BCNU) and Etoposide (VP16-213) (CBV) and CBV Plus Cisplatin (CBVP) Followed by Autologous Stem Cell Transplantation in Patients with Hodgkin’S Disease

Home , Carmustine, CBV (chemotherapy)

Bone Marrow Transplantation, (1999) 23, 1131–1138  1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Regimen-related toxicity and non-relapse mortality with high-dose cyclophosphamide, carmustine (BCNU) and etoposide (VP16-213) (CBV) and CBV plus cisplatin (CBVP) followed by autologous stem cell transplantation in patients with Hodgkin’s disease

DE Reece1, TJ Nevill1, A Sayegh1, JJ Spinelli1, DA Brockington1, MJ Barnett1, H-G Klingemann1, JM Connors2, SH Nantel1, JD Shepherd1, HJ Sutherland1, NJS Voss3, RN Fairey3, SE O’Reilly2 and GL Phillips1

1Leukemia/Bone Marrow Transplantation Program of British Columbia: Division of Hematology, Vancouver General Hospital, British Columbia Cancer Agency and the University of British Columbia; 2Division of Medical Oncology, British Columbia Cancer Agency; and 3Division of Radiation Oncology, British Columbia Cancer Agency and the University of British Columbia, Vancouver, BC, Canada

Summary: High-dose chemotherapy (HDCT) followed by autologous hematopoietic stem cell transplantation (ASCT) produces This analysis compares the regimen-related toxicity long-term progression-free survival (PFS) in up to 50% of (RRT) and overall non-relapse mortality (NRM) in patients with Hodgkin’s disease that has progressed after Hodgkin’s disease patients conditioned with either CBV primary chemotherapy.1–18 Since the prior chest radio- (cyclophosphamide, BCNU (carmustine), and VP16–213 therapy that many patients have received predisposes to (etoposide)) (26 patients) or CBVP (CBV + cisplatin) (68 interstitial pneumonitis when total body irradiation is patients) followed by autologous stem cell transplan- used,3,19 most centers have used conditioning regimens con- tation (ASCT). CBVP included a continuous infusion taining only chemotherapy. One of the most commonly rather than intermittent doses of etoposide, a lower used conditioning regimens is CBV (cyclophosphamide, BCNU dose and the addition of cisplatin. RRT and BCNU (carmustine) and VP16–213 (etoposide)). The term NRM were determined for each regimen and compared; CBV does not describe a single regimen, however, since risk factors for each were examined by multivariate the doses and schedules of the agents utilized have varied analysis. Grade IV (fatal) RRT occurred in five patients among transplant centers.1,2,9,12,14–16,18 (pulmonary in two, cardiac in two, and central nervous Since 1985, we have utilized CBV regimens at doses system in one). Eighteen patients experienced grade II– higher than those originally described1 and currently used III pulmonary RRT, consistent with BCNU damage in by some transplant centers9,14–16 in an attempt to achieve 15. Prior nitrosourea exposure was the main risk factor greater tumor cytoreduction.5, 12, 13 Patients were entered on for pulmonary RRT. Grade II mucosal and hepatic to transplant protocols at the first evidence of disease pro- RRT occurred less often after CBVP vs CBV (P = 0.031 gression, if possible, since we postulated that such higher and 0.0003, respectively). In addition, three other early dose CBV regimens would be better tolerated in patients and eight late non-relapse deaths were seen. Median fol- with less prior conventional therapy. These protocols used low-up of the entire group is 5.1 (range 2.8–10.2) years. specific brief cytoreductive reductive measures, without The probability of overall NRM was 26% (95% confi- further restaging, prior to HDCT and ASCT as part of the dence interval (CI) 13–50%) with CBV vs 23% (95% overall transplant program. CI 12–41%) with CBVP (P = 0.40). The progression-free Between March 1985 and May 1992, 94 patients with survival and relapse rates were similar. Although the Hodgkin’s disease were treated with one of two condition- rates of fatal RRT, pulmonary RRT and overall NRM ing regimens – CBV or CBV with the addition of cisplatin were similar with CBV or CBVP, CBVP produced less (CBVP), followed by ASCT.12, 13 This analysis will focus mucosal and liver RRT with a comparable antitumor on the incidence of RRT and overall NRM among the 94 effect. As many autografted patients are cured, future patients who received these conditioning regimens. efforts should include measures to decrease NRM. Keywords: Hodgkin’s disease; toxicity; non-relapse mortality; autotransplantation Patients and methods

Eligibility criteria Eligibility criteria included the following: (1) progressive Correspondence: Dr DE Reece, University of Kentucky Blood and Mar- row Transplant Program, Markey Cancer Center, 800 Rose Street, Lexing- Hodgkin’s disease after multi-agent chemotherapy given ton, KY 40536, USA with curative intent; (2) presumed low probability of cure Received 15 September 1998; accepted 4 January 1999 with further conventional therapy, specifically excluding CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1132 patients with localized recurrences potentially curable with (range 1–6) prior regimens. As shown in Table 2, the radiotherapy; (3) age Ͻ61 years; (4) no evidence of marrow majority of patients then received specific conventional involvement with Hodgkin’s disease at the time of marrow cytoreductive therapy before HDCT, which included two harvest, or adequate peripheral blood stem cell (PBSC) col- cycles of mechlorethamine, vinblastine, procarbazine and lection; (5) absence of severe co-morbid illness; (6) infor- prednisone (MVPP),21 involved-field radiation therapy (IF- med consent. These patients were treated on protocols RT), or both modalities as previously described.12,13 There- approved by the Institutional Review Boards at the British fore, the median number of prior regimens at the time of Columbia Cancer Agency, University of British Columbia ASCT was two (range 2–7). Patients were not formally and Vancouver Hospital and Health Sciences Center. restaged after conventional cytoreductive therapy, so that chemosensitivity status was not determined. All patients entered on to protocol proceeded to ASCT. Exclusions Two patients with active bone marrow involvement Five patients did not receive CBV or CBVP per protocol received PBSC transplants using cells collected after con- during the study period. The reasons for their exclusion ventional chemotherapy while eight patients who had failed included protocol violation (one patient), interruption of induction chemotherapy or who had active marrow involve- CBV conditioning due to problems with marrow viability ment at the time of disease progression received high-dose 2 (one patient), brief participation in another multi-insti- cyclophosphamide (7 g/m ) (HD-CY) over 4 days, with or tutional study20 (two patients) and omission of BCNU due without GM-CSF, as priming therapy for PBSC collection. to severe pre-existing pulmonary dysfunction (one patient). Transplant features and conditioning regimens Patient characteristics Drug doses were calculated on the basis of the lower value of ideal or actual body weight. The first 26 patients received The characteristics of the 94 patients entered on to the CBV, consisting of cyclophosphamide 1.8 g/m2 intra- transplant protocol are listed in Table 1. Seventy-six venously (i.v.) over 2 h daily on days −7to−4; VP16–213 patients had received combinations of MOPP/ABV(D) as 0.4 g/m2 i.v. over 1 h every 12 h on days −7, −6 and −5; initial chemotherapy, and all but 10 had received MOPP and BCNU 0.6 g/m2 i.v. over 2 h on day −3 (Figure 1). and ABV(D) at some time. Thirty-five patients had received Beginning in January 1988, this regimen was modified, in prior radiation therapy, which included the chest in 34 an attempt to decrease mucosal and lung toxicity while pre- patients. Most patients were entered on to protocol at the serving the antitumor effect. Sixty-eight patients received time of first relapse and had received a median of one the new CBVP regimen in which the same dose of cyclophosphamide was administered on days −6to−3. The total dose of VP16–213 remained the same but the drug was given as a continuous 34-h i.v. infusion beginning on day Table 1 Patient characteristics −7 as previously described.20 Cisplatin was added in conventional doses of 50 mg/m2 i.v. on days −7, −6 and −5 CBV CBVP P value 2 = = while the dose of BCNU was reduced to 0.5 g/m (Figure n 26 n 68 1). Cryopreserved stem cells were thawed and reinfused on day 0, except when volume considerations necessitated Age, median (range) 28 29 0.013 Sex, male 16 39 0.71 division of reinfusion on days 0 and 1. Initial stage The transplant-related characteristics of all 94 patients I or II 11 27 0.77 III 11 26 IV 4 15 B 14 18 0.032 Table 2 Transplant-related features Prior therapy Median number drugs 7 7 0.78 Median number regimens (range) 1 (1–3) 1 (1–6) 0.72 CBV CBVP P value = = Prior nitrosoureas 3 4 0.35 n 26 n 68 Prior radiotherapy 11 24 0.31 Prior chest radiotherapy 12 22 0.099 Protocol cytoreduction therapy before 0.33 History of bleomycin lung toxicity 1 4 0.69 conditioning Disease status at protocol entry 0.022 MVPP 6 28 1st untested relapse 13 43 IF-RT 9 4 у2nd untested relapse 3 9 MVPP + IF-RT 7 22 2nd complete remission 0 1 High-dose cyclophosphamide 0 8 Resistant relapse 4 1 None 4 6 Induction failure 6 14 Hematopoietic stem cell source 0.31 Karnofsky performance status 0.24 Bone marrow 25 58 90–100% 13 46 Peripheral blood 1 9 р80% 13 22 Both 0 1 B symptoms at ASCT 7 22 0.61 Post-transplant growth factors 0 26 0.0002

Characteristics at the time of protocol entry of 94 patients receiving CBV Transplant-related characteristics of 94 patients conditioned with CBV or or CBVP and ASCT for Hodgkin’s disease. CBVP before ASCT. CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1133 CBV Conditioning regimen (n = 26) method was also utilized to evaluate differences in RRT between the two conditioning regimens in each of the eight Day organ systems assessed. Logistic regression was used to Agent Dose Route –7 –6 –5 –4 –3 –2 –1 0 determine prognostic factors for hepatic and mucosal RRT. Cyclophosphamide 1.8 g/m2 i.v. •••• B The Cox proportional hazards model was used in assessing the relationship of variables to pulmonary RRT and NRM. 2 •• •• •• VP16-213 0.4 g/m i.v. M In addition to conditioning regimen, variables examined BCNU 0.6 g/m2 i.v. • T included age, sex, disease status, initial stage, initial chemotherapy, prior number of drugs and regimens, extranodal disease at diagnosis and protocol entry, ‘B’ symptoms and Karnofsky performance status at diagnosis and protocol CBVP Conditioning regimen (n = 68) entry, prior radiotherapy, relapse in the irradiated ﬁeld, type Day of pre-ASCT conventional cytoreduction, type of stem cell Agent Dose Route –7 –6 –5 –4 –3 –2 –1 0 support, and the use of post-ASCT growth factors. The

VP16-213 2.4 g/m2 CIVI B probability of pulmonary RRT, NRM, relapse and PFS were calculated from the day of transplantation (day 0) Cisplatin 50 mg/m2 i.v. •• • M using the product-limit method of Kaplan and Meier,24 and 95% conﬁdence intervals (CI) were based on the log (log) Cyclophosphamide 1.8 g/m2 i.v. •••• T transformation of the survival estimate. BCNU 0.5 g/m2 i.v. •

Figure 1 Dose and schedule of agents in CBV and CBVP. CIVI = continuous i.v. infusion. Results

Comparison of patient characteristics are summarized in Table 2. Post-transplant hematopoietic growth factors were given to 26 patients treated after 1990. The comparative characteristics of patients receiving CBV or CBVP and shown in Tables 1 and 2. The only significant differences were that more CBVP patients were Ͼ40 years Supportive care of age (P = 0.013), had ‘B’ symptoms at diagnosis = Patients were hospitalized in single rooms with high- (P 0.032), were given hematopoietic growth factors post- efficiency particulate air (HEPA) filtration. An indwelling BMT (P = 0.0002), and were less frequently in a chemo- central venous silastic catheter was placed in all patients. resistant relapse at the time of protocol entry (P = 0.022). Blood products were irradiated to у1500 cGy and were у given to maintain a hemoglobin level 90 g/l and platelet RRT grading count у20 × 109/l. Empiric broad-spectrum antibiotics, anti-fungal therapy and total parenteral hyperalimentation The RRT grade by organ system and regimen is summar- were used as indicated. Uroprophylaxis consisted of ized in Table 3. The incidence of life-threatening or fatal hyperhydration. toxicity (grades III or IV) did not differ by conditioning regimen. The only differences observed were in the incidence of grade II mucosal and liver toxicity as outlined Definitions below. Pulmonary RRT, characterized by several distinc- Regimen-related toxicity (RRT) due to the CBV or CBVP tive clinical features, was the most common potentially conditioning regimen was graded using the criteria of Bear- serious toxicity seen with both regimens. man et al,22 except that the evaluation of pulmonary toxicity was extended to day 150 post-transplant to encompass Fatal RRT: Grade IV RRT was seen in five patients (Table the development of delayed toxicity, similar to our previous 3). Two recipients of CBV died of RRT, one on day 18 description.23 In general, grade I RRT was minor and due to CNS toxicity in the form of unexplained cerebral reversible without specific therapy; grade II toxicity edema and the other on day 133 due to progressive inter- required medical intervention; grade III toxicity was life- stitial lung disease. Three patients given CBVP died of threatening; and grade IV toxicity was fatal. Non-relapse either cardiac failure (two patients on days 0 and 3) or pul- mortality (NRM) referred to death from any cause other monary hemorrhage (one patient on day 16). None of these than Hodgkin’s disease, including RRT, and was calculated patients had received HD-CY as priming therapy. from the date of transplant. PFS was calculated from the date of transplant until the date of progression, death from Life-threatening RRT: As shown in Table 3, grade III tox- non-relapse causes, or if progression-free, to the date of last icity was relatively uncommon during the early post-trans- follow-up. plant period. Two CBVP patients had grade III renal RRT (dialysis-dependent renal failure) and died of other toxicity (cardiac RRT in one and pulmonary hemorrhage in the Statistical analysis other). Four patients had grade III pulmonary RRT (three Characteristics of the patients receiving CBV or CBVP after CBV and one after CBVP). Among these cases, three were compared using the Pearson chi-square test. This patients initially had reversible interstitial changes CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1134 Table 3 Regimen-related toxicity

Organ system Grade P value

0 I II III IV

Heart CBV 22 (84) 0 (0) 3 (12) 1 (4) 0 (0) NS CBVP 60 (89) 3 (4) 3 (4) 0 (0) 2 (3) Bladder CBV 23 (88) 3 (12) 0 (0) 0 (0) 0 (0) NS CBVP 59 (87) 4 (6) 5 (7) 0 (0) 0 (0) Kidneys CBV 24 (92) 1 (4) 1 (4) 0 (0) 0 (0) NS CBVP 62 (91) 1 (2) 3 (4) 2 (3) 0 (0) Lungs CBV 20 (76) 0 (0) 2 (8) 3 (12) 1 (4) NS CBVP 54 (79) 0 (0) 12 (17) 1 (2) 1 (2) Liver CBV 8 (30) 9 (35) 9 (35) 0 (0) 0 (0) 0.0003 CBVP 49 (72) 15 (22) 4 (6) 0 (0) 0 (0) CNS CBV 25 (96) 0 (0) 0 (0) 0 (0) 1 (4) NS CBVP 68 (100) 0 (0) 0 (0) 0 (0) 0 (0) Mucosa CBV 1 (4) 11 (42) 14 (54) 0 (0) 0 (0) 0.031 CBVP 21 (31) 29 (43) 18 (26) 0 (0) 0 (0) Gut CBV 23 (88) 3 (12) 0 (0) 0 (0) 0 (0) NS CBVP 45 (66) 22 (30) 1 (4) 0 (0) 0 (0)

Grade of toxicity by organ system in patients receiving conditioning with CBV (n = 26) or CBVP (n = 68) before ASCT. The number of affected patients is listed for each category, with the percentage in parenthesis.

described more fully below, while the fourth, who suc- transplant and was grade II in two patients and grade IV cumbed to CNS toxicity as described above, required intu- in one patient. bation for lung toxicity as well as for obtudation. Of note, Seventeen patients manifested a different clinical pattern grade уIII hepatic toxicity was not seen in this analysis. of pulmonary toxicity, consisting of a syndrome of dyspnea, associated with restrictive changes on pulmonary RRT by organ system function testing and/or radiographic infiltrates, which developed a median of 57 (range 32–150) days after ASCT. Mucosal RRT: The incidence of Grade II mucositis was This relatively delayed form of pulmonary RRT was grade 54% in patients conditioned with CBV compared to 26% II in 14 patients, grade III in two patients and grade IV in in patients receiving CBVP (P = 0.031) (Table 3). Multi- one patient. All patients with grades II and III toxicity had variate analysis confirmed the importance of conditioning initial improvement in their pulmonary signs and symp- regimen in this difference (P = 0.03) and also identified toms. The pattern of toxicity was consistent with previously both prior chest radiation therapy (P = 0.0001) and stage described BCNU lung damage.26, 27 Four of these patients III or IV disease at diagnosis (P = 0.007) as additional had been conditioned with CBV, while 13 had received independent risk factors. CBVP. The incidence of grades II–IV pulmonary RRT among Hepatic RRT: Although hepatic toxicity was not life- all patients was 21% and did not differ significantly threatening in any patient, 14% of all patients experienced between the two conditioning regimens (P = 0.98). How- grade II hepatotoxicity; only two patients (2%) met the ever, multivariate analysis identified prior nitrosourea more stringent definition (‘Jones’ criteria) for hepatic veno- exposure as the strongest risk factor for pulmonary RRT occlusive disease.25 Rather, most patients with hepatotoxic- (P = 0.001); extranodal disease at diagnosis was identified ity had reversible elevations of transaminases with or with- as another predictive factor (P = 0.028). There was no sig- out weight gain. Thirty-five percent of the patients treated nificant relationship between pulmonary RRT and prior ble- with CBV developed grade II hepatic RRT compared to omycin lung toxicity or chest radiotherapy given either only 6% of patients given CBVP (P = 0.003) (Table 3). prior to protocol entry or as part of pre-conditioning cyto- Multivariate analysis indicated that the use of the CBV reductive therapy. The incidence of pulmonary RRT was regimen was the only significant risk factor for grade II 20.5% in patients who had never received chest radio- hepatic toxicity. therapy compared with 17.6% in those who had received chest radiotherapy in the past and 27.3% in those who had Pulmonary RRT: According to the Bearman criteria, grade been given IF-RT to the chest just before ASCT (P = 0.68). II pulmonary RRT is defined as chest radiograph changes Ͼ Ͼ or 10% decrease in p02 not requiring 50% supplemental Management of pulmonary RRT 02 or mechanical ventilation, while grade III RRT consists of interstitial changes requiring mechanical ventilator sup- All the patients experiencing early pulmonary hemorrhage port or Ͼ50% oxygen by mask. Three patients developed or ARDS received steroids; one recovered, one died of early pulmonary hemorrhage and/or ARDS of uncertain eti- CNS RRT and one succumbed to pulmonary failure. The ology. This occurred a median of 8 (range 8–15) days post- management of the 17 patients with BCNU lung toxicity CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1135 became more aggressive during the 7-year study period. In Discussion our hands, the optimal therapy for these patients involved the prompt institution of prednisone at a dose of 2 mg/kg In our experience, approximately 45% of all patients who per day at the first sign of dyspnea, after an infectious etiol- have undergone ASCT for recurrent or relapsed Hodgkin’s ogy had been excluded. Steroids were then tapered over 6– disease after chemotherapy remain free of disease pro- 8 weeks. None of the initial four patients who developed gression. These patients were conditioned for ASCT with this toxicity after CBV received corticosteroids early after one of two CBV-based regimens using doses of the chemo- onset of symptoms, although all were eventually treated. therapeutic agents at the higher end of the range in current All four died of progressive pulmonary damage – one early clinical practice. The two regimens differed in the adminis- and three late post-ASCT. None of the patients who tration schedule of etoposide, the dose of BCNU and the developed BCNU lung toxicity after CBVP died of pro- use of cisplatin. gressive lung disease. Twelve of these 13 received immedi- Our analysis did not disclose any significant differences ate corticosteroids, with rapid resolution of pulmonary between the two regimens with regard to severe or fatal abnormalities. The other, who had minimal grade II find- RRT or overall NRM. However, grade II mucosal and hep- ings, recovered without specific therapy. Due to the atic RRT occurred less frequently in patients treated with improved management of pulmonary RRT over time, the the CBVP regimen. Multivariate analysis confirmed the relationship between conditioning regimen and mortality importance of the conditioning protocol in these differ- due to progressive pulmonary fibrosis after RRT could not ences. Patient tolerance of conditioning regimens, parti- be assessed accurately. cularly with respect to mucositis, has become critical with the need to reduce resource utilization and the evolution of outpatient ASCT programs.28 Although it is unclear which NRM of the changes improved the RRT profile, the administration of etoposide by continuous intravenous infusion was Three other early non-relapse deaths occurred before day likely to have been the most important factor in reducing 150 due to causes not directly classifiable as RRT – one mucositis.20 In addition, two other factors contributing to due to infection (day 7), one due to pulmonary hypertension mucosal toxicity were identified. Subclinical mucosal dam- (day 62), and one due to marantic endocarditis (day 72). age related to prior exposure to chest radiation, which com- Including the five RRT-related deaths, the overall early monly includes the cervical area in Hodgkin’s disease NRM was 8% – three of 26 patients conditioned with CBV patients as part of the mantle field, might explain the predis- (11%) and five of 68 receiving CBVP (7%). Late NRM position for mucosal toxicity in these patients. However, occurred due to pulmonary fibrosis in three patients (days the increased risk of mucositis in patients previously 152, 723, 739), pneumonia in one (day 261), a motor exposed to nitrosoureas is not readily explainable. vehicle accident in one (day 1584) and second malignancies The absence of severe hepatotoxicity after these two in three, specifically brain tumor, gastrointestinal cancer higher-dose CBV-based regimens is notable. The regimen and acute myelogenous leukemia (AML) on days 1125, from Seattle reported by Weaver et al29 used drug doses 1688 and 1783, respectively. Median follow-up of all similar to those in our CBV regimens and was associated patients was 5.1 years – 7.7 (range 4.7–10.2) years for with a high risk of grade III or IV hepatic toxicity (16%). patients conditioned with CBV and 4.7 (range 2.8–7.0) Although the order of drug administration in the condition- years in those receiving CBVP. The probability of NRM ing regimens differed, differences in patient selection prob- from all causes was 26% (95% CI 13–50%) with CBV vs ably accounted for this difference. In the Seattle experience, = 23% (95% CI 12–41%) with CBVP (P 0.397). which also included patients with other lymphoid malig- Multivariate analysis of overall NRM identified three nancies, all had received prior liver or mediastinal radiation significant risk factors – prior bleomycin lung toxicity or had a history of radiation toxicity. = = (P 0.003), prior BCNU exposure (P 0.005) and prior Pulmonary RRT was the main non-hematologic toxicity chest radiotherapy at any time before transplantation observed in our patients. Although several patients = (P 0.039). developed pulmonary hemorrhage or ARDS early post- ASCT as previously described,30 pulmonary toxicity usu- ally occurred several weeks later and was consistent with Relapse BCNU lung damage. Worrisome features of this toxicity Nine relapses have occurred among patients conditioned included its insidious onset in most patients, as well as the with CBV and 17 among those receiving CBVP. The observation that patients with grade II or III RRT who were relapse probabilities were 43% (95% CI 25–66%) and 36% not treated with steroids could initially improve but later (95% CI 25–51%), respectively (P = 0.62). develop progressive fatal pulmonary fibrosis. In contrast to the experience of others, neither prior chest radiation therapy nor prior bleomycin lung toxicity were statistically PFS associated with a significant increased risk of pulmonary RRT.29,31 It is not clear why the presence of extranodal The PFS for patients receiving CBV was 42% (95% CI disease at diagnosis was a risk factor for this toxicity. On 23–60%) compared with 50% (95% CI 38–61%) in those the other hand, the identification of prior nitrosourea receiving CBVP. These differences were not significantly exposure as a risk factor for pulmonary RRT provides different (P = 0.43). potentially useful clinical information. CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1136 The relationship of prior nitrosourea therapy and acute lower-dose CBV regimens have reported very low early lung toxicity post-ASCT is not well understood. A BCNU NRM rates, on the order of 0–5%.15–17 Better supportive dose Ͼ1050 mg/m2 given with marrow support is associa- care measures, such as the use of hematopoietic growth ted with a high incidence of pulmonary toxicity.26 Jones et factors and mobilized blood progenitor cells for autograft- al27 have described a relationship between the pharmaco- ing, as well as institutional experience, probably account dynamics of high-dose BCNU (600 mg/m2), which vary for this trend.37 Most of our patients were transplanted among patients, and acute lung damage and have postulated before 1990 and newer supportive care measures were that exposure to potentially toxic metabolites of BCNU unavailable. Nevertheless, early NRM decreased from 11% might be responsible for this complication. Also, high total with the initial patients receiving CBV to 7% in the cumulative doses of BCNU given over a relatively long subsequent patients receiving CBVP. period of time can result in late fatal pulmonary fibrosis in Our study does not address the important question of the patients with other malignancies.32 Perhaps prior conven- comparative antitumor effect of higher and lower CBV tional-dose BCNU can produce a cumulative adverse effect dose ranges, particularly in view of the somewhat greater on the lung, analogous to the effect of anthracyclines on early NRM seen with the higher doses. Specifically, there the heart, leading to susceptibility to pulmonary injury after are no data available to judge definitively whether such CBV or CBVP. higher doses produce lower relapse rates and better PFS and A reduction in the dose of BCNU in the conditioning overall survival rates in patients with similar characteristics, regimen from 600 mg/m2 to 500 mg/m2 did not, as we had such as extent of prior therapy and disease status. hoped, reduce the incidence of pulmonary toxicity. The Finally, late NRM has been increasingly recognized as finding that patients receiving high-dose BCNU may exhi- transplanted patients have been followed for longer periods bit variable pharmacodynamics27 may explain the lack of of time. In the current series, late NRM occurred in eight benefit of our dose reduction. On the other hand, compari- patients and was related to infection, pulmonary fibrosis son of CBV and CBVP toxicity is potentially constrained and secondary cancers. Other studies of ASCT in Hodg- by our sample size. The statistical power of the comparison kin’s disease studies have reported similar causes of late estimated that there was a 50% chance of detecting a 20% fatal complications. The raw incidence of late fatal infec- difference of toxicity, between the CBV and CBVP regi- tions has ranged from 0.5–3% in other series; bacterial, men. Therefore, clinically important differences in pulmon- viral and fungal pathogens have all been impli- ary toxicity could have been missed. However, the work of cated.7,9,11,14,16,17,38 The relative contributions of the known other investigators suggests that the risk of BCNU-related T cell defects in this disease39 and of post-transplant lung toxicity increases when the BCNU dose in the con- immunodeficiency40 are unclear. Many studies also report ditioning regimen exceeds 450 mg/m2,26,33,34 and it is likely the occurrence of late fatal pulmonary toxicity after ASCT that both of our conditioning regimens were over the ‘thres- in a small proportion of patients.9,11,14,17 Several factors are hold’ dose of this agent. At any rate, we feel that CBV probably involved, although in our series, this complication regimens using у500 mg/m2 doses of BCNU must be given occurred as a delayed manifestation of BCNU toxicity early with a high level of awareness of the risk of pulmonary in our experience. However, the most worrisome late com- toxicity and a low threshold for early institution of cortico- plication after ASCT has been the development of second- steroid therapy if pulmonary symptoms develop. These ary cancers. regimens should be given with great caution in patients Risk factors for the development of secondary malig- previously exposed to nitrosoureas. nancies after ASCT are not yet well-defined. The cumulat- Although prior chest radiotherapy and prior bleomycin ive incidence of secondary AML or myelodysplasic syn- lung toxicity were not predictive for the development of drome (MDS) post-ASCT has ranged from 4–14% in pulmonary RRT post-transplant, these two factors, in several studies of Hodgkin’s disease patients.41–44 One addition to prior nitrosourea exposure, were identified as recent study has evaluated patients who developed post- risk factors for overall NRM. All these factors are related ASCT MDS characterized by a cytogenetic abnormality for to the extent of prior cytotoxic therapy. The number of prior which a florescence in situ hybridization (FISH) probe was regimens did not correlate with the risk of NRM probably available. The FISH probe was then used to analyze pre- because the majority of these patients were entered into ASCT specimens of bone marrow or PBSCs. Of interest, study after only one full initial regimen and, before HDCT the majority of cases had the same cytogenetic abnormality and ASCT, an additional two cycles of brief cytoreduction. detectable by FISH before transplant.45 Thus, the trans- Other analyses of Hodgkin’s disease patients also suggest forming event was likely related to bone marrow damage that NRM is lower in patients transplanted earlier in their from prior conventional cytotoxic therapy, although the disease course35,36 and we feel that higher dose CBV transplant procedure itself may also have played a contribu- regimens are associated with less NRM when used in tory role.45 In the recently reported French registry analysis, minimally pretreated patients. the incidence of MDS or AML in patients with Hodgkin’s The overall early NRM observed in this series was 8%, disease was similar after ASCT or conventional chemo- a rate in the range of that reported by other centers using therapy, a finding in keeping with the importance of prior a number of different conditioning regimens. However, in exposure to cytotoxic agents. However, among their ASCT general, the risk of early NRM in Hodgkin’s disease patients, prior splenectomy and the use of PBSC rather than patients has a trend downwards from 10–20% in some of bone marrow as the stem cell source were associated with the earlier studies4,5,7,8,10 to 5–10%, or even lower, in more a higher risk.44 Other studies have also described a relation- recent studies.12,14,18 In particular, several series using ship between PBSC grafts and the development of MDS or CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1137 AML.42,43,46 Further studies are underway to evaluate this 5 Reece DE, Barnett MJ, Connors JM et al. Intensive chemo- possible association, as well as to define other factors that therapy with cyclophosphamide, carmustine, and etoposide might contribute to the development of MDS or AML post- followed by autologous bone marrow transplantation for ASCT. Our single patient with AML after ASCT had relapsed Hodgkin’s disease. J Clin Oncol 1991; 9: 1871–1879. received a bone marrow graft; his course was notable for 6 Gianni AM, Siena S, Bregni M et al. Prolonged disease-free survival after high-dose sequential chemo-radiotherapy and protracted alkylator therapy given in an attempt to prolong haemopoietic autologous transplantation in poor prognosis a second remission before ASCT became available in our Hodgkin’s disease. Ann Oncol 1991; 2: 645–653. institution. 7 Chopra R, McMillan AK, Linch DC et al. The place of high- Less is known about the risk of secondary solid tumors dose BEAM therapy and autologous bone marrow transplan- post-ASCT. In the French registry analysis, such tumors tation in poor-risk Hodgkin’s disease. A single-center 8-year were seen in 3.7% of patients with Hodgkin’s disease study of 155 patients. Blood 1993; 81: 1137–1145. treated with ASCT, and were significantly more common 8 Rapoport AP, Rowe JM, Kouides PA et al. One hundred auto- after ASCT compared with conventional therapy. Although transplants for relapsed or refractory Hodgkin’s disease and radiotherapy has been associated with the development of lymphoma: Value of pre-transplant disease status for pre- solid tumors after conventional therapy,47–49 no significant dicting outcome. J Clin Oncol 1993; 11: 2351–2361. 44 9 Bierman PJ, Bagin RG, Jagannath S et al. High-dose chemo- risk factors were identified among their ASCT recipients. therapy followed by autologous hematopoietic rescue in The low number of events in our series precluded detailed Hodgkin’s disease: long term follow-up in 128 patients. Ann analysis. Due to the long latency period of many of these Oncol 1993; 4: 767–773. tumors, ASCT patients will need to be observed for 10 Yahalom J, Gulati SC, Toia M et al. Accelerated hyperfrac- prolonged periods to assess the full impact of this tionated total-lymphoid irradiation, high-dose chemotherapy, complication. and autologous bone marrow transplantation for refractory and In summary, among our patients conditioned with relapsing patients with Hodgkin’s disease. J Clin Oncol 1993; higher-dose CBV regimens, less mucosal and hepatic tox- 11: 1062–1070. icities were observed in those receiving CBVP when com- 11 Crump M, Smith AM, Brandwein J et al. High-dose etoposide pared to CBV. There were no significant differences in the and melphalan, and autologous bone marrow transplantation for patients with advanced Hodgkin’s disease: importance of probabilities of NRM or overall PFS. The pulmonary tox- disease status at transplant. J Clin Oncol 1993; 11: 704–711. icity seen with these higher dose CBV regimens mandates 12 Reece DE, Connors JM, Spinelli JJ et al. Intensive therapy close communication between the patient, referring phys- with cyclophosphamide, carmustine, etoposide ± cisplatin, and ician and transplant center after the patient has returned autologous bone marrow transplantation for Hodgkin’s disease home in order to optimally treat this serious but potentially in first relapse after combination chemotherapy. Blood 1994; reversible complication. In addition, ASCT patients require 83: 1193–1199. prolonged monitoring to detect the onset of late compli- 13 Reece, DE, Barnett MJ, Shepherd JD et al. High-dose cyclo- cations. Strategies to reduce such problems are needed to phosphamide, carmustine (BCNU), and etoposide (VP16–213) ± optimize the results of ASCT in this disease. with or without cisplatin (CBV P) and autologous transplantation for patients with Hodgkin’s disease who fail to enter a complete remission after combination chemotherapy. Blood 1995; 86: 451–456. Acknowledgements 14 Nademanee A, O’Donnell MR, Snyder DS et al. High-dose chemotherapy with or without total body irradiation followed The authors wish to acknowledge the contribution of the BMT by autologous bone marrow and/or peripheral blood stem cell medical and nursing staff at the Vancouver Hospital and Health transplantation for patients with relapsed and refractory Hodg- Sciences Center and the British Columbia Cancer Agency, as well kin’s disease. Results in 85 patients with analysis of prognos- as the assistance of Trudy Winkle in the preparation of this tic factors. Blood 1995; 85: 1381–1390. manuscript. 15 Burns LJ, Daniels KA, McGlave PB et al. Autologous stem cell transplantation for refractory and relapsed Hodgkin’s disease: factors predictive of prolonged survival. Bone Marrow Transplant 1995; 16: 13–18. References 16 Horning SJ, Chao NJ, Negrin RS et al. High-dose therapy and autologous hematopoietic progenitor cell transplantation for 1 Jagannath S, Dicke KA, Armitage JO et al. High-dose cyclo- recurrent or refractory Hodgkin’s disease: Analysis of the phosphamide, carmustine, and etoposide and autologous bone Stanford University results and prognostic factors. Blood marrow transplantation for relapsed Hodgkin’s disease. Ann 1997; 89: 801–813. Intern Med 1986; 104: 163–168. 17 O’Brien ME, Milan S, Cunningham D et al. High-dose 2 Carella AM, Congiu AM, Gaozza E et al. High-dose chemo- chemotherapy and autologous bone marrow transplant in therapy with autologous bone marrow transplantation in 50 relapsed Hodgkin’s disease – a pragmatic prognostic index. advanced resistant Hodgkin’s disease patients: an Italian Study Br J Cancer 1996; 73: 1272–1277. Group report. J Clin Oncol 1986; 6: 1411–1416. 18 Wheeler C, Eickhoff C, Elias A et al. High-dose cyclophos- 3 Phillips GL, Wolff SN, Herzig RH et al. Treatment of pro- phamide, carmustine, and etoposide with autologous trans- gressive Hodgkin’s disease with intensive chemoradiotherapy plantation in Hodgkin’s disease: a prognostic model for treat- and autologous bone marrow transplantation. Blood 1989; 73: ment outcomes. Biol Blood Marrow Transplant 1997; 3: 2086–2092. 98–106. 4 Jones RJ, Piantadosi S, Mann RB et al. High-dose cytotoxic 19 Pecego R, Hill R, Appelbaum FR et al. Interstitial pneumonitis therapy and bone marrow transplantation for relapsed Hodg- following autologous bone marrow transplantation. Trans- kin’s disease. J Clin Oncol 1990; 8: 527–537. plantation 1986; 42: 515–517. CBV ± P and ASCT toxicity in Hodgkin’s disease DE Reece et al 1138 20 Harden E, Bolwell B, Fay J et al. Treatment of progressive 35 Reece D. Should high risk patients with Hodgkin’s disease be Hodgkin’s disease (HD) with cyclophosphamide (C), BCNU singled out for heavier therapeutic regimens while low risk (B), and continuous infusion etoposide (V): CBVi and auto- patients are spared such therapies? Leuk Lymphoma 1995; 15 logous marrow transplantation (AMT). Proc Am Soc Clin (Suppl. 1): 19–21. Oncol 1990; 9: 27 (Abstr.). 36 Schiffman K, Buckner CD, Maziarz R et al. High-dose busul- 21 Sutcliffe SB, Wrigley PF, Peto J et al. MVPP chemotherapy fan, melphalan, and thiotepa followed by autologous periph- regimen for advanced Hodgkin’s disease. Br Med J 1978; 1: eral blood stem cell transplantation in patients with aggressive 679–683. lymphoma or relapsed Hodgkin’s disease. Biol Blood Marrow 22 Bearman SI, Appelbaum FR, Back A et al. Regimen-related Transplant 1997; 3: 261–266. toxicity and early post-transplant survival in patients undergo- 37 Bennett CL, Armitage JL, Armitage JO et al. Costs of care ing marrow transplantation for lymphoma. J Clin Oncol 1989; and outcomes for high-dose therapy and autologous transplan- 7: 1288–1294. tation for lymphoid malignancies: results from the University 23 Nevill TJ, Barnett MJ, Klingemann HG et al. Regimen-related of Nebraska 1987 to 1991. J Clin Oncol 1995; 13: 969–973. toxicity of a busulfan-cyclophosphamide conditioning regimen 38 Sweetenham JW, Taghipour G, Milligan D et al. High-dose in 70 patients undergoing allogeneic bone marrow transplan- therapy and autologous stem cell rescue for patients with tation. J Clin Oncol 1991; 9: 1224–1232. Hodgkin’s disease in first relapse: results from the EBMT. 24 Kaplan EC, Meier P. Nonparametric estimation from incom- Bone Marrow Transplant 1997; 20: 745–752. plete observations. J Am Stat Assoc 1958; 53: 457–481. 39 Schulof RS, Bockman RS, Garofalo JA et al. Multivariate 25 Jones RJ, Lee KSK, Beschorner WE et al. Veno-occlusive analysis of T-cell functional defects and circulating serum fac- disease of the liver following bone marrow transplantation. tors in Hodgkin’s disease. Cancer 1981; 48: 964–973. Transplantation 1987; 44: 778–783. 40 Guillaume T, Rubinstein DB, Symann M. Immune reconsti- 26 Phillips GL, Fay JW, Herzig GP et al. Intensive 1,3-bis (2- tution and immunotherapy after autologous hematopoietic chloroethyl)-1-nitrosourea (BCNU), NSC # 4366650 and cry- stem cell transplantation. Blood 1998; 92: 1471–1490. opreserved autologous marrow transplantation for refractory 41 Darrington DL, Vose JM, Anderson JR et al. Incidence and cancer: a phase I-II study. Cancer 1983; 52: 1792–1802. characterization of secondary myelodysplastic syndrome and 27 Jones RB, Matthes S, Shpall EJ et al. Acute lung injury fol- acute myelogenous leukemia following high-dose chemoradi- lowing treatment with high-dose cyclophosphamide, cisplatin, otherapy and autologous stem-cell transplantation for lymph- and carmustine: pharmacodynamic evaluation of carmustine. oid malignancies. J Clin Oncol 1994; 12: 2527–2534. 42 Miller JS, Arthur DC, Litz CE et al. Myelodysplastic syn- J Natl Cancer Inst 1993; 85: 640–647. drome after autologous bone marrow transplantation: an 28 Meisenberg BR, Miller WE, McMillan R et al. Outpatient additional late complication of curative cancer therapy. Blood high-dose chemotherapy with autologous stem-cell rescue for 1994; 83: 3780–3786. hematologic and nonhematologic malignancies. J Clin Oncol 43 Traweek ST, Slovak ML, Nademanee AP et al. Clonal karyo- 1997; 15:11–17. typic hematopoietic cell abnormalities occurring after autolog- 29 Weaver CH, Appelbaum FR, Petersen FB et al. High-dose ous bone marrow transplantation for Hodgkin’s disease and cyclophosphamide, carmustine, and etoposide followed by non-Hodgkin’s lymphoma. Blood 1994; 84: 957–963. autologous bone marrow transplantation in patients with 44 Andre M, Henry-Amar M, Blaise D et al. Treatment-related lymphoid malignancies who have received dose-limiting radi- deaths and second cancer risk after autologous stem cell trans- ation therapy. J Clin Oncol 1993; 11: 1329–1335. plantation for Hodgkin’s disease. Blood 1998; 92: 1933–1940. 30 Chao NJ, Duncan SR, Long GD et al. Corticosteroid therapy 45 Radford JE Jr, Abruzzese EA, Pettenati MJ et al. Presence in for diffuse alveolar hemorrhage in bone marrow transplant pre-transplant marrow of abnormal progenitor cell clones recipients. Ann Intern Med 1991; 114: 145–146. which give rise to post-transplant myelodysplasia. Impli- 31 Wheeler C, Antin JH, Churchill WH et al. Cyclophosphamide, cations for high-dose chemotherapy with autologous hemato- carmustine, and etoposide with autologous bone marrow trans- poietic stem cell transplantation. In: Keating A, Dicke KA plantation in refractory Hodgkin’s disease and non-Hodgkin’s (eds). Proceedings Ninth International Symposium on Auto- lymphoma: a dose-finding study. J Clin Oncol 1990; 8: logous Blood and Marrow Transplantation, Dallas, TX, July 648–656. 1998. 32 O’Driscoll BR, Hasleton PS, Taylor PM et al. Active lung 46 Bhatia S, Ramsay NK, Steinbuch M et al. Malignant neo- fibrosis up to 17 years after chemotherapy with carmustine plasms following bone marrow transplantation. Blood 1996; (BCNU) in childhood. New Engl J Med 1990; 323: 378–382. 87: 3633–3639. 33 Ager S, Mahendra P, Richards EM et al. High-dose carmus- 47 Hoppe RT. Hodgkin’s disease: complications of therapy and tine, etoposide and melphalan (BEM) with autologous stem excess mortality. Ann Oncol 1997; 8 (Suppl. 1): 5115–5118. cell transplantation: a dose toxicity study. Bone Marrow 48 De Vita VT Jr. Late sequelae of treatment of Hodgkin’s dis- Transplant 1996; 17: 335–340. ease. Curr Opin Oncol 1997; 9: 428–431. 34 Rubio C, Hill ME, Milan S et al. Idiopathic pneumonia syn- 49 van Leeuwen FE, Klokman WJ, Hagenbeek A et al. Second drome after high-dose chemotherapy for relapsed Hodgkin’s cancer risk following Hodgkin’s disease: a 20-year follow-up disease. Br J Cancer 1997; 75: 1044–1048. study. J Clin Oncol 1994; 12: 312–325.