Sequential adjuvant chemoradiotherapy-induced diffuse alveolar haemorrhage in a patient with breast cancer successfully treated with corticosteroid plus recombinant human soluble thrombomodulin

Abstract

The effective treatment for corticosteroid (CS)-refractory sequential chemoradiotherapy (CRT)-induced lung injury has not been established. We report a case of sequential CRT-induced diffuse alveolar haemorrhage (DAH) successfully treated with CS therapy plus recombinant human soluble thrombomodulin (rhTM). A 69-year-old woman was treated with sequential adjuvant CRT for early-stage breast cancer. After sequential CRT, she suffered from progressive dyspnoea. Chest CT scan showed consolidations in the irradiation field and diffuse ground-glass attenuations in the non-irradiation regions. We suspected sequential CRT-induced DAH because of increased haemosiderin-laden macrophages in bronchoalveolar lavage fluid. Her clinical conditions did not improve with high-dose CS therapy. Therefore, rhTM was added, and her disease and serum high-mobility group box-1 levels improved rapidly. Therefore, rhTM plus CS might be a safe and effective treatment for sequential CRT-induced lung injury, although further study is necessary to validate these findings.

Background

In recent years, a multidisciplinary approach involving surgical resection, radiation therapy, and chemotherapy has improved the prognosis of patients with solid tumours. Sequential adjuvant chemoradiotherapy (CRT) has established a notable benefit for the multidisciplinary management of early-stage breast cancer.¹ However, previous studies have shown that sequential CRT caused pulmonary toxicity in 1.3% of patients with breast cancer.² Sequential CRT-induced lung injury has many clinical patterns, ranging from benign infiltrates without any clinical symptoms to life-threatening acute respiratory distress syndrome. Corticosteroid (CS) therapy is recommended for the treatment of chemotherapy and/or radiation-induced lung injury (RILI).^{3 4} However, advanced treatment options for CS-refractory CRT-induced lung injury have not been established.

A novel biological agent, recombinant human soluble thrombomodulin (rhTM) (Recomodulin, Asahi Kasei Pharma Corporation, Tokyo, Japan), has been approved for clinical use in Japan for the treatment of disseminated intravascular coagulation. This molecule has been shown to have anti-inflammatory and anticoagulation effects and is effective for the treatment of acute exacerbation of idiopathic pulmonary fibrosis (AE-IPF).^{5 6} There have been no previous reports demonstrating the effectiveness of rhTM for sequential CRT-induced lung injury; we now present a case of severe sequential CRT-induced lung injury with diffuse alveolar haemorrhage (DAH) that was successfully treated with CS therapy plus rhTM.

Case presentation

A 69-year-old woman underwent left breast-conserving surgery for adenocarcinoma of the left breast (pT1cN0M0) at our hospital. There was no axillary lymph node metastasis. However, cancer cells were detected in the surgical margin, suggesting that the tumour had not been completely resected. The patient did not have any underlying pulmonary disease, and a preoperative respiratory function test and CT scan did not show any abnormality. Postoperatively, she was referred for adjuvant radiotherapy of the left conserved breast. She underwent adjuvant whole breast radiotherapy (50 Gy, 4 MV photons) with a 10 Gy boost (9 MV electrons) in 30 fractions (figure 1). After the final irradiation, she underwent adjuvant chemotherapy with three cycles of docetaxel (75 mg/m²) with cyclophosphamide (600 mg/m²). CT scan after the first chemotherapy cycle did not detect any abnormality (figure 2), and after the second chemotherapy cycle, there were no respiratory symptoms.

Figure 1.

(A) The digitally reconstructed beam's eye view. (B) Radiotherapy dose distribution.

Figure 2.

High-resolution CT scan findings. The patient's CT imaging after the first chemotherapy cycle showing normal upper lung (A) and middle lung (B) fields. CT imaging on admission showing the upper lung (C) and middle lung (D) fields. Patchy ground-glass attenuations (black arrowheads) were detected in the whole lung fields. In addition, consolidation was confirmed in the irradiated field (black arrow).

However, 9 days after the third administration of chemotherapy (3 months after the final irradiation) the patient experienced exertional dyspnoea for the first time. Five days later, the patient was referred to our respiratory disease centre for progressive dyspnoea and dry cough without bloody sputum or chest pain. There was no evidence of purpura, neurological symptoms, limb oedema or pain caused by pressure. Fine crackling was audible in the left anterior part of the chest. The patient had fever of 37.7°C, and her peripheral oxygen saturation was 87%, under room-air conditions. Her respiratory rate at rest in the supine position was 24 breaths/min. She showed severely laboured breathing and required oxygen at 5 L/min, despite walking a few metres only. Laboratory examinations revealed that she had elevated serum lactate dehydrogenase (333 U/L), C reactive protein (11.3 mg/dL) and surfactant protein D (136 ng/mL). D-dimer (1.8 µg/mL) and fibrin degradation products (5.6 µg/mL) were slightly elevated. Her serum thrombomodulin (TM) concentration (3.0 FU/mL), platelet count, brain natriuretic peptide (BNP), sialylated carbohydrate antigen and β-D glucan levels were normal. Renal function, liver function and urinalysis were normal. Staining for serum myeloperoxidase-antineutrophil cytoplasmic antibodies, proteinase-antineutrophil cytoplasmic antibodies, antiglomerular basement membrane antibody and antinuclear antibody were negative. Her arterial oxygen and carbon dioxide pressures were 74.2 mm Hg and 31.2 mm Hg, respectively, when using oxygen at 2 L/min.

Chest radiography revealed left-side dominant diffuse ground-glass attenuations (GGAs). Chest CT scan showed consolidations in the subpleural zone of the left lower lobe, corresponding to the irradiation fields. Moreover, we detected patchy GGAs in the whole lung field (figure 2). Echocardiographic evaluation of the ventricular function and tricuspid regurgitant jet, which is associated with pulmonary hypertension, were normal. A bronchoalveolar lavage (BAL) was performed on the right middle lobe, which was outside the irradiation field, to determine the cause of the diffuse GGAs. The total cell count of the BAL fluid was 5.8×10⁵/mL, which consisted of 37.0% macrophages, 52.3% lymphocytes, 9.6% neutrophils and 1.1% eosinophils. The proportion of haemosiderin-laden macrophages was ≥20%, which fulfilled the commonly used BAL fluid criterion for DAH. In BAL fluid samples, there was no evidence of infectious pathogens or malignant cells. BAL fluid TM (<1.0 FU/mL) and D-dimer (0.3 µg/mL) levels were normal.

Differential diagnosis

We excluded a diagnosis of pulmonary artery embolism or heart failure because of the low D-dimer level, normal echocardiographic findings and normal BNP levels. We did not suspect deep venous thrombosis according to the findings of the physical examination. Infectious disease was ruled out by laboratory analysis and BAL fluid findings. Based on the auscultatory findings, elevated surfactant protein D, and CT scan findings, interstitial pulmonary disease was suggested. Drug-induced lymphocyte stimulation tests (DLST) for docetaxel and cyclophosphamide, which were performed before the administration of CS, were negative. Finally, we suspected RILI and/or drug-induced interstitial lung disease (DILD) caused by docetaxel or cyclophosphamide.

Treatment

She was administered high-dose CS pulse therapy (methylprednisolone, 1000 mg/day for 3 days). On the second day of CS therapy, D-dimer levels were re-evaluated (1.1 µg/mL), but they were unchanged. However, the patient's lung infiltrations worsened progressively (figure 3). In addition, during rest, she needed oxygen at 4 L/min, which was administered via a nasal cannula (figure 4). Therefore, we started rhTM administration at 380 IU/kg/day for 5 days, in addition to CS therapy. CS therapy was tapered after pulse therapy (1 mg/kg/day).

Figure 3.

Chest radiography findings. The changes on chest radiographs are shown. (A) Left-side dominant ground-glass attenuations (GGAs) were detected on admission. (B) Despite high-dose corticosteroid therapy, infiltration and GGAs worsened. (C) After the additional administration of recombinant human soluble thrombomodulin (rhTM), GGAs were improved producing scarring.

Figure 4.

Clinical course. First, the patient was treated with steroid pulse therapy. However, her oxygenation, symptoms, and HMGB-1 levels worsened rapidly. The administration of rhTM plus CS ameliorated her oxygenation, HMGB-1, SP-D, and CRP levels. Her haemoglobin level had also improved. CRP, C reactive protein; CS, corticosteroid; Hb, haemoglobin; HMGB-1, high-mobility group box-1; mPSL, methylprednisolone; PSL, predonisolone; rhTM, recombinant human soluble thrombomodulin; SP-D, surfactant protein D.

Outcome and follow-up

Eventually, the patient's oxygenation and lung infiltration levels improved (figures 3 and 4). C reactive protein, surfactant protein D and high-mobility group box 1 (HMGB-1) levels decreased after the administration of rhTM. The patient's haemoglobin levels also recovered (figure 4).

Discussion

In the present case, it was suspected that sequential adjuvant CRT induced DAH. Although high-dose CS therapy was initiated, the patient's oxygenation and lung infiltrations worsened progressively. Since CS-refractory sequential CRT-induced lung injury was suspected, rhTM was administered in addition to the CS therapy, resulting in an improvement in her clinical condition.

Sequential CRT-induced lung injury is associated with radiation and/or administered drugs. In the present case, the lung infiltration was most intense in the irradiation field. Therefore, we initially suspected RILI. However, radiation recall pneumonitis has been known to occur after previous lung irradiation in response to treatment with certain chemotherapeutic agents, including taxanes.⁷ Moreover, the risk of cyclophosphamide-induced pulmonary toxicity can be increased by the concomitant use of radiation therapy.⁸ In addition, lymphocytic BAL fluid was reported to correspond to both RILI and DILD.^{9 10} In the present case, although DLST for these antineoplastic agents was negative, the sensitivity of DLST (66.9%) was not sufficient to detect drug-induced pneumonitis.¹¹ Therefore, there was a possibility that the administration of docetaxel and cyclophosphamide influenced the development of RILI in the present case. Finally, we considered that the most likely diagnosis was a radiation reaction influenced by docetaxel and/or cyclophosphamide, resulting in the patient's respiratory symptom. This consideration was supported by the CT scan findings, which showed consolidations corresponding with the irradiation fields and bilateral diffuse GGAs.

We initiated high-dose CS therapy according to the recommendations for the treatment of RILI and DILD,^{3 4} although the patient's condition worsened and bronchoscopy showed the presence of ≥20% haemosiderin-laden macrophages in the BAL fluid taken from outside of the irradiated region, indicating the involvement of DAH.¹² There were no indications of vasculitis. ‘Bland’ pulmonary haemorrhage was ruled out because of normal BNP levels, echocardiographic findings and renal function. Therefore, in the present case, the most reasonable cause of DAH was diffuse alveolar damage (DAD) caused by RILI and/or DILD. In fact, the histopathological findings in CS-refractory RILI with DAH have been reported to be mostly caused by DAD.¹³ DAD with DAH has a significantly poorer prognosis compared with DAD without DAH.¹⁴ Therefore, we considered that additional treatment was necessary for her sequential CRT-induced life-threating DAH. Although treatment for CS-refractory sequential CRT-induced lung injury has not been established, rhTM was reported to be effective for DAD induced by AE-IPF and DILD.^{5 6 15} Therefore, in the present case, we administered rhTM, which significantly improved the patient's clinical condition.

In the present case, serum HMGB-1 levels were elevated despite high-dose CS therapy. HMGB-1 is an important proinflammatory mediator that contributes to acute lung injury.¹⁶ This cytokine induces an accumulation of neutrophils and causes protein exudation into the alveolar space.¹⁶ HMGB-1-related severe inflammation was associated with poor prognosis in patients with AE-IPF,¹⁷ whereas decreased serum HMGB-1 levels were associated with improved prognosis in these patients.¹⁸ The anti-inflammatory action of rhTM is mediated via its ability to adsorb and degenerate HMGB-1.¹⁹ In the present case, elevated serum HMGB-1 levels were significantly decreased after the administration of rhTM. Therefore, as an anti-inflammatory, rhTM might have played a crucial role in the improvement of the sequential CRT-induced lung injury.

In a recent study, TM and D-dimer were increased in the BAL fluid of patients with AE-IPF,⁵ which reflects the presence of intra-alveolar activated coagulation. Moreover, rhTM had anticoagulant effects mediated by activated protein C-dependent mechanisms.²⁰ Therefore, in the present case, we predicted severe coagulation abnormality in the intra-alveolar regions and expected an anticoagulant effect of rhTM therapy. However, TM and D-dimer were normal in the patient's BAL fluid. Therefore, it could be presumed that the main action of rhTM in the present case was anti-inflammatory.

CS therapy is mostly effective for DILD and RILI.^{3 4} However, some cases progress to life-threating conditions despite CS therapy.¹³ One case report demonstrated the efficacy of cyclosporine A for the treatment of CS-refractory RILI,²¹ although it took a few weeks to obtain sufficient clinical effect.²² The present case demonstrated the potential benefits of rhTM in addition to CS therapy for severe sequential CRT-induced lung injury. However, further studies are required to determine the efficacy of CS plus rhTM for the treatment of CS-refractory sequential CRT-induced lung injury.

In conclusion, we present a case of CS-refractory sequential CRT-induced DAH, possibly caused by DAD. The patient recovered after treatment with CS therapy plus rhTM. This combination therapy may be effective for severe sequential CRT-induced lung injury because of its anti-inflammatory effects, although further studies are needed to validate these findings.

Learning points.

• There are no advanced treatment strategies for corticosteroid therapy-refractive sequential chemoradiation-induced lung injury.

• In the present case, recombinant human soluble thrombomodulin might have played a crucial role in the improvement of sequential chemoradiotherapy-induced lung injury.

• Further study is necessary to confirm the efficacy of recombinant human soluble thrombomodulin for corticosteroid-refractory sequential chemoradiotherapy-induced lung injury.