Functional Outcome in Sarcomas Treated With Limb-Salvage Surgery or Amputation

Purpose. In all patients treated at the Centre for Bone and Soft Tissue Sarcomas of Aarhus the functional outcome is prospectively evaluated by use of the Enneking system for the functional evaluation after surgical treatment of tumours of the musculoskeletal system. This system has been accepted by the Musculoskeletal Tumour Society and the International Symposium on Limb Salvage. Patients/methods. In the present study the functional outcome after limb-salvage surgery (89 patients) and amputation (58 patients) was compared. In the limb-salvage group the treatment was surgery alone in 50% and surgery combined with either radiotherapy in 39% or chemotherapy in 11%. Inclusion criteria were: Deep seated extremity sarcomas, age >14 years, more than 1 year post-treatment follow-up time and alive at the end of the study. Median age was 49 years (range 14–88 years). Median tumour diameter was 8 cm (range 1–20 cm), median follow-up time was 4.8 years (range 1–11 years). Wilcoxon and χ2-tests were used for statistical analyses. Results. The two groups were comparable according to age, sex, size of tumour, type of tumour, location of tumour, as well as post-treatment follow-up time. The functional scores were significantly higher after limb-salvage surgery as compared to amputation, the median scores being 85 and 47, respectively (p<0.001). A similar difference was observed if the Enneking scores were subdivided into general health-related scores and extremity-related scores. No association was found between functional scores and the following factors by use of univariate analysis: size of tumour, radiation therapy, localization of tumour and surgical margin. Discussion. We conclude that this study indicates that limb-salvage surgery is associated with a better functional outcome than that observed after amputation. However, whether this also indicates a difference in quality of life needs further studies.


Introduction
Previously, patients with extrem ity tumours were routinely treated with am putation. W ithin the last 15± 20 years the preferred treatment has shifted, and today the m ajority of patients are offered lim bsalvage surgery. Several factors have played a role in this shift of treatment strategy, the most im portant being: the developm ent within the ® eld of adjuvant therapies in the form of radiation therapy (soft tissue tum ours) and chemotherapy (bone tumours), im proved im age diagnostics such as m agnetic resonance im aging (M RI) and com puted tomograph y (C T) as well as im proved techniques of excision and reconstruction. 1,2 Am putation is an extensive and often invalidating procedure, but in spite of this the patient can often be discharged and m obilized after a relatively short period of time. In contrast, lim b-salvag e surgery m ay cause long hospitalization and result in a not optim ally functioning lim b. D ue to individual factors, such as age, social and personal relations as well as the nature of the tum our, it is im portant to be able to offer each patient an individual treatm ent. Since sarcom a is a rare disease and the patients should be offered a m ulti-disciplinary treatm ent, the m anagem ent of these patients should only be perform ed in centres with expertise in treatment of sarcom as. 3,4 In order to apply the m ost suitable treatment for each patient, it is of great importance to be able to evaluate a given treatment. In this perspective, a standardized, validity-tested system of evaluation of function is necessary. Several system s of evaluation have previously been used. However, these system s have focused on the function of the operated extremity rather than the general condition of the patient. Furtherm ore, these system s have often been based on the doctor' s judgem ent, and studies con® rm ing that the data obtained by these m ethods agree with that of the patient are lacking.
A system for functional evaluation following tum our surgery described by Enneking has been ® eld tested in 1989 by the M usculoskeletal Tum our Society (M ST S) and adopted by the M ST S and the International Syrnposium on Limb-salvag e. l T he system is based on a questionnaire and a sim ple clinical test.
The aim of the present study was to investigate the functional outcom e after am putation and lim bsalvage surgery of sarcom as by use of the Enneking system . In addition, the in¯uence of different factors on the functional results was exam ined.

P atients and m aterials
F or all patients surgically treated for sarcomas at the C entre for Soft Tissue and Bone Sarcom as at the U niversity Hospital in Aarhus during the period 1983± 1995, inform ation concerning treatm ent and tum our have been registered. After treatment, all patients were regularly seen in the outpatient clinic for observation of possib le recurrence. T he postoperative functional outcome was prospectively exam ined by use of the Enneking system for functional evaluation, w hich consists of a questionnaire and a sim ple clinical test. T he questionnaires were ® lled out independently by the patients, and the clinical test was carried out by the doctor present in the outpatient clinic. T he system assigns num erical values (0± 5) for each of six categories: pain, function and emotional acceptance in the upper and in the lower extrem ity, respectively. In addition, in the lower extremity, supports, w alking as well as gait, and in the upper extrem ity, hand positioning, dexterity as well as lifting ability were assigned. To allow a comparison of results, num erical scores and percentage ratings were calculated. T he system has been ® eld tested, and is accepted and recomm ended by the MST S. 1 Patients were selected for the study based on the following criteria of inclusion: age 14 years or above; m alignant, deeply localized extremity tum ours; at least 1 year of post-treatment follow-up time; alive at the end of the study; am putation or lim b-salvage surgery. T he general principles for perform ing either am putation or lim b-salva ge surgery w ere as follows. In the ® rst part of the period, com partm ental resection or am putation were aim ed at for highly m alignant tum ours. If this w as not achievable due to the localization or dissem ination of the tum our, resection with the largest possib le m argin followed by radiation therapy w as carried out. Later, this principle was changed to com bined surgery and radiotherapy in m ost patients.
The criteria of inclusion were ful® lled by 147 patients. T he patients were divided in tw o groups according to the type of operation that was perform ed; lim b-salvag e surgery or am putation.

The limb-salvage surgery group
Eighty-nine (46 females, 43 males) patients had limbsalvage surgery. Median age was 50 years (range 14± 87 years). The median diameter of the tumours was 8 cm (range 1± 20 cm). Twenty tumours were , 5 cm in diameter and 69 were > 5 cm. Twenty-one patients had bone tumours and 68 soft tissue tumours. Twenty-seven tumours were localized in the upper extremity and 62 in the lower extremity. Twothirds of the bone tumours were localized in tibia, humerus and femur, while more than half of the soft tissue tumours were localized in the thigh. Chondroand osteosarcomas dominated in the bone tumours accounting for more than 90%. In the soft tissue tumours, liposarcomas and malignant fibrous histiocytoma (MFH) accounted for more than 50%.
The histopathogical grades of the tumours were determ ined on basis of m icroscopy of the rem oved tum our. 5 Thirty patients had grade I tum ours, 20 grade II, 16 grade IIIA, 11 grade IIIB and 12 patients had m alignant tum ours whose histological grades could not be described m ore precisely. F orty-® ve patients in this group received limb-salvag e surgery as the only treatment, while 34 patients were treated with com bined lim b-salva ge surgery and radiation therapy, nine patients with com bined lim b-salva ge surgery and chem otherapy, and one patient with lim b-salvag e surgery as w ell as radiation therapy and chem otherapy. In general, chem otherapy w as given to patients with bone tum ours, and radiation therapy to patients with soft tissue tum ours. In this context, the term`com bined' m eans that radiation therapy or chemotherapy were given at som e point in the course of treatment, not necessarily adjuvant to the prim ary surgical treatm ent. Yet for 91% of the patients the radiation therapy was given adjuvant to the primary surgical treatment.

The amputation group
Fifty-eight patients were treated with am putation (40 m ales, 18 fem ales). M edian-age w as 47 years (range 14± 88 years). T he m edian tum our size w as 8 cm (range 1± 20 cm). T en tum ours were , 5 cm in diam eter and 48 were > 5 cm .
Twenty-® ve patients had bone tumours and 33 patients had soft tissue tum ours. Forty-two (72%) tum ours were localized to the lower extremities and 16 (28%) to the upper extrem ities. Two-thirds of the bone tumours w ere localized in the fem ur and the tibia, w hile more than the half of the soft tissue tum ours were localized in the thigh, knee and the lower leg. Among the bone tum ours, osteo-and chondrosarcom as dom inated. M FH com prised onethird of the soft tissue sarcomas.
The distribution of the tumour grades was as follows: six patients had grade I tumours, seven grade II, 23 grade IIIA, 12 grade IIIB and 10 patients had m alignant tumours where the histological grade could not be described m ore precisely. Forty-one patients in this group received surgery as the only treatm ent. F our patients were treated w ith com bined surgery and radiation therapy, while 13 patients were treated with com bined surgery and chem otherapy.

Statistics
T o com pare the post-operative functional results between the two groups, W ilcoxon' s range-sum test w as used. T he c 2 -test was used to test if tw o groups w ere com parab le according to patient characteristics.

Results
The two groups were comparable according to age, sex, time of observation and tumour size, including the distribution of small and large tumours ( , or > 5 cm). Also the distribution of bone and soft tissue tumours as well as the distribution of tumours localized in the upper and lower extremities were comparable in the two groups (Table 1).
Patients treated w ith lim b-salvag e surgery (n 5 89) had a signi® cantly higher functional score com pared to the group of patients undergoing am putation (n 5 58) ( Table 2). T he functional m edian score follow ing lim b-salvage surgery was 85 (range 10± 100) as com pared to 47 (range 13± 87) after amputation (p , 0.001). T he Enneking system for functional evaluation can be subdivided into a general health-related score and an upper and a lower extremity-related score. W e found that both the general health score and the extremity-related scores were signi® cantly higher in the lim b-salvage group com pared to the am putation group (Table 2).
Since the outcom e of low-and high-grade tum ours, might differ, the analysis w as also perform ed after excluding grade I tum ours in both the groups. How ever, also in this group of high-grade tum ours, we found a signi® cantly higher functional score in the group of patients treated with lim b-salvage surgery (n 5 59) com pared to the group of patients operated w ith am putation (n 5 52). T he m edian functional score following limb-salvag e surgery w as 83 (range 10± 100) com pared to 47 (range 13± 87) after am putation (p , 0.001).
The in¯uence of different factors on the functional score was tested. In neither the limb-salvag e group nor the am putation group did the localization of tum our result in a difference in functional scores.
For the lim b-salva ge group the functional score according to the type of tum our (soft tissue or bone) was exam ined (Table 3). Patients with soft tissue tum ours (n 5 68) scored signi® cantly higher than patients w ith bone tum ours (n 5 21). An endoprothesis w as used in 10/21 patients who had bone tum ours. Those patients had a lower functional score com pared to patients w ith soft tissue tum ours. N o signi® cant difference in functional score was found between patients with soft tissue tum ours and patients w ith bone tumours having lim b-salvage surgery without the use of an endoprothesis.
In the group of patients treated with lim b-salvage surgery, there was a tendency to a better functional score am ong patients with tumours , 5 cm (n 5 20) com pared to patients with tumours . 5 cm (n 5 69), the m edian score being 90 (range 60± 100) and 82 (range 10± 100), respectively (p 5 0.07).
In the lim b-salva ge group, surgery was com bined to radiation therapy in 34/89 patients. N o signi® cant correlation was found between radiation therapy and the functional score, the m edian score being 87 (range 23± 100) for patients receiving radiation therapy and 83 (range 10± 100) for patients who were not treated with radiation therapy. Table 4 shows the functional scores in relation to the surgical m argin achieved after perform ing lim bsalvage surgery. The patients were divided into four subgroups according to surgical stages. 6,7 T he subgroups were com parable according to age, size of tum our as well as the distribution of low (grade I) and high-grade tumours (grade II± III). N o signi® cant difference in functional scores w as found between the four subgroups treated with differently surgical m argins. Also there was no signi® cant difference in functional score if patients treated with intralesional, m arginal and w ide m argins were combined into one group and com pared to patients treated w ith com partm ental m argin.

D iscussion
T he present study showed a higher functional score in the lim b-salva ge group compared to the am putation group using the Enneking functional system. T he sam e result was found when studying the group of high-grade tum ours only. Several studies have evaluated the functional outcom e following lim bsalvage procedures, 8± 13 but to our knowledge only a single study has com pared the functional results after lim b-salvage and am putation with use of the Enneking system of functional evaluation. Rougraff et al. 14 com pared the functional results am ong patients with lower extrem ity osteosarcom as using the Enneking system for functional evaluation, and sim ilarly to our study they observed a better func-tional outcom e follow ing lim b-salvag e. On the other hand, they failed to ® nd any difference between the groups when analyzing only the scores related to the patient' s general conditionÐ a difference which had been observed in other studies. Sugarbake r et al. 15 demonstrated that there was no difference in psychosocial adjustm ent to illness or in quality of life between patients who were treated with lim b-salvage surgery and patients treated with am putation. Sim ilarly, W eddington et al. 16 w ere unable to detect a signi® cant difference in psych osocial outcom e between patients w ho underwent am putation and lim b-salva ge.
In the present study, both the general healthrelated scores and the extrem ity-related scores were signi® cantly higher in the limb-salvag e group compared to the am putation group. In the limb-salvag e group the general health-related score and the extremity-related scores were of equal size. In contrast, the general health-related score was higher than both the upper and the lower extremity-related scores in the am putation group. This suggests that the low functional score in the am putation group is especially due to a functional reduction of the operated extrem ity rather than an im pairm ent of the patients' general condition. This m ay support the ® ndings of other authors 15,16 w ho also failed to demonstrate any difference in quality of life after am putation and lim b-salvag e. Several studies 8,13 have reported that radiation therapy m ay have a great im pact on the functional results. At least in the present study, w e were unable to dem onstrate that radiation therapy had any effect on the functional outcom e.
Among patients who underwent lim b-salva ge, our results showed a signi® cant lower functional score in patients w ith soft tissue tumours than in patients with bone tum ours. The fact that half of the patients with bone tum ours had an endoprothesis could explain this observation, because those were the patients that actually had a low score compared to patients with soft tissue tum ours. In the group of patients with bone tum ours that did not have an endoprothesis, there was no difference in functional score compared to those patients with soft tissue tum ours.
In the present study, a number of subanalyses of the in¯uence of different factors on the functional outcom e were perform ed. D ivision of the groups according to, for instance, size, localization, type of tum our and level of am putation (e.g. above or below the knee) in som e cases resulted in a relatively few num ber of patients in each subgroup. The risk of m aking errors of type I and II m ay thus be increased. F or example, instead of studying the functional outcom e of the whole group of patients, it could be argued that only patients w ith sim ilar tum our locations in the tw o groups should be compared. However, the num ber of patients in each of these subgroups was too sm all to allow a m eaningful com parison. T herefore, a different study including m ore patients and thus larger subgroups could produce different results. As a result of the above and the fact that the present study was a non-randomized study, the results are valid only for the patients included in this w ork and cannot directly be applied to another population of patients. O n the other hand, the present study clearly indicates the im portance of m easuring functional outcome in future studies and especially in studies with an expected possib le difference in functional outcom e. Such a study w ith the necessary num ber of patients is in progress.
O n the basis of the present results, we conclude that lim b-salva ge surgery is asso ciated with a better functional outcom e com pared to am putation. Based on the Enneking system for functional evaluation also, the patients' general condition after limb-salvage is better com pared to the patients' general condition after amputation. However, a m ore complex examination of the patients' quality of life is needed in order to conclude that lim b-salva ge is also asso ciated with a better quality of life com pared to am putation.

A cknowledgem ents
T he study w as supported by the C linical Research U nit, D anish Cancer Society, O ncologic Center, Aarhus U niversity H ospital, D enm ark.