Bone Marrow Transplantation (2001) 28, 399–403  2001 Nature Publishing Group All rights reserved 0268–3369/01 $15.00 www.nature.com/bmt Post-transplant complications Pulmonary toxicity syndrome following CDEP (cyclophosphamide, dexamethasone, etoposide, cisplatin) chemotherapy

A Fassas1,3, I Gojo1, A Rapoport1, M Cottler-Fox1,3, B Meisenberg2, JC Papadimitriou2 and G Tricot1,3

1Department of Medicine, Division of Bone Marrow and Stem Cell Transplantation, Greenebaum Cancer Center; and 2Department of Pathology, University of Maryland, Baltimore, MD, USA

Summary: development of this pulmonary toxicity syndrome (PTS). This is further supported by the lack of correlation between We report on three patients with multiple myeloma who severity of PTS in case of BCNU-based high-dose chemo- developed drug-induced pneumonitis 1–2. months fol- therapy with age, tobacco use or baseline pulmonary func- lowing maintenance (post autologous transplantation) tion, suggesting that other factors play a role in PTS, in chemotherapy with CDEP (cyclophosphamide, dexame- addition to chemotherapy exposure.3 thasone, etoposide, cisplatin) and 6–20 months after From this perspective, it was intriguing to observe three exposure to carmustine (BCNU) 300 mg/m2, used in patients, with a history of either proven (two) or suspected combination with melphalan 140 mg/m2, as pre-trans- (one) fungal pneumonia, who developed typical clinical, plant conditioning regimen. All patients had either a radiographic and histologic features of PTS 5–11 weeks proven (two) or suspected (one) fungal pneumonia and after treatment with CDEP (cyclophosphamide, dexame- were treated with liposomal amphotericin B. Dyspnea, thasone, etoposide, cisplatin) chemotherapy, a fairly com- fever and cough were the prominent clinical symptoms, monly used regimen which has not so far been associated while air-space disease with ground glass appearance with non-infective pulmonary toxicity. These patients had was seen radiographically. Histologic features typical a dramatic response to a brief course of corticosteroids. for drug-induced lung injury were detected. All patients There was no evidence of fungal infection at the time of had a dramatic, clinical and radiographic response to a PTS diagnosis. brief course of corticosteroids. Although CDEP-induced pneumonitis appears to be a rare complication, its early recognition and prompt treatment, as well as its possible Patients and methods association with preceding fungal infection may have important clinical implications. Bone Marrow Transplan- Between August 1997 and February 2000, 231 patients tation (2001) 28, 399–403. underwent autologous peripheral blood stem cell transplan- Keywords: non-infective pulmonary toxicity; CDEP tation (APBSCT), at our institution, for a variety of hemato- chemotherapy; fungal infection logic malignancies (multiple myeloma, 87; non-Hodgkin lymphoma, 64; Hodgkin’s disease, 11; others, 14) and solid tumors (breast carcinoma, 46; others, nine). In the vast majority (200 patients), the same conditioning regimen of Drug-induced pulmonary toxicity, not associated with BCNU at 300 mg/m2 and melphalan at 140 mg/m2 infection, is a well-described complication of different (BCNU/MEL) was applied, followed by post-transplant chemotherapeutic agents. Among them, carmustine chemotherapy starting 3 months after the APBSCT and (BCNU), cyclophosphamide, bleomycin, busulphan, mel- administered every 3 months × four courses during the year 1 phalan and total body irradiation have been implicated. following transplant (or until disease relapse). The regi- Furthermore, combinations of various agents (even the ones mens used for the post-transplant chemotherapy were rarely associated with the syndrome) may potentially lead cyclophosphamide 300 mg/m2 by continuous i.v. infusion 2 to lung injury through drug interactions. Additionally, over 24 h for 4 days, dexamethasone 40 mg p.o. daily for unrecognized host–drug interactions may contribute to the 4 days, etoposide 30 mg/m2 by continuous i.v. infusion over 24 h for 4 days and cisplatin 15 mg/m2 by continuous i.v. infusion over 24 h for 4 days (CDEP), for courses 1 and Correspondence: Dr A Fassas, Myeloma and Transplantation Center, Uni- 3, alternating with dexamethasone 40 mg p.o. daily for 4 versity of Arkansas for Medical Sciences, 4301 W Markham Street, Little days, taxol 135 mg/m2 by i.v. infusion over 6 h on day 2 Rock, AR, 72205, USA 2 3Current address: Myeloma and Transplantation Research Center, Univer- and cisplatin 75 mg/m by i.v. infusion over 24 h on day sity of Arkansas for Medical Sciences, Little Rock, AR, USA 3 of the cycle, for courses 2 and 4. The feasibility and Received 25 September 2000, accepted 15 May 2001 toxicity data on 50 patients with multiple myeloma treated Non-infectious pulmonary toxicity after CDEP chemotherapy A Fassas et al 400 with such an approach have been previously reported.4 Nystatin was discontinued in mid-November 1999, after no Eight patients (4%) developed non-infective, pulmonary further change was documented by serial computed tom- toxicity syndrome (seven lung-biopsy proven, one ography (CT) scans. Of note, the patient remained severely suspected). All had been exposed to BCNU during the pre- anemic and thrombocytopenic during the entire post-trans- transplant conditioning. While in five patients the syndrome plant period, requiring frequent red cell and platelet trans- occurred in temporal proximity to BCNU exposure (32– fusions. Despite the stormy post-transplant course and the 98 days after administration), the remaining three patients good control of his disease (with bone marrow plasma- experienced this complication 6–20 months following the cytosis р5%) maintenance chemotherapy consisting of pre-transplant conditioning and interestingly enough within CDEP was administered between 10 and 14 December 1–2. months after treatment with CDEP. The clinical 1999. Additional stem cells were infused at the completion history of these patients is presented. of this chemotherapy. Prophylaxis with itraconazole was initiated, but no prophylaxis against PCP was given. On 10 Patient 1 January 2000, the patient presented with new onset dys- pnea, malaise, non-productive cough and severe hypoxemia

A 56-year-old man was diagnosed with nonsecretory mul- (PO2: 55 mmHg at room air). No blood products were tiple myeloma, stage II-A, in January 1999. He had a 20 transfused during the preceding 48 h. No fever was docu- pack-year history of smoking, but quit smoking approxi- mented. Chest CT showed a new infiltrate (Figure 1a). mately 17 years prior to diagnosis. Following two courses Bronchoscopy showed scant clear secretions in otherwise of pulse dexamethasone and mobilization chemotherapy normal airways. BAL failed to reveal any pathogens while consisting of cyclophosphamide 4500 mg/m2 and etoposide the transbronchial biopsy was consistent with chemo- 2000 mg/m2 (CTX/VP16), he received APBSCT on 27 May therapy-induced pneumonitis. Rapid and dramatic sympto- 1999, after conditioning with BCNU/MEL. Delayed matic and objective improvement (as evidenced by resol- engraftment, despite an excellent collection of stem cells, ution of dyspnea, weaning off supplemental oxygen, prompted the infusion of additional stem cells on day +19, correction of hypoxemia and resolution of the radiographic with ANC finally reaching 1500/␮l on day +31. Evaluation picture) occurred after a brief course of oral corticosteroids. for persistent neutropenic fevers revealed pulmonary nod- Due to the recent history of fungal pneumonia, a relatively ules. Bronchoalveolar lavage (BAL) was negative for low dose of methylprednisolone (total 24 mg/day) was pathogens (Pneumocystis carinii pneumonia (PCP), chosen and symptoms recurred on fast tapering. Reinsti- Legionella, acid fast bacilli, viral and fungal pathogens). tution of corticosteroids and much slower tapering led again Empiric treatment with liposomal amphotericin B was to complete resolution of the symptoms. The patient experi- initiated. A repeat bronchoscopy with BAL was performed enced no subsequent reactivation of fungal pneumonia. for evaluation of hemoptysis on 6 August 1999; silver stains of cytologic specimens were positive for fungal Patient 2 elements consistent with aspergillus. Liposomal nystatin A 53-year-old woman, without prior history of smoking, was introduced with improvement in the radiologic picture. was diagnosed with IgA ␬ multiple myeloma, stage III-

Figure 1 Improvement of pulmonary infiltrates following treatment with corticosteroids. Upper panel, pre-treatment; lower panel, post-treatment.

Bone Marrow Transplantation Non-infectious pulmonary toxicity after CDEP chemotherapy A Fassas et al 401 B, in March 1998. Following three courses of VAD and mobilization chemotherapy with CTX/VP16, she received APBSCT on 6 August 1998, after conditioning with BCNU/MEL. She achieved complete hematologic recovery and a partial remission of the myeloma. In November 1998, she was given CDEP chemotherapy, which was compli- cated by pneumonia with extensive bilateral infiltrates. Bronchoscopy with BAL was performed; cultures of the specimens grew Candida tropicalis only. The same patho- gen, ie Candida tropicalis, was also identified in blood cul- tures. Further maintenance chemotherapy was therefore held. Relapsing disease in September 1999, prompted the administration of a second APBSCT on 14 October 1999, after conditioning with melphalan 200 mg/m2. Prophylaxis with itraconazole was initiated prior to the transplantation and continued over the next several months. No prophylaxis against PCP was given. Thalidomide was started after (timely) hematologic recovery. She achieved a complete remission (with Ͻ2% bone marrow plasmacytosis and undetectable monoclonal protein). On 20 January 2000, the patient received another course of CDEP chemotherapy. On 31 March 2000, she presented with fever, malaise, non- productive cough, dyspnea and mild hypoxemia (PO2:84 mmHg at room air). No blood products were transfused during the preceding 48 h. Chest CT scan showed a diffuse ground glass pattern, predominantly involving the lower lungs (Figure 1b). Bronchoscopy showed moderate mucoid secretions in otherwise normal airways. No pathogens (Pneumocystis carinii pneumonia, Legionella, acid fast bacilli, viral and fungal pathogens) were identified by BAL and the transbronchial biopsy was consistent with chemo- therapy-induced pneumonitis (Figure 2a, b). Corticosteroid treatment was initiated (total dose of methylprednisolone 60 mg/day, tapered over the next 3 weeks) with rapid resol- ution of the clinical symptomatology and normalization of the radiographic picture. The patient experienced no subsequent reactivation of fungal pneumonia.

Patient 3 A 65-year-old woman, without prior history of smoking, ␬ was diagnosed with IgA multiple myeloma, stage III-A Figure 2 Histologic findings in lung biopsies of patient 2 (a, b) and in January 1999. Following two courses of pulse dexame- patient 3 (c). All biopsies show thickening of the alveolar septa caused thasone and mobilization chemotherapy with CTX/VP16, by interstitial fibrosis and chronic inflammation. The higher power magni- she received APBSCT on 20 May 1999, after conditioning fication (b) shows the markedly hyperplastic and atypical pneumocytes with BCNU/MEL. Her post-transplant course was compli- type II (arrow). Intra-alveolar macrophages and edema are seen in different degrees in all cases. (Original magnification, × 10 a, c; × 40 b). cated by Candida parapsilosis fungemia, Staphylococcus epidermidis and Enterococcus faecium bacteremia, Clostri- dium difficile colitis, cytomegalovirus reactivation (CMV

antigenemia) and right lung pneumonia (for which she hypoxemia (PO2: 50 mmHg at room air). No blood pro- underwent bronchoscopy and BAL with negative results). ducts were transfused during the preceding 48 h. Chest CT Liposomal amphotericin B was given for 3 weeks along scan showed extensive, diffuse air-space disease with a with appropriate antibacterial and antiviral coverage with ground glass appearance (Figure 1c). Bronchoscopy resolution of fever and blood cultures becoming negative. showed scant mucopurulent secretions in otherwise normal Repeat chest radiographs were also clear. She achieved a airways. No pathogens (Pneumocystis carinii pneumonia, near complete remission (with Ͻ2% bone marrow plasma- Legionella, acid fast bacilli, viral and fungal pathogens) cytosis and IgA only detectable by immunofixation). On 25 were identified by BAL and the transbronchial biopsy was August 1999, she was given CDEP chemotherapy with consistent with chemotherapy-induced pneumonitis (Figure good tolerance and timely hematologic recovery. No 2c). Corticosteroid treatment was initiated (total dose of prophylaxis against PCP was given. On 16 November 1999, methylprednisolone 80 mg/day, tapered over the next 3 she presented with dyspnea, low grade fever and severe weeks) with dramatic resolution of the symptoms and the

Bone Marrow Transplantation Non-infectious pulmonary toxicity after CDEP chemotherapy A Fassas et al 402 radiographic picture. The patient experienced no sub- BCNU-based conditioning, developed pulmonary symp- sequent reactivation of fungal pneumonia. toms with a median onset of 10.3 weeks (range 2.4–16.6 weeks). Finally, in a retrospective study of 10 women with Pathology breast cancer and biopsy-proven pulmonary toxicity after high-dose BCNU, all became symptomatic at a mean time The findings in the transbronchial biopsies (Table 1) are of 48 ± 14 days (range, 21–62 days) after the initiation of young, connective tissue growth in the alveolar septa, high-dose treatment.7 Furthermore, the dose of BCNU used enlargement and atypia of type II pneumocytes and in our treatment program was only 300 mg/m2, much lower accumulation of intra-alveolar edema and macrophages. A than the dose used in the previously mentioned studies. fifth feature, less consistently observed, was pulmonary Therefore, it is rather unlikely that a BCNU effect can arterial branch changes consisting of intimal edema and explain the whole picture. fibrosis and endothelial cell injury. It may be also argued that the toxicity observed in our The results were assessed semiquantitatively by one of patients was caused by cyclophosphamide or etoposide. the authors (JCP). The grades assigned were 0–4+ (0, nor- Several case reports of suspected cyclophosphamide- mal; 4+, most pathological). induced lung toxicity have been described.8 Both the timing In all three cases the presence of infectious agents of the reaction and the cumulative dose of the drug were (viruses, fungi) was ruled out with hematoxylin and eosin widely variable. More importantly, however, in the over- and special stains. The observed findings of pulmonary whelming majority of related cases, confounding variables injury expressed by the pneumocyte II atypia, the probably contributed to the lung findings. In a review from accompanying young fibrosis and intra-alveolar accumu- the Mayo Clinic,9 the authors identified only six patients, lation of edematous fibroid and macrophages are – in the over a 20-year period, in whom cyclophosphamide was the absence of any infectious etiology – compatible with only identifiable factor for lung toxicity. Most patients (5/6) drug toxicity. developed late-onset pneumonitis that was frequently asso- ciated with pleural thickening and did not respond to corticosteroids, features not observed in our patients. Eto- Discussion poside, administered orally for lung cancer management, has been implicated in causing drug-induced pneumonitis Three patients (1.5%) developed typical clinical radio- in two patients. In both reports (in Japanese), the clinical graphic and histologic features of PTS 5–11 weeks after and pathologic features were similar to the ones described treatment with CDEP (cyclophosphamide, dexamethasone, in our patients. etoposide, cisplatin) chemotherapy. None had received Pretreatment with BCNU may have ‘primed’ our patients prior radiation treatment for local disease control and only for further lung injury by the following cyclophosphamide- one had a limited, rather remote, exposure to tobacco. All based regimen. In a recent report10 standard induction patients had a remote (у6 months) exposure to BCNU 300 chemotherapy with CAF (cyclophosphamide, doxorubicin, mg/m2 and melphalan 140 mg/m2 with history of either pro- 5-fluorouracil) produced asymptomatic pulmonary dysfunc- ven (two) or suspected (one) fungal pneumonia. The clini- tion in women with high-risk breast cancer (as evidenced cal and radiographic picture resolved with a brief course by significant decrease in the diffusing capacity of the lungs of corticosteroids. No evidence of reactivation of fungal for carbon monoxide (DLCO)). Further decrease in DLCO infection or transfusion-related lung toxicity was found. and subsequent PTS were observed with the BCNU-based It is possible that the PTS seen in our patients merely pre-transplant conditioning regimen. Another study11 (with reflects a delayed BCNU toxic effect. Pulmonary toxicity a design similar to ours) employed post-autologous trans- attributed to BCNU has been reported up to 17 years post plantation ‘maintenance’ chemotherapy with paclitaxel in a exposure.5 However, 23 out of 26 patients who developed group of 10 women with metastatic breast cancer. BCNU idiopathic pneumonia syndrome following BCNU-based was used in the conditioning regimen at the dose of 450 high-dose chemotherapy with autologous bone marrow mg/m2. No toxicity was reported with a median follow-up transplantation (ABMT) for relapsed Hodgkin’s disease, of 2 years post transplant. Although this ‘priming’ effect experienced symptoms during the first 6 months post-treat- can not be ruled out (as no serial pulmonary function tests ment (with the remaining three patients diagnosed between were performed in our patients), the low incidence of this 6 and 12 months post-chemotherapy).6 In another report,3 complication makes this possibility less likely. From this 26 women with breast cancer, who underwent ABMT after point of view, it is also instructive that patient no. 2 did not have any appreciable decline in DLCO performed prior to her first and second transplantation and yet she Table 1 Pathological findings in the transbronchial biopsies developed the PTS following the administration of the CDEP course after her second APBSCT and Ͼ1. years Alveolar Atypical Intra-alveolar Intra-alveolar after exposure to BCNU. septal type II edema macrophages fibrosis pneumocytes Finally, the interesting possibility of a prior infectious ‘priming’ to PTS needs to be considered. Two of our Patient 1 + + ++ + patients (patients 1 and 2) had a history of proven pulmon- Patient 2 ++ +++ + ++ ary mycosis (which in the latter case deterred the adminis- Patient 3 ++ ++ + +++ tration of further maintenance chemotherapy). The third patient (patient 3) had pulmonary infiltrates in the context

Bone Marrow Transplantation Non-infectious pulmonary toxicity after CDEP chemotherapy A Fassas et al 403 of Candida parapsilosis fungemia, for which she was 2 Zimmerman MS, Ruckdeschel JC, Hussain M. Chemotherapy- treated with amphotericin B and promptly responded. It is induced interstitial pneumonitis during treatment of small cell highly unlikely that active infection was present at the time anaplastic lung cancer. J Clin Oncol 1984; 2: 396–405. of PTS, given the stability of the clinical and radiographic 3 Wilczynski SW, Erasmus JJ, Petros WP et al. Delayed pul- picture at the time of chemotherapy administration, the fail- monary toxicity syndrome following high-dose chemotherapy and bone marrow transplantation for breast cancer. Am J ure to identify fungal (or other infectious) elements by BAL Respir Crit Care Med 1998; 157: 565–573. and the rapid improvement on corticosteroids. Therefore, it 4 Tricot G, Fassas A, Rapoport A et al. Post-transplantation is possible that a preceding (albeit treated and well (PT) intensive chemotherapy is feasible and results in prompt controlled) fungal pneumonia may somehow (in the setting and complete hematologic recovery in patients with multiple of CDEP chemotherapy and remote BCNU exposure) have myeloma and low grade lymphoma. ASH 1999; 1485: 332a contributed to PTS. Alternatively, the antifungal treatment (Abstr.). (liposomal amphotericin) which has good lung tissue pen- 5O’Driscoll BR, Hasleton PS, Taylor PM et al. Active lung etration might be a factor in the subsequent development fibrosis up to 17 years after chemotherapy with carmustine of the PTS (possibly interaction with the chemotherapy (BCNU) in childhood. New Engl J Med 1990; 323: 378–382. agents). 6 Rubio C, Hill ME, Milan S et al. Idiopathic pneumonia syn- If the above-mentioned hypothesis is correct, it has to be drome after high-dose chemotherapy for relapsed Hodgkin’s emphasized that the presumption of the fungal etiology of disease. Br J Cancer 1997; 75: 1044–1048. 7 Todd NW, Peters WP, Ost AH et al. Pulmonary drug toxicity the pulmonary symptoms and radiologic findings in patients in patients with primary breast cancer treated with high-dose with a similar clinical presentation needs to be avoided. combination chemotherapy and autologous bone marrow Furthermore, transbronchial biopsy should always be transplantation. Am Rev Respir Dis 1993; 147: 1264–1270. encouraged. 8 Spector JI, Zimbler H, Ross JS. Early-onset cyclophospham- In conclusion, non-infective pulmonary toxicity syn- ide-induced interstitial pneumonitis JAMA 1979; 242: 2852– drome may rarely complicate the course of a commonly 2854. used chemotherapy regimen, such as CDEP, possibly in 9 Malik SW, Myers JL, DeRemee RA, Specks U. Lung toxicity association with previous, treated, fungal pneumonia. Rec- associated with cyclophosphamide use. Two distinct patterns. ognition of the syndrome, timely establishment of diagnosis Am J Respir Crit Care Med 1996; 154: 1851–1856. and prompt treatment with steroids are essential for a 10 Bhalla KS, Wilczynski SW, Abushamaa AM et al. Pulmonary favorable clinical outcome. toxicity of induction chemotherapy prior to standard or high- dose chemotherapy with autologous hematopoietic support. Am J Respir Crit Care Med 2000; 161:17–25. References 11 Rahman Z, Kavanagh J, Champlin R et al. Chemotherapy immediately following autologous stem-cell transplantation in 1 Kreisman H, Wolkove N. Pulmonary toxicity of antineoplastic patients with advanced breast cancer. Clin Cancer Res 1998; therapy. Semin Oncol 1992; 19: 508–520. 4: 2717–2721.

Bone Marrow Transplantation