Granulocyte colony-stimulating factor exacerbates the acute lung injury and pulmonary fibrosis induced by intratracheal administration of bleomycin in rats

Exp Toxic Pathol 2002; 53: 501–510 URBAN & FISCHER http://www.urbanfischer.de/journals/exptoxpath 1

Toxicology Laboratory, Chugai Pharmaceutical Co., Ltd., Shizuoka, Japan Department of Veterinary Pathology, Graduate School of Agricultural and Life Sciences, The University of Tokyo, Japan 3 SLA Research Inc., Tokyo, Japan 4 Prefectural Aichi Hospital, Okazaki, Japan 2

Granulocyte colony-stimulating factor exacerbates the acute lung injury and pulmonary fibrosis induced by intratracheal administration of bleomycin in rats KENJI ADACHI1,2, MASAMI SUZUKI1, TETSUROU SUGIMOTO1, SHIGEO SUZUKI1, RIKIO NIKI3, ATSUSHI OYAMA4, KOJI UETSUKA2, HIROYUKI NAKAMAYA2, and KUNIO DOI2 With 6 figures and 1 table Received: August 29, 2001; Revised: November 2, 2001; Accepted: November 12, 2001 Address for correspondence: Dr. KENJI ADACHI, Department of Veterinary Pathology, Graduate School of Agricultural and Life Sciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan; Tel and Fax: +81-3-5841-8185, e-mail: [email protected] Key words: Granulocyte colony-stimulating factor; bleomycin; diffuse alveolar damage; pulmonary fibrosis; neutrophil.

Summary

Introduction

We investigated the effects of granulocyte colony-stimulating factor (G-CSF) on lung injury induced by intratracheal administration of bleomycin (BLM, 2 mg/200 µl) in rats. In experiment 1, G-CSF (10, 30 and 100 µg/kg/day, s.c.) was administered to rats treated with BLM or saline for 7 days starting immediately after BLM administration. In rats receiving G-CSF alone, a large number of neutrophils were noted in the pulmonary capillaries, although there were no lung lesions. In rats receiving BLM alone, diffuse alveolar damage was observed. The administration of G-CSF to BLM-treated rats increased the total lung lesion per unit of pulmonary parenchyma (total lung lesion %) along with increases in the peripheral neutrophil count and the number of neutrophils infiltrating in the pulmonary lesion in a dose-dependent fashion. In experiment 2, 100 µg/kg/day of G-CSF was administered to rats treated with BLM or saline for up to 28 days starting immediately after BLM administration. The administration of 100 µg/kg/day of G-CSF to BLM-treated rats showed no effects at 14 days but it increased the lung lesion % and the score of lung fibrosis along with the increase in the number of neutrophils infiltrating in the pulmonary lesion at 28 days. These findings suggest that G-CSF administration to BLM-treated rats influenced and exacerbated the BLM-induced acute lung injury, and also exacerbated pulmonary fibrosis in a dose-dependent fashion. The exacerbation of lung injury coincided with the marked increase in the peripheral neutrophil count and the number of neutrophils infiltrating in the pulmonary lesion.

Granulocyte colony-stimulating factor (G-CSF) is a hematopoietic growth factor that enhances the differentiation and proliferation of neutrophil progenitor cells (DEMETRI and GRIFFIN 1991). Furthermore, this factor enhances neutrophil mobilization from the bone marrow to peripheral blood and neutrophil functions, including superoxide production, chemotaxis and phagocytosis (WELT et al. 1987). G-CSF has been used clinically to treat granulocytopenia secondary to cancer chemotherapy and bone marrow transplantation. It is effective in reducing the occurrence of fever and infection associated with granulocytopenia and reduces hospital stay for patients with cancer (BRONCHUD et al. 1987; MORSTYN et al. 1988; GABRILOVE et al. 1988a; GABRILOVE et al. 1988b). Furthermore, the use of GCSF has made possible the delivery of high-dose chemotherapy. Although G-CSF is generally known to have few side effects, several studies have reported that G-CSF caused interstitial pneumonia and pulmonary fibrosis in rare cases. In 1993, IKI et al. initially reported that G-CSF treatment following the administration of pneumotoxic agents such as bleomycin (BLM) might induce interstitial pneumonia in patients with malignant lymphoma. Thereafter, some studies indicated that G-CSF might 0940-2993/02/53/06-501 $ 15.00/0

501

potentiate the lung toxicity of pneumotoxic agents (KATOH et al. 1993; MATTHEWS 1993; NIITSU et al. 1997). DIRIX et al. (1994) evaluated a patient with ovarian cancer, and speculated that a rapid increase in the neutrophil count and neutrophil activation related to G-CSF administration had exacerbated BLM-induced lung injuries. Furthermore, G-CSF can exert an untoward effect on the lung even in the absence of known pneumotoxic drugs (EISENBEIS et al. 2001). In contrast to the studies quoted previously, BASTION et al. (1994a) conducted 2 randomized comparative studies and examined the effects of combination therapy using BLM and G-CSF in patients with non-Hodgkin’s lymphoma. The results indicated no significant differences in the number of patients with pulmonary toxicity between the G-CSF-treated and placebo groups. Furthermore, some studies have reported that G-CSF administration improved the respiratory function in patients with leukemia complicated by adult respiratory distress syndrome (ARDS) (HEYLL et al. 1991). Several studies have investigated the effects of pretreatment with G-CSF on the endotoxin-induced acute lung injury model. Lung injury was not exacerbated in the model receiving intravenous administration of endotoxin (KANAZAWA et al. 1992; KING et al. 1995; KOIZUMI et al. 1993; KOIZUMI et al. 1997; INANO et al. 1998), but lung injury was exacerbated in the model receiving intratracheal administration of endotoxin (TERASHIMA et al. 1994). Thus, the relationship between G-CSF and lung injury has not been clearly explained. Furthermore, there are few data focused on the proliferating and fibrotic stages of lung injury following GCSF treatment in lung injury models. Therefore, additional investigations into the mechanisms of G-CSF-induced lung toxicity are needed (BASTION et al. 1994a; BASTION et al. 1994b; SAXMAN et al. 1997; EISENBEIS et al. 2001). In this study, we investigated the effects of G-CSF on lung injury induced by a single intratracheal administration of BLM in non-neutropenic rats in order to clarify whether G-CSF exacerbates BLM-induced lung injury. The lungs of rats treated with BLM had diffuse alveolar damage (DAD) and subsequent pulmonary fibrosis (FUKUDA et al. 1987; LAZENBY et al. 1990; LINDENSCHMIDT et al. 1986; THRALL et al. 1979). Clinically, the administration period of G-CSF is commonly 5 to 7 days after the reduction of peripheral neutrophil count induced by anticancer chemotherapy. DIRIX et al. (1994) speculated that a rapid increase in the neutrophil count and neutrophil activation related to GCSF administration exacerbated BLM-induced lung injury. Therefore, in experiment 1, we examined the effects of short-term (7-day) administration of G-CSF on the BLM-induced acute lung injury. In addition, in experiment 2, we examined the effects of longer term and highdose administration of G-CSF on the BLM-induced pulmonary fibrosis. 502

Exp Toxic Pathol 53 (2002) 6

Material and methods Animals: Ten-week-old male Sprague-Dawly rats (Japan SLC Inc., Shizuoka) were used (body weight range: 324.8 to 376.0 g). The animals were housed in cages (3 to 4 animals/cage) in an animal room maintained at a temperature of 23 ± 2 °C and a humidity of 55 ± 10% with 14 to 16 air changes per hour, and a 14-hr light and 10-hr dark cycle. They were fed pelleted chow (CE-2; Clea Japan Inc., Tokyo) and tap water ad libitum. Preparation of BLM-induced lung injury model: BLM (Nippon Kayaku Co., Tokyo, Japan) was dissolved in saline to a concentration of 10 µg/ml. To prepare the BLMinduced lung injury model, the cervical ventral median line of rats was incised under ether anesthesia, and then 2 mg of BLM in 200 µl of saline was intratracheally administered. As a control agent, saline was used. Administration of G-CSF: Recombinant human GCSF (freeze-dried preparation, 100 µg/vial) manufactured by Chugai Pharmaceutical Co., Ltd. (Tokyo, Japan) was used. This agent was dissolved in 1 ml of distilled water for the 100 µg/kg group. In addition, it was also dissolved in saline to prepare 10 and 30 µg/ml solution for the 10 and 30 µg/kg group, respectively. G-CSF was subcutaneously injected into the back of rats once a day. As a control vehicle, the base solution, which consists of phosphate buffered saline containing 0.1% human serum albumin (Green Cross Co., Osaka, Japan), 5% D-Mannitol (Nikko Chemicals Co., Ltd., Tokyo, Japan) and 0.01% Tween 20 (Nikko Chemicals Co., Ltd., Tokyo, Japan), was administered.

Fig. 1. Study design and experimental protocol of experiment 1 (a) and experiment 2 (b). Open arrow, a single intratracheal administration of saline; closed arrow, a single intratracheal administration of BLM; open column, subcutaneous administration of the vehicle; light gray column, subcutaneous administration of 10 µg/kg/day of G-CSF; dark gray column, subcutaneous administration of 30 µg/ kg/day of G-CSF; closed column, subcutaneous administration of 100 µg/kg/day of G-CSF; closed triangle, times of sacrifice; BLM, bleomycin.

Experimental design: In experiment 1, as shown in figure1a, three treatment groups were administered with 10, 30 and 100 µg/kg/day of G-CSF, respectively, for 7 days starting immediately after BLM administration (BLM + 10, 30 and 100 µg groups). Figure 1a also shows corresponding control groups and necropsy was performed at 7 days after BLM administration. In experiment 2, as shown in figure 1b, one group was administered with 100 µg/kg/day of GCSF for up to 28 days starting immediately after BLM administration (BLM + 100 µg group), and necropsy was performed at 14 and 28 days after BLM administration. Figure 1b also shows the corresponding control groups. Eight animals were assigned to each BLM-treated group, while 6 animals to each group not treated with BLM. All animals were weighed twice a week throughout the experimental period. Blood was collected from each animal through the dorsal metatarsal vein the day after the final administration of G-CSF. The leukocyte count (electrical resistance detection method, DC detection method) was measured using a multiparameter auto-cell counter (CC-780: Toa Medical Electronic Co., Ltd. Hyogo, Japan). Using blood smears stained by Wright’s stain, the differential blood count was determined by calculating the number of neutrophils in peripheral blood. After blood collection, the animals were sacrificed by exsanguination under ether anesthesia, and necropsied. Pathological examination: The lungs were removed from each animal and weighed. Relative lung weight was calculated by dividing absolute lung weight by body weight and expressed as mg/100 g body weight. The lung was perfused via the trachea with 10% neutral buffered formalin at a water pressure of 20 cm and immersed in the same fixative. Specimens were collected from the right posterior lobe and from the middle region of the left lung (total 2 specimens/rat). According to the standard method, these specimens were paraffin-embedded and cut into 3 µm sections. Thereafter, these sections were stained with hematoxylin and eosin (HE). Azan’s stain was used on some sections. Histopathological evaluations including determination of the number of neutrophils in the pulmonary lesion per unit of pulmonary parenchyma, the total lung lesion per unit of pulmonary parenchyma (total lung lesion %), and the score of lung fibrosis. The number of neutrophils in the pulmonary lesion was counted as follows. The HE-stained sections were viewed using a microscope at a total magnification of ×400. A 10 × 10 squared grid eyepiece was used to designate the sample field (0.0625 mm2). The total number of neutrophils in the pulmonary lesion including pulmonary capillaries, interstitium, and alveoli was recorded from 10 randomly chosen microscope fields of a typical lung lesion on each lung slide (20 fields/rat). The average number of neutrophils per field was divided by a conversion factor of 0.0625 mm2, and the number of neutrophils in the pulmonary lesion was calculated. The square of the graticule was always placed over an area that only contained a lung lesion. Measurement of the total lung lesion % was based on the method described by MAUTZ et al. (1988). The HE-stained sections were viewed using a microscope at a total magnification of ×100. A 10 × 10 squared grid eyepiece was used to designate the sample field (1 mm2), and 10 randomly chosen microscope fields of each specimen (20 fields/animal) were scored. Within every field, each of the 100 boxes

(“points”) were scored as one of the followings: 0 indicates nonparenchyma; 1 indicates alveolar septa and alveolar ducts of normal septal thickness and cellularity; 2 indicates slightly to moderately injured alveolar tissue with increased cellularity and thickness of the alveolar wall but no disruption of normal alveolar shape; 3 indicates markedly to severely injured alveolar tissue with clear distortion of normal alveolar architecture and a more marked increase in septal cellularity and thickness than in 2. The total lung lesion % was determined by dividing the number of points by the total number of points that were classified as parenchyma. Score of lung fibrosis was assessed semi-quantitatively according to the method of ASHCROFT et al. (1988). Briefly, lung fibrosis was scored on a scale from 0 to 8 by examining all parenchyma in a lung section (2 sections/animal) at a magnification of ×100. A 10 × 10 squared grid eyepiece (1 mm2) was used to designate the sample field. Criteria for grading lung fibrosis were as follows: grade 0, normal lung; grade 1, minimal fibrous thickening of alveolar or bronchiolar walls; grade 3, moderate thickening of walls without obvious damage to lung architecture; grade 5, increased fibrosis with definite damage to lung structure and formation of fibrous bands or small fibrous masses; grade 7, severe distortion of structure and large fibrous areas; “honeycomb lung” is placed in this category; grade 8, total fibrous obliteration of lung. If there was any difficulty in deciding between two odd-numbered categories, the field would be given the intervening even-numbered grade. The score of lung fibrosis was expressed as a mean score of fibrosis for each animal. Whether G-CSF exacerbated lung injury was judged from the lung lesion % and the score of lung fibrosis. The dorsal subcutaneous sites, where G-CSF or vehicle was administered, were also stained with HE and examined histopathologically. Statistical analysis: Results of survival rate, body weight, peripheral neutrophil count, absolute and relative lung weight and histopathological factors were analyzed using the statistical software package, Statcel (Excel 2000 for Windows; Seiun-shya, Tokyo, Japan). Each value was expressed as the mean ± standard deviation (SD) except for survival rate. Significant differences in the survival rate between BLM-treated groups were determined using the analysis of Logrank test followed by Kaplan Meier method. In experiment 1 and 2, significant differences between the Saline + 0 µg group and the Saline + 100 µg group or the BLM + 0 µg group were sought using the analysis of Student’s t-test for isovariance, Welch’s t-test for unisovariance, or Mann-Whitney’s U test for abnormal distribution. In experiment 1, dose-response relationship was analyzed by simple regression analysis in the BLM-treated groups. If a dose-response relationship was observed, significant differences between the BLM + 0 µg and the BLM + G-CSF groups were sought using the analysis of Student’s t-test for isovariance, Welch’s t-test for unisovariance, or MannWhitney’s U test for abnormal distribution. In experiment 2, significant differences between the BLM + 0 µg and the BLM + 100 µg groups were sought using the analysis of Student’s t-test for isovariance, Welch’s t-test for unisovariance, or Mann-Whitney’s U test for abnormal distribution. Group differences with p values less than 0.05 were considered significant. Exp Toxic Pathol 53 (2002) 6

503

Results Experiment 1: The effects of short-term administration of G-CSF on the BLM-induced acute lung injury Survival rate and body weight changes: As shown in table 1, there was one death in all the BLM + G-CSF groups. In the BLM + 0 µg group, weight loss or reduced weight gain was noted between 3 and 7 days after BLM administration. During the study period, body weight decreased compared with the Saline + 0 µg group. In the BLM-treated groups, the body weight decreased in dosedependent fashion (y = –0.37 x + 318.72, p < 0.05). However, there were no differences in the body weight among the BLM-treated groups. In the Saline + 0 and the Saline + 100 µg groups, changes in body weight were similar to those in the untreated group. Peripheral blood neutrophil count: As shown in figure 2a, the neutrophil count markedly increased in the G-

CSF-treated groups. Furthermore, in the BLM-treated groups, the neutrophil count increased in a dose-dependent fashion (y = 1.1 x + 20.4, p < 0.01). When values in the BLM + 0 µg group were regarded as 100%, the percentages in the BLM + 10, 30 and 100 µg groups were 277%, 356% and 876%, respectively. There were no changes in the neutrophil count in the BLM + 0 µg group and the Saline + 0 µg group. Lung weight: As shown in figure 3a, the absolute and relative lung weights increased notably in the BLM + 0 µg group. However, no dose-response relationship was found among the BLM-treated groups. In the Saline + 100 µg GCSF group, the absolute and relative lung weights slightly increased compared with the Saline + 0 µg group. In the Saline + 0 µg group, there were no abnormal changes. Histopathological evaluation of the lungs: In the Saline + 0 µg group, there were no abnormalities in the lung (fig. 4a). In the Saline + 100 µg group, an increased number of neutrophils was noted in the pulmonary capillaries, although there were no lung lesions (fig. 4b).

Table 1. Survival rate and body weight change in rats treated with or without granulocyte colony-stimulating factor for 7, 14 or 28 days following a single intratracheal administration of saline or bleomycin. Experiment 1

Experiment 2

Day 0a

Day 7

Day 0

Day 7

Day 14

Day 28

Survival rateb Untreated Saline + 0 µg Saline + 100 µg BLM + 0 µg BLM + 10 µg BLM + 30 µg BLM + 100 µg

100.0 (6/6) 100.0 (6/6) 100.0 (6/6) 100.0 (8/8) 100.0 (8/8) 100.0 (8/8) 100.0 (8/8)

100.0 (6/6) 100.0 (6/6) 100.0 (6/6) 100.0 (8/8) 87.5 (7/8) 87.5 (7/8) 87.5 (7/8)

100.0 (12/12) 100.0 (12/12) 100.0 (12/12) 100.0 (16/16) NE NE 100.0 (16/16)

100.0 (12/12) 100.0 (12/12) 100.0 (12/12) 100.0 (16/16) NE NE 100.0 (16/16)

100.0 (12/12) 100.0 (12/12) 100.0 (12/12) 93.8 (15/16) NE NE 87.5 (14/16)

100.0 (6/6) 100.0 (6/6) 100.0 (6/6) 82.0 (7/8) NE NE 87.5(7/7)

Body weightc Untreated Saline + 0 µg Saline + 100 µg BLM + 0 µg BLM + 10 µg BLM + 30 µg BLM + 100 µg

346.6 ± 10.4 341.4 ± 6.2 343.1 ± 8.7 347.0 ± 9.4 344.5 ± 7.5 347.5 ± 13.2 344.9 ± 10.1

372.9 ± 10.8 364.2 ± 5.8 370.6 ± 13.3 321.2 ± 31.0&,$$ 323.1 ± 44.3& 293.3 ± 19.3& 285.1 ± 41.8&

346.4 ± 12.6 343.5 ± 7.3 343.2 ± 8.0 345.7 ± 12.0 NE NE 347.0 ± 12.3

373.9 ± 13.6 368.8 ± 9.6 366.9 ± 9.7 299.1 ± 41.2$$ NE NE 275.9 ± 29.0

401.5 ± 16.5 394.5 ± 9.3 390.6 ± 17.5 312.5 ± 77.8$$ NE NE 273.3 ± 65.1

436.9 ± 21.9 436.7 ± 14.6 443.0 ± 16.3 381.1 ± 86.2 NE NE 305.6 ± 97.2

Definition of abbreviations: BLM = bleomycin; 0 µg = subcutaneous administration of the vehicle; 10 µg = subcutaneous administration of 10 µg/kg/day of granulocyte colony-stimulating factor (G-CSF); 30 µg = subcutaneous administration of 30 µg/kg/day of G-CSF; 100 µg = subcutaneous administration of 100 µg/kg/day of G-CSF; NE = not examined. a Days after the admnistration of saline or BLM. b Each value represents the survival rate (%) and parenthesized number represents the number of survived animal/number of examined animal. #p < 0.05, ##p < 0.01, comparing between the BLM-treated groups (Logrank test followed by Kaplan Meier method). c Each value represents the mean ± standard deviation (n = 6 in BLM-untreated groups, n = 7–8 in BLM-treated groups). *p < 0.05, **p < 0.01, comparing the Saline + 0 µg group and the Saline + 100 µg group (Student's t-test). &p < 0.05, &&p < 0.01, dose-response relationship in BLM-treated groups (simple regression analysis). #p < 0.05, ##p < 0.01, comparing the BLM + 0 µg group and the BLM + G-CSF groups (Student's t-test or Mann-Whitney’s U test for abnormal distribution). $ p < 0.05, $$p < 0.01, comparing the Saline + 0 µg group and the BLM + 0 µg groups (Student's t-test or Mann-Whitney’s U test for abnormal distribution). 504

Exp Toxic Pathol 53 (2002) 6

Fig. 2. Neutrophil counts in peripheral blood of untreated, saline- or BLM-treated rats with or without G-CSF for 7, 14 or 28 days in experiment 1 (a) and experiment 2 (b). Each value represents the mean ± standard deviation (SD) (n = 6 in BLMuntreated groups, n = 7–8 in BLM-treated groups). *p < 0.05, **p < 0.01, comparing the Saline + 0 µg group and the Saline + 100 µg group (Welch’s t-test). &p < 0.05, && p < 0.01, dose-response relationship in BLM-treated groups (simple regression analysis). #p < 0.05, ##p < 0.01, comparing the BLM + 0 µg group and the BLM + GCSF groups (Welch’s t-test or MannWhitney’s U test for abnormal distribution). $p < 0.05, $$p < 0.01, comparing the Saline + 0 µg and BLM + 0 µg (Student’s t-test for isovariance, Welch’s t-test for unisovariance). n.s.: Not significant.

Fig 3. Lung weight of untreated, salineor BLM-treated rats with or without GCSF for 7, 14 or 28 days in experiment 1 (a) and experiment 2 (b). Each value represents the mean ± SD (n = 6 in BLM-untreated groups, n = 7–8 in BLM-treated groups). *p < 0.05, **p < 0.01, comparing the Saline + 0 µg group and the Saline + 100 µg group (Student’s t-test for isovariance or Welch’s t-test for unisovariance). & p < 0.05, &&p < 0.01, dose-response relationship in BLM-treated groups (simple regression analysis). #p < 0.05, ##p < 0.01, comparing the BLM + 0 µg group and the BLM + G-CSF groups (Student’s t-test for isovariance or Mann-Whitney’s U test for abnormal distribution). $p < 0.05, $$p < 0.01, comparing the Saline + 0 µg and BLM + 0 µg (Welch’s t-test for unisovariance or Mann-Whitney’s U test for abnormal distribution). n.s.: Not significant. Exp Toxic Pathol 53 (2002) 6

505

Fig. 4. Lung sections of rats treated with 100 µg/kg/day of G-CSF (b, d) or vehicle (a, c) for 7 days after intratracheal administration of BLM (c, d) or saline (a, b). (a) Saline + 0 µg group. There are no abnormalities in the lung. HE. ×370. (b) Saline + 100 µg group. There is an increase in the number of neutrophils found in the pulmonary capillaries, but there are no lung lesions. HE. ×370. (c) BLM + 0 µg group. Swelling, degeneration and desquamation of the alveolar epithelium, infiltration of inflammatory cells composed mainly of neutrophils and macrophages are noted. HE. ×370. (d) BLM + 100 µg group. Enhancement of neutrophil infiltration in the pulmonary lesion is observed in addition to the aforementioned findings of BLM + 0 µg group. HE. ×370.

In the BLM + 0 µg group, DAD was observed. Degeneration and desquamation of alveolar epithelia, infiltration of inflammatory cells composed mainly of neutrophils and macrophages (fig. 4c), migration of myofibroblasts, hemorrhage and edema in alveolar spaces, degeneration and desquamation of bronchial epithelia were noted. In the alveolar duct, bronchiolization was detected. In addition to these lesions, very slight pulmonary fibrosis with proliferation of interstitial cells and deposition of fibrous tissues was observed in some animals. In the BLM + G-CSF groups, more neutrophils were observed in pulmonary lesion than in the lesion caused by BLM alone (fig. 4d). The comparison of the numbers of neutrophils in the pulmonary lesion, the total lung lesion % and the scores of lung fibrosis between the BLM + 0 µg group and the 506

Exp Toxic Pathol 53 (2002) 6

BLM + G-CSF groups are shown in figure 5a. In the BLM + G-CSF groups, the numbers of infiltrating neutrophils and the total lung lesion % increased in a dosedependent fashion (the number of neutrophils: y = 0.65 x + 12.45, p < 0.01; the total lung lesion %: y = 0.32 x + 64.80, p < 0.01). Furthermore, the number of neutrophils in the BLM + G-CSF groups and the total lung lesion % in the BLM + 100 µg group increased compared with the BLM + 0 µg group. Histopathological findings of the dorsal subcutis: Although slight hemorrhage, infiltration of inflammatory cells and formation of granulation tissue were detected in the dorsal subcutis where G-CSF or vehicle was administered, there were no differences between the G-CSFtreated and the vehicle-treated groups.

Experiment 2: The effects of 28-day and high-dose administration of G-CSF on the BLM-induced pulmonary fibrosis Survival rate and body weight changes: As shown in table 1, there were one or two deaths in all the BLMtreated groups at days 14 and 28. However, there were no differences in the survival rate among the BLM-treated groups. In the BLM + 0 µg group, weight loss or reduced weight gain was noted between 3 and 17 days after BLM administration. Compared with the BLM + 0 µg group, there were no differences in the body weight between the BLM-treated groups. In the Saline + 0 and the Saline + 100 µg groups, changes in body weight were similar to those in the untreated group. Peripheral blood neutrophil count: As shown in figure 2b, the neutrophil count markedly increased in the GCSF-treated groups. When values in the BLM + 0 µg group at each time point were regarded as 100%, the percentages in the BLM + 100 µg group were as follows: 755% at 14 days after BLM administration, and 2,027% at 28 days after BLM administration, respectively. In the BLM + 0 µg group, the neutrophil count slightly decreased at 28 days after BLM administration. In the Saline + 0 µg group, there were no changes in the neutrophil count. Lung weight: As shown in figure 3b, the absolute and relative lung weights increased notably in all the BLMtreated groups at 14 and 28 days after BLM administration. However, there were no significant differences between the BLM + 0 µg group and the BLM + 100 µg group at 14 and 28 days. In the Saline + 100 µg group, the absolute and relative lung weights slightly increased compared with the Saline + 0 µg group at 14 days. In the Saline + 0 µg group, there were no abnormal changes at 14 and 28 days. Fig. 5. Histopathological evaluation of the number of neutrophils in the pulmonary lesion, the lung lesion % per pulmonary parenchyma and the score of lung fibrosis for each BLM-treated rat with or without G-CSF for 7, 14 or 28 days in experiment 1 (a) and 2 (b). Each value represents the mean ± SD (n = 7–8). &p < 0.05, &&p < 0.01, dose-response relationship in BLM-treated groups (simple regression analysis). #p < 0.05, ##p < 0.01, comparing the BLM + 0 µg group and the BLM + G-CSF groups (Student’s t-test for isovariance, Welch’s t-test for unisovariance or Mann-Whitney’s U test for abnormal distribution). n.s.: Not significant.

Histopathological evaluation of the lungs: In the Saline + 0 µg group, there were no abnormalities in the lung. In the Saline + 100 µg group, an increased number of neutrophils was observed in the pulmonary capillaries, although there were no lung lesions. In the BLM + 0 µg group, DAD and subsequent pulmonary fibrosis were observed. At 14 days after BLM administration, edema in the alveolar space, infiltration of inflammatory cells composed mainly of neutrophils, eosinophils and macrophages, proliferation of myofibroblasts, hemorrhage and marked fibrosis were noted. The fibrosis and inflammation occurred mainly around the bonchi and bronchioles and near the pleura. Pulmonary alveoli adjacent to the severely fibrotic areas showed dilation and distortion. In some animals, honeycomb lung formation was also observed (fig. 6). At 28 days after BLM administration, inflammatory cell infilExp Toxic Pathol 53 (2002) 6

507

Fig. 6. Lung sections of rats treated with vehicle for 28 days after intratracheal administration of BLM. Severe distortion of structure and large fibrosis are observed. The pulmonary parenchyma adjacent to the severe fibrotic lesion has the typical appearance of “honeycomb lung”. HE. ×70.

tration in the alveolar space and proliferation of myofibroblasts were reduced, and fibrotic lesions were mainly observed. In the BLM + 100 µg group, increased numbers of neutrophils were observed than in the BLM + 0 µg group. The comparison of the numbers of neutrophils in the pulmonary lesion, the total lung lesion % and the scores of lung fibrosis between the BLM + 0 µg group and the BLM + 100 µg group are shown in figure 5b. In the BLM + 100 µg group, compared with the BLM + 0 µg group, the total lung lesion % and the score of lung fibrosis increased at 28 days after BLM administration along with the increase in the number of neutrophils in the pulmonary lesion. On the other hand, there were no statistical differences in the total lung lesion % and the score of lung fibrosis between the BLM + 100 µg group and the BLM + 0 µg group at 14 days after BLM administration. Histopathological findings of the dorsal subcutis: As in experiment 1, there were no histopathological differences between G-CSF-treated and vehicle-treated groups in the dorsal subcutis where G-CSF or vehicle was administered.

Discussion In the present study, we examined the relationship between G-CSF administration and lung injury using a common lung injury model induced by intratracheal administration of BLM. It is known that BLM damages the alveolar capillary endothelium, alveolar epithelium and epithelial basement membrane, and causes inflammation 508

Exp Toxic Pathol 53 (2002) 6

(KAWAMOTO and FUKUDA 1990; LAZENBY et al. 1990; USUKI and FUKUDA 1995). Marked damage in the epithelial basement membrane may subsequently cause pulmonary fibrosis (KAWAMOTO and FUKUDA 1990; LAZENBY et al. 1990; USUKI and FUKUDA 1995). BLM-treated animals in the present study showed the lung histopathology such as DAD and subsequent pulmonary fibrosis which are consistent with those previously reported (CHANDLER et al. 1983; THRALL et al. 1979; ZAPOL et al. 1979). In experiment 1, the effects of G-CSF on the BLMinduced acute lung injury were examined. Ten, 30 and 100 µg/kg/day were employed as doses of G-CSF. These doses corresponded to about 2–20 fold of the clinical dose in humans. As a result, a dose-dependent increase in the peripheral blood neutrophil count was observed in the BLM + G-CSF-treated groups, and this may be related to the significant pharmacological activity of G-CSF (DEMETRI GD and GRIFFIN JD 1991). The total lung lesion % in the BLM-treated groups increased in a dosedependent fashion, and the administration of 100 µg/ kg/day of G-CSF for 7 days to BLM-treated rats increased the total lung lesion % along with the increase in the number of neutrophils in the pulmonary lesion. This suggests that a short-term administration of G-CSF increased the severity of DAD in the lung of BLM-rats in a dose-dependent fashion, and the exacerbation of lung injury coincided with the marked increase in the peripheral neutrophil count and in the number of neutrophils in the pulmonary lesion. INANO et al. (1998) reported that an intravenous injection of G-CSF caused a rapid neutropenia and neutrophil sequestration within the microvasculature of the lung but did not induce neutrophil emigration or albumin leakage into the alveolar space in rabbits. They concluded that an injection of G-CSF did not induce lung injury. In the Saline + 100 µg group in the present study, a large number of neutrophils were detected in the alveolar capillaries along with a marked increase in the peripheral neutrophil count at 7, 14 and 28 days. However, lung injury was not detected by histopathological observation. This indicates that high-dose administration of G-CSF for up to 28 days did not induce injury in normal lung tissue. In experiment 2, the effects of high-dose and 28-day administration of G-CSF on the pulmonary fibrosis in BLM-rats were examined. The administration of 100 µg/ kg/day of G-CSF to BLM-rats for 28 days increased the total lung lesion % and the score of lung fibrosis along with the increase in the number of neutrophils in the pulmonary lesion. This suggests that high-dose and 28-day administration of G-CSF increased the severity of lung fibrosis in BLM-rats. It is known that G-CSF enhances the differentiation and proliferation of neutrophil progenitor cells and the function of mature neutrophils (WELT et al. 1987). Neutrophil activation is important for protecting the body through phagocytosis of foreign substances. However, in some cases, this factor may enhance inflammatory reac-

tion and injure tissues (HOGG 1987). OGINO et al. (1996) have examined the actions of G-CSF on the inflammatory reaction using the carrageenin-induced pleuritis model. Their results have shown that the increased neutrophil chemotaxis from peripheral blood to local inflammatory lesions, as well as plasma exudation, resulted in exacerbation of inflammatory reaction when G-CSF administration increased the peripheral blood neutrophil count above the normal value. In the present study, the BLM + G-CSF groups enhanced the number of infiltrating neutrophils in association with a marked increase in the peripheral neutrophil count. Therefore, enhanced neutrophil infiltration into the local inflammatory lesion may be involved in exacerbation of lung injury. In conclusion, G-CSF administration to BLM-treated rats influenced and exacerbated the BLM-induced acute lung injury as well as pulmonary fibrosis in a dose-dependent fashion. The exacerbation of lung injury coincided with the marked increase in the peripheral neutrophil count and the number of neutrophils in the pulmonary lesion. Acknowledgments: We thank Ms. YUKIKO TABATA, Ms. YUMIE OGAWA and Ms. NORIKO NOGUCHI at Chugai Pharmaceutical Co., Ltd. for their skillful technical assistance.

References ASHCROFT T, SIMPSON JM, TIMBRELL V: Simple method of estimating severity of pulmonary fibrosis on a numerical scale. J Clin Pathol 1988; 41: 467–470. BASTION Y, REYES F, BOSLY A, et al.: Possible toxicity with the association of G-CSF and bleomycin. Lancet 1994; 343: 1221–1222. BASTION Y, COIFFIER B: Pulmonary toxicity of bleomycin: is G-CSF a risk factor? Lancet 1994; 344: 474. BRONCHUD MH, SCARFFE JH, THATCHER N, et al.: Phase I/II study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer. Br J Cancer 1987; 56: 809–813. CHANDLER DB, HYDE DM, GIRI SN: Morphometric estimates of infiltrative cellular changes during the development of bleomycin-induced pulmonary fibrosis in hamsters. Am J Pathol 1983; 112: 170–177. DEMETRI GD, GRIFFIN JD: Granulocyte colony-stimulating factor and its receptor. Blood 1991; 78: 2791–2808. DIRIX LY, SCHRIJVERS D, DRUWÈ P, et al.: Pulmonary toxicity and bleomycin. Lancet 1994; 344: 56. EISENBEIS CF, WINN D, POELMAN S, et al.: A case of pulmonary toxicity associated with G-CSF and doxorubicin administration. Ann Hematol 2001; 80: 121–123. FUKUDA Y, ISHIZAKI M, MASUDA Y, et al.: The role of intraalveolar fibrosis in the process of pulmonary structural remodeling in patients with diffuse alveolar damage. Am J Pathol 1987; 126: 171–182. GABRILOVE JL, JAKUBOWSKI A, SCHER H, et al.: Effect of granulocyte colony-stimulating factor on neutropenia

and associated morbidity due to chemotherapy for transitional-cell carcinoma of the urothelium. N Engl J Med 1988; 318: 1414–1422. GABRILOVE JL, JAKUBOWSKI A, FAIN K, et al.: Phase I study of granulocyte colony-stimulating factor in patients with transitional cell carcinoma of the urothelium. J Clin Invest 1988; 82: 1454–1461. HEYLL A, AUL C, GOGOLIN F, et al.: Granulocyte colonystimulating factor (G-CSF) treatment in a neutropenic leukemia patient with diffuse interstitial pulmonary infiltrates. Ann Hematol 1991; 63: 328–332. HOGG JC: Neutrophil kinetics and lung injury. Physiological Reviews 1987; 67: 1249–1295. IKI S, YOSHINAGA K, OHBAYASHI Y, et al.: Cytotoxic druginduced pneumonia and possible augmentation by GCSF – clinical attention. Ann Hematol 1993; 66: 217–218. INANO H, KAMEYAMA S, YASUI S, et al.: Granulocyte colony-stimulating factor induces neutrophil sequestration in rabbit lungs. Am J Respir Cell Mol 1998; 19: 167–174. KANAZAWA M, ISHIZAKA A, HASEGAWA N, et al.: Granulocyte colony-stimulating factor dose not enhance endotoxin-induced acute lung injury in guinea pigs. Am Rev Respir Dis 1992; 145: 1030–1035. KATOH M, SHIKOSHI K, TAKADA M, et al.: Development of interstitial pneumonitis during treatment with granulocyte colony-stimulating factor. Ann Hematol 1993; 67: 201–202. KAWAMOTO M, FUKUDA Y: Cell proliferation during the process of bleomycin-induced pulmonary fibrosis in rats. Acta Pathol Jpn 1990; 40: 227–238. KING J, DEBOISBLANC BP, MASON CM, et al.: Effect of granulocyte colony-stimulating factor on acute lung injury in rat. Am J Respir Crit Care Med 1995; 151: 302–309. KOIZUMI T, KUBO K, SHINOZAKI S, et al.: Granulocyte colony-stimulating factor dose not exacerbate endotoxin-induced lung injury in sheep. Am Rev Respir Dis 1993; 148: 132–137. KOIZUMI T, KUBO K, KOYAMA S, et al.: Neutrophils pretreated with granulocyte colony-stimulating factor (GCSF) are not related to the severity of endotoxin-induced lung injury. Exp Lung Res 1997; 23: 393–404. LAZENBY AJ, CROUCH EC, MCDONALD JA, et al.: Remodeling of the lung in bleomycin-induced pulmonary fibrosis in the rat – an immunohistochemical study of laminin, type IV collagen, and fibronectin. Am Rev Respir Dis 1990; 142: 206–214. LINDENSCHMIDT RC, TRYKA AF, GODFREY GA, et al.: Intratracheal versus intravenous administration of bleomycin in mice: acute effects. Toxicol Appl Pharmacol 1986; 85: 69–77. MATTHEWS JH: Pulmonary toxicity of ABVD chemotherapy and G-CSF in Hodgkin’s diseases: possible synergy. Lancet 1993; 342: 988. MAUTZ WJ, KLEINMAN MT, PHALEN RF, et al.: Effects of exercise exposure on toxic interactions between inhaled oxidant and aldehyde air pollutants. J Toxicol Environ Health 1988; 25: 165–177. MORSTYN G, CAMPBELL L, SOUZA LM, et al.: Effect of granulocyte colony stimulating factor on neutropenia induced by cytotoxic chemotherapy. Lancet 1988; 1: 667–671. Exp Toxic Pathol 53 (2002) 6

509

NIITSU N, IKI S, MUROI K, et al.: Interstitial pneumonia in patients receiving granulocyte colony-stimulating factor during chemotherapy: survey in Japan 1991–96. Br J Cancer 1997; 76: 1661–1666. OGINO M, MAJIMA M, KAWAMURA M, et al.: Increased migration of neutrophils to granulocyte-colony stimulating factor in rat carrageenin-induced pleurisy: roles of complement, bradykinin, and inducible cyclooxygenase-2. Inflamm Res 1996; 45: 335–346. SAXMAN SB, NICHOLS CR, EINHORN LH. Pulmonary toxicity in patients with advanced-stage germ cell tumors receiving bleomycin with and without granulocyte colony stimulating factor. Chest 1997; 111: 657–660. TERASHIMA T, KANAZAWA M, SAYAMA K, et al.: Granulocyte colony-stimulating factor exacerbates acute lung injury

510

Exp Toxic Pathol 53 (2002) 6

induced by intratracheal endotoxin in guinea pigs. Am J Respir Crit Care Med 1994; 149: 1295–1303. THRALL RS, MCCORMICK JR, JACK RM, et al.: Bleomycininduced pulmonary fibrosis in the rat – inhibition by indomethacin. Am J Pathol 1979; 95: 117–130. USUKI J, FUKUDA Y: Evolution of three patterns of intraalveolar fibrosis produced by bleomycin in rats. Pathology International 1995; 45: 552–564. WELT K, BONILLA MA, GABRILOVE JL, et al.: Recombinant human granulocyte-colony stimulating factor: in vitro and in vivo effects on myelopoiesis. Blood Cells 1987; 13: 17–30. ZAPOL WM, TRELSTAD RL, COFFEY JW, et al.: Pulmonary fibrosis in severe acute respiratory failure. Am Rev Respir Dis 1979; 119: 547–554.