣IFN, irrespective of age and other risk References factors, varies from 3.5 to 5.5 years. The patients in our series did not differ significantly from these series in terms 1 Clift RA, Buckner CD, Thomas ED et al. Treatment of CGL of median age or modified Sokal score. in chronic phase by allogeneic BMT. Lancet 1982; ii: 621– Actuarial survival for patients in our series at 3.5 years 623. post autograft is 75%. This is comparable to the 87% sur- 2 Biggs JC, Szer J, Crilley P et al. Treatment of CML with vival reported by the Genoa group.10 Survival from diag- allogeneic BMT after preparation with BuCy2. Blood 1992; nosis for autografted patients is 74% at 5 years; again, this 80: 1352–1357. compares favourably with other similar studies.9,10 The sur- 3 Coulombel L, Kalousek DK, Eaves CJ et al. Long term marrow culture reveals chromosomally normal haemopoietic pro- vival durations, on crude observation, compare favourably + ␣ genitor cells in patients with Ph ve CML. New Engl J Med with those reported for patients failing to respond to IFN 1983; 308: 1493–1498. 13 who do not undergo autologous transplantation. Survival 4 Hoyle C, Gray R, Goldman J et al. Autografting for CGL in in our series also compares favourably with other studies chronic phase: an update. Br J Haematol 1994; 86: 76–81. of autologous transplantation in CML using unmanipulated 5 Barnett MJ, Eaves CJ, Phillips GL et al. Autografting with progenitors. Five-year survival following autologous trans- cultured marrow in chronic myeloid leukaemia: results of a plantation with unmanipulated buffy coat or bone marrow pilot study. Blood 1994; 84: 724–732. in unselected patients varies from 55 to 60%.4,6 These 6 McGlave PB, De Fabritiis P, Deisseroth A et al. Autologous patients have not been selected by previous exposure to transplants for chronic myelogenous leukaemia: results from ␣IFN, making direct comparison difficult. The Vancouver eight transplant groups. Lancet 1994; 343: 1486–1488. 7 Chalmers EA, Franklin IM, Kelsey SM et al. Mobilisation of group have achieved significantly better results with in vitro Ph-negative progenitors into peripheral blood in chronic purging, although at the expense of delayed neutrophil myeloid leukaemia. Br J Haematol 1994; 96: 627–634. 5 engraftment (29 days vs 18 days). These data suggest that 8 Carella AM, Cunningham I, Lerma E et al. Mobilisation and in vivo or ex vivo purging of Ph-positive progenitors prior transplantation of Philadelphia negative peripheral blood pro- to stem cell rescue does confer a survival benefit. genitor cells early in chronic myelogenous leukaemia. J Clin Overall survival for all 29 patients in this series at 5 Oncol 1997; 15: 1575–1582. years from diagnosis is 70% and median survival 7.3 years. 9 Singer IO, Franklin IM, Clark RE et al. Autologous tranplant- However, these data need to be considered independently ation in chronic myeloid leukaemia using peripheral blood as they contain patients who may have selected themselves stem cells. Short report. Br J Haematol 1998; 102: 1359–1362. out as good responders to ␣IFN and therefore may never 10 Carella AM, Lerma E, Corsetti MT et al. Autografting with Philadelphia chromosome-negative mobilised haematopoietic need a transplant. Interestingly, the revised Sokal risk score progenitor cells in chronic myelogenous leukaemia. Blood at diagnosis had no obvious impact on survival in this 1999; 93: 1534–1539. protocol, although the numbers are small. 11 Talpaz M, Kantarjian HM, McCredie K et al. Haematological The majority of patients who were transplanted are alive remission and cytogenetic improvement induced by recombi- and remain stable on standard therapy with a median fol- nant human interferon alpha in CML. New Engl J Med 1986; low-up of 4.7 years from diagnosis. The majority of 314: 1065–1069. patients experienced a brief period of conversion to partial 12 Tura S, Baccarani M, Zuffa E for the Italian Cooperative or complete Philadelphia negativity following autografting. Study Group on Chronic Myeloid Leukaemia. Interferon alfa- One patient has had a major cytogenetic response (Ͼ90% 2a as compared with conventional chemotherapy for the treat- Ph-negativity) to ␣IFN, persisting 23 months post PBSCT, ment of chronic myeloid leukaemia. New Engl J Med 1994; 330: 820–825. and two achieved a major response. The non-sustainability 13 Allan NC, Richards SM, Shepherd PCA. UK Medical of a durable cytogenetic remission following PBSCT is in Research Council randomised multicentre trial of interferon ␣ accordance with other studies, although the Vancouver for chronic myeloid leukaemia: improved survival irrespective group have shown patients to continue with predominantly of cytogenetic response. Lancet 1995; 345: 1392–1397. Ph-negative haematopoiesis for more than 3 years follow- 14 Kantarjian HM, Smith TL, O’Brien S et al. Prolonged survival

Bone Marrow Transplantation Autologous transplantation in CML NC McBride et al 1172 in chronic myelogenous leukaemia after cytogenetic response 18 Carella AM, Simonsson B, Link H et al. Mobilisation of Phil- to interferon-␣ therapy. Ann Intern Med 1995; 122: 254–261. adelphia negative peripheral blood progenitor cells with 15 Hasford J, Pﬁrrmann M, Hehlmann R et al. A new prognostic chemotherapy and rhuG-CSF in chronic myelogenous leu- score for survival of patients with chronic myeloid leukaemia kaemia patients with a poor response to interferon-alpha. Br treated with interferon alfa. Writing Committee for the Collab- J Haematol 1998; 101: 111–118. orative CML Prognostic Factors Project Group. J Natl Cancer 19 Carella AM, Frassoni F. ICE, mini ICE or high-dose hydrox- Inst 1998; 90: 850–858. yurea to mobilise Philadelphia negative PBPC in chronic mye- 16 Guilhot F, Chastang C, Michallet M et al for the French CML logenous leukaemia. Br J Haematol 1996; 95: 212–216. Study Group. Interferon alfa-2b combined with cytarabine vs 20 Hehlmann R, Heimpel H, Hasford J and the German CML interferon alone in chronic myeloid leukaemia. New Engl J Study Group. Randomized comparison of interferon-alpha Med 1997; 337: 223–271. with busulfan and hydroxyurea in chronic myelogenous leuke- 17 Hughes TP, Morgan GJ, Martiat P, Goldman JM. Detection mia. Blood 1994; 84: 4064–4077. of residual leukaemia after bone marrow transplant for chronic myeloid leukaemia: role of polymerase chain reaction in pre- dicting relapse. Blood 1991; 77: 874–878.

Bone Marrow Transplantation