| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Report |
1 Department of Radiology, Memorial Sloan-Kettering Cancer Center, 1275 York
Ave., New York, NY 10021.
2 Weill Medical College of Cornell University, New York, NY 10021.
3 Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY
10021.
Received October 28, 2003;
accepted after revision January 30, 2004.
Presented at the 2004 annual meeting of the American Roentgen Ray Society,
Miami, FL.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. In 33 patients with proven tumorlets, a nodule was visible on CT in the same region as that of the resected specimen. Despite its ominous-sounding name, a pulmonary tumorlet represents benign tissue that may manifest as a subcentimeter pulmonary nodule and should be considered in the differential diagnosis of small pulmonary nodules visible on CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Bennett and Chew [11] reported on a patient in whom multiple pulmonary tumorlets were shown on CT and subsequently identified at histopathologic examination. The radiologic literature contains no other mention of the CT appearance of pulmonary tumorlets, to our knowledge. The purpose of this study was to determine the frequency with which pulmonary tumorlets can be seen on CT and to describe their appearance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Of the 35 patients, 28 (80%) were women and seven (20%) were men (mean age, 68.8 years; range, 51–83 years). Thirty-one patients had a diagnosis of cancer: lung cancer (n = 15), breast cancer (n = 6), carcinoid tumor (n = 5), lymphoma (n = 3), renal cell cancer (n = 2). One patient each had melanoma, prostate cancer, endometrial cancer, esophageal cancer, and thymic cancer; thus, five of the 31 patients had two types of cancer (lung and breast, n = 2; lung and lymphoma, n = 1; lung and prostate, n = 1; melanoma and endometrium, n = 1). Four patients had no known cancer diagnosis.
The CT scans of these 35 patients were retrospectively reviewed by two radiologists (one chest radiologist faculty member and one of two body imaging fellows). Interpretations were made in consensus. The radiologists were told at the time of review which lobe of the lung was partially or fully resected to direct their attention to that lobe. The dominant nodules or masses for which lung resection was performed were disregarded in this review, as were any incidental calcified granulomas; the sizes and numbers of any additional nodules in the specified lobe were noted.
Subsequently, these CT interpretations were correlated with the findings in the official pathology reports. These reports indicated that in addition to the lesions for which the surgical resection was performed, pulmonary tumorlets were present in the resection specimens. Multiple tumorlets were reported in the specimens from nine patients. Tumorlet size was not mentioned in every pathology report, but sizes specifically reported ranged from 0.5 to 5 mm.
The CT scans had been obtained 4–86 days (mean, 30.6 days) before pulmonary resection, with slice thicknesses ranging from 3.75 to 8 mm. Nineteen of these CT examinations were performed in our institution on a single-detector helical scanner or a 4-MDCT scanner (HiSpeed or Lightspeed, respectively; GE Healthcare); film images from the 16 other CT examinations that were performed at outside facilities were digitized into our PACS. Each scan was reviewed at lung window settings on a dedicated PACS workstation.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
CT did not reveal any pulmonary nodules in the remaining two patients. One of these patients had diffuse infiltrates in the lung that were confirmed at pathology to be interstitial fibrosis and diffuse alveolar damage, which probably hampered visualization of any nodules on CT. No nodule in the lobe in which a 3-mm tumorlet was found at histopathologic examination was seen on CT in the other patient.
In the 26 patients with a single tumorlet mentioned in their pathology reports, 17 (65%) had a single nodule visible on CT and seven (27%) had multiple visible nodules; the remaining two patients had no nodules visible on CT. Among the nine patients with multiple tumorlets mentioned in their pathology reports, eight (89%) had multiple nodules visible on CT; in the remaining patient, only one nodule was visible on CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The relationship between pulmonary tumorlets and carcinoid tumors has not been definitively established. It has been suggested that carcinoids can evolve from tumorlets. Tumorlets are substantially more common than carcinoid tumors, and most do not progress beyond the tumorlet stage [13–15], although the natural history of most tumorlets is unknown. The mechanism for any such progression may involve critical genetic alterations causing progressive growth deregulation [16]. Allelic imbalance of the int-2 gene has been studied and may be a useful discriminatory marker for pulmonary neuroendocrine neoplasia in the future. Int-2 allelic imbalance has been significantly associated with carcinoid tumor formation, whereas tumorlets showed allelic balance in patients having tumorlets or both tumorlets and carcinoids. Tumorlets do not appear to represent intramucosal spread from a carcinoid tumor because they are of different genotypes [11]. A histologic relationship among tumorlets, bronchial carcinoid tumors, and large cell carcinoma has been proposed, with the cell of origin in each being the bronchial counterpart of the intestinal argentaffin cell [17]. In one study [18], six of 84 patients who underwent resection for bronchial carcinoid tumors also had tumorlets associated with the primary tumor; these findings suggest that the presence of tumorlets, atypical carcinoid histology, and lymph node involvement portend a more malignant potential of the carcinoid tumor.
Interestingly, most of our study patients were women. A recent study [19] reported that almost a third of the patients with resected peripheral carcinoid tumors also had diffuse neuroendocrine cell hyperplasia and obliterative bronchiolitis; most of those patients were also women.
The term "tumorlet" was once used to refer to small tumors in other organs, including plexiform tumorlet of the uterus (a rare variant of epithelioid leiomyoma) [5, 6] and Wilms' tumorlet (a small deposit of Wilms' tumor) [7]. However, those nonneuroendocrine tumors are not related to pulmonary tumorlet and should not be confused with it.
Detection of one or more pulmonary nodules on CT can create a dilemma in patient treatment, especially in a patient with known cancer. Such nodules can be benign or malignant (primary or metastatic), each with different treatment options. The list of differential diagnostic possibilities for a pulmonary nodule is long and includes infectious, inflammatory, neoplastic, vascular, traumatic, and developmental causes. Most pulmonary nodules are of benign origin, even in patients with cancer [20]. To our knowledge, pulmonary tumorlet is typically not included as a cause of a pulmonary nodule in the radiology literature. However, our study shows that tumorlets can usually be detected on CT when they are large enough to be mentioned in pathology reports.
Our study has several limitations. The tumorlets identified at histopathologic examination could not be proven unequivocally to represent the specific nodules noted on CT; a point-by-point CT–pathologic correlation would be preferable in each case, but such data were not available in this retrospective study. Nevertheless, we believe that the combination of a pathology report mentioning a dominant lung nodule or mass and nearby tumorlets in the resection specimen, a CT examination of the same region showing a dominant nodule or mass and nearby tiny nodules, and a pathology report mentioning no other cause for tiny pulmonary nodules constitute compelling evidence that the pathologist and radiologist were describing the same entities in each case. Moreover, a tumorlet measuring a few millimeters in diameter should be visible on CT (assuming a favorable anatomic location in the lung). Therefore, we believe that it is highly likely that most nodular opacities seen on CT in our study actually represented the tumorlets found at histopathologic examination. Due to their small size, some tumorlets could have remained undetected on CT. The variable slice thicknesses used in these CT scans may have affected the number of nodules detected on CT.
Pathologists consider tumorlets to be incidental findings and may not have mentioned their presence in some cases, thus affecting our estimate of the frequency with which they are observable on CT. Also, pathologists may not search diligently for tumorlets because they are considered incidental findings, usually of no clinical importance. It is probable that a dedicated histopathologic examination of the resection specimens focusing on the presence of tumorlets would reveal even more tumorlets than our pathologists reported. It is also probable that pathologists would detect and report larger tumorlets more often than smaller ones; these larger lesions would also be more likely to be detected on CT. Because the patients in this study had pathologically documented tumorlets, we cannot estimate the prevalence of tumorlets that would be visible on CT in a different patient population. A prospective study is needed to determine the actual prevalence of pulmonary tumorlets visible on CT.
Despite its ominous-sounding name, a pulmonary tumorlet represents benign tissue that may manifest as a subcentimeter pulmonary nodule and should be considered in the differential diagnosis of small pulmonary nodules visible on CT.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
