| Japanese Journal of Clinical Oncology | Pages |
Primary Unknown Cancer in Pulmonary Hilar Lymph Node with Spontaneous Transient Regression: Report of a Case
Introduction
Case Report
Discussion
Acknowledgment
References
Primary Unknown Cancer in Pulmonary Hilar Lymph Node with Spontaneous Transient Regression: Report of a Case
INTRODUCTION
There have been some reports of cases in which metastasis was the only presentation and careful evaluation failed to disclose the primary lesion of the tumor. These cases have been categorized as primary unknown cancer (1-4) and are not rare. To our knowledge, however, there have been only 11 primary unknown cases in the literature in which pulmonary hilar or mediastinal nodes were involved (5-13). We report a case of pulmonary hilar lymph node cancer without evident primary lesion, which regrew two years after spontaneous regression.
CASE REPORT
A 69-year-old man was referred to our hospital in December 1993 because an abnormal mass had been detected in the right pulmonary hilum on a routine chest roentgenogram and a transbronchial lung biopsy at a local hospital was diagnosed as suspicious adenocarcinoma.
Chest roentgenogram at his first visit showed a right hilar mass measuring 6.8 × 3.2 cm (Fig. 1). Computed tomography (CT) revealed a hilar lymph node swelling measuring 2.8 × 2.6 cm between the right middle and lower lobe bronchi, and an adherent tumor measuring 3.2 × 3.0 cm in the right ventrobasal segment (S8) (Fig. 2a and b). The serum examination showed an elevated level of carcinoembryonic antigen (CEA), 29.9 ng/ml, the normal value being <5.0 ng/ml. In January 1994, a bronchoscopy showed extrinsic compression and deformity of the right B8b bronchus, but the bronchial mucosa was intact. Transbronchial biopsy specimen showed no malignant cells. He was admitted to our hospital for further evaluation of the tumor in February 1994. Chest roentgenogram on 3 February 1994, however, showed no evident tumor in the right lung field (Fig. 1). The serum CEA level decreased to 7.8 ng/ml, but was above the normal range. As there was neither pathological nor radiological evidence of malignancy, he was discharged without any therapy and was followed up on an outpatient basis every three months. There had been neither detectable new lesion nor relapse of the tumor, and the serum CEA level stayed between 5.8 and 11.5 ng/ml for two years.
Figure 1. Chest roentgenogram on 14 December 1993 (a), 3 February 1994 (b) and 13 March 1996 (c). Figure 2. Chest CT tomography in December 1993 revealed a swollen hilar lymph node (a) and an adherent tumor in right S8 (b). Chest CT in March 1996 revealed a swollen hilar lymph node (c), but there was no adherent tumor in right S8 (d). In March 1996, the mass in the right pulmonary hilum reappeared on chest roentgenogram (Fig. 1). Chest CT revealed a hilar lymph node measuring 3.0 × 2.8 cm between the right middle and lower lobe bronchi, but there was no tumor in the right S8 (Fig. 2c and d). The serum CEA level was elevated to 19.9 ng/ml. He was admitted to our hospital to evaluate the tumor on 10 May 1996. Physical examination on readmission revealed mild dementia attributable to Alzheimer's disease and moderate impairment of hearing; height 158 cm and weight 55 kg. An examination of the chest and abdomen showed no abnormalities. The blood cell count was normal except for moderate lymphocytopenia (1295 cells/µl). The serum CEA level was 27.3 ng/ml. He had gastritis when he was 40 years old and was cured medically. He had a smoking history of 20 cigarettes per day for the past 48 years. His father had died of gastric cancer, and his older brother had died of lung cancer. Brain and abdominal CT showed no abnormal findings, except for a cystic lesion measuring 0.5 cm in diameter in the left adrenal gland, which was radiologically diagnosed as a benign cyst. The patient underwent video-assisted thoracoscopy for diagnosis of the tumor on 17 May 1996. There was neither pleural effusion nor adhesion in the right thorax. The tumor was located between and adherent to the middle and basal bronchi of the right lung, and invaded the adjacent pulmonary parenchyma. The tumor appeared to involve a swollen inferior interlobar lymph node. Intraoperative needle biopsy of the node revealed cancer cells. As the tumor appeared to invade both the middle and lower lobes, we performed a right middle and lower bilobectomy with mediastinal dissection. Macroscopically, a stony hard tumor was located between the right middle and lower lobe bronchi, and measured 4.2 × 3.0 × 2.4 cm in size. The cut surface of the tumor was heterogeneous with white and gray areas (Fig. 3). The tumor was well encapsulated and was separated distinctly from the bronchi. Microscopically, the tumor cells replaced most of the lymph node, and a small area of lymphatic tissue was preserved in the marginal area of the node. The tumor showed a squamoid stratified pattern with central necrosis, but without definitive squamous differentiation, such as cancer pearl formation or intercellular bridges (Fig. 4). A part of the tumor cells grew in an organoid pattern and showed rosette-like structures mimicking large cell neuroendocrine carcinoma. Immunostaining using anti-chromogranin A and anti-neural cell adhesion molecule (Lu-243)14 antibodies showed diffuse granular staining in the cytoplasm and cell membrane, respectively, indicating the neuroendocrine nature of the neoplastic cells. The tumor cells had a moderate amount of eosinophilic foamy cytoplasm and large vesicular nuclei with prominent nucleoli. It was noteworthy that many macrophage aggregation foci were identified in the tumor nests. Some of the tumor nests were almost completely replaced by macrophages and lymphocytes (Fig. 5). Figure 3. Macroscopic appearance of the cut surface of the resected right lung. Figure 4. Microscopic findings of the resected tumor. The tumor showed a squamoid stratified pattern with central necrosis, but without definitive squamous differentiation. Figure 5. Some of the tumor nests were recognized as coagulation necrosis and some of the tumor nests were almost completely replaced by macrophages. The resected middle and lower lobes were serially sliced at 5 mm thickness and were examined thoroughly. However, there were no tumors in the lobes except for a coagulation necrosis measuring 1.5 cm in diameter in the S8 visceral to the B8b bronchus, corresponding to the site where a tumor shadow had been depicted on the CT image in December 1993. Microscopically, abundant cellular debris was noted in the coagulation necrosis (Fig. 6). We diagnosed the disease as metastatic large cell carcinoma with neuroendocrine features in the pulmonary interlobar lymph node, and T0N1M0 pathologically by the UICC classification (15). We concluded that the coagulation necrosis might be the primary site of the tumor, which had regressed spontaneously and regrew in the metastatic interlobar node. Figure 6. Microscopic findings of the coagulation necrosis in S8. Debris of degenerated cells was noted in the coagulation necrosis. The postoperative course was uneventful. He was discharged on the 34th postoperative day in good condition. The serum CEA level decreased to 3.7 ng/ml two months after the operation (Fig. 7). The patient is well without any signs of recurrence 20 months after the operation. Figure 7. Time course of serum CEA level. Primary unknown cancer is not an unusual phenomenon. It has been reported that patients with primary unknown cancer constitute 0.5-3.3% of all cancer patients (1,2,4). Greager et al. (3) studied 286 cases of primary unknown cancer and reported that the most common site of the detected cancer was the lymph node. The locations of the lymph node were cervical (15.1%), supraclavicular (9.1%), axillary (4.2%) and inguinal (2.8%) regions. However, carcinoma detected only in mediastinal or hilar lymph nodes is rare. To our knowledge, 11 cases of primary unknown cancer detected only in the pulmonary hilar or mediastinal lymph nodes have been reported (5-13). The involved nodes were located in the hilum in five cases and in the mediastimum in six. Histologically, four cases were small cell carcinoma, three were adenocarcinoma, two were squamous cell carcinoma and two were large cell carcinoma. Several explanations for the development of primary unknown cancer are possible. One possibility is that the primary site is the lymph node. Benign epithelial inclusion has been occasionally reported to exist in the lymph node (16-19), and there has been a report describing a carcinoma which originated from benign epithelial inclusion in an axillary lymph node (19). While it is necessary to prove epithelial inclusion in the lymph node histologically to demonstrate that a carcinoma originates in the node, there have been no cases in which epithelial inclusion was proven in the resected lymph nodes among the reported 11 cases of primary unknown cancer of the thoracic nodes (5-13). The present case did not show ectopic epithelial inclusion in the resected lymph nodes. Although this hypothesis did not agree with our case, it cannot be ruled out, because benign epithelial inclusion might have been replaced by the tumor. Holmes and Fouts (1) suggested that primary tumors had been too small to be found at either clinical examination or autopsy, or that primary tumors had been inadvertently removed or destroyed earlier by physician or sloughing by host in the primary unknown cancer cases. They also considered a possibility that the host defense mechanism had destroyed primary tumors, which they called spontaneous regression of cancer. Everson and Cole (20) defined spontaneous regression of cancer as a partial or complete disappearance of a malignant tumor without any treatment thought to have the potential to cure cancer. They reported that spontaneous regression occurred no more than once in 60 000-100 000 cancer cases. Spontaneous regression is rare, especially in lung cancer. In a review of the literature, we found only nine such cases (20-29). In our case, a pulmonary hilar shadow on chest roentgenogram in December 1993, which consisted of a swollen lymph node and an adherent tumor in S8, regressed without any therapy. As we failed to confirm the malignant histology of the disease before the regression, it could have been a benign lesion. However, we detected clusters of degenerated nuclei in the coagulation necrosis in S8 in the resected specimen. In addition, some of the tumor nests in the hilar lymph node were almost completely replaced by macrophages and lymphocytes; this findings resembling the coagulation necrosis in S8. By the way, the location of the coagulation necrosis in the resected specimen was the same as the tumor shadow in S8b on the chest CT image on his first visit. Furthermore, the level of serum CEA paralleled the tumor size on the chest roentgenograms. Therefore, we assume that a CEA-producing primary tumor in S8b metastasized to the hilar lymph node, that both the primary and metastatic tumor regressed spontaneously, and that the metastatic node alone relapsed two years later, while the primary site remained as a coagulation necrosis. This work was supported in part by Grant-in-Aids for Cancer Research from the Ministry of Health and Welfare.
DISCUSSION
Acknowledgment
References
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 24 Jul 1998
Copyright©Japanese Journal of Clinical Oncology, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This Article 
Abstract
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Request Permissions
Google Scholar
Articles by Kawasaki, H
Articles by Nishiwaki, Y
Search for Related Content
PubMed
PubMed Citation
Articles by Kawasaki, H
Articles by Nishiwaki, Y
Social Bookmarking
What's this?