High-dose therapy for diffuse large-cell lymphoma in first remission

Annals of Oncology 9 (Suppl. I): S9-S14. 1998. © 1998 Khmer Academic Publishers. Printed in the Netherlands.

Review High-dose therapy for diffuse large-cell lymphoma in first remission A. R. Perry & A. H. Goldstone University College London Hospitals. London, UK

Summary

Introduction

The optimal timing of any treatment for any disease depends on which outcomes can be achieved, when, and at what cost. Since diffuse large-cell lymphoma (DLCL) is curable in some patients, improvements in therapy must as a priority be aimed at increasing the overall number of patients that are cured. However, this should be achieved at low risk to the patients in terms of short and long-term morbidity and mortality and - as a secondary consideration - at lowfinancialcost. On theoretical grounds, high-dose therapy (HDT) with autologous stem cell support has the potential to increase cure rates, for DLCL is a chemosensitive disease that displays a steep dose-response curve. However, the risks of HDT are not negligible, and the treatment carries considerable resource implications.

Definitions DLCL is the most common single histological diagnosis among patients with non-Hodgkin's lymphoma (NHL), accounting for about a third of all NHL [1]. Collating evidence from the literature on the treatment of DLCL is made difficult by the fact that many reports do not differentiate between histological subtypes of high-grade NHL and, when they do, the definition of DLCL may vary. As far as possible, this discussion will focus on DLCL, to include diffuse large cell and immunoblastic B-cell subtypes (Working Formulation categories G and H). Although DLCL is clearly somewhat heterogeneous

Key words: autologous transplantation, diffuse large-cell lymphoma, high-dose therapy, lymphoma, peripheral blood stem cell transplantation, prognostic factors

in its response to treatment, these two categories have been found to have very similar remission and relapse rates [2]. Where the evidence applies less specifically to broader categories of NHL, this will be noted in the text. The timing of HDT DLCL has been treated for several decades with combination chemotherapy regimens, the original of which was CHOP. Despite attempts at using regimens such as MACOP-B that include a greater number of drugs, no single regimen has consistently proved to be better than full-dose CHOP [3-7]. Approximately 50%-60% of patients with DLCL achieve and maintain complete remission after first-line therapy; 30%-40% relapse and 10% have refractory disease. HDT almost certainly benefits some of the patients in the worst group - those with refractory disease; it also has a significant survival advantage over conventional salvage therapy for those with relapsed disease [8]. What is not clear is whether HDT should be witheld until the patient relapses, or used upfront as part offirst-linetherapy, particularly for patients who have achieved remission. Theoretical arguments for the latter approach include a potential increase in overall cure rates and, possibly, a reduced toxic morbidity and mortality for HDT when undertaken in remission. In an analysis of EBMT data (unpublished), almost 30% of patients with DLCL are transplanted in first CR, and toxic deaths do indeed appear to be lower in remission 4.4% in those transplanted in CR, but 10.6% in those transplanted not in remission.

Downloaded from http://annonc.oxfordjournals.org/ at Florida International University on July 18, 2015

Diffuse large-cell lymphoma (DLCL) is curable by first-line conventional chemotherapy in 50%-60% of patients. High-dose therapy makes no contribution to this group of patients and, if applied indiscriminately as first-line consolidation therapy, is likely to unnecessarily increase overall morbidity and mortality. Instead, recent interest has been directed towards (a) the identification of a group of patients with a poor prognosis, and (b) the intensification of first-line treatment for such patients with high-dose therapy and allied regimens. Many prognostic

factors have now been standardised, while studies are progressing in the identification of newer prognostic factors, such as the molecular markers. Multi-centre randomised trials are currently in progress to determine the appropriate level of treatment for prognostic subsets, with the value of high-dose therapy being assessed for those in the worst prognostic groups.

10 Table 1 Poor prognostic factors for patients with diffuse large-cell lymphoma. Ref.

Histology

Stage

n

Poor prognostic factors

[22]

DLCL

I & II

147

Increased age Wide tumour extent Raised LDH

[23]

Working Formulation F, G, H

161

Bcl-2 expression present

DLCL

151

Bcl-2 expression high

[24]

(WF:G,H) DLCL/undifferenciated

III & IV

113

Large tumour size B symptoms Extra-nodal sites

[26]

DLCL

II-IV

151

Bulky mediastinal or retropentoneal disease Raised LDH Increased age

[27]

DLCL with BM involvement

50

Large-cell lymphoma in bone marrow (not small cleaved)

[28]

DLCL

95

Age Actual relative dose intensity CHOP Cyclophosphamide dose Stage III to IV

[29]

DLCL

115

Relative dose intensity adriamycin Raised LDH Performance status 3 or 4

[30]

DLCL

[31]

Aggressive NHL (50% DLCL)

[32]

DLCL

[33]

[26]

Advanced

180

Raised LDH Bone marrow involvement High tumour burden

737

Raised LDH Stage III-IV Two or more extra-nodal sites Tumour mass > 10 cm

II-IV

121

Performance status 2 or more Two or more extra-nodal sites Tumour bulk > 1 0 c m

DLCL

Advanced

126

B symptoms Two or more extranodal sites Age > 60 Three or more nodal sites

DLCL

III & IV

105

High tumour burden Raised LDH

Theoretical considerations Supposing that all patients with DLCL are treated conventionally in the first instance, with HDT being reserved for first relapse. For those who achieve remission, the chance of cure with first-line therapy is about 60%, and the chance of cure with HDT after relapse is approximately 30% 40% [9-11], in some reports, greater

Poor prognostic factors Table 1 shows some of the factors that have been associated with poor prognoses in DLCL/high-grade NHL patients. One criticism of the analyses is that prognosis is often assessed from diagnosis, with the adverse factors affecting the chance of achieving remission, as well as the chance of relapsing. What is perhaps more appropriate for decision-making is the probability of relapse once remission is achieved. Frequently identified prognostic associations include stage of disease: generally, poor prognosis patients are those with stage III or IV disease - although there may be a tiny subset of stage II patients who likewise will fare poorly [6]. Several attempts to rationalise some of the associations into prognostic categories have been made (Table 2a). In 1993, an endeavour to reach international consensus resulted in the development of the International Prognostic Index, defined as in Table 2b [12]. This index appears to be fairly robust when compared to other indices, and carries the advantage of its relative simplicity in terms of clinical application [6]. Once again, however, the features corresponding to an increased risk of relapse were also associated with a decreased likelihood of obtaining an initial CR, making decision-making in remission difficult. Nevertheless, those patients who fall into the worst groups have less than 50% chance of long-term survival, and even those that achieve CR have a lower relapsefree survival than those in better risk groups (Table 2b), making experimental approaches - including HDT in first CR - justified. The question remains whether HDT can improve the survival in such poor prognosis pa-

Downloaded from http://annonc.oxfordjournals.org/ at Florida International University on July 18, 2015

[25]

than 50%. The overall curative potential of this schedule (ignoring, for the moment, non-disease-related mortality) is thus about 75%. HDT given to all patients in first remission would have to improve on this cure rate. Supposing, also, that the early toxic mortality for salvage chemotherapy and HDT is 10% in relapsed patients. Then, for all those who achieve remission, the overall chance of dying a transplant-related death if HDT is witheld until relapse is 4%. HDT may indeed be safer in patients in first remission, but to carry an advantage (in terms of toxic deaths) over waiting until the patient relapses, the procedure-related mortality for HDT in first CR should be less than 4%. Furthermore, if HDT is used as consolidation therapy, rather than for patients with relapsed/refractory lymphoma, over half of patients would undergo the procedure unnecessarily, with the attendant cost implications that this would entail. In fact, because of such considerations, HDT is not generally recognised as justifiable consolidation treatment for all patients with DLCL. Rather, attention has been paid to the selection of those patients with DLCL who have a poor prognosis - where HDT may, perhaps, have an impact on survival. This approach depends firstly on the appropriate selection of patients.

Table 2a. Stratification of diffuse large-cell lymphoma into prognostic subgroups - examples. [22]

Stagel&lIDLCL

71 = 147

1. Raised LDH 2. Tumour extent

- Neither (1) nor (2) -d)or(2) -d)and(2)

[31]

Aggressive NHL 50% DLCL at diasnosis

7i = 737

1. 2. 3. 4.

[34]

DLCL III & IV

n = 105

1. High tumour burden 2. Raised LDH

[32]

DLCL with BM invt

- No factors 3yS (88%) 3yS(71%) - (1) only, or two others - (1) and (2. 3 or 4) or (2. 3 and 4) 3yS(41%) - No factors - (I) and (2)

5yS (87%) 5yS (48%) 5yS (20%) 5yS (68%) 5yS (55%) 5yS (24%)

71= 121

1. Performance status 2+ 2. 2+ extra-nodal sites 3. Tumour bulk > 10 cm

- No factors - ( 2 ) or (3) - (1) or (2 and 3)

n = 161

Bcl-2 expression

- No -Yes

7i = 50

Small-cleaved-cell lymphoma in BM Large-cell lymphoma in BM

10yDFS(69%) 10yDFS(35%) 5yS (79%) 5yS(12%)

Table 2b. The International Prognostic Index (age-adjusted). [12]

Sixteen cooperative groups and institutions in Europe, the US and Canada Significant pre-treatment factors for 1274 patients aged 60 years or less: Stage (I or II vs. Ill or IV) ECOG performance status (0 or I vs. > / ) LDH (normal vs. greater than normal)

Low risk (0-1 risk factors) Low intermediate risk (one factor) High intermediate risk (two factors) High risk (three factors)

Distribution

CR

Five-year DFS for those in CR

Five-year survival

22% 32% 32% 14%

92% 78% 57% 46%

86% 66% 53% 58%

83% 69% 46% 32%

tients, or whether their poor-prognostic characteristics mean that all treatments are equally likely to fail. HDT for poor prognosis patients A number of studies have used HDT as consolidation therapy for poor prognosis patients, summarised in Tables 3 and 4. There are two problems in attempting meta-analysis: firstly, the various studies differ in terms of disease characteristics included, in terms of the prognostic criteria employed, and in terms of the high-dose therapy used. Secondly, in some studies categorisation into risk-groups (and, in several cases, randomisation) takes place at diagnosis rather than on achieving remission. As a consequence not all the patients are necessarily in CR at the time of transplant. However, some of the data do suggest an improvement in survival for HDT over controls or historical comparisons - thus, HDT may have a role in improving the outlook for poor prognosis patients. Two of the trials merit more detailed discussion. The Groupe d'Etude des Lymphomes de TAdulte (GELA) recruited 1043 patients into the LNH87-2 study,

from 35 participating centres [13]. Patients were 16-55 years with intermediate and high-grade NHL and at least one adverse prognostic factor. About 65% of patients had DLCL. Patients received induction chemotherapy according to a separate randomisation, and patients who achieved complete remission were then randomised to receive either sequential chemotherapy consolidation, or intensive consolidation with the CBV regimen followed by autologous bone marrow transplantation. 614 patients achieved CR, and there was no difference in disease-free or overall survival between the two consolidative arms, even after stratification by induction regimen. However, when a subgroup of 236 high-risk patients with two or three risk factors was analysed, a significant advantage of high-dose therapy was seen in terms offive-yeardisease-free survival (59% vs. 39%, P - 0.01) although no overall survival difference has yet been observed (Table 3). A randomised trial, with cross-over option, of highdose sequential therapy including a myeloablation/stem cell transplantation phase - versus MACOP-B chemotherapy - has also recently been reported [14]. Patients were 17-60 years, with DLCL (Working Formulation G and H) and bulky disease or stage III-IV. More patients

Downloaded from http://annonc.oxfordjournals.org/ at Florida International University on July 18, 2015

[27]

Large cell lymphoma

Raised LDH Stage III-IV Two + extranodal sites Tumour mass > 10 cm

10yS(79%) 10yS(62%) 10yS(44%)

12 Table 4. Non-randomised studies of high-dose therapy in first-line therapy for poor prognosis patients.

HDTfor patients in firsl CR [13] n = 236 (representing a subgroup of the LNH87-2 trial) with intermediate or high-grade NHL (68% DLCL) and two or more of: performance status 2-4. tumour burden > 10 cm, bone marrow or CNS involvement. Burkitt's or lymphoblastic subtypes Patients identified at diagnosis, and randomised if CR achieved

[36]

n - 12 patients with intermediate/high-grade NHL (six with DLCL) and either bulk disease or raised LDH at diagnosis Patients selected for HDT if CR achieved Cyclophosphamide/TBI high-dose therapy Transplant-related mortality 17% Median follow-up 32 months

DFS 67% at 32 months

Randomised to: high-dose therapy with CBVconditioning patient sequential chemotherapy

[37]

n = 20 patients with aggressive NHL (nine with DLCL) and either raised LDH and one of three other adverse factors (tumour bulk > 10 cm, stage III-IV, two or more extranodal sites), or normal LDH and all other factors, present at diagnosis Patients selected for HDT if CR achieved Treated with HDT including total body irradiation in 17 patients Transplant-related mortality 0% Median follow-up 34 months

DFS 84% at 13 months

[38]

// = 30 patients with aggressive NHL (17 with DLCL) and one of five adverse factors (stage IV, tumour bulk > 10 cm. two or more extranodal sites. NHL after treatment for Hodgkin's disease, CNS involvement) at diagnosis Patients selected for HDT if CR achieved High-dose melphalan +/-etoposide/TBI Transplant-related mortality 0% Median follow-up 44 months

EFS83%at three years

[39]

n = 33 patients with high grade NHL (20 with DLCL) and in high- or high-intermediate risk groups according to the 1PI Patients identified at diagnosis Busulphan/cyclophosphamide high-dose therapy Transplant-related mortality 6% Median follow-up 31 months

EFS 61% at two years

[15]

n = 34 patients with DLCL and stage II disease with B symptoms or bulk disease, or stage III-IV at diagnosis Patients received HDT in partial or complete remission BEAM high-dose therapy Transplant-related mortality 0% Median follow-up 25 months

EFS 70% at two years

[40]

n = 50 patients with DLCL. with high tumour burden and raised LDH or with large cell bone marrow involvement Patients identified at diagnosis BEAM high-dose therapy Transplant-related mortality 6% Complete remission 72% Median follow-up 32 months

DFS 69% at 32 months for those achieving CR

[41]

n = 18 patients with high-risk intermediate or high-grade NHL. according to the I PI Patients identified at diagnosis CBV + TBI high-dose therapy Transplant-related mortality 22% Median follow-up 15 months

EFS 56% at 15 months

Median follow-up 54 months

Toxic deaths

Five-year DFS

HDT Sequential chemotherapy Difference

0% 0%

59% 39% P = 0.01

versus out-

HDT as part of induction therapy [14]

n = 98 patients with D L C L and bulk disease or stage III —IV at diagnosis. Patients identified at diagnosis

Randomised to' high-dose sequential therapy (melphalan-TBIor mitoxanthrone-melphalan myeloablation) versus MACOP-Bfor 12 weeks Median follow-up 55 months

Toxic deaths

CR

Seven-year EFS

HD sequential MACOP-B Difference

8% 6%

96% 70% P = 0.001

76% 49% P = 0.004

[35] H = 124 patients with aggressive NHL with stage 11—III and one adverse factor, or stage IV. Patients identified at diagnosis Randomised to: VACOP-B plus ABMT versus VACOP-B, followed by radiotherapy or second-line chemotherapy as necessary No difference in three-year OS (65% vs. 67%), DFS or PFS. High dropout rate prior to HDT

in the HDT arm achieved remission, and freedom-fromprogression of disease and event-free survival rates were also significantly greater in this arm. Again, differences in overall survival do not reach significance at this stage of reporting. Notwithstanding the fairly low procedure-related mortality in some reports, it should be noted that most studies have not looked at chronic morbidity, although one study identifies incomplete long-term platelet recovery as a potential problem [15]. In addition, follow-up for most studies remains short. Even where a survival plateau appears to have been reached, it is possible, for example, that those patients who received HDT will have a higher late complication rate than those who did not. Randomised trials are currently on-going to examine this entire issue further. One such UK trial, run by the British National Lymphoma Institute, compares three courses of CHOP followed by autologous stem cell transplant, with six courses of CHOP then stop, in poor prognosis patients (stage III or IV disease, raised LDH

or WHO performance status 2-4). Recruitment so far stands at more than 250 patients, and the outcome is awaited. Other ongoing trials include the EORTC-20901

Downloaded from http://annonc.oxfordjournals.org/ at Florida International University on July 18, 2015

Table 3 Randomised studies of high-dose therapy in first-line therapy for poor prognosis patients

13 trial, randomising patients with aggressive NHL to chemotherapy followed by BEAC HDT and radiotherapy, or to chemotherapy followed by radiotherapy alone; and the LY02 EBMT trial, which randomises high-risk patients to chemotherapy versus high-dose therapy.

asTaxol, and the introduction of drug resistance inhibitors into the treatment schedule [20]. Any perceived advantages of HDT today must be reconsidered in the light of future developments in the treatment of DLCL. Conclusions

Intensive sequential therapy with PBSC support

Alternatives to HDT It may be that HDT is neither the only, nor the best, option for improving survival beyond remission in DLCL patients. One alternative approach is to intensify standard first-line regimens with the aid of colony stimulating factors, as described, for example, by Shipp et al. [17]. In this dose-finding study of patients with high-risk aggressive NHL, escalating doses of doxorubicin and cyclophosphamide were used in a CHOP-regimen, with G-CSF support. At the maximum tolerated dose (MTD), 86% of 22 patients achieved CR, 79% of whom were progression-free at median 20-month follow-up. It should additionally be noted that thrombocytopenia was the dose-limiting toxicity in this study, and the emergence of newer growth factors such as thrombopoietin may allow even higher doses to be tolerated. Another phase I trial [18] in a similar disease group examined the effects of simultaneous dose-escalation of four agents in the ProMACE-CytaBOM regimen, in conjunction with GM-CSF, and showed a MTD of 200%, and the Southwest Oncology Group are currently undertaking a phase II trial of dose-intensified CHOP versus dose-intensified ProMACE-CytaBOM. However, when dose intensification becomes high enough to necessitate prolonged hospitilisation for neutropenia or gut toxicity, there is arguably minimal difference between this and high-dose therapy, and arguably no benefit, in terms of morbidity or economic considerations, of G-CSF over PBSC support. Adjuvant radiotherapy is a further strategy which may improve outcome over conventional therapy [19], although it should be noted that advanced stage NHL tends to relapse outside sites of prior disease, unlike limited stage NHL, and radiotherapy may not therefore be the most appropriate consolidative treatment. Other approaches include the exploration of newer agents such

In the absence of a large international randomised trial with uniform inclusion criteria, the merit of HDT in DLCL must be assessed on the basis of the smaller trials, published and on-going, with the usual caveats involved in the interpretation of trials with small sample size [21]. It seems unlikely at present that HDT will be deemed appropriate for indiscriminate application to all patients with DLCL in first remission. However, it seems probable that there does exist a group of poor prognosis patients who would benefit from HDT in first CR, given the alternative options available at the present time. The difficulty then comes in identifying these patients systematically, so that the high dose therapy 'net' is neither cast too widely or too narrowly. The International Prognostic Index (IPI) has helped to standardise the approach, but with time and a greater understanding of the heterogeneity of DLCL, other prognostic markers (such as bcl-2 expression) may become important. This leads to the philosophical question of how far therapy should be tailored to the individual disease of the individual patient, at the expense of adherence to a uniform approach and the creation of an evidence-base for future improvements. Since, at present, the evidence-base is not substantial, it is perhaps better to persist with a standardised approach. On these grounds, patients who fall into the IPI categories 'high-risk'or 'high-intermediate' risk should currently be offered the option of HDT in first remission unless there is an appropriate randomised trial available; meanwhile, entry into on-going trials should be encouraged.

References 1. The NHL Classification Project. A National Cancer Institute sponsored study of classification of NHL. Summary and description of a Working Formulation for clinical usage. Cancer 1982; 49:2112-35. 2. Kwak LW, Wilson M. Weiss LM et al. Clinical significance of morphological subdivision in diffuse large-cell lymphoma. Cancer 1991: 68: 1988-93. 3. Armitage JO. The place of third generation regimens in the treatment of adult aggressive non-Hodgkin's lymphoma. Ann Oncol 1991;2(Suppl 1): 37-41. 4. Vose JM. Armitage JO. The present status of therapy for patients with aggressive non-Hodgkin's lymphoma. Ann Oncol 1991: 2 (Suppl 2): 171-6. 5. Matthews JR, Cooper IA, Matthews JP. Ding JC. Failure of intensive therapy in poor prognosis non-Hodgkin's lymphoma. Aust N Z J Med 1992: 22: 123-8. 6. Bartlett NL. Kwak LW, Horning SJ et al. MACOP-B + / - radiation therapy for diffuse large-cell lymphoma. Analysis of the Stanford results according to prognostic indices. Cancer 1993; 71: 4034-42.

Downloaded from http://annonc.oxfordjournals.org/ at Florida International University on July 18, 2015

Recently, Stoppa et al. [16] reported a feasibility study of repeated cycles of intensive therapy with PBSC support. Of 20 NHL patients, 13 had DLCL, and all were in the high- or high-intermediate-risk group as determined by the age-adjusted international prognostic index. The patients were mobilised with GCSF following an intensified CHOP regimen in each of three cycles; apheresed PBSCs were then returned after cycles 4, 5 and 6, each consisting of intensified CHOP plus etoposide and cisplatin. This schedule proved fairly toxic, with one toxic death; 65% of patients achieved CR, of whom 86% had maintained two-year disease-free survival.

14 7

8.

9.

10.

11

13.

14.

15.

16.

17.

18.

19.

20. 21. 22.

23.

24.

25.

26. Straus DJ. Wong G, Yahalom J et al. Diffuse large cell lymphoma. Prognostic factors with treatment Leukaemia 1991: 5 (Suppl 1): 32-7. 27. Robertson LE, Redman JR. Butler JJ et al. Discordant bone marrow involvement in diffuse large cell lymphoma: A distinct clinical-pathological entity associated with a continuous risk of relapse. J Clin Oncol 1991: 9: 236-42. 28. Epelbaum R, Faraggi D et al. Survival of diffuse large-cell lymphoma. A multivariate analysis including dose intesity variables Cancer 1990:66: 1124-9. 29. Kwak LW, Halpern J, Olshen RA el al. Prognostic significance of actual dose intensity in diffuse large-cell lymphoma. Results of a tree-structured survival analysis. J Clin Oncol 1990; 8: 963-77. 30. Resegotti L, Vitolo U, Bertini M et al. MACOP-B for advanced stage DLCL. More is better'' Italian Multiregional Cooperative Study Group (MCSGL). Leukaemia 1991; 5 (Suppl 1): 86-9. 31. Coiffier B. Gisselbrecht C. Vose JM et al. Prognostic factors in aggressive malignant lynphomas: Description and validation of a prognostic index that could identify patients requiring a more intensive therapy The Groupe d'Etudes des Lymphomes Agressifs. J Clin Oncol 1991; 9: 211-9. 32. Shipp MA, Harrington DP, Jochelson MS et al. Identification of major prognostic subgroups of patients with large-cell lymphoma treated with M-BACOD or M-BACOD. Ann Intern Med 1986; 104: 757-65. 33. Hoskins PJ, Ng V, Spinelli JJ et al Prognostic variables in patients with diffuse large-cell lymphoma treated with MACOP-B. J Clin Oncol 1991; 9: 220-6. 34. Gianni AM, Bregm M, Siena S et al. High-dose chemotherapy and autologous bone marrow transplantation compared with MACOP-B in aggressive B-cell lymphoma N Engl J Med 1997; 336(18): 1290-7. 35. Santini G, Salvagno L, Chiesesi Tet al.VACOP-B vs. VACOP-B + autologous BM transplantation for aggressive non-Hodgkin's lymphoma. Bone Marrow Transplant 1997; 19 (Suppl 1): S154. 36. Baro J, Richard C, Calavia J et al. Autologous bone marrow transplantation as consolidation therapy for non-Hodgkin's Lymphoma patients with poor prognostic features. Bone Marrow Transplant 1991; 8: 283-9. 37. Nademanee A, Schmidt GH, O'Donnell MR et al. High-dose chemoradiotherapy followed by autologous bone marrow transplantation as consolidation therapy during first complete remission in adult patients with poor risk aggressive lymphoma: A pilot study. Blood 1992, 80: 1130-4. 38. Jackson GH, Lennard AL. Taylor PRA et al. Autologous bone marrow transplantation in poor risk high-grade non-Hodgkin's lymphoma in first complete remission. Br J Cancer 1994; 70' 501-5. 39 Pettengell R. Radford JA. Morgenstern GR et al. Survival benefit from high dose therapy with autologous blood progenitor cell transplantation in poor prognosis non-Hodgkin's lymphoma J Clin Oncol 1996: 14. 586-92. 40. Vitolo U, Cartellazzo S, Liberati AM et al. Intensified and high dose chemotherapy with granulocyte colony stimulating factor and autologous stem cell transplantation support as first-line therapy in high-risk diffuse large cell lymphoma. J Clin Oncol 1997; 15:491-8. 41. Milpied N, Delwail V, Sensebe L et al. sequential high-dose therapy for high-risk intermediate and high-grade adult patients with non-Hodgkin's lymphoma. A pilot study of Goelams group. Bone Marrow Transplant 1997; 19 (Suppl 1): SI54.

Correspondence to: Dr A. H. Goldstone Department of Clinical Haematology Unhersity College Hospital Grafton Way London WC1E6AU UK

Downloaded from http://annonc.oxfordjournals.org/ at Florida International University on July 18, 2015

12.

Savarese DMF. Hsieh C, Stewart FM. Clinical impact of chemotherapy dose escalation in patients with hematologic malignancies and solid tumours. J Clin Oncol 1997; 15: 2981-95. Philip T. Guglielmi C. Hagenbeek A et al. Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive NHL. N Engl J Med 1995: 333: 1540-5. Philip T. Armitage JO. Spitzer G et al. High dose therapy and ABMT after failure of conventional therapy in adults with intermediate grade and high-grade non-Hodgkin's lymphoma. N Engl J Med 1987 316: 1493-8. Gribben JG, Goldstone AH, Linch DC et al. Effectiveness of high-dose combination chemotherapy and autologous bone marrow transplantation for patients with non-Hodgkin's lymphomas who are still sensitive to conventional dose therapy. J Clin Oncol 1989; 7: 1621-9. Gisselbrecht C. Autologous stem cell transplantation in aggressive non-Hodgkin's lymphoma. Hematol Cell Ther 1996; 38: 297304. The International Non-Hodgkin's Lymphoma Prognostic Factors Project. A predictive model for aggressive non-Hodgkin's lymphoma. N Engl J Med 1993; 329- 987-94. Haioun C, Lepage E. Gisselbrecht C et al Benefit of ABMT over sequential chemotherapy in poor risk aggressive non-Hodgkin's lymphoma: Updated results of the prospective study LNH87-2. J Clin Oncol 1997; 15. 1131-7. Jagannath S, Velasquez WS, Tucker SL et al. Tumor burden assessment and its implication for a prognostic model in advanced diffuse large-cell lymphoma. J Clin Oncol 1986; 4 (6): 859-65. Rossi A. Cortelazzo S, Bellavita P et al. Long-term haematological reconstituton following BEAM and autologous transplantation of circulating progenitor cells in non-Hodgkin's lymphoma Br J Haematol 1997; 96: 620-6. Stoppa AM. Bouabdallah R, Chabannon C et al. Intensive sequential chemotherapy with repeated blood stem-cell support for untreated poor-prognosis non-Hodgkin's lymphoma. J Clin Oncol 1997; 15: 1722-9. Shipp MA, Neuberg D, Janicek M et al. High-dose CHOP as initial therapy for patients with poor-prognosis aggressive nonHodgkin's lymphoma: A dose-finding pilot study. J Clin Oncol 1995; 13: 2916-23. Gordon LI, Andersen J. Habermann TM et al Phase I trial of dose escalation with growth factor support in patients with previously untreated diffuse aggressive lymphomas: Determination of the maximum tolerated dose of ProMACE-CytaBOM J Clin Oncol 1996; 4: 1275-81. Aviles A, Delgado S. Nambo MJ et al. Adjuvant radiotherapy to sites of previous bulky disease in patients with stage IV diffuse large-cell lymphoma. Int J Radiat Oncol Biol Phys 1994; 30: 799803. Press OW. Fisher R, Dalla-Favera R. Aggressive Non-Hodgkin's Lymphomas. ASH Educational Program 1996. Siu LL, Tannock IF. Chemotherapy dose escalation: Case unproven. J Clin Oncol 1997: 15: 2765-8 (Editorial). Velasquez WS. Fuller LM. Jagannath S. Stages I and II diffuse large cell lymphomas: Prognostic factors and long term results with CHOP-Bleo and radiotherapy. Blood 1991. 77. 942-7. Hill ME, MacLennan KE. Cunningham DC et al. Prognostic significance of hcl-2 expression and bcl-2 major breakpoint region rearrandgement in diffuse large cell lymphoma. A British National Lymphoma Investigation study. Blood 1996. 88: 1046 51. Hermine O. Haiuon C, Lepage E et al. Prognostic significance of bcl-2 protein expression in aggressive non-Hodgkin's lymphoma. Groupe d'Etude des Lymphomes de l'Adulte (GELA). Blood 1996: 87: 265-72. Guglielmi C, Amadori S. Martelli M et al. The F-MACHOP sequential combination chemotherapy regimen in advanced diffuse aggressive lymphomas: Long-term results. Ann Oncol 1991; 2: 365 71.