Segmentectomy for giant pulmonary sclerosing haemangiomas with high serum KL-6 levels

Abstract

We describe a 61-year old female patient with a giant pulmonary sclerosing haemangioma (PSH) and an extremely high preoperative serum KL-6 level. During an annual health screening, the patient showed a posterior mediastinal mass on chest radiography. Chest computed tomography and magnetic resonance imaging revealed a well-circumscribed 60 mm diameter nodule with a marked contrast enhancement in the left lower lobe. The preoperative serum KL-6 level was elevated to 8204 U/ml. We performed a four-port thoracoscopic basal segmentectomy and lymph node sampling for diagnosis and therapy. The postoperative diagnosis showed PSH. The serum KL-6 level decreased dramatically with tumour resection. To the best of our knowledge, this is the first report of a patient with PSH showing a high serum KL-6 level.

INTRODUCTION

Pulmonary sclerosing haemangioma (PSH) is a rare pulmonary neoplasm. It is defined as a benign tumour or a very low-grade malignancy. PSH was first reported by Liebow and Hubbel in 1956 as a vascular disorder [1]. Immunohistochemical (IHC) analyses strongly suggest that PSH originates from a primitive respiratory epithelium [2]. KL-6 is produced and secreted by alveolar pneumocytes (type II).

We describe a patient who underwent four-port thoracoscopic segmentectomy for PSH and had a high serum KL-6 level. We believe that alveolar pneumocytes produce KL-6. Such a case of PSH has not yet been reported.

CASE REPORT

The patient was a 61-year old woman. At the age of 51 years, a posterior mediastinal mass was observed on chest radiography during an annual health screening. The mass was 48 mm in diameter in January 2004 (Fig. 1a), and it gradually increased in size. The patient initially refused to undergo tumour resection. However, she finally accepted four-port thoracoscopic surgery because the tumour had increased in size to 60 mm. The doubling time was ∼1950 days. The serum KL-6 level, which was measured routinely before operation, elevated to 8204 U/ml (normal, <499 U/ml). Bilateral lungs did not show interstitial pneumonia or emphysematous changes on computed tomography (CT). All tumour markers were negative. Systematic imaging showed the absence of any other malignant diseases. Preoperative magnetic resonance imaging (MRI) and CT showed that this giant mass extended into the lung parenchyma and not the mediastinum. It formed a 70 mm, solid and well-circumscribed nodule that compressed the aorta, oesophagus and branches of the bronchus in the lower lobes (Fig. 1b).

Figure 1:

(a) A Chest X-ray taken in January 2004 revealed a mass in the left lower field, the greatest axis was 48 mm in size. (b) Enhanced MRI showed the left intrapulmonary mass shadow which was non-homogenous with a smooth enhanced margin. A tumour compressing the surrounding organs such as the oesophagus and aorta was detected. (c) No invasion and strict adhesion was observed during surgery (d) The gross specimen showed a non-homogenous mass with focal area of haemorrhage.

Left basal segmentectomy or left lower lobectomy was planned via four-port thoracoscopic surgery. Although the tumour was found to be compressing the adjacent organs, no invasion and strict adhesion was observed during surgery (Fig. 1c). We judged that the preservation of V6 and S6 lung parenchyma was possible during the operation. Macroscopic findings of hilar and peribronchial lymph nodes (LNs) strongly suggested no cancer metastasis. Four-port thoracoscopic left basal segmentectomy and LN sampling were accomplished. The operative time was 151 min and intraoperative bleeding was 120 ml. The serum KL-6 level was 4347 and 2322 U/ml at 7 and 20 days after surgery, respectively. The patient showed no complications and was discharged from hospital on the ninth postoperative day.

Macroscopic findings of the tumour showed a miscellaneous pattern (Fig. 1d). Histological analyses showed that the tumour consisted of heterogeneous features (haemorrhage > solid > sclerosic > papillary; Fig. 2a–d). Surface and round cells were positive for thyroid transcription factor-1 (TTF-1) and vimentin; this provided additional evidence of the epithelial nature of this tumour (Fig. 2e–h). Surface cells stained positively for pancytokeratin (AE1/AE3) but round cells did not. Hyperplasia of collagen fibres was identified using the Elastica van Gieson (EVG) stain (Fig. 2i). The final pathological diagnosis was sclerosing haemangioma of the lung.

Figure 2:

Sclerosing haemangioma is characterized by four architectural patterns. (a) Haemorrhagic, (b) solid, (c) sclerotic and (d) papillary. IHC staining of sclerosing haemangioma. (e and g) TTF-1 is positive in both types of cells and detected in nucleus. (f and h) Vimentin is positive in both types of cells and detected in cytoplasm diffusely. (i) EVG is negative in the collagen fibres. The hyperplasia of collagen fibres is detected.

DISCUSSION

Surgical resection is the only treatment of choice for PSH. The necessity of systemic LN dissection in patients with PSH remains controversial because of the possibility of regional LN metastasis in 2–4% of cases [2, 3]. Kim et al. [4] and Yano et al. [5] reported that solitary lesions of diameter >3 cm are indicators of LN metastasis. However, some authors reported that regional LN metastasis (even if it includes multiple) does not affect prognosis [4–6]. Situ et al. [6] reported the outcome of 24 patients and reached two main conclusions: (i) systemic LN clearance cannot be recommended as a routine treatment for PSH and (ii) the prognosis of PSH is good and most patients can survive after successful resection. Four-port thoracoscopic left basal segmentectomy with regional LN sampling diagnosed as negative by freezing method was appropriate in this case.

KL-6 is a mucin-like high-molecular-weight glycoprotein produced and secreted by alveolar pneumocytes (type II) [7]. In this case, the KL-6 level was high but other laboratory data including tumour markers were unremarkable. To our knowledge, PSH with a high serum KL-6 level has never been reported. In addition, the KL-6 level decreased dramatically with PSH resection. The KL-6 level is dependent on the number of regenerating alveolar epithelial cells and integrity of the alveolar–capillary membrane. KL-6 may also have chemotaxis against fibroblasts and play a functional role in fibrosis. KL-6 reflects a relatively easy shift from tissues into the blood, unlike other tumour markers. IHC analysis in the largest series provided data that strongly suggest that PSH originates from the primitive respiratory epithelium [2]. On histological analyses, PSH is morphologically distinct, consisting of two epithelial cell types (surface and round cells) as well as four architectural patterns (papillary, sclerotic, solid and haemorrhagic). In this case, IHC analysis suggested that surface and round cells exhibited nuclear TTF-1 expression. Hyperplasia of collagen fibres was also identified by EVG staining. It is interesting that there may be a correlation between PSH and an increased KL-6 level. We know that KL-6 is a marker for lung adenocarcinoma as well as that for pulmonary fibrosis. This is understandable because KL-6 is a glycoprotein expressed in type II alveolar cells and such cells with nature of type II pneumocytes are growing in both lung adenocarcinoma and lesions of pulmonary fibrosis. We should bear in mind that PSH comprises cells of type II cell nature.

In the present case, an unusual feature of IHC staining was found. Surface and round cells exhibited strong positivity for vimentin in the cytoplasm. Reduction in epithelial cell–cell adhesion via transcriptional repression of cadherins in combination with mesenchymal transformation is a key determinant of epithelial–mesenchymal transition (EMT). The EMT is associated with the invasion of tumour cells and recurrence of cancer. Vimentin is a representative marker for mesenchymal properties [8]. However, vimentin expression was ∼100% in surface and round cells and Ki-67 staining was <1%. We consider the tumour to be benign tumours. We believe that the tumour acquired heterogeneous aspects (i.e. fusion of round and surface cells) because of its growth over a long period.

Conflict of interest: none declared.