Bone Marrow Transplantation (2000) 25, 309–313  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt Pulmonary toxicity following -based preparative regimens and autologous peripheral blood progenitor cell transplantation in hematological malignancies

EP Alessandrino1, P Bernasconi1, A Colombo1, D Caldera1, G Martinelli1, P Vitulo2, L Malcovati1, C Nascimbene2, M Varettoni1, E Volpini2, C Klersy3 and C Bernasconi1

1Centro Trapianti di Midollo Osseo, Istituto di Ematologia, 2Divisione di Pneumologia, 3Biometry-Scientific Direction IRCCS, Policlinico S Matteo, Pavia, Italy

Summary: the use of BCNU is suspected to be caused by damage to the glutathione system.2 BCNU has been included in high- Sixty-five patients with hematological malignancies (25 dose conditioning regimens for gliomas, breast cancer, multiple myeloma, 18 Hodgkin’s disease, 22 non-Hodg- Hodgkin’s disease, non-Hodgkin’s lymphomas, multiple kin’s lymphomas) who received a carmustine-based myeloma. Toxic pulmonary reactions have been recognized regimen followed by autologous PBPC transplantation, in as many as 16–64% of patients;3–13 onset generally were studied retrospectively to evaluate the incidence of occurs within 1 year of starting BCNU. Events occurring post-transplant non-infective pulmonary complications up to 17 years later have been reported.14 The drug seems (NIPCs), risk factors predictive of NIPCs, and response to cause pulmonary damage in a dose-related manner:5,10,11 to steroids. Carmustine (BCNU) given i.v. at a dose of Phillips et al6 reported a 9.5% incidence of fatal pulmonary 600 mg/m2 was combined with etoposide and cyclophos- toxicity in patients receiving BCNU doses as high as 1.200 phamide in 40 patients (BCV regimen) and with etopo- mg/m2 as a single agent; another report suggests no pul- side and melphalan in 25 patients (BEM regimen). Sev- monary toxicity at doses of less than 1000 mg/m2.7 The enteen of 65 patients (26%) had one episode of NIPCs. threshold dose for BCNU toxicity is still controversial: lung The median time to NIPCs was 90 days (52–289). Fac- toxicity may occur at 600 mg/m2 when the drug is associa- tors that increased the risk of developing NIPCs on ted with other toxic agents like CY.1,15 multivariate analysis were female sex (P Ͻ 0.001) and This report deals with patients with hematological malig- BCV regimen (P Ͻ 0.05). All patients with NIPCs nancies who received a carmustine-based preparative regi- received prednisone at a dose of 1 mg/kg body weight men; the drug was combined with and for 10 days then tapered by 5 mg every two days; com- etoposide or with etoposide and melphalan and both regi- plete response to steroids was achieved in 15 of 17 mens were followed by autologous peripheral blood pro- patients; one unresponsive patient died of interstitial genitor cell (PBPC) transplantation. The aim of this retro- . BCNU given at the dose of 600 mg/m2 is spective analysis was to characterize the incidence of well tolerated when associated with melphalan and eto- pulmonary toxicity in patients treated with two diverse car- poside. In females and in patients receiving BCNU with mustine-containing regimens, and also to determine predic- cyclophosphamide, a BCNU dose reduction may be tive risk factors and response to steroid therapy. advisable. Bone Marrow Transplantation (2000) 25, 309–313. Keywords: carmustine; pulmonary toxicity; autologous Methods peripheral blood progenitor cell transplantation The medical records of 65 patients undergoing autologous PBSC transplantation who had received a BCNU-contain- ing preparative regimen were reviewed to assess the inci- There is evidence that several agents used in the preparative dence of pulmonary complications occurring within 12 regimens for both autologous and allogeneic bone marrow months from transplant and the risk factors for their devel- transplantation may induce lung injury without evidence of opment; response to steroid therapy was also evaluated. All infection. Agents that have been implicated include carmus- data concerning the total dose of cyclophosphamide, bleo- tine (BCNU), busulphan, cyclophosphamide (CY), and mycine, history of radiation and mediastinal involvement, 1 total body irradiation. BCNU causes toxic lung reactions history of smoking, previous history of or characterized by chronic interstitial fibrosis, cough, dyspnea other lung diseases were taken into account. Episodes of and decrease in lung diffusing capacity; toxicity related to pulmonary toxicity occurring after the first 12 months were excluded. The study includes only patients with normal Correspondence: EP Alessandrino, Centro Trapianti di Midollo Osseo, pulmonary function tests before transplant. Istituto di Ematologia, IRCCS Policlinico S Matteo, 27100 Pavia (I), Italy All pulmonary complications occurring within 12 months Received 15 May 1999; accepted 28 September 1999 from transplant were recorded. They were classified as fol- Pulmonary complications following carmustine-based regimens EP Alessandrino et al 310 lows: (1) minor complications: reduction by 20% of single- Kaplan–Meier estimates of pulmonary complications were breath carbon monoxide diffusing capacity (DLCO) with- computed and curves drawn for selected variables. A P out any other sign or symptoms; (2) major complications: value of 0.05 was considered statistically significant. Stata reduction of the above-reported parameter with symptoms 5.0 (College Station, TX, USA) was used for computation. (cough, dyspnea, fever) and signs (tachypnea, tachycardia, crackles or wheezes on auscultation) associated or not asso- ciated with evidence of interstitial pneumonia on chest Results radiograph. Patients with minor or major complications without evidence of bacterial, viral or fungal infection Patients’ characteristics despite extensive examination, were classified as having non-infective pulmonary complication (NIPC). Patients Sixty-five patients undergoing autologous PBPC transplan- with NIPC received prednisone therapy orally at a dose of tation who had received a BCNU-containing preparative 1 mg/kg body weight for 10 days, then tapered by 5 mg regimen were identified. Their main characteristics are every other day. shown in Table 1. The median age was 40 years (range 13– 61). Thirty–three were male and 32 female. Thirteen had a history of smoking. Previous pulmonary complications Preparative regimens were recorded in two cases; both had bacterial pneumonia. Patients received as conditioning regimen a combination of Eighteen patients with HD had received bleomycine BCNU, etoposide, melphalan (BEM regimen) or BCNU, together with adriamycin, vincristine and dacarbazine etoposide and cyclophosphamide (BVC regimen). The (ABVD scheme). Twenty-three patients had had medias- BEM regimen was given to 25 patients with multiple myel- tinal involvement and 12 of these had received radiotherapy oma: it consisted of BCNU 600 mg/m2/day on day −4, eto- to the mediastinum with a median dose of 35 Gy (range 25– poside 1000 mg/m2/continous infusion on day −3, mel- 40 Gy). Twenty-five patients with MM received the BEM phalan 200 mg/m2 on day −2. With the BVC regimen, regimen, while 40 patients with advanced HD or NHL had BCNU was given at a dose of 600 mg/m2 on day −5; it the BCV regimen (Table 2). PBPC were mobilized by 2 was administered with etoposide administered at a dose of cyclophosphamide given at a dose of 4000 mg/m followed ␮ 1000 mg/m2 by 1 day continuous infusion on day −4 and by G-CSF (5 g/kg/day) in all cases. One or two leukapher- with cyclophosphamide 4000 mg/m2 over 2 consecutive days (day −3 and day −2). Eighteen patients with HD and 22 with NHL received a transplant after the BVC regimen. Table 1 Main characteristics On day 0 all patients received a minimum dose of × 6 + 4 10 /kg autologous CD34 peripheral cells. Patients were Parameters No. cases (%) nursed in single rooms with laminar flow. All patients received the same antiseptic and anti-fungal mouth care and Total number of cases 65 irradiated blood product support. Median age (years) 40 (13–61) years Pulmonary function evaluation Sex Male 33 (50.8) Pulmonary function tests including spirometry and DLCO Female 32 (49.2) were performed before transplant in all patients. Patients Diagnosis were routinely followed monthly for 3 months and every NHL 22 (33.8) 3 months thereafter. Patients with cough, fever, or chest HD 18 (27.7) MM 25 (38.5) tightness underwent post-transplant chest radiography and Conditioning regimen additional spirometry with DLCO. In some cases broncho- BEM 25 (38.5) alveolar lavage (BAL) and trans-bronchial biopsy were per- BCV 40 (61.5) formed; patients were selected for these procedures if they Smokers 13 (20) had major complications and gave their informed consent. Previous pulmonary complications 2 (3.1) Previous bleomycine therapy 18 (27.7) Statistics Mediastinal involvement at diagnosis 23 (35.4) Descriptive statistics have been calculated for all variables. Radiotherapy on mediastinal field 12 (18.5) Data on patient ages, sex, diagnosis, previous radiation Pulmonary complications 24 (36.9) treatment, history of smoking, mediastinal involvement, Infection-related complications 7 (10.8) mediastinal field radiotherapy, pretransplant pulmonary Non-infection-related complications 17 (26.1) complications, previous treatment with bleomycine, and Major 11 (16.9) conditioning regimen were analyzed to assess by a Cox Minor 6 (9.2) model their predictive value for the occurrence of non- Median time of pulmonary complication and 90 (52–289) range (days) infective pulmonary complications: all the potential risk Ͻ factors (P 0.1) identified on univariate analysis were NHL = non-Hodgkin’s lymphoma; HD = Hodgkin’s disease; MM = mul- included in a multivariate Cox model. In the event of collin- tiple myeloma; BEM = carmustine, etoposide, melphalan; BCV = carmus- earity, choice of variable was based on clinical importance. tine, cyclosphosphamide, etoposide.

Bone Marrow Transplantation Pulmonary complications following carmustine-based regimens EP Alessandrino et al 311 Table 2 Pulmonary complications according to preparative regimen diagnosis appear to be significant prognostic factors on univariate analysis, while on multivariate analysis, con- Parameters BCV BEM ditioning regimen and sex are the only independent prog- regimen regimen nostic factors (Table 3) (Figure 1).

No. cases 40 25 Age (median) years 36 50 (13–61) (31–61) Discussion Diagnosis HD/NHL 40 — Cytotoxic chemotherapy and radiotherapy given as prep- MM — 25 aration for autologous hemopoietic stem cell transplantation Smoker 6 7 16 Mediastinal involvement 23 — have been associated with the development of NIPC. Radiotherapy 12 — BCNU, increasingly used in several high-dose chemo- Previous pulmonary complications 1 1 therapy combinations, induces NIPC in as many as 15–42% Bleomycine therapy 18 — of patients with HD, NHL or MM;5,9,10,17 its lung toxicity Post-transplant pulmonary seems to increase with the total dose given and the toxicity complications 17 7 Infection-related complications 3 4 threshold is higher in patients receiving the drug as a single Non-infection related agent than in those receiving BCNU along with cyclophos- complications 14 3 phamide or other chemotherapeutic agents.5–7,12,15 In the Major 10 1 present study, 65 patients with hematological malignancies Minor 4 2 2 Incidence of non-infective related received BCNU at the dose of 600 mg/m together with complications (events per person etoposide and cyclophosphamide (BCV) in 40 cases and per year observation) (95% 0.53 0.15 with melphalan and etoposide (BEM) in 25 cases. The confidence interval) (0.32–0.91) (0.05–0.47) higher incidence of major NIPC observed in patients treated with BCV compared with those who received BEM (35% vs 12%) supports the hypothesis that CY may augment the eses were performed to collect a minimum of 4 × 106/kg detrimental effect of BCNU on the lungs; comparable CD34-positive cells. results were observed in breast cancer patients receiving BCNU associated with CY and cisplatin.3 In a dose toxicity Pulmonary complications study of 72 patients receiving a modified BEM regimen Ager et al5 observed that BCNU given at the dose of 600 Pulmonary complications occurring within 1 year from mg/m2 to seven patients (five lymphomas, one ANLL, one transplant occurred in 24 of 65 patients; seven had infec- MM), was associated with an unacceptable incidence of tion-related pulmonary complications, 17 non-infectious lung toxicity (42%); Rubio et al4 reported idiopathic pneu- pulmonary complications (NIPCs) (Table 1). NIPCs were monia in 28% of patients with relapsed HD receiving a classified as major pulmonary complications in 11 patients; BEM regimen, while in our study patients receiving BEM six patients only had a reduction of DLCO, classified as a had minimal lung toxicity; they all had MM. minor complication. One NHL patient with major NIPCs Diagnosis, previous chemotherapy and status at trans- died of acute respiratory distress on day +52 from trans- plant may have a role in inducing NIPCs; in our study, plant. The median time for pulmonary complication presen- however, multivariate analysis did not demonstrate the tation was 90 days (range 52–289 ). Broncho-alveolar lav- impact of diagnosis and previous chemotherapy in pre- age was performed in seven cases; five underwent a trans- dicting NIPC (Table 3). bronchial biopsy which showed evidence of interstitial Risk factors favoring NIPCs may be previous lung dis- pneumonia (two cases) and signs of lymphocytic alveolitis ease, a history of smoking,18 previous treatment with bleo- and septal fibrosis (three cases). The incidence of NIPCs mycine or chest radiotherapy.19 In our study, however, none given as event per person per year of observation was of these factors reached statistical significance (Table 3). higher in patients receiving the BCV regimen (Table 2). BCV conditioning regimen and female sex were the only All patients with major or minor complications started ster- risk factors for the development of NIPC on multivariate oids at the dosage of 1 mg/kg day for 10 days which were analysis. The increased toxicity noticed in females has been then tapered by 5 mg every other day. Complete resolution previously reported; there is, however, no explanation for was achieved in eight cases within 3 weeks, with improve- this fact;4,20 the use of oral contraceptive pills may perhaps ment in functional tests, signs and symptoms. Response to be involved. treatment was reached at a median of 8 weeks in seven Lung injury induced by BCNU can be reversed by ster- patients. One unresponsive patient died of interstitial pneu- oids which may reduce symptoms and improve monia. Another patient had only slight improvement; he is DLCO.18,21,22 Early reports, however, suggested no bene- still receiving steroids at the small dose of 0.1 mg/kg of fit.23 Jones and co-workers21 reported symptomatic prednisone on day +900 from transplant. improvement in 12 of 13 breast cancer patients treated with prednisone. Some authors advise immediate initiation of Prognostic factors prednisone therapy to reduce need for long-term treatment and fatal outcome.22 In our study, patients received predni- Results from univariate and multivariate analysis are sum- sone promptly at the onset of symptoms or at least when marized in Table 3. Age, sex, conditioning regimens and a reduction of DLCO was noticed. This approach induced

Bone Marrow Transplantation Pulmonary complications following carmustine-based regimens EP Alessandrino et al 312 Table 3 Risk factors for non-infective pulmonary complications

With complication HR 95% CI Model P value (%)

Univariate analysis Conditioning regimen BCV 14 (35) 3.33 0.97–11.11 0.0321 BEM 3 (12) Age per year — 0.96 0.92–0.99 0.0374 Age class р40 13 (40.6) 3.33 1.1–10 0.0209 Ͼ40 4 (12.9) Sex Female 13 (40.6) 4.54 1.49–14.29 0.0034 Male 4 (12.1) Mediastinal involvement No 10 (23.8) 0.4595 Yes 7 (30.4) 1.45 Radiotherapy No 12 (22.6) 0.55–3.80 0.2110 Yes 5 (41.7) 0.49 0.17–1.41 Diagnosis NHL 9 (40.9) HD 5 (27.8) 0.54 0.18–1.61 0.267 MM 3 (12.0) 0.23 0.06–0.84 0.026 Lymphoma 14 (35) 3.33 0.97–12.5 0.0321 Myeloma 3 (12)

Multivariate analysisa HR 95% CI Model P value = 0.0003 Age per year 0.98 0.94–1.02 0.274 Sex Female vs male 7.14 2.17–25 0.001 Conditioning regimen BCV vs BEM 4.17 1.04–16.67 0.043

HR = hazard ratio; 95% CI = 95% confidence interval. aDiagnosis has been omitted from the model due to collinearity with conditioning regimen.

1.00 ment and fatal complications; a reduction in BCNU dosage P = 0.04 is probably advisable in females and in patients receiving BCNU along with cyclophosphamide where the incidence 0.75 of NIPC was particularly high.

0.50 References BCV 0.25 1 Kreisman H, Wolkove N. Pulmonary toxicity of antineoplastic therapy. Semin Oncol 1992; 19: 508–520. BEM 2 Stahl W, Eisenbrand G. Comparative study on the influence 0.00 of two 2-chloroethylnitrosoureas with different carbamoylat- 0 100 200 300 400 ing potential towards glutathione and glutathione-related enzymes in different organs of the rat. Free Rad Res Comms Days after transplant 1991; 14: 271–278. 3 Cagnoni PJ, Nieto Y, Shpall EJ et al. High-dose chemotherapy Figure 1 Actuarial probability of NIPCs according to different con- with autologous hemopoietic progenitor-cell support as part of ditioning regimens. combined modality therapy in patients with inflammatory bre- ast cancer. J Clin Oncol 1998; 16: 1661–1668. 4 Rubio C, Hill ME, Milan S et al. Idiopathic pneumonia syn- response in 16 of 17 patients (94%). Despite the good drome after high dose chemotherapy for relapsed Hodgkin’s results observed in our study, the lack of a control group disease. Br J Cancer 1997; 75: 1044–1048. renders the conclusions tentative. In some reports late 5 Ager S, Mahendra P, Richards EM et al. High dose carmus- administration of steroids may explain the high incidence of tine, etoposide and melphalan (‘BEM’) with autologous stem fatal NIPC.4,5 Prophylactic use of steroids concurrent with cell transplantation: a dose–toxicity study. Bone Marrow BCNU to prevent pulmonary toxicity or progression of Transplant 1996; 17: 335–340. toxicity is still debatable:6,7,24 its use should be tested in a 6 Phillips GL, Fay JW, Herzing GP et al. Intensive 1,3-bis(2- randomised trial. chloroethyl)-1- (BCNU), NSC 436650 and cryo- preserved autologous marrow transplantation for refractory In our study the main NIPC risk factors were BCV com- cancer: a phase I–II study. Cancer 1983; 52: 1792–1802. bination and female sex. Patients receiving high-dose 7 Selker RG, Jacobs SA, Moore PB et al. 1,3-bis(2-chloroethyl)- BCNU should have close follow-up with frequent DLCO 1-nitrosourea (BCNU) induced pulmonary fibrosis. Neurosur- determinations to allow prompt initiation of steroid therapy gery 1980; 7: 560–565. which may be of value in reducing need for long-term treat- 8 Wilczynski SW, Erasmus JJ, Petros WP et al. Delayed pul-

Bone Marrow Transplantation Pulmonary complications following carmustine-based regimens EP Alessandrino et al 313 monary toxicity syndrome following high dose chemotherapy 16 Quabeck K. The lung as a critical organ in marrow transplan- and bone marrow transplantation for breast cancer. Am J tation. Bone Marrow Transplant 1994; 14: S19–S28. Respir Crit Care Med, 1998; 157: 565–573. 17 Valteau D, Hartmann O, Benhamou E et al. Nonbacterial non- 9 Reece DE, Connors JM, Spinelli JJ et al. Intensive therapy fungal interstitial pneumonitis following autologous bone mar- with cyclophosphamide, carmustine, etoposide, cisplatin and row transplantation in children treated with high dose chemo- autologous bone marrow transplantation for Hodgkin’s disease therapy without TBI. Transplantation 1988; 45: 737–740. in first relapse after combination chemotherapy. Blood 1994; 18 Demirer T, Weaver CH, Buckner CD et al. High dose cyclo- 83: 1193–1199. phosphamide, carmustine and etoposide followed by allo- 10 Weaver CH, Appelbaum FR, Petersen FB et al. High dose geneic bone marrow transplantation in patients with lymphoid cyclophosphamide, carmustine, and etoposide followed by malignancies who had received prior dose limiting radiation autologous bone marrow transplantation in patients with therapy. J Clin Oncol 1995; 13: 596–602. lymphoid malignancies who have received dose limiting radi- 19 Todd NW, Peters WP, Ost AH et al. Pulmonary drug toxicity ation therapy. J Clin Oncol 1993; 11: 1329–1335. in patients with primary breast cancer treated with high dose 11 Wheeler C, Antin JH, Churchill WH et al. Cyclophosphamide, combination chemotherapy and autologous bone marrow carmustine and etoposide with autologous bone marrow trans- transplantation. Am Rev Respir Dis 1993; 147: 1264–1270. plantation in refractory Hodgkin’s disease and non Hodgkin’s 20 Lund MB, Kongerud J, Boe J et al. Cardiopulmonary sequelae lymphoma: a dose finding study. J Clin Oncol 1990; 8: after treatment for Hodgkin’s disease: increased risk in 648–656. females? Ann Oncol 1996; 7: 257–264. 12 Seiden MV, Elias A, Ayash L et al. Pulmonary toxicity asso- 21 Jones RB, Matthes S, Shpall EJ et al. Acute lung injury fol- ciated with high dose chemotherapy in the treatment of solid tumors with autologous marrow transplant: an analysis of four lowing treatment with high dose cyclophosphamide, cisplatin, chemotherapy regimen. Bone Marrow Transplant 1992; 10: and carmustine: pharmacodynamic evaluation of carmustine. 57–63. J Natl Cancer Inst 1993; 85: 640–647. 13 Chap L, Shpiner R, Levine N et al. Pulmonary toxicity of high 22 Kalaycioglu M, Kavuru M, Tuason L, Bolwell B. Empiric dose chemotherapy for breast cancer: a non-invasive approach prednisone therapy for pulmonary toxic reaction after high- to diagnosis and treatment. Bone Marrow Transplant 1997; dose chemotherapy containing carmustine (BCNU). Chest 20: 1063–1067. 1995; 107: 482–487. 14 O’Driscoll BR, Hasleton PS, Taylor PM et al. Active lung 23 Weiss RB, Poster DS, Penta JS. The and pulmon- fibrosis up to 17 years after chemotherapy with carmustine ary toxicity. Cancer Treat Rev 1981; 8: 111–125. (BCNU) in childhood. New Engl J Med 1990; 323: 378–382. 24 Aronin PA, Mahaley MS, Rudnick SA et al. Prediction of 15 Durant JR, Norgord MJ, Murad TM et al. Pulmonary toxicity BCNU pulmonary toxicity in patients with malignant gliomas: associated with bis-chloroethyl nitrosourea (BCNU). Ann assessment of risk factors. New Engl J Med 1980; 303: Intern Med 1979; 90: 191–194. 183–188.

Bone Marrow Transplantation