Fatal pulmonary fibrosis associated with BCNU: the relative role of
platelet-derived growth factor-B, insulin-like growth factor I, transforming
growth factor-b1 and cyclooxygenase-2
Y-C Shen
1, C-F Chiu
1,5, K-C Chow
2, C-L Chen
3,4, Y-C Liaw
4and S-P Yeh
1,4,51Division of Hematology and Oncology, Department of Internal Medicine, China Medical University Hospital, Taichung, Taiwan; 2Cancer Research Center, Department of Medical Research, China Medical University Hospital, Taichung, Taiwan;3Department of
Pathology, China Medical University Hospital, Taichung, Taiwan;4Stem Cell Research Lab, Department of Medical Research,
China Medical University Hospital, Taichung, Taiwan; and5China Medical University, Taichung, Taiwan
Summary:
Pulmonary fibrosis is a severe complication associated with bis-chloronitrosourea (BCNU) therapy. However, the pathogenetic mechanism has never been well investi-gated. We report here a 26-year-old female with diffuse large B-cell lymphoma who died of severe pulmonary fibrosis 81 days after the administration of high-dose BCNU (600 mg/m2). Thoracoscopic wedge resection of
left upper lung performed 10 days before patient’s death showed severe pulmonary fibrosis with prominent hyper-plasia of alveolar macrophages and type II pneumocytes. We further used immunohistochemistry (IHC) to examine the relative role of platelet-derived growth factor-B (PDGF-B), insulin-like growth factor I (IGF-I), trans-forming growth factor-b1 (TGF-b1) and cyclooxygenase-2 (COX-cyclooxygenase-2) in the pathogenesis of BCNU-related pulmo-nary fibrosis. Strong expressions of PDGF-B and IGF-1 on alveolar macrophages and type II pneumocytes were clearly demonstrated, but in contrast, the expressions of TGF-b1 and COX-2 were almost undetectable. In conclusion, pulmonary fibrosis can develop early and progress rapidly after the administration of high-dose BCNU. The markedly increased expression of fibrogenic factors PDGF-B and IGF-1 on hyperplastic alveolar macrophages and hyperplastic type II pneumocytes may play an important role in the fibrogenesis of this disease. These novel findings may offer specific therapeutic targets in the treatment of BCNU-associated pulmonary fibrosis. Bone Marrow Transplantation (2004) 34, 609–614. doi:10.1038/sj.bmt.1704616
Published online 2 August 2004
Keywords: BCNU; pulmonary fibrosis; PDGF; IGF; COX-2
Pulmonary fibrosis is a severe complication typically developed months or years after repetitive doses of BCNU and the risk is increased when the cumulative dose is greater than 1200 mg/m2.1,2 More recently, acute lung
injury after single high-dose BCNU given as conditioning chemotherapy for hematopoietic stem cell transplantation (HSCT) is increasingly recognized.3Some risk factors, such
as history of irradiation to lung, cigarette smoking, female gender and a dosage greater than 475 mg/m2, are known to
be associated with BCNU-related pulmonary injury.4,5
However, the pathogenetic mechanism of BCNU-asso-ciated pulmonary fibrosis has never been studied.
Current concepts suggest that macrophages play an important role in idiopathic pulmonary fibrosis (IPF) by virtue of their ability to release many cytokines promoting fibrogenesis,6,7 and the final result of ‘fibrosis’ involves
increased activity and exaggerated responses by a spectrum of profibrogenic mediators with a concomitant loss of counter-regulative factors that protect against fibrogenesis. Many studies have shown that PDGF is a potent mitogen and chemoattractant of fibroblasts. Expression of PDGF-B, an isoform of PDGF, from adenoviral vectors or administration of recombinant human PDGF-B, delivered intratracheally to rat lung, produces histological features of fibrosis.8 IGF-1 stimulates proliferation of fibroblasts
where it may act synergistically with other fibrogenic growth factors, and is also a potent inducer of collagen synthesis.9,10 Increased IGF-1 messenger RNA (mRNA)
has been detected in lung tissues following the induction of bleomycin-induced pulmonary fibrosis in mice.7TGF-b1 is
also a potent chemoattractant of fibroblasts and is consistently found to be upregulated at sites of fibrogenesis, including pulmonary fibrosis.11,12
Pulmonary fibrosis and collagen synthesis are regulated by a complex interaction between many profibrogenic media-tors (mentioned above) and many antifibrogenic mediamedia-tors. Of these, PGE2 was best studied and has been showed to potently inhibit fibroblast proliferation and collagen synth-esis.13,14 Decreased capacity to upregulate COX-2 mRNA
expression and COX-2-derived PGE2 synthesis in the presence of increasing levels of profibrogenic mediators leads to unopposed fibroblast proliferation and collagen
Received 13 October 2003; accepted 13 April 2004 Published online 2 August 2004
Correspondence: Dr S-P Yeh, Division of Hematology and Oncology, Department of Internal Medicine, China Medical University Hospital, 2 Yuh Der Road; Taichung 404, Taiwan;
E-mail: [email protected]
synthesis.15 Recent data show that PGE2 synthesis and
suppression of fibroblast proliferation by alveolar epithelial cells is COX-2 dependent.16Would PDGF-B, IGF-1,
TGF-b1 and COX-2 also play some roles in BCNU-associated pulmonary fibrosis? We report here a 26-year-old female receiving autologous peripheral blood stem cell transplanta-tion (APBSCT) for diffuse large B-cell lymphoma who died of rapidly progressing BCNU-associated pulmonary fibrosis. Furthermore, we also examined the relative role of PDGF-B, IGF-1, TGF-b1 and COX-2 in the pathogenesis of BCNU-associated pulmonary fibrosis.
Methods Case history
APBSCT was performed on a 26-year-old female of diffuse large B-cell lymphoma at second complete remission. Treatments before transplantation included seven courses of CHOP (cyclophosphamide, doxorubicin, vincristine and prednisolone), six courses of intrathecal chemotherapy with methotrexate, cytarabine and dexamethasone for menin-geal relapse, and two courses of high-dose cytarabine-based salvage chemotherapy. PBSC harvest was performed during the recovery phase after the second course of salvage chemotherapy. The pretransplant chest X-ray and vital organ function were normal. The conditioning chemotherapy for APBSCT consisted of BCNU 600 mg/ m2i.v. on day4, etoposide 800 mg/m2i.v. on day –3 and
cyclophosphamide 4 gm/m2i.v. daily on days 2 and –1.
The date of PBSC infusion is defined as day 0. Early post transplantation course was uneventful, and recovery of neutropenia occurred on day 10. However, fever, non-productive cough and pulmonary infiltration developed on day 34. She was initially treated for pneumonia although no microorganism was isolated from repeated cultures of sputum and blood. Prednisolone 1 mg/kg/day was added on day 51 for increasing dyspnea and infiltration on chest X-ray (Figure 1a). Unfortunately, her clinical condition and the pulmonary infiltrates on chest X-ray deteriorated rapidly after the use of steroid (Figure 2b, chest X-ray on day 63) and so video-assisted thoracoscopy with wedge resection of left upper lobe was performed on day 64. The patient became ventilator dependent after this procedure and finally died of respiratory failure on day 76.
Routine pathological studies
The tissue specimens obtained from thoracoscopic wedge resection of left upper lobe were fixed with 10% formalde-hyde and embedded in paraffin, from which 6-mm-thick sections were cut. Sections of each specimen were stained with hematoxylin and eosin (H&E) for routine histopatho-logic examination, fast stain for the detection of acid-fast microorganism and silver stain for the detection of pneumocystis carinii.
Immunohistochemical studies
Serial sections, prepared from blocks of lung tissue obtained from thoracoscopic wedge resection, were fixed,
deparaffinized, and then incubated with 0.3% hydrogen peroxide to remove endogenous peroxidase activity fol-lowed by reblock with normal horse serum to reduce nonspecific binding. The lung sections were then stained with mouse monoclonal antibody (Ab) to human macro-phage (DAKO CD68, DAKO, Kyoto, Japan), rabbit polyclonal Ab to PDGF-B (Santa Cruz, Santa Cruz, CA, USA), mouse monoclonal Ab to IGF-1 (Upstate, Lake Placid, NY, USA), rabbit polyclonal Ab to TGF-b1 (Santa Cruz, Santa Cruz, CA, USA) and COX-2 (Oxford Biomedical Res., Oxford, MI, US), and mouse lyophilized monoclonal antibody to late antigen of cytomegalovirus (Novocastra Lab Ltd, Newcastle upon Tyne, UK). The lung tissues were then incubated with biotinylated anti-mouse/rabbit IgG Ab. Subsequent color development was achieved by incubating the sections with either aminoethyl carbazole (AEC; DAKO, Carpinteria, CA, USA) for CD68, TGF-b1 and COX-2 or 3,30-diaminobenzidine
tetrahydrochloride (DAB; Zymed Lab. Inc., South San Francisco, CA, USA) for PDGF-B, IGF-1, and late antigen of cytomegalovirus. Positive staining was recog-nized under a microscope as crimson color for AEC and brown color for DAB.
Results
Lung biopsy showed severe pulmonary fibrosis and no microorganism including cytomegalovirus was identified. Marked interstitial expansion by fibrous tissue with few inflammatory cells was found and the alveoli spaces were filled by mononuclear cells (Figure 2a). The strongly
Figure 1 Chest X-ray before and after the treatment of prednisolone. (a): Chest X-ray on day 51. The date of PBSC infusion is defined as day 0. Prednisolone 1 mg/kg i.v. injection daily was started on this day. (b) Chest X-ray on day 63, 12 days after the treatment of prednisolone.
positive staining with monoclonal anti-CD68antibody indicated that these alveolar mononuclear cells were macrophages (Figure 2b). In addition to the markedly
increased foaming alveolar macrophages, prominent hyperplasia of type II pneumocytes could also be found at higher magnification (Figure 2c). Immunohistochemical studies showed that PDGF-B and IGF-1 were positively stained in the alveolar macrophages and type II pneumo-cytes (Figure 3a, b). In contrast, the expressions of TGF-b1 and COX-2 were almost undetectable (Figure 3c, d). With higher magnification, the expressions of PDGF-B and IGF-1 were more prominent in the hyperplastic type II pneumocytes as compared to the alveolar macro-phages (Figure 3e, f), besides, IGF-1 was also strongly expressed in the interstitial fibroblasts and endothelial cells (Figure 3f).
Discussion
BCNU-associated pulmonary fibrosis was first reported by Holoye et al in 1976.17Over the past 10 years, high-dose
BCNU (300–800 mg/m2) with HSCT has been used
increasingly to treat patients with lymphoma. In this setting, BCNU-related acute lung injury had been reported in 10–40% of patients after HSCT.18Despite the fact that
BCNU is long recognized as a pulmonary toxin, the pathophysiology of BCNU-associated lung injury has not been well investigated. In contrast, considerable advances in the understanding of the pathogenesis of IPF have been achieved recently. In a recent consensus statement issued jointly by the American Thoracic Society and the European Respiratory Society, the roles of ‘various cytokines and growth factors’ are described as ‘critical’ to the process of fibrosis.19In the current study, although only one patient’s
specimen was examined, the overexpression of PDGF-B and IGF-1 is demonstrated and localized predominantly to alveolar macrophages and type II pneumocytes. In normal adult lung, PDGF-B and IGF-1 proteins are expressed moderately in alveolar macrophages only.20This indicates
that both PDGF-B and IGF-1 played an important role in the fibrogenic process of this patient. In addition, the more intense expression of PDGF-B and IGF-1 in type II pneumocytes as compared to alveolar macro-phage indicates that type II pneumocyte may be a more important source of these profibrogenic mediators than alveolar macrophage. The type II pneumocyte is traditionally recognized as the ‘defender’ of the alveolus, a concept challenged recently.21 Our findings suggest
that type II pneumocytes may contribute, at least partially, to the pathogenesis of BCNU-associated pulmonary fibrosis.
TGF-b1 overexpression was found in many animal models of pulmonary fibrosis. Surprisingly, the expression of TGF-b1 was almost undetectable in this patient, and suggests that BCNU-associated lung injury may be mediated via a TGF-b1-independent pathway. However, since the lung tissue we examined was taken 13 days after prednisolone treatment, we cannot exclude the possibility that the decreased expression of TGF-b1 is the result of a suppressive effect of corticosteroid on TGF-b1 produc-tion.22COX-2 expression also was almost undetectable in
this patient. Whether the decreased expression of COX-2 is due to a BCNU-related biological effect or the suppressive
Figure 2 Interstitial fibrosis with marked hyperplasia of type II pneumocytes and alveolar macrophages. (a) Marked interstitial expansion by fibrous tissue with few inflammatory cells infiltration. The alveoli were filled by mononuclear cells. (b) By immunohistochemistry, CD68was strongly expressed in the mononuclear cells within alveoli. (c) With higher magnification, markedly increased foaming alveolar macrophages and prominent hyperplasia of type II pneumocytes were found.
Figure 3 Expression of PDGF-B, IGF-1, TGF-b1 and COX-2 in the lung. (a/b) By immunohistochemistry, PDGF-B (Figure 3a) and IGF-1 (Figure 3b) were expressed in the alveolar macrophages and type II pneumocytes. (c/d) Expression of TGF-b1 (Figure 3c) and COX-2 (Figure 3d) were almost undetectable. (e/f) With higher magnification, the expression of PDGF-B (Figure 3e) and IGF-1 (Figure 3f) were more prominent in type II pneumocytes as compared to alveolar macrophages. Besides, IGF-1 was also strongly expressed in the endothelial cells (arrow head) and interstitial fibroblasts (arrow). 612
effect23 of previous prednisolone treatment, the decreased
activity of this ‘protective’ enzyme may also contribute to fibrogenesis of this patient.
Many important issues, however, are not answered in this study. The lung tissues we examined showed the late stage of pulmonary injury. Is there an ‘initial inflammatory’ stage after BCNU exposure? The cytokine profiles we examined in this study may not represent a pathogenetic mechanism but rather may be a consequence of the inflammatory process. Why were PDGF-B and IGF-1 overexpressed? Was this from the effect of BCNU itself, or mediated by other chemokines or cytokines? What is the precise role of type II pneumocyte? Defender or offender? What is the effect of concomitant cyclophosphamide chemotherapy on BCNU-associated lung injury? The lack of appropriate animal models limits our understanding of the pathophysiology of BCNU-associated lung injury in detail, and we believe these questions will remain unan-swered in the recent future. In addition to these, the decreased expression of COX-2 and the refractoriness to corticosteroid treatment of this patient should not be interpreted as ineffectiveness of corticosteroid in the treatment of BCNU-associated lung injury. Lung injury after high-dose BCNU may include a severe form, that is, idiopathic pneumonia syndrome (IPS), and a less severe form, that is, delayed pulmonary toxicity syndrome (DPTS). Rubio et al5 reported that nine (17%) of 53
patients with relapsed Hodgkin’s disease treated with BCNU higher than 475 mg/m2 died of IPS early (median
74 days) after transplantation and many were cortico-steroid refractory. In contrast, Wilczynski et al reported that 41 (91%) of 45 patients with breast cancer treated with BCNU 600 mg/m2had a significant decline in single breath
diffusing capacity (DLco) after transplantation. Corticos-teroid was given to 26 symptomatic patients and a subsequent improvement in DLco of 17.1% was found. Most important of all, no patient died of pulmonary injury in this cohort and the authors attributed this favorable result to early recognition of DPTS and rapid institution of treatment with corticosteroid.24Whether the
DPTS seen in these patients is the ‘early stage’ of the spectrum of IPS or a different disease entity still needs to be clarified.25However, it is highly likely that in case of frank
pulmonary fibrosis, as seen in our patient we presented here, the result of corticosteroid treatment would be disappointing not only because of the persistent expression of profibrogenic growth factors such as PDGF-B and IGF-1 but also due to the potentially detrimental effect of corticosteroid on COX-2 expression. A more specific antifibrogenic agent such as pirfenidone, which can suppress the synthesis of PDGF,26 may deserve a trial in
this situation.
In conclusion, pulmonary fibrosis can develop early after the administration of high-dose BCNU and may progress rapidly to death. The increased expression of fibrogenic growth factors PDGF-B and IGF-1 on the markedly hyperplastic type II pneumocytes and alveolar macro-phages may play an important role in the fibrogenesis of this disease. Nevertheless, examination of more tissue specimens from other patients with BCNU-associated pulmonary fibrosis is needed to confirm our findings.
References
1 Aronin PA, Mahaley Jr MS, Rudnick SA et al. Prediction of BCNU pulmonary toxicity in patients with malignant gliomas: an assessment of risk factors. N Engl J Med 1980; 303: 183–188. 2 Weiss RB, Poster DS, Penta JS. The nitrosoureas and
pulmonary toxicity. Cancer Treat Rev 1981; 8: 111–125. 3 Jones RB, Matthes S, Kemme D et al. Cyclophosphamide,
cisplatin, and carmustine: pharmacokinetics of carmustine following multiple alkylating-agent interactions. Cancer Che-mother Pharmacol 1994; 35: 59–63.
4 Alessandrino EP, Bernasconi P, Colombo A et al. Pulmonary toxicity following carmustine-based preparative regimens and autologous peripheral blood progenitor cell transplantation in hematological malignancies. Bone Marrow Transplant 2000; 25: 309–313.
5 Rubio C, Hill ME, Milan S et al. Idiopathic pneumonia syndrome after high-dose chemotherapy for relapsed Hodg-kin’s disease. Br J Cancer 1997; 75: 1044–1048.
6 Homma S, Nagaoka I, Abe H et al. Localization of platelet-derived growth factor and insulin-like growth factor I in the fibrotic lung. Am J Respir Crit Care Med 1995; 152 (6 Part 1): 2084–2089.
7 Cao B, Guo Z, Zhu Y et al. The potential role of PDGF, IGF-1, TGF-beta expression in idiopathic pulmonary fibrosis. Chin Med J (Engl) 2000; 113: 776–782.
8Yoshida M, Sakuma J, Hayashi S et al. A histologically distinctive interstitial pneumonia induced by overexpression of the interleukin 6, transforming growth factor beta 1, or platelet-derived growth factor B gene. Proc Natl Acad Sci USA 1995; 92: 9570–9574.
9 Goldstein RH, Poliks CF, Pilch PF et al. Stimulation of collagen formation by insulin and insulin-like growth factor I in cultures of human lung fibroblasts. Endocrinology 1989; 124: 964–970.
10 Harrison NK, Cambrey AD, Myers AR et al. Insulin-like growth factor-I is partially responsible for fibroblast prolifera-tion induced by bronchoalveolar lavage fluid from patients with systemic sclerosis. Clin Sci (London) 1994; 86: 141–148. 11 Border WA, Noble NA. Transforming growth factor beta in
tissue fibrosis. N Engl J Med 1994; 331: 1286–1292.
12 Wang Q, Hyde DM, Gotwals PJ, Giri SN. Effects of delayed treatment with transforming growth factor-beta soluble receptor in a three-dose bleomycin model of lung fibrosis in hamsters. Exp Lung Res 2002; 28: 405–417.
13 Goldstein RH, Polgar P. The effect and interaction of bradykinin and prostaglandins on protein and collagen production by lung fibroblasts. J Biol Chem 1982; 257: 8630–8633.
14 McAnulty RJ, Hernandez-Rodriguez NA, Mutsaers SE et al. Indomethacin suppresses the anti-proliferative effects of transforming growth factor-beta isoforms on fibroblast cell cultures. Biochem J 1997; 321 (Part 3): 639–643.
15 Keerthisingam CB, Jenkins RG, Harrison NK et al. Cyclo-oxygenase-2 deficiency results in a loss of the anti-proliferative response to transforming growth factor-beta in human fibrotic lung fibroblasts and promotes bleomycin-induced pulmonary fibrosis in mice. Am J Pathol 2001; 158: 1411–1422.
16 Lama V, Moore BB, Christensen P et al. Prostaglandin E2 synthesis and suppression of fibroblast proliferation by alveolar epithelial cells is cyclooxygenase-2-dependent. Am J Resp Cell Mol Biol 2002; 27: 752–758.
17 Holoye PY, Jenkins DE, Greenberg SD. Pulmonary toxicity in long-term administration of BCNU. Cancer Treat Rep 1976; 60: 1691–1694.
18Jones RB, Matthes S, Shpall EJ et al. Acute lung injury
following treatment with high-dose cyclophosphamide,
cisplatin, and carmustine: pharmacodynamic evaluation of carmustine. J Natl Cancer Inst 1993; 85: 640–647.
19 The American Thoracic Society and the European Respiratory Society: idiopathic pulmonary fibrosis: diagnosis and treat-ment. International consensus statetreat-ment. Am J Respir Crit Care Med 2000; 161 (2 Part 1): 646–664.
20 Homma S, Nagaoka I, Abe H et al. Localization of platelet-derived growth factor and insulin-like growth factor I in the fibrotic lung. Am J Respir Crit Care Med 1995; 152 (6 Part 1): 2084–2089.
21 Fehrenbach H. Alveolar epithelial type II cell: defender of the alveolus revisited. Resp Res 2001; 2: 33–46.
22 Wen FQ, Kohyama T, Skold CM et al. Glucocorticoids modulate TGF-beta production. Inflammation 2002; 26: 279–290.
23 Newton R, Seybold J, Kuitert LM et al. Repression of cyclooxygenase-2 and prostaglandin E2 release by dexametha-sone occurs by transcriptional and post-transcriptional me-chanisms involving loss of polyadenylated mRNA. J Biol Chem 1998; 273: 32312–32321.
24 Wilczynski SW, Erasmus JJ, Petros WP et al. Delayed pulmonary toxicity syndrome following high-dose chemother-apy and bone marrow transplantation for breast cancer. Am J Resp Crit Care Med 1998; 157: 565–573.
25 Folz RJ. Mechanisms of lung injury after bone marrow transplantation. Am J Resp Cell Mol Biol 1999; 20: 1097–1099. 26 Gurujeyalakshmi G, Hollinger MA, Giri SN. Pirfenidone inhibits PDGF isoforms in bleomycin hamster model of lung fibrosis at the translational level. Am J Physiol 1999; 276 (2 Part 1): L311–8.
