Simple guidelines for efficient referral of soft-tissue sarcomas: a population-based evaluation of adherence to guidelines and referral patterns.

BACKGROUND: Optimal treatment of soft-tissue sarcoma requires multidisciplinary management at a sarcoma center. However, these rare tumors are often misinterpreted as benign and many are inadequately treated outside a sarcoma center, with an increased risk of local recurrence that often requires further extensive surgical treatment. To improve referral and centralization of soft-tissue sarcoma management in the southern Sweden health care region, an open-access outpatient clinic at our sarcoma center and simple referral guidelines have been established for the past thirty years. The guidelines call for referral of all deep-seated soft-tissue tumors and of all ≥5-cm superficial tumors before open biopsy or surgery. We evaluated adherence to these guidelines and characterized referral patterns. We also studied the consequences of our strategy with regard to the relative numbers of benign and malignant diagnoses among referred patients. METHODS: Adherence to guidelines, referral pathways, and time to referral to the sarcoma center were analyzed in a population-based series of 100 consecutive patients with soft-tissue sarcoma in the extremities or trunk wall. We also analyzed diagnosis and management of benign and malignant tumors in a second cohort consisting of 464 consecutive patients referred to the sarcoma center because of a soft-tissue tumor. RESULTS: Ninety-seven of the 100 patients with soft-tissue sarcoma were referred to the sarcoma center. All fifty-eight of the deep-seated soft-tissue sarcomas and twenty-eight of the forty-two superficial tumors were referred before open biopsy or surgery. Three-quarters of the patients with soft-tissue sarcoma first presented to a general practitioner. One-quarter of these patients were directly referred to the sarcoma center, which cut the referral time in half compared with patients initially referred to a local hospital. One-quarter of all patients referred to the outpatient clinic were diagnosed with a malignancy, with the majority of the malignancies being soft-tissue sarcoma. CONCLUSIONS: Our simple referral guidelines and open-access outpatient clinic resulted in nearly complete referral of patients with soft-tissue sarcoma to the sarcoma center. The "excess work" associated with referral of benign tumors according to our strategy was limited to the diagnosis of three benign tumors for each malignant tumor. We consider this surplus evaluation of benign tumors acceptable and probably necessary to achieve a high referral rate of soft-tissue sarcoma before initial surgery. LEVEL OF EVIDENCE: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

Deep-seated ordinary and atypical lipomas: histopathology, cytogenetics, clinical features, and outcome in 215 tumours of the extremity and trunk wall.

Deep-seated lipomas are often atypical histologically and are considered by some to have a high risk of recurrence after excision. We reviewed 215 deep-seated lipomas of the extremities and trunk wall with reference to histology, cytogenetics, clinical features and local recurrence. We classified tumours with atypical features and/or ring chromosomes as atypical lipomas. These were more common in men, larger than ordinary lipomas and more often located in the upper leg. The annual incidence was estimated as ten per million inhabitants and the ratio of atypical to ordinary lipomas was 1:3. In total, six tumours (3%), recurred locally after a median of eight years (1 to 16); of these, four were classified as atypical. The low recurrence rate of deep-seated lipomas of the extremity or trunk wall, irrespective of histological subtype, implies that if surgery is indicated, the tumour may be shelled out, that atypical lipomas in these locations do not deserve the designation well-differentiated liposarcoma, and that routine review after surgery is not required.

Identification of low-risk tumours in histological high-grade soft tissue sarcomas.

In more than one-third of patients with a histological high-grade malignant soft tissue sarcoma metastasis develops despite local control of the primary tumour. Hence, adjuvant chemotherapy is increasingly used for these relatively chemoresistant tumours which requires improved prognostication to exclude low-risk patients from overtreatment. We assessed the value of stepwise prognostication in a series of 434 histological high-grade STS of the extremity and trunk wall. Vascular invasion was used as the first discriminator whereafter the risk factors tumour necrosis, size (>8cm) and infiltrating growth pattern were used to discriminate high- and low-risk tumours. We identified a high-risk group with a cumulative incidence of metastasis >0.4 at 5 years, and a low-risk group, comprising half of the tumours, with a cumulative incidence of metastasis <0.15. The model was validated in an independent material of 175 patients. This model improved prognostication in STS and is of value for identifying patients who probably should not receive adjuvant chemotherapy.

Giant-cell tumour of the knee: the condition of the cartilage after treatment by curettage and cementing.

We reviewed nine patients at a mean period of 11 years (6 to 16) after curettage and cementing of a giant-cell tumour around the knee to determine if there were any long-term adverse effects on the cartilage. Plain radiography, MRI, delayed gadolinium-enhanced MRI of the cartilage and measurement of the serum level of cartilage oligomeric matrix protein were carried out. The functional outcome was evaluated using the Lysholm knee score. Each patient was physically active and had returned to their previous occupation. Most participated in recreational sports or exercise. The mean Lysholm knee score was 92 (83 to 100). Only one patient was found to have cartilage damage adjacent to the cement. This patient had a history of intra-articular fracture and local recurrence, leading to degenerative changes. Interpretation of the data obtained from delayed gadolinium-enhanced MRI of the cartilage was difficult, with variation in the T1 values which did not correlate with the clinical or radiological findings. We did not find it helpful in the early diagnosis of degeneration of cartilage. We also found no obvious correlation between the serum cartilage oligomeric matrix protein level and the radiological and MR findings, function, time after surgery and the age of the patient. In summary, we found no evidence that the long-term presence of cement close to the knee joint was associated with the development of degenerative osteoarthritis.

Increased risk of malignancies in a population-based study of 818 soft-tissue sarcoma patients.

Soft-tissue sarcomas (STS) have been associated with various rare cancer syndromes and occur at increased frequencies in survivors of childhood cancer. Also adult patients with STS have been suggested to be at an increased risk of additional malignancies. After exclusion of syndrome-associated and radiation-induced sarcomas, we studied multiple primary malignancies in a population-based cohort of 818 patients with primary STS of the extremities and the trunk wall. In total, 203 other malignancies developed in 164 (20%) patients median 10 (0-32) years before and median 4 (0-35) years after the sarcoma diagnosis. Standardised morbidity ratios (SMRs) were determined for primary malignancies following a STS. Hereby individuals who had developed a STS were identified to be at increased risk of second primary malignancies (SMR for all malignant tumours=1.3; 95% CI=1.0-1.5; P=0.02) with STS being the only specific tumour type that occurred at an increased risk (SMR=17.6; 95% CI=8.1-33.5; P<0.001). Hence, this population-based series demonstrates a high frequency of second primary tumours among STS patients and indicates a particularly increased risk of developing a new STS.

Treatment of local recurrences of giant cell tumour in long bones after curettage and cementing. A Scandinavian Sarcoma Group study.

We retrospectively studied local recurrence of giant cell tumour in long bones following treatment with curettage and cementing in 137 patients. The median follow-up time was 60 months (3 to 166). A total of 19 patients (14%) had at least one local recurrence, the first was diagnosed at a median of 17 months (3 to 29) after treatment of the primary tumour. There were 13 patients with a total of 15 local recurrences who were successfully treated by further curettage and cementing. Two patients with a second local recurrence were consequently treated twice. At the last follow-up, at a median of 53 months (3 to 128) after the most recent operation, all patients were free from disease and had good function. We concluded that local recurrence of giant cell tumour after curettage and cementing in long bones can generally be successfully treated with further curettage and cementing, with only a minor risk of increased morbidity. This suggests that more extensive surgery for the primary tumour in an attempt to obtain wide margins is not the method of choice, since it leaves the patient with higher morbidity with no significant gain with respect to cure of the disease.

Prognostic information in soft tissue sarcoma using tumour size, vascular invasion and microscopic tumour necrosis-the SIN-system.

We have earlier devised a system for soft tissue sarcoma (STS), based on three negative prognostic features: large tumour size, vascular invasion, and microscopic tumour necrosis, the SIN-system. Tumours which exhibit 2 or 3 of these features are categorised as high-risk, the others as low-risk. We have now tested this system for reproducibility both as regards recognition of its components, and as regards prognostic strength in patients from another institution. We have also compared it with the American Joint Committee on Cancer (AJCC) system. 200 patients with STS were analysed, all had been treated by surgery, in 97 patients combined with radiotherapy. The median follow-up for the 117 survivors was 10 (1.5-27) years. Without knowledge of the clinical data, three groups of pathologists independently reviewed original slides from all of the tumours. Based on the factors, the tumours were classified as high-risk or low-risk. The prognostic strength was compared using the results obtained by the different observers. Concordance in recognition of vascular invasion, tumour necrosis, and overall grading was seen in 156 (78%), 154 (77%), and 167 (84%) of the 200 tumours, respectively. Based on the different observers' grading, the cumulative 5-year metastasis-free survival rate (MFSR) varied for patients with low-risk tumours between 0.85 and 0.80, and for patients with high-risk tumours between 0.48 and 0.43. The Kappa-value for grading between all three groups of observers was 0.77. The SIN-system gave more clinically useful prognostic information than the AJCC system. Useful prognostic information in STS can be obtained by using tumour size, vascular invasion and microscopic tumour necrosis. This system provides two distinct prognostic groups, and has a high reproducibility.

Clinical impact of molecular and cytogenetic findings in synovial sarcoma.

Synovial sarcoma is an aggressive soft-tissue tumor that accounts for up to 10% of soft-tissue sarcomas. Cytogenetically, synovial sarcoma is characterized by the t(X;18)(p11;q11), found in more than 95% of the tumors. This translocation results in rearrangements of the SYT gene in 18q11 and one of the SSX1, SSX2, or SSX4 genes in Xp11, creating a SYT/SSX1, SYT/SSX2, or SYT/SSX4 chimeric gene. It has been shown that patients with SYT/SSX1 fusion genes have a shorter metastasis-free survival than do patients with SYT/SSX2. Previous studies have also suggested that clonal evolution may be associated with disease progression. In the present study, RT-PCR analysis showed that all 64 examined synovial sarcomas from 54 patients had SYT-SSX chimeric genes. SYT/SSX1 was found in 40 tumors from 33 patients, SYT/SSX2 in 23 tumors from 20 patients, and SYT/SSX4 in one case. Two patients had variant SYT/SSX2 transcripts, with 57 bp and 141 bp inserts, respectively, between the known SYT and SSX2 sequences. Patients with tumors with SYT/SSX1 fusions had a higher risk of developing metastases compared to those with SYT/SSX2 fusions (P = 0.01). The reciprocal transcripts SSX1/SYT and SSX2/SYT were detected using nested PCR in 11 of the 40 samples with SYT/SSX1 and 5 of the 23 samples with SYT/SSX2, respectively. Among 20 blood samples, SYT/SSX1 and SYT/SSX2 were detected in one sample each. The t(X;18), or variants thereof, was found cytogenetically in all patients but three. Among 32 primary tumors, the t(X;18) or a variant translocation was the sole anomaly in 10. In contrast, of the seven metastatic lesions that were investigated prior to radiotherapy, only one had a t(X;18) as the sole anomaly; all other tumors displayed complex karyotypes. Cytogenetic complexity in primary tumors was, however, not associated with the development of metastases. Tumors with SYT/SSX2 less often (4/12 vs. 7/15) showed complex karyotypes than did tumors with SYT/SSX1, but the difference was not significant. Combining cytogenetic complexity and transcript data, we found that the subgroup of patients with tumors showing simple karyotypes and SYT/SSX2 fusion had the best clinical outcome (2/8 patients developed metastases), and those with tumors showing complex karyotypes together with SYT/SSX1 fusion the worst (6/7 patients developed metastases). This corresponded to 5-year metastasis-free survival rates of 0.58 and 0.0, respectively (P = 0.02).

Clinicopathologic re-evaluation of 100 malignant fibrous histiocytomas: prognostic relevance of subclassification.

PURPOSE: Malignant fibrous histiocytoma (MFH) has been regarded as the most common soft tissue sarcoma (STS) in adults. Yet its true nature and the validity of this diagnostic concept have increasingly been questioned. Available data suggest that most patients with MFH can be subclassified into specific STS types, but the clinical relevance of such categorization has been argued. In a retrospective study, we reclassified 100 tumors of the extremity and trunk wall primarily diagnosed as MFH and analyzed the outcome. PATIENTS AND METHODS: Patients were adults (median age, 70 years; range, 32 to 94 years). The median tumor size was 8 cm (range, 1 to 30 cm), and the thigh was the most common tumor location (n = 31). Median follow-up was 8 years (range, 3 to 16 years). The overall 5-year metastasis-free survival rate was 0.64. The tumors were reanalyzed histologically, immunohistochemically, and, where available, ultrastructurally, and were classified according to strict diagnostic criteria. Patients were staged according to the American Joint Committee on Cancer system, and prognoses were compared among different groups of the reclassified diagnoses, paying special attention to myogenic tumors. RESULTS: In 84 of 100 tumors, a specific line of differentiation was either proved or strongly suggested. The most common diagnoses were myxofibrosarcoma (n = 22) and leiomyosarcoma (n = 20). Overall, 30 tumors could be grouped as some form of myogenic sarcoma. These tumors had a worse prognosis, even within the same American Joint Committee on Cancer stage, and a shorter time to metastasis than nonmyogenic tumors. CONCLUSION: This retrospective study confirms that most so-called MFH can be subclassified by defined criteria; it provides evidence that such classification is clinically important. Specifically, pleomorphic STS showing myogenic differentiation are significantly more aggressive, a finding that allows planning future therapeutic trials.

Monitoring referral and treatment in soft tissue sarcoma: study based on 1,851 patients from the Scandinavian Sarcoma Group Register.

This report is based on 1.851 adult patients with soft tissue sarcoma (STS) of the extremities or trunk wall diagnosed between 1986 and 1997 and reported from all tertiary referral centers in Norway and Sweden. The median age at diagnosis was 65 years and the male-to-female ratio was 1.1:1. One third of the tumors were subcutaneous, one third deep, intramuscular and one third deep, extramuscular. The median size was 7 (1-35) cm and 75% were high grade (III-IV). Metastases at presentation were diagnosed in 8% of the patients. Two thirds of STS patients were referred before surgery and the referral practices have improved during the study. The preoperative morphologic diagnosis was made with fine-needle aspiration cytology in 81%, core-needle biopsy in 9% and incisional biopsy in 10%. The frequency of amputations has decreased from 15% in 198688 to 9% in 1995-1997. A wide surgical margin was achieved in 77% of subcutaneous and 60% of deep-seated lesions. Overall, 24% of operated STS patients had adjuvant radiotherapy. The use of such therapy at sarcoma centers increased from 20% 1986-88 to 30% in 1995-97. Follow-up has been reported in 96% of the patients. The cumulative local recurrence rate was 0.20 at 5 years and 0.24 at 10 years. The 5-year metastasis-free survival rate was 0.70.

Recurrent chromosome aberrations in fibrous dysplasia of the bone: a report of the CHAMP study group. CHromosomes And MorPhology.

The nosologic status of fibrous dysplasia (FD), a well-known and relatively common bone lesion, is controversial. Information collected by the CHromosomes And MorPhology (CHAMP) study group on published and unpublished cases of fibrous dysplasia shows the presence of clonal chromosome changes in at least a proportion of these lesions. The chromosome aberrations found in FD lesions have been quite variable and have included both structural and numerical changes. Two of the three cases investigated at the study group had trisomy 2 as the sole acquired anomaly. Combined with previously published data, +2 and rearrangements involving chromosome band 12p13 have each been detected in 3 of 8 cases with abnormal karyotype of 11 in which chromosomal analysis has been performed, suggesting that FD is a neoplastic lesion rather than a "dysplastic" process, as has been generally believed and as implied by its very name.

Cytogenetic-morphologic correlations in aneurysmal bone cyst, giant cell tumor of bone and combined lesions. A report from the CHAMP study group.

Aneurysmal bone cyst and giant cell tumor of bone are relatively rare bone tumors that sometimes coexist. We examined the karyotypes of 3 aneurysmal bone cysts, 12 giant cell tumors, and 3 combined lesions. All aneurysmal bone cysts showed involvement of chromosome segments 17p11-13 and/or 16q22. In addition, in 1 of the 3 giant cell tumors with secondary aneurysmal bone cyst, both chromosome bands were rearranged as well, albeit not in a balanced translocation. Seven out of 12 giant cell tumors were characterized by telomeric associations. One giant cell tumor showed a dup(16)(q13q22), suggesting the presence of a (minor) secondary aneurysmal bone cyst component, despite the absence of histological proof. Our results, combined with literature data further substantiate that segments 16q22 and 17p11-13 are nonrandomly involved in at least some aneurysmal bone cysts, irrespective of subtype (primary, secondary, intra/extraosseous, solid or classic). These findings strongly suggest that some aneurysmal bone cysts are true neoplasms. In addition, telomeric associations are the most frequent chromosomal aberrations in giant cell tumor of bone, the significance of which remains elusive. In combined giant cell tumor/aneurysmal bone cyst each component seems to retain its own karyotypic abnormality.

Cytogenetic, clinical, and morphologic correlations in 78 cases of fibromatosis: a report from the CHAMP Study Group. CHromosomes And Morphology.

Whether fibromatoses are neoplastic or reactive lesions has long been controversial and the relationship, if any, between the superficial and deep forms (desmoid tumors) are poorly understood. Clinical, pathologic, and cytogenetic data of 78 cases of fibromatosis were analyzed and correlated with each other. The results demonstrate that clonal chromosome aberrations are a common feature of this entity, being present in 46% of desmoid tumors, although less frequent in the superficial types (10%). In the deep-seated extra-abdominal fibromatoses, trisomies 8 and 20 and loss of 5q material were the only recurrent features. No correlation between +8 and local recurrence was found. Our findings provide additional evidence for the neoplastic nature of fibromatoses.

Cytogenetic evidence of clonality in cutaneous benign fibrous histiocytomas: a report of the CHAMP study group.

AIMS: Cutaneous benign fibrohistiocytic tumours are among the most common soft tissue lesions. Their biological nature, in particular whether they are neoplastic or reactive, has long been disputed. Some morphological subtypes can be confused with sarcoma. Since available karyotypic data in these lesions are scarce, this study was undertaken to determine whether their cytogenetic analysis might demonstrate clonality and might help in differential diagnosis. METHODS AND RESULTS: Thirteen karyotyped benign cutaneous fibrous histiocytomas (BFH) were morphologically reassessed and classified as ordinary BFH (eight cases), cellular BFH (four cases), and one ankle-type lesion. Five cases (38%) showed clonal cytogenetic changes, although the aberrations varied and did not correlate with histological subtypes. Karyotypic aberrations were more common in cellular BFH (3/4) than in the ordinary BFH (2/8). CONCLUSIONS: The demonstration of clonal chromosome abnormalities, in at least some cases, supports the neoplastic nature of cutaneous BFH. The karyotypic changes identified are different from those in dermatofibrosarcoma, with which cellular BFH is often confused histologically.

Comparative cytogenetic study of spindle cell and pleomorphic leiomyosarcomas of soft tissues: a report from the CHAMP Study Group.

Leiomyosarcomas (LMS) of soft tissues frequently show complex karyotypic changes, and no specific aberration has been identified. The aim of this study was to search for recurrent chromosome aberrations in soft tissue LMSs and to correlate these, if present, with morphological and clinical parameters. From a series of soft tissue sarcomas thoroughly reexamined cytogenetically and histopathologically, 45 LMSs were retrieved; 35 were classified microscopically as spindle cell, 3 as epithelioid, and 7 as pleomorphic. Clonal chromosome changes were present in 14, 3, and 3 cases, respectively. This series was combined with 11 previously published, karyotypically abnormal pleomorphic LMSs for cytogenetic-clinico-histopathological correlations. The breakpoints were widely scattered, with no predilection of any of the recurrent breakpoints and losses to any of the morphologic subtypes. Combining numerical and unbalanced structural changes, the most frequently lost segments were 3p21-p23 (11 cases), 8p21-pter, 13q12-q13, 13q32-qter (10 cases each), 1q42-qter, 2p15-pter, 18p11 (9 cases each), 1p36, 11q23-qter (8 cases each), and 10q23-qter (7 cases). The most frequent gain was 1q12-q31 (6 cases). There was a greater frequency of losses in 1p and 8p and a lower frequency of losses in 10q and 13q in tumors that had metastasized than in localized tumors. We conclude that LMSs with clonal abnormalities display highly complex karyotypic changes and extensive heterogeneity. No significant correlation exists between these changes and age and sex of the patients, or with depth of tumor, topography, microscopic subtype, or tumor grade. Losses in 1p36 and 8p21-pter may be associated with increased risk of metastases. Comparison of our findings in soft tissue LMS with those previously reported in LMS in other locations suggest that the karyotypic profile is more dependent on site of origin than on microscopic features.

Radiation-associated sarcomas are characterized by complex karyotypes with frequent rearrangements of chromosome arm 3p.

Ionizing radiation is a well-known risk factor for sarcoma development. To investigate whether radiation-associated sarcomas are characterized by chromosome aberrations that distinguish them from de novo sarcomas, we identified those patients in our series of more than 500 cytogenetically abnormal sarcomas that fulfilled the following criteria: (1) each patient should have been irradiated for another malignancy at least 3 years prior to the sarcoma diagnosis, and (2) the sarcoma should have developed within the field of radiation. Ten patients fulfilling these criteria could be retrieved (median age at sarcoma diagnosis was 55 years, range 17-79; median latency period between primary tumor and radiation-associated sarcoma was 9 years, range 4-30). The diagnoses were typical for radiation-associated sarcomas: 2 each of malignant fibrous histiocytoma, leiomyosarcoma, and pleomorphic sarcoma, and 1 each of osteosarcoma, fibrosarcoma, myxofibrosarcoma, and spindle cell sarcoma. All 10 cases had relatively complex karyotypes with multiple, mostly unbalanced, structural rearrangements, similar to what has been reported in de novo sarcomas of the corresponding histologic subtypes. The only cytogenetic features that were unusually frequent among the radiation-associated sarcomas were the finding of unrelated clones in 3 cases, and loss of material from chromosome arm 3p, in particular 3p21-3pter, in 8 cases. Loss of the same chromosome segment has been described in 4 of the 8 previously published cases of radiation-associated sarcomas that have been analyzed after short-term culturing, which makes this imbalance significantly (P < 0.001) more frequent among radiation-associated sarcomas (12 of 18 cases) than among unselected cases of the corresponding histologic subtypes (74 of 282 cases). In contrast to the cytogenetic results, no 3p deletions were detected among the 6 cases of the present series that could be analyzed by comparative genomic hybridization (CGH). The most frequent imbalance detected by CGH was gain of 15cen-q15 (3 cases), followed by loss of chromosome 13 and gain of 5p, and 7cen-q22, each detected in 2 cases.

Cytogenetic characterization of peripheral nerve sheath tumours: a report of the CHAMP study group.

The findings of characteristic, sometimes pathognomonic, chromosome aberrations in several types of soft tissue tumours have not only added to our understanding of the mechanisms behind the genesis of these tumours, but have also revealed the importance of cytogenetic analysis as a diagnostic tool. For many soft tissue tumours, including peripheral nerve sheath tumours, the number of analysed cases is, however, still very low, precluding evaluations of the clinical or biological significance of different chromosomal patterns. As part of an ongoing project aiming at identifying clinical-histopathological-cytogenetic correlations among soft tissue tumours, a series of 46 benign, the vast majority of which were located in the extremities, and 20 malignant peripheral nerve sheath tumours (BPNSTs and MPNSTs, respectively) that had been successfully analysed by chromosome banding techniques were evaluated with regard to clinical, morphological, and cytogenetic features. Clonal chromosome aberrations were found in 20 BPNSTs, with abnormal karyotypes being significantly more frequent among Schwannomas than among neurofibromas. Recurrent aberrations, all of which were confined to the Schwannoma subtypes, included loss of 22q material, loss of a sex chromosome, and trisomy 7. The results show that the cytogenetic features of Schwannomas are not dependent on the site of origin. The MPNSTs, all of which had clonal chromosome aberrations, displayed complex karyotypes with numerous structural and numerical changes, except in two cases showing +7 and -22, respectively, as the sole changes. None of the recurrent imbalances was restricted to either NF1-associated or sporadic MPNST, nor was any of the imbalances significantly associated with clinical outcome. The presence of a triploid or tetraploid clone was, however, associated with grade 3 tumours and a poor prognosis. The cytogenetic findings in peripheral nerve sheath tumours show that the karyotype is a good discriminator between BPNSTs and MPNSTs, and that the pattern of aberrations among the latter may add prognostic information.

Clinical significance of genetic imbalances revealed by comparative genomic hybridization in chondrosarcomas.

DNA copy number changes were studied by comparative genomic hybridization (CGH) in 50 chondrosarcoma samples from 45 patients. Mean number of genetic aberrations in primary tumors was 4.8 +/- 1.8. The most frequently gained regions were 20q12-qter (37%), 20q (32%), 8q24.1-qter (27%), 20p (24%), and 14q24-qter (24%). Losses were 5.5 times less frequent than gains and observed mainly at Xcen-q21, 6cen-q22, and 18cen-q11.2 (11% each). Recurrent and metastatic tumors showed a mean of 4.0 +/- 2.2 aberrations per sample. The most frequently gained regions were chromosome 7 (4 cases), 5q14-q32 (4 cases), 6p (3 cases), and 12q (3 cases). Losses of DNA sequences were 3.4 times less frequent than gains. Histological tumor grade was significantly associated with metastasis-free survival (P = .002) and overall survival (P = .003), being the strongest prognostic factor tested. A statistically significant correlation was found between gain at 8q24.1-qter and shorter overall survival (P = .01) but not with local recurrence or metastasis-free survival. Gain at 14q24-qter was associated with a trend to shorter overall survival (P = .05) but neither with an increased risk for local recurrence nor with metastasis-free survival. In a multivariate analysis, only the tumor grade associated with overall survival (P = .02). In a multivariate analysis together with the tumor grade, gain at 8q24.1-qter did not retain its significance (P = .44), indicating that this imbalance is not an independent prognostic factor.