| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1
Department of Radiology B1D502G, University of Michigan Hospitals, 1500
Medical Center Dr., Ann Arbor, MI 48109-0030.
2
Department of Surgery, University of Michigan Hospitals, Ann Arbor, MI
48109-0030.
Received July 23, 1999;
accepted after revision September 8, 1999.
Presented at the annual meeting of the American Roentgen Ray Society, New
Orleans, May 1999.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. We reviewed the medical records and CT examinations obtained at the time the recurrence was diagnosed and tabulated data for all patients. Data for patients with high-grade malignancies were compared with those of patients with low-grade malignancies to determine whether there were differences in the interval between initial tumor resection and recurrence. We also compared CT appearances to determine patterns of recurrent disease.
RESULTS. Twenty-five of 33 recurrences were detected within 2 years of initial surgery. Only 16 patients had symptoms, and when present, most symptoms were nonspecific. In 28 (85%) patients, recurrent tumor was in the abdomen at the time of diagnosis. In nine patients, the largest detectable abdominal tumor was less than 5 cm in diameter. Interval to recurrence was similar for patients with low- and high-grade tumors. Although the CT appearance was similar for both grades, distant metastases were identified only in patients with high-grade malignancies.
CONCLUSION. Primary retroperitoneal malignancies frequently recur within 2 years of initial surgical resection. For asymptomatic patients, diagnosis is typically made during routine follow-up CT. Most patients have abdominal recurrences that may be small when first detected.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The most effective treatment of recurrent retroperitoneal sarcomas is additional surgical resection [4, 7]. Because surgery plays such a central role in the treatment of recurrent cancers, it is beneficial for recurrences to be identified when still small in size, improving the likelihood of complete resection. Therefore, radiologists should be familiar with the cross-sectional imaging appearances of recurrences to maximize the chances of early detection.
To date, there has been a paucity of information on the imaging of recurrent retroperitoneal sarcomas. For this reason, we evaluated the CT appearance of a variety of recurrent retroperitoneal tumors over a 5-year period. We determined the interval between initial surgery and recurrence, the CT characteristics and patterns of recurrence, and the efficacy of subsequent treatment.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The recurrent tumor cell type, grade, and location were recorded, and the interval between initial resection and recurrence was calculated. At the time of recurrence, patient symptoms, if any, were identified. The treatment for the recurrence (if any) and the survival rate after treatment was noted.
Each CT revealing recurrence was independently evaluated by two experienced radiologists. CT scans were evaluated for each of the following: location of recurrence (subdivided into three patterns: in or near the original tumor bed, in the abdomen but located outside of the original tumor bed, or at distant extraabdominal sites); maximum diameter of the largest recurrence; and internal CT attenuation characteristics (the presence or absence of grossly visualized fat and heterogeneity).
CT technique varied because some of the CT examinations were performed on 9800 scanners (General Electric Medical Systems, Milwaukee WI) and others were performed at outside hospitals on different equipment. Thirty patients received oral contrast material, and 27 patients received IV contrast material (100-200 ml of 60% by weight iodinated contrast media). CT scans were obtained using 5-mm contiguous slices in two patients, 7-mm slices in seven patients, and 10-mm slices in 24 patients.
In nine patients, CT scans obtained at the time of recurrence were compared with initial CT scans. We compared the CT scans to determine whether the appearance of the recurrence was similar to the appearance of the initial tumor. Unfortunately, because most patients were referred to our institution after receiving their original diagnoses, their initial CT scans were unavailable for review. However, we were able to determine the initial tumor location by reviewing the operative note or the CT report from the outside imaging study.
Data for all patients were analyzed as a group. Subsequently, data for patients with low-grade tumors were compared with those of patients with high-grade tumors to determine whether there were differences in the interval between initial tumor resection and recurrence. We also compared the imaging characteristics of recurrent disease. We used the Mann-Whitney test to compare the interval between initial tumor resection and recurrence for patients with high- and low-grade tumors. A nonparametric test was used because a normal distribution could not be assumed. The Fisher's exact test was used to compare the locations of abdominal recurrences (local or nonlocal) for patients with high- and low-grade tumors.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cell Type and Grade of Initial Tumor
Eleven patients' tumors were originally diagnosed as liposarcomas, nine as
leiomyosarcomas, six as malignant fibrous histiocytomas, two as
paragangliomas, one as embryonal rhabdomyosarcoma, one as spindle cell tumor,
one as desmoplastic round cell tumor, one as hemangiopericytoma, and one as
chondrosarcoma.
Grades of the original retroperitoneal tumors were available for 28 patients. We identified six grade I tumors: four liposarcomas, one hemangiopericytoma, and one leiomyosarcoma. We identified four grade II tumors: two liposarcomas, one leiomyosarcoma, and one chondrosarcoma. We identified 18 grade III tumors: six malignant fibrous histiocytomas, five liposarcomas, five leiomyosarcomas, one embryonal rhabdomyosarcoma, and one desmoplastic round cell tumor. A grade was not assigned for two paragangliomas, two leiomyosarcomas, and one spindle cell tumor.
Cell Type and Grade of Recurrent Tumor
In two of 33 patients, the cell type of initial tumors differed from that
of recurrent tumors. In one patient, the original mass, located in the
retroperitoneum adjacent to the left kidney, was diagnosed as a myxoid tumor.
Eleven months later, when a mass in the operative bed was biopsied during
repeat surgery, the recurrence was classified as a rhabdomyosarcoma. In
another patient, a large right retroperitoneal mass was initially diagnosed as
liposarcoma, and a pathologic specimen obtained 12 months later was diagnosed
as malignant fibrous histiocytoma.
In 26 of 28 patients for whom initial tumor grade was available, the grade of the recurrence was the same as the grade of the original tumor. In one patient, liposarcoma was originally classified as grade II, with scattered foci that were grade III; however, the recurrent tumor, appearing 6 years after initial surgery, was exclusively grade III. In another patient, grade II liposarcoma recurred 25 months later containing grade III components.
Interval Between Initial Resection and Recurrence
Twenty-five (76%) of 33 recurrences were detected within 2 years of initial
surgical resection (Fig. 1).
Eight recurrences were detected more than 2 years after surgery (five were
diagnosed between 2 and 5 years and three were diagnosed more than 5 years
after initial surgery). There was no statistically significant difference in
the intervals between initial surgical resection and recurrence when patients
with grade I tumors were compared with patients with grade III tumors
(Fig. 2). In 17 patients,
recurrences were detected on the first CT after surgery; however, 16 patients
had prior negative CT examinations after surgery.
|
|
Symptoms at the Time of Recurrence
Sixteen of 33 patients had symptoms at the time their recurrences were
diagnosed, including 14 of 25 patients who recurred less than 2 years and two
of eight patients who recurred more than 2 years after initial surgery. In
most patients, the symptoms, when present, were nonspecific consisting of
abdominal pain (eight patients), weight loss and fatigue (five patients),
fevers (two patients), and headache (one patient). Seven patients complained
of abdominal fullness or the sensation of a mass. Recurrence was diagnosed on
routine followup CT in all asymptomatic and in most symptomatic patients.
Location of Recurrence
In most patients (28/33; 85%), recurrent tumors were present in the abdomen
at the time of diagnosis (Table
1), and local recurrences were present in nearly half (n
= 15) of the patients. In 10 patients with local recurrences, disease was
present only locally (Fig. 3).
Local and nonlocal recurrences were present in four patients. Nonlocal
recurrences were found in the mesentery, peritoneum, retroperitoneum, and
liver. One patient with a local recurrence also had distant (lung)
metastases.
|
|
In 13 patients, abdominal recurrences were located only outside of the operative bed. Ten of these patients had mesenteric or peritoneal masses, six had masses in the retroperitoneum not local to the primary (Fig. 4), and seven had liver metastases. One patient (with a recurrent grade III liposarcoma) also had distant (lung) metastases. Four of seven patients with liver metastases had leiomyosarcomas.
|
There was no significant difference in the distribution of abdominal disease between patients with high- and low-grade tumors.
In five of seven patients with extraabdominal disease, recurrent disease was located only at distant sites. Four of these patients had lung metastases, one had a tumor in the anterior aspect of the right thigh, and one patient (who also had lung metastases) had metastatic disease in the supraclavicular lymph nodes. Of the five patients who only had distant metastases, three had grade III tumors, one had a grade II tumor, and one had an ungraded tumor. None of the patients with extraabdominal metastases had grade I tumors.
Size of Recurrence
We measured the size of extrahepatic tumors only because identification of
small hepatic metastases is straightforward on CT; however, identification of
small retroperitoneal, mesenteric, or peritoneal masses can be more difficult.
Of 22 patients with recurrent extrahepatic abdominal tumors, nine had masses
that measured less than 5 cm in maximum diameter. In one of these patients,
the largest detectable recurrent mass measured less than 2 cm. Therefore, in
several instances, the recurrent tumor was very small (Fig.
5A,5B).
|
|
CT Attenuation Characteristics
In 14 of 22 patients with extrahepatic abdominal disease, tumors appeared
heterogeneous (containing areas of soft-tissue, near-water, or fat
attenuation) on CT. Seven patients had recurrent liposarcomas, four had
recurrent leiomyosarcomas, two had recurrent malignant fibrous histiocytomas,
and one had recurrent rhabdomyosarcoma.
In eight patients, the CT appearance of recurrent liposarcoma was compared with that of the original tumor, and in four, the CT attenuation characteristics had changed. Although all liposarcomas contained recognizable fat on initial CT scans, only four of the recurrent tumors contained identifiable fat. The remaining four recurred as masses containing areas of soft-tissue and near-water attenuation only (Fig. 6A,6B).
|
|
In two of four fatty attenuation liposarcoma recurrences, the presence of fat in the tumor made identification more difficult. The recurrence was identified in these patients because its appearance differed subtly from that of normal surrounding retroperitoneal fat. In one patient, the fat in the tumor was slightly more heterogeneous than retroperitoneal fat (because it contained multiple tiny linear areas of soft-tissue attenuation) (Fig. 7). In another patient, the fat in the tumor was only minimally heterogeneous but had slightly increased attenuation when compared with that of the normal retroperitoneal fat (Fig. 8A,8B).
|
|
|
Eight patients with extrahepatic abdominal recurrences had masses that appeared homogeneous on CT. In two of these patients, IV contrast material was not administered. The other six patients had masses that were homogeneous on contrast-enhanced CT scans, including two with malignant fibrous histiocytomas, one with liposarcoma, one with leiomyosarcoma (Fig. 9), one with spindle cell tumor, and one with hemangiopericytoma. The homogeneous liposarcoma recurrence was comprised exclusively of soft-tissue attenuation.
|
Treatment and Follow-Up
Recurrent tumors were reresected in 16 patients without distant metastatic
disease and in the one patient with a distant metastatic recurrence in the
thigh. Reresection was not performed in any of the patients who recurred
exclusively with distant disease. All patients with exclusive distant disease
had multiple lung nodules, and pulmonary metastasectomy would not have
improved patient survival. Nine patients received chemotherapy, one received
palliative radiation, and two entered hospice programs. Four patients were
lost on follow-up.
Repeated rerecurrences developed in 10 of 17 patients in whom reresections were performed with eight recurring again within 2 years of repeat surgery (at 3, 5, 5, 7, 9, 12, 13, and 18 months). Two of these rerecurrences occurred at 27 and 60 months. All 10 rerecurrences developed in the abdomen and six were local to the region of the original recurrence.
Of the seven patients who underwent reresection for repeated recurrent tumors and in whom no recurrence developed, one patient died from postoperative sepsis, and six are still being followed up (only one patient is 2 or more years past reresection).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our study, recurrences typically appeared within 2 years of initial surgery. This was true for low- and high-grade tumors. These data are similar to those presented previously. For example, in a study by Glenn et al. [2], 75% of tumors recurred in the first 2 years. In a study by Storm and Mahvi [21], 72% of patients who underwent complete resection developed recurrence within 5 years. Unfortunately, these researchers did not assess the frequency of disease recurrences at intervals less than 5 years.
In our study, some tumors of both low and high grades did not recur until more than 5 years after the initial surgery. In several patients, the diagnosis was made only after the patient became symptomatic after presumed cure. Delay in the appearance of some sarcoma recurrences is well known [20]. In the experience of Heslin et al. [20], 40% of patients who were disease free at 5 years developed recurrences within 10 years after initial surgery.
Although some patients with recurrent retroperitoneal sarcomas have distant metastatic disease (usually in the lungs), most have some evidence of recurrence in the abdomen [7, 9, 11, 12, 15, 22,23,24,25,26]. In our series, 28 of 33 patients had abdominal recurrences. The distribution of disease was almost evenly split between recurrences in the primary tumor bed and recurrences exclusively elsewhere in the abdomen. Recurrences that were not in the primary tumor bed were usually elsewhere in the retroperitoneum, peritoneum, or mesentery. In a review by McGrath [22], up to 75% of first recurrences occurred at the site of the original tumor. Similarly, a study by Van Doorn et al. [7] found local recurrences in 23 of 32 patients with primary retroperitoneal sarcomas.
Primary retroperitoneal neoplasms typically originate as large masses. In one series [27], the average maximum diameter at the time of original diagnosis ranged from 11 to 20 cm (the variation in size was caused by differences in cell type). However, in our study, recurrences were frequently much smaller at the time that they were detected. In nine of 22 patients with extrahepatic abdominal recurrences, the largest tumor measured less than 5 cm in diameter.
In our study, most recurrent tumors were heterogeneous on CT (an appearance that frequently helped facilitate their detection); however, the recognition of the subgroup of recurrent liposarcomas was more difficult. CT attenuation characteristics of some tumors differed from that of the primary tumor. Fatty attenuation components could not be recognized in four of eight recurrent liposarcomas that originally contained identifiable fat. In addition, two small fatty attenuation recurrent tumors were difficult to distinguish from normal or uninvolved retroperitoneal fat because they blended with surrounding uninvolved retroperitoneal fat. For these reasons, radiologists must carefully scrutinize the operative bed and the abdominal cavity in patients for whom a retroperitoneal liposarcoma has been removed. New soft-tissue attenuation masses may represent recurrent tumor, even for patients in whom the primary liposarcomas contained fat. Also, subtle differences in the attenuation of fat and areas of fat-containing increased soft tissue must be viewed with suspicion for possible recurrence.
Our study had several limitations. First, the number of patients was relatively small. Second, in many instances the initial evaluation and, sometimes, the initial surgery were performed at outside institutions. Data from these institutions were not always complete, and copies of the initial cross-sectional imaging studies were frequently unavailable. Third, our study probably has selection bias because many of our patients were referred to us from other facilities. However, despite these limitations, our results are significant and valuable.
In summary, regardless of tumor cell type or grade, detectable recurrences of primary retroperitoneal sarcomas typically occur in the abdomen within the first 2 years after surgical resection. Additionally, primary retroperitoneal sarcomas are typically detected on routine follow-up studies of asymptomatic patients. Follow-up cross-sectional imaging must be examined with a high level of suspicion, especially when the primary neoplasm is a liposarcoma that contains grossly visualized fat. At our institution, all patients undergo initial imaging (with abdominal and pelvic CT) 4-6 months after the initial surgical resection. Subsequent patient follow-up is then individualized for each patient. However, in general, patients with grade I tumors undergo annual imaging and patients with grade II or III tumors undergo semiannual imaging for 5 years. Thereafter, all patients undergo annual imaging because recurrences of low- and high-grade tumors may appear late in some patients.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
