Fax: (713) 745-1921
Surgical management of soft tissue sarcomas of the hand and foot
Article first published online: 31 JUL 2002
Copyright © 2002 American Cancer Society
Volume 95, Issue 4, pages 852–861, 15 August 2002
How to Cite
Lin, P. P., Guzel, V. B., Pisters, P. W. T., Zagars, G. K., Weber, K. L., Feig, B. W., Pollock, R. E. and Yasko, A. W. (2002), Surgical management of soft tissue sarcomas of the hand and foot. Cancer, 95: 852–861. doi: 10.1002/cncr.10750
- Issue published online: 31 JUL 2002
- Article first published online: 31 JUL 2002
- Manuscript Accepted: 29 MAR 2002
- Manuscript Revised: 18 FEB 2002
- Manuscript Received: 12 DEC 2001
- soft tissue sarcoma;
- connective and soft tissue;
- local neoplasm recurrence
Soft tissue sarcomas of the hand and foot present unique management challenges. The purpose of the current study study was to determine oncologic outcome, particularly with respect to factors affecting local recurrence, distant recurrence, and disease-specific survival.
A retrospective study was performed on 115 patients with soft tissue sarcomas of the hand or foot who were evaluated, treated, and followed at the authors' institution between 1980 and 1998. The medical records and radiographs were reviewed. Kaplan–Meier analysis was used to assess patient survival.
Most patients (95%) were referred after previous surgery. The majority of tumors (75%) were T1 lesions (less than 5 cm), and most tumors (81%) were high grade. Patients who were treated by definitive, wide re-excision (n = 43) had a 10 year local recurrence-free survival of 88%, which was significantly better than the corresponding rate of 58% for patients who did not have re-excision (n = 40, P = 0.05). Radiation improved local control in patients who did not undergo re-excision (n = 17, P = 0.02). However, radiation did not improve local control in patients who had definitive re-excision with negative margins (n = 13, P = 0.51). The disease-specific survival at 5 and 10 years was 76% and 65%, respectively, for patients who presented with localized disease. Disease-specific patient survival was significantly worse for patients who had regional or distant metastasis. Radical amputation as initial surgical treatment did not decrease the likelihood of regional metastasis and did not improve disease-specific patient survival. The presence of distant metastasis at presentation was an independent predictor of local recurrence.
Limb sparing treatment is possible in many patients with soft tissue sarcomas of the hand and foot. Re-excision to achieve microscopically negative surgical margins is an effective method of achieving a high rate of local control in appropriately selected patients who present after unplanned excision of the primary tumor. There does not appear to be a survival benefit to immediate radical amputation, which should be reserved for cases where surgical excision or re-excision with adequate margins cannot be performed without sacrifice of functionally significant neurovascular or osseous structures. Cancer 2002;95:852–61. © 2002 American Cancer Society.
The concept of wide surgical excision has been validated for soft tissue sarcomas of the extremities.1–3 In many situations, wide excision can be combined with radiation to produce a high rate of local control.4–8 Limb sparing approaches do not appear to increase significantly the risk of distant relapse and death.9–14
In certain locations, anatomic constraints make it difficult to achieve wide margins. The hand and foot are two regions where there is very little readily expendable soft tissue that can be removed with the tumor to provide satisfactory oncologic margins. In addition, there are nerves, blood vessels, tendons, and other soft tissue structures at risk that are important for normal function. Sacrifice of these structures may result in significant functional loss.
There is limited published data pertaining to the surgical treatment of soft tissue sarcomas in the hand and foot.15–23 It is not certain that the accepted tenets of sarcoma treatment necessarily apply to hand and foot sarcomas. The optimal treatment for tumors in these locations has yet to be defined. It is therefore important to determine whether limb-sparing surgery can be justified for these tumors. Does limb-sparing surgery result in a high rate of local recurrence and is the survival of patients compromised? Another important issue is the role of radiation treatment. Although radiation is beneficial for local control of large, high grade soft tissue sarcomas, it is not always performed for small lesions, and it is not certain whether radiation is indicated in the hand and foot.
We reviewed our experience with these rare tumors over the past 20 years. To our knowledge, the current study represents one of the largest reported series of sarcomas of the hand and foot. The factors affecting local and distant recurrence were analyzed, particularly with regard to tumor characteristics and the different surgical treatments that were employed. The goal of the study was to determine whether limb-sparing surgery was indicated for these tumors and to determine whether there are any prognostic factors to guide treatment.
MATERIALS AND METHODS
During the period from 1980 to 1998, 192 patients were evaluated for a soft tissue sarcoma of the hand or foot at our institution. The patients were identified by searching the hospital computer database for all primary tumors of the hand and foot and then searching this subset for all diagnoses that qualified as a soft tissue sarcoma. Seventy seven patients were seen only in consultation and were not included in the current series. This left 115 patients who were evaluated, treated, and followed at this institution. These 115 patients form the cohort for the current study. No sarcomas of bone, skin cancers, or fibromatoses were included. Primary tumors extending to the level of the wrist and ankle joints were included, but tumors located proximal to the wrist or ankle skin creases were excluded. Patients were followed for a mininum of 12 months unless they died of disease. The median time of followup was 61 months (range, 2–241).
A retrospective review of medical records and radiographs was performed. The diagnosis was verified by reviewing the reports by the pathologists at our institution. The stage was established at the time of the patient's presentation at this institution. Tumors were staged according to the American Joint Committee on Cancer system.24 Tumors were divided into low and high grades according to histologic findings.
Treatment was based on multidisciplinary case reviews and may have included surgery, radiation, and/or chemotherapy. Radical amputations were defined as those occurring above the wrist or ankle joint, which removed the entire hand or foot. Partial amputations included digital, ray, transmetatarsal, and Syme amputations.
Local recurrence-free survival, distant relapse-free survival, disease-free survival, and disease-specific survival were assessed. Local recurrences were defined as tumors arising in the vicinity of the original tumor bed and surgical incision. These were distinguished from regional metastases, which occurred in the soft tissues, bones, or lymph nodes of the same extremity remote from the original tumor. Regional metastases were considered distant metastases for the purposes of outcome analysis. The analysis of risk factors for local recurrence was performed for the set of 95 patients who did not undergo radical amputation as the definitive surgical treatment.
Kaplan–Meier survivorship and other statistical tests were performed with the SPSS 10.1 statistical program. The log rank test was used to compare survival curves for univariate analysis. Patients who died of causes unrelated to the sarcoma or its treatment were censored for purposes of Kaplan-Meier analysis of disease-specific survival. The chi-square and Fisher exact test were also used to determine whether there was non random distribution of nonparametric factors. Multivariate analysis was performed using the Cox proportional hazards model. A parsimonious multivariate model was determined by selecting only covariates that contributed significantly to the final model. The chi-square statistic was the criterion for selecting covariates. Statistical significance was defined as P ≤ 0.05. Marginal statistical significance was defined as 0.05 < P ≤ 0.10.
Patient Demographics and Tumor Characteristics
The study included 55 females and 60 males. The mean age was 41 years (range, 5–81 years). There were 83 tumors of the foot and 32 tumors of the hand. Synovial sarcoma was the most common diagnosis. The relative frequency of different tumor histologies is shown in Table 1. The mean tumor size was 4.0 cm (median, 3.0 cm; range, 1–22 cm). Most of the tumors (75%) were T1 lesions (5 cm or less), and most (81%) were high grade. Twenty eight patients had regional or distant metastases at the time of referral, and 87 patients had localized disease.
|Diagnosis||No. of cases||Percentage|
|Malignant fibrous histiocytoma||20||17%|
|Clear cell sarcoma||12||10%|
Of the 115 patients, 111 patients had their initial surgery elsewhere. In six patients, the pre-referral surgery consisted of an incisional biopsy. In 72 patients the pre-referral surgery consisted of an unplanned type of local excision. In 21 cases a wide excision was reported. Twelve patients underwent amputation prior to referral: two digital, six below knee, one above knee, one above elbow, and two below elbow.
Sixty five patients underwent surgery after referral. Twenty two patients underwent radical amputation (16 below knee and 6 below elbow); 10 underwent partial amputation of the hand or foot; and 33 underwent wide re-excision of soft tissues. One below knee amputation was performed for a non healing wound without evidence of tumor recurrence.
Pathologic analysis of the re-excised specimens showed microscopic evidence of tumor in 70% (37 out of 53) of cases that did not have grossly recurrent tumors. This did not include the nine cases that presented with recurrent tumors after previous treatment.
The 5 and 10 year continuous local recurrence-free survivals for the cohort of 83 patients who did not undergo radical amputation as the definitive surgical treatment were 77% and 74%, respectively. The analysis of risk factors for local recurrence was performed for this set of 83 patients. Regional recurrences, which pertain to relapses outside of the original tumor bed and incision, are analyzed separately below.
The two primary predictive factors for local recurrence on univariate analysis were the presence of metastasis at presentation and whether wide surgical excision was performed after referral. Of the 83 patients who underwent an attempt at limb salvage, 11 presented with distant metastases. These patients were more likely to develop local recurrence by Kaplan-Meier analysis. Three of the 11 patients developed local recurrence, and all of these occurred within 36 months (P = 0.02).
Of the 83 patients who did not immediately undergo radical amputation, a subset of 43 patients underwent surgical treatment at this institution in an effort to preserve the limb. These patients underwent wide re-excision of the previous tumor bed (42 patients who had previous surgery) or primary wide excision of the tumor (1 patient who had no previous surgery). For the purposes of the current study, wide surgical excision included 9 cases of limited amputations (digital or ray amputation) following pre-referral surgery; 1 patient who underwent a digital amputation without previous surgery elsewhere; and 33 patients who underwent local soft tissue re-excision following pre-referral surgery.
The continuous local recurrence-free survival was 88% at both 5 and 10 years for the 43 patients who underwent wide excision, as defined above (Fig. 1). In comparison, the corresponding rates were 64% at 5 years and 58% at 10 years for the 40 patients who did not have re-excision (P = 0.05). Of these 40 patients, there were 12 patients (including 2 with digital amputations) who were reported by the outside surgeon to have undergone a limb-sparing, wide excision of the primary tumor prior to referral. These patients did not undergo re-excision of the wound. It was not possible to verify independently whether the patients indeed had a wide excision. These patients were therefore not included in the group who underwent re-excision or primary wide excision at this institution. The 5 and 10 year continuous local recurrence-free survivals for these 12 patients were 74% and 55% respectively.
The continuous local recurrence-free survival rate was 100% for the 10 patients who had a limited amputation after referral compared to 85% for the 33 patients who underwent re-excision of soft tissue only, but the difference was not significant (P = 0.29).
There were positive margins in 4 of 33 (12%) soft-tissue re-excisions, and 2 of 10 (20%) limited amputations (P = 0.59). Positive margins were associated with a greater risk of local recurrence (P = 0.05).
Radiation to the primary site was performed on 38 patients. The mean dose was 54 Gy (median, 60 Gy; standard deviation, 14.8 Gy; range, 17.5–66 Gy). Radiation improved local control in the group of patients who did not undergo re-excision (Fig. 2). The continuous local recurrence free survival rate at five years was 73% for the patients who had radiation compared to 28% for the patients who did not have radiation (P = 0.02). However, in the group of patients who underwent wide re-excision of the primary tumor, radiation did not have a significant impact on local recurrence-free survival (P = 0.51).
There were no other significant predictive factors for local recurrence on univariate analysis, including patient gender, tumor size, tumor grade, site of tumor, use of adjuvant chemotherapy, and histologic diagnosis. Adjuvant chemotherapy was used in 10 patients who had not previously undergone radical amputation. None of these 10 patients experienced local recurrence, but this did not reach statistical significance because of the small number of patients (P = 0.13). There were no local recurrences in the 10 cases of clear cell sarcoma, but this also did not reach statistical significance (P = 0.18).
On multivariate analysis of risk factors for local recurrence, the only covariate that emerged as an independent predictor was the presence of distant metastasis at the time of presentation (Table 2). In this analysis, only the 83 patients who did not undergo immediate radical amputation were included. Factors that were not selected into the final model included age, gender, site, size, grade, radiation treatment, surgical re-excision, limited amputation, adjuvant chemotherapy, histologic subtype, and microscopic surgical margin. Of the 11 patients who presented with metastasis, 3 patients developed local recurrence within 36 months. Kaplan-Meier analysis showed that the five year continuous local recurrence-free survival rate was 80.0% for patients without metastasis vs. 40.0% for patients with metastasis at presentation (P = 0.002).
|Factor||P Value||Wald Statistic||Regression coefficient (B)||Relative risk (eB)||95% CI|
|Metastasis at presentation||<0.001||15.6||2.0||7.1||2.7-18.9|
At the time of last followup, 37 patients had died of disease; 5 patients had died of other causes; 8 were alive with disease; and 65 patients had no evidence of disease. For the entire cohort of 115 patients, the disease-specific survival was 68% and 58% at 5 and 10 years, respectively. The distant relapse-free survival was 56% and 54% at 5 and 10 years, respectively. The continuous disease free survival was 49% and 47% at 5 and 10 years, respectively.
A strong predictive factor for survival was metastasis at the time of presentation (Fig. 3). Disease-specific survival at 5 and 10 years was 76% and 65% respectively for the 84 patients who presented with localized disease. In contrast, the disease-specific survival at 5 and 10 years was 24% and 16% for the 31 patients who presented with metastatic disease (P < 0.0001).
Local recurrence in the original tumor bed was not associated with statistically significantly worse disease-specific patient survival (Fig. 4). The disease-specific survival at 5 years was 78% for patients who did not have a local recurrence and 70% for patients who had local recurrence (P = 0.85). The results imply that the development of local recurrence may not reflect the likelihood of distant spread.
In sharp contrast, regional metastasis was associated with poor outcome and survival (Fig. 5). The disease-specific survival at five years was 86% for patients who did not develop regional metastasis and 38% for patients who developed regional metastases (P < 0.0001). The type of surgical treatment did not appear to influence the development of regional metastases, and radical amputation did not prevent regional metastases from occurring (Table 3). There was no significant association between histologic diagnosis and the development of lymph node or other regional metastasis.
|Surgical treatment||Regional metastasis||Total no. of cases||percentage|
Radical amputation was not associated with superior disease-specific or continuous disease-free survival (Fig. 6). In fact, radical amputation was associated with a worse prognosis among patients with localized disease, but this was not statistically significant. For patients who presented with localized disease, the disease-specific survival at 5 and 10 years was 80% and 72%, respectively, for patients who did not undergo radical amputation. The disease-specific survival at 5 and 10 years was 76% and 46% respectively for patients who underwent radical amputation (P = 0.12, Fig. 6A). Likewise, the continuous disease-free survival at 5 and 10 years was 66% and 63% respectively for patients who did not have radical amputation and 36% and 36% respectively, for patients who had radical amputation (P = 0.08, Fig. 6B).
Other factors were examined for their effect on survival. Males had a significantly worse prognosis compared to females; the five year disease-specific survival was 51% for men and 84% for women (P = 0.001). A greater number of males presented with metastases. Twenty of 59 males had metastases compared with 8 of 56 females at the time of presentation (P = 0.02).
There were no other factors that had a significant impact on survival. Specifically, the histologic diagnosis, grade, size, and location were not significant prognostic factors for survival.
Soft tissue sarcomas of the hand and foot are usually small lesions that are detected at a relatively early stage. Despite their small size, sarcomas of the hand and foot tend to be aggressive, and the observation has been previously made that these tumors may carry a worse prognosis than similarly sized tumors in other anatomic locations.22 Part of this may be due to the fact that most of the tumors are high grade lesions with a high potential for metastasis.
It is not surprising that most of the sarcomas had undergone some form of unplanned surgery prior to referral to an oncologist. Many tumors have an innocuous appearance, and they are often treated without biopsy or imaging studies. Sarcomas are rare, and benign soft tissue lesions outnumber malignant tumors by approximately 100:1.25 Although sarcomas may undergo unplanned excision in other anatomic locations, the implications in the hand and foot may be more serious. There is much more potential for contaminating vital anatomic structures, such as tendons, digital arteries, and nerves, than for comparable-sized lesions located elsewhere. It has previously been reported that unplanned excisions of tumors could make it more difficult to achieve wide margins with re-excision.26
The results of the current study support limb-sparing surgical re-excision for treatment of soft tissue sarcomas arising in the hand and foot. In 70% of the resected specimens, there was residual, microscopic tumor. Patients who underwent wide re-excision of the surgical bed had an 88% rate of local tumor control. Limb sparing surgery did not adversely affect disease-specific or disease-free survival.
Surgical re-excision may include limited amputations, such as digital amputations and ray amputations. The goal is to preserve the majority of the end organ without having to sacrifice the entire hand or foot. The function after digital or ray amputation in the hand is generally good. In contrast, complete ablation of the hand is quite debilitating and results in significant functional impairment.
Among patients who did not have re-excision, radiation treatment improved local control. However, it was not clear whether radiation provided a beneficial effect for patients who underwent wide re-excision with negative margins. Since the number of patients in the current study was relatively small, the lack of a statistically significant difference in local recurrence rates for radiated and nonradiated patients should not be considered strong evidence that radiation does not increase local control or decrease the need for radical amputation. Our previously reported experience with radiation treatment in the hand and foot showed a positive effect with little functional impairment and modest morbidity.27 It therefore seems reasonable to include radiation when the potential for local recurrence is perceived to be elevated.
Positive margins were associated with a greater risk of local recurrence. This finding is in agreement with previous studies.9, 10, 13 Postoperative radiation should be considered when margins are close or positive. Although there is potential for severe contractures to occur in the hand and foot following radiation, attention to rehabilitation during and following radiation can minimize the risk of post radiation complications. Most patients treated with high dose radiation to the hand and foot are able to attain excellent function without chronic pain.27
There was no survival benefit to immediate radical amputation. This parallels previously published findings for soft tissue sarcomas of the extremities.1 In the current study, prognosis was not improved by radical amputation. As anticipated, patients who were treated with radical amputation were still at risk for both regional metastases and distant metastases.
It is not possible to determine from this retrospective study whether wide re-excision and limb sparing surgery could be applied to patients with large tumors in unfavorable locations in the hand and foot. Tumors that are not covered by normal, expendable soft tissue may not be appropriate for attempts at limb preservation. Wide excision may be indicated only for relatively smaller lesions in favorable locations, such as the hypothenar muscles or the phalanges, where a ray amputation may allow for a wide margin without sacrifice of the entire hand. Thus, one must interpret the results of the current study with caution.
It was interesting that in the multivariate analysis of risk factors for local recurrence, the only covariate that emerged as an independent predictor was the presence of metastasis at presentation. This may reflect a more aggressive biologic potential in the tumors that metastasize early. Such tumors may be more likely to fail local treatment. However, it would not be necessarily appropriate to conclude that these patients should uniformly undergo radical amputation. The overall prognosis of the patients is unfavorable, and it is not at all certain whether an immediate radical amputation would improve the chances for survival.
Soft tissue sarcomas encompass a large number of different histopathologic entities. Although to our knowledge this is one of the largest reported series of soft tissue sarcomas of the hand and foot, the number of cases of any particular diagnosis is still small. Hence, it is not surprising that significant differences could not be found in the behavior of different tumors. At the present time, these tumors must still be grouped together, and the recommendations apply to the group as a whole. In the future, it is hoped that more specific recommendations could be made for individual histologic subtypes of soft tissue sarcoma.
In summary, soft tissue sarcomas of the hand and foot are frequently not suspected at initial presentation. Optimally, the initial surgery would be performed to achieve wide margins while preserving limb function. Wide re-excision is an effective method of achieving a high rate of local control and limb preservation after inadequate surgical excision. Radiation is indicated when margins are close or positive. There is not an inherent benefit to immediate radical amputation, and this should be reserved for cases where re-excision is not feasible with acceptable functional results.
- 20Synovial sarcoma of the foot and ankle. Clin Orthop. 1999; 364: 220–216., , .
- 24American Joint Committee on Cancer staging manual. Philadelphia: J. B. Lippincott, 1997., , , .