Pleomorphic Carcinoma of the Lung: Relationship Between CT Findings and Prognosis

Transcription

1 Cardiopulmonary Imaging Original Research Fujisaki et al. CT of Pleomorphic Carcinoma of the Lung Cardiopulmonary Imaging Original Research Akitaka Fujisaki 1 Takatoshi Aoki 1 Takahiko Kasai 2,3 Shunsuke Kinoshita 1 Yoshinori Tomoda 4 Fumihiro Tanaka 5 Kazuhiro Yatera 6 Hiroshi Mukae 6 Yukunori Korogi 1 Fujisaki A, Aoki T, Kasai T, et al. Keywords: CT, lung cancer, pleomorphic carcinoma, prognosis DOI: /AJR Received August 30, 2015; accepted after revision February 1, Department of Radiology, University of Occupational and Environmental Health School of Medicine, 1-1 Iseigaoka, Yahatanishi-ku, Kitakyushu , Japan. Address correspondence to T. Aoki (a-taka@med.uoeh-u.ac.jp). 2 Department of Pathology, National Hospital Organization Kinki-Chuo Chest Medical Center, Sakai, Japan. 3 Department of Pathology and Oncology, University of Occupational and Environmental Health School of Medicine, Kitakyushu, Japan. 4 Cancer Therapy Center, Tobata Kyoritsu Hospital, Kitakyushu, Japan. 5 Second Department of Surgery, University of Occupational and Environmental Health School of Medicine, Kitakyushu, Japan. 6 Department of Respiratory Medicine, University of Occupational and Environmental Health School of Medicine, Kitakyushu, Japan. AJR 2016; 207: X/16/ American Roentgen Ray Society Pleomorphic Carcinoma of the Lung: Relationship Between CT Findings and Prognosis OBJECTIVE. The objective of our study was to assess the radiologic and clinical findings of pleomorphic carcinoma (PC) of the lung and to evaluate whether there are any characteristic features that can be used to predict prognosis. MATERIALS AND METHODS. Forty-four consecutive patients whose diagnosis of PC was histologically confirmed through resection of the lung tumor were included in this study. The clinical and CT findings of these patients were retrospectively reviewed. Two thoracic radiologists evaluated the CT findings including the size, location, internal characteristics, and margin characteristics of the tumors and the presence of chest wall invasion, mediastinal invasion, and surrounding lung abnormalities. A multivariate analysis by the Cox proportional hazards regression model was used to identify variables that can be used to predict overall survival and disease-free survival. RESULTS. In the patients with PC, a central low-attenuation area or cavity (40/44, 91%), chest wall invasion (19/44, 43%), and pulmonary emphysema (30/44, 68%) were frequently observed on CT. On multivariate analysis, a massive central low-attenuation area or cavity (> 25% of the lesion) on CT indicating necrosis was the only significant independent factor for overall survival and disease-free survival (p < 0.05). Clinical findings, the presence of lymph node metastasis at surgery, and postoperative pathologic stage were not significant predictors of overall survival and disease-free survival. CONCLUSION. A massive central low-attenuation area or cavity on CT was the only predictor of overall survival and disease-free survival in patients with lung PC. P leomorphic carcinoma (PC) is a subtype of sarcomatoid carcinoma that contains a component of sarcoma or sarcomalike (spindle cells, giant cells, or both) differentiation. According to the criteria of the World Health Organization classification [1], PC of the lung is defined as a poorly differentiated non small cell lung carcinoma (NSCLC) containing spindle cells, giant cells, or both or a carcinoma consisting of only spindle and giant cells. At least 10% of the carcinoma should be composed of spindle cells, giant cells, or both for it to be classified as a PC. This tumor is rare, and its incidence has been reported to be % of all lung malignancies [2, 3]. Patients with PC tend to present at a more advanced stage and to have a poorer prognosis than those with a common type of NSCLC [3 5]. Martin et al. [6] compared the 5-year survival rate of 63 sarcomatoid lung tumor patients with that of propensity score matched patients with other types of NSCLC and showed that the 5-year survival rate of the sarcomatoid lung tumor patients (24.5%) was significantly less than that of the patients with other types of NSCLC (46.3%). Although several clinicopathologic studies of lung PC have been reported in the literature, there are few studies about its radiologic features [7, 8]. Moreover, to our knowledge, the correlation of CT findings with prognosis has not been described to date. The purpose of our study was to assess the radiologic and clinical findings of lung PC and to evaluate whether there are any characteristic features that predict the prognosis. Materials and Methods Our institutional review board approved this study, and informed consent was waived for retrospective review of patient records and images. Patients and Clinical Findings The medical records of 59 Asian patients with a diagnosis of lung PC at our hospital between AJR:207, August

2 Fujisaki et al. June 1995 and December 2013 were retrospectively reviewed, and patients who had undergone surgical resection of the lung tumor, had undergone contrast-enhanced CT before surgery, and had not been given any anticancer drugs before surgery were included in this study. PET/CT was performed in 27 of the 59 patients and detected occult metastasis in five patients. Fifteen patients were excluded from the study for the following reasons: 11 patients received neoadjuvant chemotherapy before CT, two patients underwent needle biopsy without surgical resection for PC diagnosis, and two patients did not undergo contrast-enhanced CT. Thus, 44 patients were included in this study. The time interval between the final preoperative CT examination and surgery ranged from 2 to 26 days. Surgical resection in the form of a wedge resection (n = 1), segmentectomy (n = 5), lobectomy (n = 35), or pneumonectomy (n = 3) was performed. Eight patients had undergone CT more than 2 months before surgery in addition to the preoperative CT study; therefore, tumor growth could be evaluated using the previous CT examination and the preoperative CT examination. CT observation periods of the eight patients ranged from 70 to 329 days. The medical record of each patient was reviewed by two of the authors for the following: age, sex, smoking habits, treatment, and long-term clinical status after surgery (i.e., recurrence, metastasis, or survival). A CT Analysis Scanning of the whole lungs was performed on a 4-, 16-, 32-, or 64-MDCT unit or on a single-detector helical CT unit. For the MDCT examinations, the following parameters were used: 2.0-mm section width with 2.0-mm reconstruction interval, pitch (ratio of table travel per rotation to total beam width) of 15, 120 kvp, and 300 ma. For the single-detector CT examinations, the following parameters were used: 10-mm section thickness, 120 kvp, 150 ma, and a table speed of 10 mm/s. For additional scanning of the tumor using the single-detector CT unit, the following parameters were used: 2.0-mm section thickness, 120 kvp, and 250 ma. All images were reviewed on an ultra-high-resolution gray-scale monitor (20.8 inches [27.4 cm], pixels; Coronis 3MP, BARCO Display Systems) using standard lung window settings (window width, 1600 HU; window level, 600 HU) and mediastinal C D Fig year-old man with pleomorphic carcinoma (PC) of lung. Liver metastasis was detected 7 months after surgery, and patient died of disease 9 months after surgery. A, CT image obtained before surgery shows grossly irregular nodule and centrilobular emphysema. B, CT image obtained before surgery shows central low-attenuation area; this finding indicates necrosis is present within tumor. C, Low-power photomicrograph of histologic specimen shows extensive necrosis in central portion of tumor. D, High-power photomicrograph shows predominantly atypical multinucleated giant cells (arrows). B 290 AJR:207, August 2016

3 CT of Pleomorphic Carcinoma of the Lung window settings (window width, 350 HU; window level, 50 HU). Two radiologists with 24 and 10 years of experience in interpreting thoracic CT, respectively, evaluated the CT examinations for the size, location, internal characteristics (i.e., central low-attenuation area or cavity and calcification), and margin characteristics of the tumors and for the presence of chest wall invasion, mediastinal invasion, and surrounding lung abnormalities (i.e., emphysema or interstitial pneumonia). A central low-attenuation area or cavity was defined as an internal area with low attenuation relative to the attenuation of the surrounding musculature. We used unenhanced CT images for differentiating a tumor with a central low-attenuation area or cavity from a tumor with rim enhancement. The percentage of the central low-attenuation area or cavity was semiquantitatively calculated by dividing the area of the lowattenuation area or cavity by the total tumor area in the transaxial plane and was classified into two groups: 0 25% of the lesion and more than 25% of the lesion. The reviewers interpreted the images separately, and a final decision was reached by consensus if their interpretations differed. In the eight patients who had undergone a previous CT examination before the preoperative CT examination, the method originally described by Schwartz [9] was used to calculate tumor doubling time: Tdt = tlog2 / 3log(Dt / D0), where Tdt is the tumor doubling time, t is the time lapse between the two measurements, Dt is the mean tumor diameter at the final measurement, and D0 is the mean tumor diameter at the initial measurement. Histopathologic Analysis All surgical specimens were fixed in the inflated state by transpleural and transbronchial infusion of formalin. The specimens were stained with H and E. Histopathologic findings (i.e., vessel invasion, lymph node metastasis, and pathologic stage) were reviewed in the surgical specimens by a lung pathologist for this study. Pathologic staging was performed according to the classification of the Union for International Cancer Control [10]. The internal characteristics of the tumors seen on CT scans were compared with those seen at pathologic examination of the specimens. The correlations were decided by consensus of the pathologist and one radiologist. TABLE 1: Thin-Section CT Findings of 44 Patients With Lung Pleomorphic Carcinoma CT Findings Statistical Method Overall survival and disease-free survival were calculated according to the Kaplan-Meier method. Univariate analyses were performed to determine the influence of age, sex, smoking habits, CT findings, vessel invasion, lymph node metastasis, and pathologic stage. Variables with p values < 0.05 by univariate analyses were chosen as the variables for the multivariate logistic regression analysis. A multivariate analysis by the Cox proportional hazards regression model was used to identify variables that can be used to predict prognosis. Interobserver agreement for the CT findings was analyzed by computing the intraclass correlation coefficient (ICC). The strength of agreement was considered slight for an ICC of 0.40 or less, fair for an ICC of , moderate for an ICC of , and excellent for an ICC of 0.81 or greater. Results Clinical Features Of the 44 patients in this study, 36 were men and eight were women. The age of the patients at the time of PC diagnosis ranged from 36 to 91 years, with an average of 67.1 years. Thirty-nine patients (89%) were smokers, and 34 (77%) were heavy smokers (> 20 pack-years). Twenty-two of the 44 patients died months (mean, 27.5 months) after surgery, and 22 patients were alive 5 80 months (mean, 31.5 months) at the time of the most recent follow-up. Fourteen of the 22 surviving patients were disease-free, and the remaining eight had disease progression. CT Findings The results of the CT findings in all patients with lung PC are summarized in Table 1. A central low-attenuation area or cavity No. (%) of Patients Tumor size 30 mm 18 (40.9) > 30 mm 26 (59.1) Internal characteristics cavity 40 (90.9) > 25% of the lesion 27 (61.4) Calcification 0 (0) Margin characteristics Sharp and smooth 0 (0) Some irregular undulations 27 (61.4) Grossly irregular with spiculations 17 (38.6) Chest wall invasion 19 (43.2) Mediastinal invasion 11 (25.0) Pulmonary emphysema 30 (68.2) Interstitial pneumonia 3 (6.8) (40/44, 91%), tumor margin with some irregular undulations (27/44, 61%), chest wall invasion (19/44, 43%), and pulmonary emphysema (30/44, 68%) were frequently observed on CT (Figs. 1 3). In the majority of cases with a central low-attenuation area or cavity, the low-attenuation area or cavity was greater than 25% of the lesion (27/44, 61%). Interobserver agreement (kappa value) for the CT findings was excellent ( ). The tumor doubling times (n = 8) based on CT ranged from 53 to 139 days, with a mean of 86.7 days. Histopathologic Findings The pathologic stage of the PC was IA in five patients, IB in nine, IIB in 10, IIIA in 11, IIIB in six, and IV in three. Vessel invasion was observed in 26 (59%), and lymph node metastasis was seen in 17 (39%) patients. Pathologic analysis showed that the areas of tumor necrosis with hemorrhagic foci corresponded to the massive central low-attenuation area or cavity on CT in most patients and that the areas of myxoid degeneration also corresponded to the low-attenuation area or cavity on CT in two patients. Correlation Clinical and CT Features With Prognosis A univariate analysis of the prognostic factors influencing overall survival and disease-free survival is summarized in Table 2. A massive central low-attenuation area or cavity on CT and advanced pathologic stage AJR:207, August

4 Fujisaki et al. TABLE 2: Univariate Analysis of Prognostic Factors Influencing Overall Survival and Disease-Free Survival of Patients With Lung Pleomorphic Carcinoma Prognostic Factors Overall Survival A B Fig year-old man with pleomorphic carcinoma of lung. Patient was alive without evidence of recurrence at 3-year follow-up. A and B, CT images obtained before surgery show mass with some irregular undulations (A) and chest wall and mediastinal invasion (B). p Disease-Free Survival Age Sex Smoking index > 20 pack-years Tumor size > 30 mm cavity > 25% of the lesion Margins grossly irregular with spiculations Chest wall invasion Mediastinal invasion Pulmonary emphysema or interstitial pneumonia Vessel invasion Lymph node metastasis Stage (I II vs III IV) TABLE 3: Multivariate Analysis of Prognostic Factors Influencing Overall Survival and Disease-Free Survival of Patients With Lung Pleomorphic Carcinoma Prognostic Factors Relative Risk 95% CI p Overall survival cavity > 25% of the lesion Lymph node metastasis Stage (I II vs III IV) Disease-free survival cavity > 25% of the lesion Lymph node metastasis Stage (I II vs III IV) (stage III IV) predicted poorer overall survival (p < 0.05; Fig. 4). A massive central low-attenuation area or cavity on CT, lymph node metastasis, and advanced stage (stage III IV) predicted poorer disease-free survival (p < 0.05; Fig. 5). A multivariate analysis of the prognostic factors influencing overall survival and disease-free survival is summarized in Table 3. A massive central low-attenuation area or cavity on CT indicating necrosis was the only significant independent factor for predicting prognosis (p < 0.05). Discussion The results of our study showed that a massive central low-attenuation area or cavity on CT was the only significant independent factor for predicting prognosis (p < 0.05); pathologic stage was not a significant predictor. Surgical resection alone is therefore insufficient in patients with this finding on CT even if the pathologic stage is not advanced, and the combination of extensive surgical intervention with aggressive postoperative chemotherapy, radiotherapy, or both needs to be explored. Given that CT is routinely performed of most patients with lung PC, this imaging study is readily available and requires no additional cost. A massive central low-attenuation area or cavity on CT may help in selecting a therapeutic strategy for patients with lung PC. In most cases in our study, the massive central low-attenuation area or cavity seen on contrast-enhanced CT scans corresponded to areas of tumor necrosis in pathologic specimens. Tumor necrosis represents a paradoxical relationship whereby evidence of in- Fig year-old woman with pleomorphic carcinoma of lung. Lung metastasis was detected 7 months after surgery, and patient died of disease 19 months after surgery. CT image obtained before surgery shows necrotic cavity within mass. 292 AJR:207, August 2016

5 CT of Pleomorphic Carcinoma of the Lung 1.0 of lesion ( ) 1.0 Cumulative Survival Survival (mo) creased tumor cell death indicates that the tumor is a more aggressive tumor. This relationship can be explained by rapid tumor growth to a size at which the tumor has outgrown its blood supply [11]. Hypoxia is a characteristic of invasive cancers that can lead to the development of an aggressive phenotype through a mechanism that is mediated mainly by hypoxia-inducible factor (HIF) 1 and that includes cell immortalization and dedifferentiation, ph regulation, autocrine growth and survival, angiogenesis, invasion and metastasis, and resistance to chemotherapy [12 14]. We speculate that rapid tumor growth leads to inadequate blood supply to the central area of the tumor, which results in ischemic changes. Extensive necrosis due to ischemic changes was the most prevalent cause of cavity formation and has been reported to be an independent histologic factor in predicting prognosis of patients with lung PC [4, 15]. Because lung PC is a rare type of lung tumor, the studies in the peer-reviewed literature on its clinical features are relatively limited. However, several clinical features that are unique to lung PC have been reported: PC shows prevalence in male smokers who have a history of heavy tobacco consumption, and the average age at presentation is 60 years [3, 4, 16]. Similar to these results, our results also showed a male predominance (male-female ratio, 4.5:1) and a clear association with a smoking habit (89% had a history of smoking, the majority whom were heavy smokers [i.e., > 20 packyears]). The high prevalence of pulmonary emphysema (68%) surrounding the lung p = of lesion (+) Fig. 4 Graph shows overall survival according to CT finding of central lowattenuation area or of lesion. Cumulative Survival of lesion ( ) p = of lesion (+) Survival (mo) Fig. 5 Graph shows disease-free survival according to CT finding of central lowattenuation area or of lesion. PC in our study may be explained by these unique clinical features. The imaging findings of lung PC have been reported in only a few articles to date. Kim et al. [7] retrospectively evaluated the CT features of 10 patients with lung PC and reported that lung PCs preferentially manifest as large peripheral lung neoplasms with a central lowattenuation area (80%) and frequently invade the pleura or chest wall (70%). Another group of researchers, Kim et al. [8], also assessed the CT features of surgically resected lung PC in 30 patients, and a central low-attenuation area or cavity was observed in 50% of the patients in their series. In our larger series, a low-attenuation area or cavity and chest wall invasion were observed in 91% and 43% of patients, respectively, and these results concur with those of previous reports. These results suggest that lung PC tends to be necrotic, cavitary, and locally invasive. Rapid growth of lung PC has been sporadically reported [3, 15, 17]; in some cases that were measurable on CT using the Schwartz method, the tumor doubling times were less than 30 days [3, 15, 17]. In our cases, most of the tumors grew as rapidly as those in the past reports, and the tumor doubling times were shorter than those of the common types of NSCLC [18]. Although the characteristic of rapid growth has not fully been explained, the sarcomatoid elements of lung PC have a high proliferative activity (MIB1 cell proliferation marker index) and may be related to the rapid growth [15]. From another point of view, the expression of HIF-1α and microvessel density (MVD), which strongly affect angiogenesis through unregulation of vascular endothelial growth factor, are significantly greater in lung PC than in lung adenocarcinoma [19]. Because angiogenesis is essential for tumor growth, the overexpression of HIF- 1α and the increase of MVD of lung PC are also considered causes of rapid growth. The current study has several limitations. First, this study included a relatively small number of Asian patients because lung PC is rare. An additional prospective study in a broader population would yield more comprehensive results. Second, our study population did not reflect the entire spectrum of lung PC because we could include only patients with surgically resected lung PC and because we excluded patients who were treated with neoadjuvant chemotherapy. These factors may have given rise to selection bias. Third, the difference in surgical management and the progression of chemotherapy over a relatively long study period may have had an impact on survival. Fourth, in the eight patients who had undergone a previous CT examination before the preoperative CT examination, we assessed the CT findings on the preoperative CT examination (i.e., the later CT examination) because we correlated CT features and postoperative prognosis. Given the aggressive nature of lung PC, the use of the preoperative (later) images may have biased the CT findings. Despite these limitations, we believed that it is important to determine the CT findings that can be used to predict prognosis in patients with lung PC. Finally, CT software was not used in this study for the calculation of tu- AJR:207, August

6 Fujisaki et al. mor doubling time because of the retrospective nature of this study. Further study of a greater number of cases of lung PC with CT software that can perform 3D volume measurements is likely necessary to characterize tumor growth of lung PC. In conclusion, a central low-attenuation area or cavity in PC that occupies greater than 25% of the tumor is associated with decreased overall survival and decreased disease free-survival, more so than pathologic stage and lymph node metastases of NSCLC. Recognition of this prognostic factor may have important clinical implications, and further large-scale study is encouraged to determine the appropriate treatment strategy based on CT findings for patients with lung PC. References 1. Corrin B, Chang YL, Rossi G, et al. Sarcomatoid carcinoma. In: Travis WD, Brambilla E, Müller- Hermelink HK, Harris CC, eds. World Health Organization classification of tumors: pathology and genetics of tumours of the lung, pleura, thymus and heart. Lyon, France: IARC Press, 2004: Chang YL, Lee YC, Shih JY, Wu CT. Pulmonary pleomorphic (spindle) cell carcinoma: peculiar clinicopathologic manifestations different from ordinary non small cell carcinoma. Lung Cancer 2001; 34: Fishback NF, Travis WD, Moran CA, Guinee DG Jr, McCarthy WF, Koss MN. Pleomorphic (spindle/giant cell) carcinoma of the lung: a clinicopathologic correlation of 78 cases. Cancer 1994; 73: Rossi G, Cavazza A, Sturm N, et al. Pulmonary carcinomas with pleomorphic, sarcomatoid, or sarcomatous elements: a clinicopathologic and immunohistochemical study of 75 cases. Am J Surg Pathol 2003; 27: Nakajima M, Kasai T, Hashimoto H, Iwata Y, Manabe H. Sarcomatoid carcinoma of the lung: a clinicopathologic study of 37 cases. Cancer 1999; 86: Martin LW, Correa AM, Ordonez NG, et al. Sarcomatoid carcinoma of the lung: a predictor of poor prognosis. Ann Thorac Surg 2007; 84: Kim TH, Kim SJ, Ryu YH, et al. Pleomorphic carcinoma of lung: comparison of CT features and pathologic findings. Radiology 2004; 232: Kim TS, Han J, Lee KS, et al. CT finding of surgically resected pleomorphic carcinoma of the lung in 30 patients. AJR 2005; 185: Schwartz M. A biomathematical approach to clinical tumor growth. Cancer 1961; 14: Sobin LH, Gospodarowicz MK, Wittekind C, eds. International Union Against Cancer (UICC) TNM classification of malignant tumours, 7th ed. New York, NY: Wiley-Blackwell, Swinson DE, Jones JL, Richradson D, et al. Tumour necrosis is an independent prognostic marker in non small cell lung cancer: correlation with biological variables. Lung Cancer 2002; 37: Gilchrist KW, Gray R, Fowble B, Tormey DC, Taylor SG 4th. Tumor necrosis is a prognostic predictor for early recurrence and death in lymph node-positive breast cancer: a 10-year follow-up study of 728 Eastern Cooperative Oncology Group patients. J Clin Oncol 1993; 11: Semenza GL. HIF-1 inhibitors for cancer therapy: from gene expression to drug discovery. Curr Pharm Des 2009; 15: Hiraoka N, Ino Y, Sekine S, et al. Tumour necrosis is a postoperative prognostic marker for pancreatic cancer patients with a high interobserver reproducibility in histological evaluation. Br J Cancer 2010; 103: Fujioka S, Nakamura H, Adachi Y, et al. Pleomorphic carcinoma of the lung in which the sarcomatous element grew rapidly: a case report. Ann Thorac Cardiovasc Surg 2009; 15: Mochizuki T, Ishii G, Nagai K, et al. Pleomorphic carcinoma of the lung: clinicopathologic characteristics of 70 cases. Am J Surg Pathol 2008; 32: Ito K, Oizumi S, Fukumoto S, et al. Clinical characteristics of pleomorphic carcinoma of the lung. Lung Cancer 2010; 68: Detterbeck FC, Gibson CJ. Turning gray: the natural history of lung cancer over time. J Thorac Oncol 2008; 3: Tsubata Y, Sutani A, Okimoto T, et al. Comparative analysis of tumor angiogenesis and clinical features of 55 cases of pleomorphic carcinoma and adenocarcinoma of the lung. Anticancer Res 2015; 35: AJR:207, August 2016