We investigated whether our policy of routine re-excision of the tumour bed after an unplanned excision of a soft-tissue sarcoma was justified.
Between April 1982 and December 2005, 2201 patients were referred to our hospital with the diagnosis of soft-tissue sarcoma, of whom 402 (18%) had undergone an unplanned excision elsewhere. A total of 363 (16.5%) were included in this study. Each patient was routinely restaged and the original histology was reviewed. Re-excision was undertaken in 316 (87%). We analysed the patient, tumour and treatment factors in relation to local control, metastasis and overall survival.
Residual tumour was found in 188 patients (59%). There was thus no residual disease in 128 patients of whom 10% (13) went on to develop a local recurrence. In 149 patients (47%), the re-excision specimen contained residual tumour, but it had been widely excised. Local recurrence occurred in 30 of these patients (20%). In 39 patients (12%), residual tumour was present in a marginal resection specimen. Of these, 46% (18) developed a local recurrence. A final positive margin in a high-grade tumour had a 60% risk of local recurrence even with post-operative radiotherapy.
Metastases developed in 24% (86). The overall survival was 77% at five years. Survival was related to the grade of the tumour and the finding of residual tumour at the time of re-excision.
We concluded that our policy of routine re-excision after unplanned excision of soft-tissue sarcoma was justified in view of the high risk of finding residual tumour.
Soft-tissue sarcomas are rare primary malignant tumours arising from tissue of mesodermal origin and comprise less than 1% of all malignant tumours. They can arise at any age and at virtually any site. Delay in diagnosis is common because they remain unrecognised. Most present as a painless lump which gradually increases in size. The mean diameter of these at presentation to our unit is 9.3 cm (0.5 to 54).1 Simple guidelines have been produced and widely circulated in an attempt to make an early diagnosis.2,3 These recommend that a lump with any one of the following features should be referred to a specialist centre:2,3 a diameter of more than 5 cm; a deep-seated tumour; increasing size, and pain.1
Because sarcomas are rare and because benign lumps are very common, it is inevitable that some soft-tissue sarcomas are inadvertently and inappropriately excised before the correct diagnosis has been made histologically.4 Referrals after an inadequate initial excision account for between 19% and 53% of the new patients seen in sarcoma centres.5–7
The management of the referred patient is based on a thorough review of the clinical information from the referring surgeon, including the results of any pre-operative investigations and details of the operation itself, as well as a formal review of the histology and the margins of excision. In most cases, the options include no further surgery, re-excision of residual tumour and consideration of adjuvant chemo- and radiotherapy, either individually or in combination.
A recent paper by Fiore et al8 found that the outcome was similar for patients with soft-tissue sarcomas whether they had undergone primary or re-excision surgery. They concluded that the presence of microscopic disease at re-excision was a marker of clinical aggressiveness. Lewis et al7 found that patients who had undergone re-excision surgery at a tertiary centre fared better than those who had their primary surgery there.
For 23 years, our unit has had a policy of recommending, whenever possible, a wide re-excision of the operative field when patients are referred after inadvertent excision of a soft-tissue sarcoma. The aim of this study was to investigate the results of this policy and to identify the factors which influence the patient’s prognosis in these circumstances.
Patients and Methods
We reviewed the relevant data from our prospective tumour database and included all patients who had undergone an unplanned excision of a soft-tissue sarcoma, who had been referred to our unit within three months of diagnosis and who had more than two years’ follow-up.
Over a 21-year period between April 1982 and december 2005 we treated 2201 patients for soft-tissue sarcoma, of whom 402 (18%) had undergone an unplanned excision elsewhere. Of these, we excluded 39 patients who had either not been referred immediately and who had an obvious local recurrence and those with less than two years’ follow-up, which left 363 patients for review (Table I⇓). A total of 223 patients (61%) had a superficial tumour and 140 (39%) a deep tumour. The mean diameter of the superficial tumours was 4.8 cm (0.2 to 23) and that of the deep tumours 6.4 cm (2 to 20). Although 117 patients (32%) had small, subcutaneous tumours, all but 15 had one of the other two criteria (i.e., an increase in size or pain) which should have raised a suspicion of malignancy. Metastases were present in 15 patients (4.1%); ten with high-grade tumours, four with intermediate grade and one with low grade. All patients were routinely restaged within three months of the original surgery and, whenever possible, underwent re-excision in an attempt to obtain clear margins. Re-excision was not undertaken in 47 (13%) patients either because further surgery would cause considerable disability or because the patient declined further surgery. Consequently, 316 patients (87%) underwent re-excision. In 38 patients (12%), more than one re-excision was required as the margins were not clear after the first re-excision. The resected specimen was examined histologically and placed in one of three categories based on Enneking’s classification of margins:9 1) no residual disease (sterile); 2) residual disease excised with wide margins (wide); 3) residual disease with close or contaminated margins (marginal).
Adjuvant radiotherapy was not routinely used at the start of the series except for patients with high-grade tumours or those whose final re-excision was judged to be marginal. Subsequently, radiotherapy was used more frequently and it is now our policy to offer radiotherapy to all patients who have tumours of more than 5 cm in diameter, intermediate or high-grade tumours, all those whose final excision is marginal and those who did not undergo re-excision of their tumour. In other words, radiotherapy is only avoided for low-grade tumours excised with a wide margin. In this series, adjuvant radiotherapy was used after re-excision in 118 of the 316 patients (37%) and in 19 of the 47 patients (40%) who did not have a re-excision. The mean follow-up was 40 months (24 to 180).
We analysed patient, tumour and treatment factors in relation to local control, metastasis and overall survival using Statview software (SAS Institute Inc, Cary, North Carolina). Differences between groups were assessed using the Mann-Whitney U test, and for nominal data, the chi-squared test. Survival was estimated using Kaplan-Meier survival curves with patients censored at the time of last follow-up. Hazard ratios (HR) were estimated using a Cox model with the relevant values entered as continuous or grouped variables. Multivariate analysis was carried out using a backwards stepwise variable selection procedure to identify the set of independent predictors of treatment failure. Significance was taken as p < 0.05.
Of the 316 patients who underwent re-excision, 128 (41%) had no residual tumour and were classified as sterile: 188 had residual tumour (59%), of whom, 149 had a wide margin and 39 a marginal margin after their final re-excision.
Residual tumour was present in 72% (80) of the high-grade tumours, 55% (71) of the intermediate and 49% (37) of the low-grade tumours, and in 55% (115) of the subcutaneous and 68% (73) of the deep tumours. The bigger the primary tumour, the greater the risk of there being residual tumour, but even with tumours measuring less than 5 cm there was a 53% risk of residual tumour, while those measuring more than 5 cm had a 68% risk. The greatest risk (83%) of residual disease was in high grade, deeply-situated tumours measuring more than 5 cm. In low-grade subcutaneous tumours of less than 5 cm, the risk was 40%.
When analysed by diagnosis, the risk of finding residual tumour was: dermatofibrosarcoma protruberans (85%), myxofibrosarcoma (80%), malignant peripheral nerve sheath tumour (66%), synovial sarcoma (57%), malignant fibrous histiocytoma (57%), liposarcoma (50%) and leiomyosarcoma (37%).
Local recurrence occurred in 76 of the 363 patients (21%) included in the study. Of these, 41 had high-grade tumours, 25 intermediate grade and 10 low-grade. The recurrence rates were 32% (41 of 127) for high-grade tumours, 17% (25 of 144) for intermediate-grade and 11% (10 of 92) for low-grade. The rate of local recurrence was 10% even when the re-excision was sterile, 46% when the re-excision was marginal, 20% when the re-excision was wide, and 34% when there was no re-excision. The risk of local recurrence was 25% when there was residual tumour and 10% when there was no residual tumour (chi-squared test, p = 0.002). The risk of local recurrence was 29% if the sarcoma was deeply situated and 16% if it was superficial.
The risk of local recurrence was virtually the same in high-grade tumours whether the re-excision was sterile or wide (23% and 25% respectively) but was almost three times greater if there was a marginal re-excision (67%). The risk of local recurrence was 42% for a malignant peripheral nerve sheath tumour or myxofibrosarcoma, 27% for a malignant fibrous histiocytoma but only 5% for a dermatofibrosarcoma protruberans. The five-year local recurrence risk was 17% when no residual tumour was found in the resection specimen, 33% when residual tumour was present and 32% when there was no re-excision.
When combined in a multivariate model, the only significant factors were high tumour grade, a diagnosis of myxofibrosarcoma or malignant peripheral nerve sheath tumour and a marginal final excision margin. Patients with a high-grade tumour and a marginal excision had a 67% risk of local recurrence compared with all other patients who had a risk of 19% (HR 5.10, 95% confidence interval (CI) 2.53 to 10.28; log-rank test, p < 0.0001). The risk of local recurrence was slightly greater in the first two years than in the subsequent two, but local recurrence was still being encountered for the first time up to seven years after re-excision (Fig. 1⇓).
It proved impossible to assess the effect of radiotherapy on local recurrence as this was only used in those patients who were thought to be at greatest risk (Table II⇓). In patients at greatest risk of local recurrence (high-grade tumours with a marginal excision) local recurrence occurred in six of the ten who received radiotherapy and in four of five who did not.
In the 117 patients who had small, subcutaneous tumours which had been previously excised, residual tumour was found in 57 (49%). Of the 117 patients, 12 (10%) developed local recurrence. Five of these local recurrences arose in the 60 patients with sterle re-excisions (8.3%) and seven in the 57 who had residual tumour (12.3%).
Metastases developed in 86 (24%) patients, most commonly in those who had undergone an incomplete resection for malignant fibrous histiocytoma (10, 33%) or synovial sarcoma (17, 27%). Metastases occurred in 10% of patients who had no residual tumour after re-excision, and 28% (13) of those who had no re-excision. Metastases occurred in 45% (57) of patients with a high-grade tumour, 17% (24) with an intermediate grade tumour and 5% (5) with a low-grade tumour.
The estimated overall five-year survival was 77%. On univariate analysis survival was related to the grade and size of the tumour but not to its depth. Older patients had a slightly worse prognosis (univariate analysis, p = 0.02). Survival was also related to the presence or absence of residual tumour and local recurrence and to the adequacy of the excision margin. On multivariate analysis, only the grade of the tumour and whether or not there was residual tumour were found to be significant (high-grade tumour HR 4.49, 95% CI 1.75 to 11.51, multivariate analysis, p = 0.0018; residual tumour HR 3.07, 95% CI 1.43 to 6.58, multivariate analysis, p = 0.0037).
Because benign lumps are seen more frequently than soft-tissue sarcomas, it is hardly surprising that many authors5,7,8,10–20 have reported a significant number of excised lumps which unexpectedly turn out to be sarcomas. Most of these lumps will be small and, particularly if they are subcutaneous, little harm may come from a prior unplanned excision. If however the tumour is large or deep, then it can be difficult to identify the extent of the original tumour at re-excision thereby increasing the risk of local recurrence, and this may prejudice the patient’s survival. Our results show that even after excision of small (< 5 cm) subcutaneous tumours, residual tumour was present in 49% with a risk of local recurrence of 10%.
Lewis et al7 found malignant fibrous histiocytoma (28%) and liposarcoma (26%) as common tumour types in their series of 1092 patients who underwent primary and re-excision of soft-tissue sarcoma. Fiore et al8 found that liposarcoma (36.2%) was the most common tumour which was re-excised. Goodlad et al11 found that myxofibrosarcoma (17%), leiomyosarcoma (17%) and liposarcoma (16%) were common types of tumour at re-excision. Synovial sarcoma (17%) was the most common tumour in the present series.
The mean diameter of the initial tumour was 5.5 cm (0.2 to 23) in this series compared with a mean size of 9.3 cm (0.5 to 54) for all soft-tissue sarcomas which presented to our hospital between 1998 and 2002.1 This may be because smaller tumours were thought to be benign. Fiore et al8 found the median size of tumour was 5 cm in the re-excision group and 9 cm (5 to 15) in the primary resection group.
We can confirm the observation from previous studies5,7,8,10–14 (Table III⇓) that the unplanned excision of a soft-tissue sarcoma results in incomplete excision in the majority of cases. The incidence of residual tumour in our series is as high as that of an earlier series11 from the United Kingdom. We believe that this justifies our policy of wide re-excision wherever possible.
We found residual tumour in 188 (59%) of the 316 re-excised specimens. No residual tumour was found in 128 (41%). We were unable to identify any one diagnosis or group of patients in whom one could be confident that residual tumour was unlikely. The true incidence of residual tumour is probably higher than 59% because 13 of the 128 patients who had no identifiable residual tumour still developed a local recurrence, so at least 201 of the 316 who had a re-excision (64%) must have had a residual tumour. It is also likely that, of the remaining 115 patients who had neither residual tumour identified nor developed local recurrence, some would have had residual tumour not identified by the pathologist but controlled by the re-excision or adjuvant radiotherapy.
There is no doubt that the inadvertent excision of a lump which turns out to be a soft-tissue sarcoma will leave residual tumour cells in the majority of cases, making further treatment, ideally a wide re-excision, essential.
The incidence of local recurrence was 21% in this series. Bauer et al15 found that the cumulative local recurrence rate was 20% at five years and 24% at ten years for patients treated at sarcoma centres in Scandinavia. Others have reported similar figures (Table III⇑).
Our study also showed that, despite further re-excision, the risk of local recurrence remains high. This risk was doubled when residual tumour was found in re-excised tissue, when the tumour was of high grade, when the re-excision margin was marginal and if the tumour was deep.
The value of adjuvant therapy after an attempted wide re-excision remains unclear from this and other studies. Chemotherapy has not been shown to be of benefit, even for most high-grade, large, deep soft-tissue sarcomas.21,22 Radiotherapy is routinely used in most centres for those sarcomas measuring more than 5 cm, or after marginal excision.7,8,15,20 These are, of course, the patients at greatest risk of developing a local recurrence: our inability to show any benefit for the use of adjuvant radiotherapy is probably the result of this. Furthermore, our use of radiotherapy expanded during the course of this study and we now use it routinely for all but the smallest and lowest grade tumours.
The overall survival was 77% in the present series which is comparable with other reported series of primary or re-excised soft-tissue sarcomas (Table III⇑). The cumulative five-year survival was lower when the tumour was of high grade, when there had been no re-excision, when residual tumour was found in the resected specimen and when the resection was marginal.
Lewis et al7 found that patients who underwent a re-excision fared better than those who had their primary operation at a sarcoma centre. This was thought to be because larger tumours, and tumours in difficult anatomical sites, were referred primarily to a sarcoma centre whereas smaller and more accessible tumours were excised inadvertently and referred thereafter. Fiore et al8 did not find a statistically significant difference between patients who had undergone a re-excision and primary excision at a sarcoma centre.
The high incidence of residual tumour identified in this series suggests that any patient who has had a lump excised which subsequently turns out to be a sarcoma should be referred to a sarcoma centre. That centre should stage the patient and should normally consider wide re-excision and the possible use of adjuvant therapy to maximise local control and survival. All surgeons should be aware of the current United Kingdom guidelines3 and should refer any patient with a lump bigger than 5 cm, or any lump increasing in size or deep to the fascia, for further investigation to a sarcoma centre rather than proceeding to an unplanned excision.
Supplementary tables showing the risk factors for local recurrence on univariate analysis and the five-year risk of tumour-related death of patients are available with the electronic version of this article on our website at www.jbjs.org.uk. An extended version of Table III⇑ and a supplementary figure showing the overall survival of all the patients in this study are also available on the electronic version of this article on our website.
The authors acknowledge and thank the surgeons, pathologists, radiologists, oncologists and the members of the multidisciplinary team who were involved in the treatment of the patients.
No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.
- Received May 31, 2007.
- Accepted September 28, 2007.
- © 2008 British Editorial Society of Bone and Joint Surgery