Clinicopathological characteristics of surgically resected pulmonary pleomorphic carcinoma

Abstract

OBJECTIVES

Since few large-scale studies of patients with pleomorphic carcinoma have been documented, factors affecting survival after pulmonary resection for pleomorphic carcinoma, as well as its clinicopathological characteristics, are still unknown. For a better understanding of the patients undergoing resection of pulmonary pleomorphic carcinoma, we reviewed our experience with these patients.

METHODS

Between 2002 and 2010, 26 patients with pulmonary pleomorphic carcinoma underwent macroscopically complete pulmonary resections. Various perioperative variables were investigated retrospectively to confirm a role for pulmonary resection and to analyse prognostic factors for overall survival and disease-free survival after lung resection.

RESULTS

Twenty-four patients (92%) were male. Twenty-one patients (81%) were smokers and all of them smoked more than 30 pack-years. In 25 patients (96%), the tumour was located peripherally. Twenty-three of these 25 patients revealed the tumour touching the visceral pleura widely in the preoperative chest computed tomography. In all 26 patients, the tumour was completely resected macroscopically; however, three patients (12%) had microscopically positive surgical margins. Among them, additional irradiation was conducted in two patients and additional surgical resection was performed in one patient. Combined resections were performed in 11 patients (42%), including chest wall resections in 7 patients. Overall survival rate after pulmonary resection was 48% at 5 years. Disease-free survival rate after pulmonary resection was 33% at 5 years. Patients with tumours invading the visceral pleural surface and microscopically positive surgical margin had significantly worse overall survivals (P = 0.048 and 0.037, respectively). However, there were no significant prognostic factors for disease-free survival.

CONCLUSIONS

Despite small number of cases, we found that pleural invasion suggested a worse prognosis for resection of pulmonary pleomorphic carcinoma. Surgical strategy might be constructed to achieve not only macroscopically, but also microscopically complete resection for such large tumours with aggressive nature and peripheral preference.

INTRODUCTION

Pulmonary pleomorphic carcinoma is a rare epithelial malignant tumour and its incidence has been reported to range from 0.1 to 0.4% of all lung cancers [1]. The 1999 World Health Organization (WHO) classification identified pulmonary pleomorphic carcinoma as a specific type of lung cancer with pleomorphic, sarcomatoid or sarcomatous elements [2]. Pulmonary pleomorphic carcinoma has a more aggressive clinical course than other non-small-cell lung cancers (NSCLCs) [3–5], and recent reports also described that pleomorphic carcinoma had a worse outcome than other NSCLCs [3, 6, 7]. Although there are several reports about the survival and prognostic factors for the patients undergoing surgical resection of pulmonary pleomorphic carcinomas, the clinicopathological characteristics of pleomorphic carcinoma are not well known, and the reliable predictors of long-term survival are still controversial [3, 6, 8, 9].

We reviewed the clinical data of patients with pulmonary pleomorphic carcinoma who underwent pulmonary resection at our hospital to better understand the clinical behaviour and the prognostic predictors of survival in this histotype of pulmonary carcinoma.

PATIENTS AND METHODS

From January 2002 to December 2010, 28 patients had undergone pulmonary resection for pleomorphic carcinoma according to our medical records. Two patients were excluded from this study because the tumour was not completely resected surgically due to its invasion of the thoracic aorta in one patient and small-cell carcinoma was also detected in the resected lung in the other patient. The patients consisted of 24 males and 2 females with a median age of 69 years (range, 49–83 years). Medical records provided information on age, gender, presenting symptoms, smoking habits, tumour markers [preoperative carcinoembryonic antigen (CEA) levels], radiological findings, surgical procedures and pathological results (Tables 1–3). Basic bronchoscopies were performed preoperatively in order to diagnose the malignancy and to confirm the extent of tumour invasion. Preoperative evaluations included physical examinations, chest roentgenographies and tumour marker analyses. In addition, computed tomographic scans of the chest and abdomen were obtained routinely. Either computed tomography or magnetic resonance imaging of the brain was performed in order to rule out brain metastasis. Either bone scintigraphy or positron emission tomography was also conducted routinely. Staging was made based on the new International Staging System for Lung Cancer [10]. Standard surgical techniques were used and accompanied by routine systemic dissection or sampling of the hilar and mediastinal lymph nodes in every case. Intraoperative frozen tissue examinations were performed in order to determine the extent of resection and to assess nodal status. After surgical interventions, generally, the patients were examined in the outpatient clinic at 3–6-month intervals for at least 5 years. Pleomorphic carcinoma was diagnosed according to the 2004 WHO classification [11]. A diagnosis was made on the basis of the light microscopic findings and it was confirmed with immunohistochemical examinations, if needed. Pleomorphic carcinoma was defined as an NSCLC containing 10% sarcomatoid components. Sarcomatoid components were classified as follows: spindle cell type, giant cell type or the combination of spindle and giant cell types. Epithelial elements were also described. Pleural invasion was described using Pl factors as follows: Pl 0 (tumour within the subpleural parenchyma or invading superficially into the pleural connective tissue below the elastic layer), Pl 1 (tumour invading beyond the elastic layer), Pl 2 (tumour invading into the visceral pleural surface) and Pl 3 (tumour invading the parietal pleura). An institutional review board approved this retrospective study, and written informed consent for the surgical intervention was obtained from each patient.

The endpoint assessment consisted of overall survival (OS) and disease-free survival (DFS) after pulmonary resection. OS was defined as the length of time between the date of the pulmonary resection and the date of the last follow-up or death by any cause. DFS was defined as the length of time between the date of the pulmonary resection and the date of proven detection of recurrence or metastases.

Statistical analysis

Statistical analyses were performed with the StatView (version 4.5) software package (Abacus Concepts, Berkeley, CA, USA). The postoperative survival rate was analysed by the Kaplan–Meier method. The prognostic influence of variables on survival was analysed using the log-rank test for univariate analyses. Differences were considered significant when P < 0.05.

RESULTS

Twenty-six patients with pulmonary pleomorphic carcinoma underwent a macroscopically complete pulmonary resection. All the detailed data of each patient were shown in Tables 1 and 2. Eleven patients were asymptomatic (Table 3). Of the 15 symptomatic patients, 9 had haemoptysis, 4 had coughs and 2 had back pain. Twenty-two of the patients (85%) were smokers, and their median smoking history was 53 pack-years. The locations of the tumours were as follows: 11 cases were in the left upper lobe, 11 were in the right upper lobe, 3 were in the right lower lobe and 1 was in the left lower lobe. In 25 patients (96%), the tumour was located peripherally. Twenty-three of these 25 patients revealed the tumour touching the visceral pleura widely in the preoperative chest computed tomography. All of the patients but one did not receive the diagnosis of pleomorphic carcinoma preoperatively or intraoperatively. In one patient, a preoperative ultrasound-guided tru-cut biopsy obtained a specimen highly suggestive of pleomorphic carcinoma. Regarding the clinical staging of the tumours, 6 cases were stage IA, 10 cases were stage IB, 2 cases were stage IIA, 6 cases were stage IIB and 2 cases were stage IIIA. In contrast, the pathological staging of the tumours revealed that 2 cases were stage IA, 10 cases were stage IB, 4 cases were stage IIA, 5 cases were stage IIB and 5 cases were stage IIIA. All 26 patients underwent macroscopically complete resections; however, in three patients, pathological examinations of the resected specimens revealed the microscopic tumour invasion beyond its margin. In two patients, additional irradiation was performed for local control. In one patient, additional resection of the chest wall was performed with negative surgical margin. Lobectomies were performed in 23 cases, whereas segmentectomies were performed in two cases and wedge resection was conducted in one case. Sublobar resection was selected because of the patient's cardiopulmonary reserve. Combined resections were performed in 11 patients, and these included the chest wall with ribs in 7 patients, the diaphragm in 1 patient, the subclavian artery in 1 patient, the azygous vein in 1 patient and the pericardium in 1 patient. Adjuvant chemotherapy was recommended if the pathological staging was higher than IB, there was no contraindication for chemotherapy and the patient was physically and mentally tolerable for the treatment. In fact, several types of adjuvant chemotherapy could be performed in only eight patients (31%).

The median length of the follow-up examinations was 23 months (range, 1–99 months). The 5-year OS rate was 48% and the 5-year DFS rate was 33% (Fig. 1). Univariate analyses showed that a better OS was observed in patients without tumour invasion to the visceral pleural surface (P = 0.048; Fig. 2), and patients with microscopically complete resections (P = 0.037, Table 4 and Fig. 3). In contrast, univariate analyses did not show any significant prognostic factors for DFS.

The first sites of relapse after the pulmonary resections were locoregional in two patients, distant in seven patients and both in two patients (Table 5). Locoregional relapse included tumour recurrence at the surgical margin and pleural dissemination. Distant relapsed sites included bone in three cases, the gastrointestinal tract in two cases, the axillary lymph nodes in two cases, the adrenal gland in two cases, the liver in one case and the tonsils in one case. As shown in Table 4, 4 of 12 patients (33%) with Pl < 2 showed recurrence after surgery, whereas 7 of 14 patients (50%) with Pl ≥ 2 presented tumour recurrence. On the other hand, distant relapse was seen in 4 of 12 patients with Pl < 2 and 5 of 14 patients with Pl ≥ 2, but there was no locoregional relapse in patients with Pl < 2. Furthermore, two of four patients with Pl < 2 showing distant relapse had pN2 disease, whereas one of five patients with Pl ≥ 2 presented with pN2 disease. In two relapsed patients with Pl 3, who were both pN0, tumour relapse was detected very early both locoregionally and distantly at the same time (3 and 5 months) after surgery.

DISCUSSION

In 1999, pleomorphic carcinoma was recognized as a neoplasm with pleomorphic, sarcomatoid or sarcomatous elements in the category of carcinomas by the WHO classification [2]. Before 1999, pleomorphic carcinoma was considered as a variant of other well-known lung carcinomas because of its biphasic appearance and its frequent association with other histologic types [12]. Diagnosis of this histotype was problematic and confusing because of the lack of uniform diagnostic criteria. Thus, few studies of a large series of patients with pulmonary pleomorphic carcinoma have been described in the international literature [3, 6, 8, 9], and we therefore decided to report our experience with pulmonary resections in patients with pleomorphic carcinoma in order to evaluate the role of resection in their treatment with a focus on OS and DFS. The 5-year OS rate for patients who underwent a pulmonary resection for pleomorphic carcinoma was ∼40% in studies with a relatively large number of patients [3, 9], which was consistent with our results.

Several favourable prognostic factors have been reported, such as no nodal involvement [3, 6, 8, 9], complete resection [6], lower pathologic stages [9] and massive necrosis [9]. According to Yamamoto et al. [6], pleural involvement was not a significant adverse prognostic factor, but they did suggest that patients with no pleural involvement had a tendency for a better OS. In our study, we confirmed that patients with tumours that did not invade the visceral pleura had a better OS. This result suggests the aggressive feature of this malignant tumour, which favours peripheral location. In NSCLCs, visceral pleural invasion was also significantly associated with a higher frequency of lymph node involvement [13, 14] and was observed significantly more frequently in tumours with factors indicative of tumour aggressiveness and invasiveness [13]. In addition, microscopically incomplete resection, which was seen in only patients with Pl 3, was also an adverse prognostic factor. This also implied that pleural invasion might jeopardize the patients due to the incomplete local control of the disease.

In terms of the relationship between the first relapsed site and tumour pleural invasion, two of the four distantly relapsed patients with Pl < 2 presented with nodal metastasis, whereas one of the five distantly relapsed patients with Pl ≥ 2 had nodal metastasis. In addition, two relapsed patients with Pl ≥ 2, who developed simultaneous locoregional and distant metastasis relatively soon after surgery, did not show nodal metastasis. Despite the small number of cases, these facts partly support the idea that pleural invasion might significantly affect OS. In the current study, nodal involvement was not a significant adverse prognostic factor, which might be because there were only two patients with pN2 disease in our series of patients. Furthermore, one patient with pN2 (patient 10) survived 99 months after surgery, probably because he presented only one station pN2 disease.

Pulmonary pleomorphic carcinoma often presents in symptomatic male smokers as a large peripheral lesion [9]. It is of note that our series of patients also exhibited this feature, which was seen as part of the presenting symptoms of 58% of the patients, 92% of the males, 85% of the smokers, 38% of patients with tumours more than 50 mm and there was a peripheral location in all patients but one.

In terms of adjuvant therapies, there has yet been no consensus on the treatment of pleomorphic carcinoma even among specialists. Since few studies have reported on the use of chemotherapy for pulmonary pleomorphic carcinoma, treatment regimens remain controversial [4, 9]. Therefore, we have not designed any fixed criteria or regimen as an adjuvant setting in this series of patients. In fact, there was no significant effect on outcomes of adjuvant chemotherapy in our study. However, Kaira et al. [15] described that a mutation of the epidermal growth factor receptor was recognized in ∼20% of patients with pleomorphic carcinoma, and the use of a molecular targeting drug might improve the outcomes of the patients with pleomorphic carcinoma. The use of radiation therapy for pleomorphic carcinoma of the lung also remains controversial [6]. In our series, we performed adjuvant radiation therapy on two patients with microscopically incomplete resections. One patient relapsed with a tumour soon after the irradiation, whereas the other patient was still alive without an evident relapse 15 months after surgery.

There are several limitations to our analysis. The retrospective design is the most practical way of addressing our question because of the low incidence of pleomorphic carcinoma, but the results should be interpreted with caution. That is, the adverse prognostic factors suggested in our study should be reconfirmed by a prospective study. Furthermore, our results were based on a small number of patients at one institution. We recognize that this is the biggest limitation of our study and, therefore, a prospective, large-scale study at multiple institutions is needed in the future to confirm the current results.

In conclusion, despite small number of cases, we found that pleural invasion of the visceral pleural surface suggested a worse prognosis after resection of pulmonary pleomorphic carcinoma. Furthermore, microscopically incomplete resection was also an adverse prognostic factor. Therefore, surgical strategy might be constructed to achieve not only macroscopically, but also microscopically complete resection for such large tumours with aggressive nature and peripheral preference; however, an accumulation of cases is necessary to accurately evaluate prognostic factors and to determine the selection criteria for resection.

Conflict of interest: none declared.

REFERENCES