Fax: (905) 521-1703
A call to arms (and legs)
Article first published online: 10 AUG 2009
Copyright © 2009 American Cancer Society
Volume 115, Issue 18, pages 4046–4054, 15 September 2009
How to Cite
Barr, R. D. and Wunder, J. S. (2009), Bone and soft tissue sarcomas are often curable—But at what cost?. Cancer, 115: 4046–4054. doi: 10.1002/cncr.24458
This article is based on an invited presentation to the Musculo-Skeletal Tumor Society during the 75th Annual Meeting of the American Academy of Orthopedic Surgery, San Francisco, California, March 5-9, 2008.
Dedicated to the memory of our late colleague, Dr. Nigel Colterjohn, who died on March 31, 2008 at the age of 47.
- Issue published online: 4 SEP 2009
- Article first published online: 10 AUG 2009
- Manuscript Accepted: 3 FEB 2009
- Manuscript Revised: 29 DEC 2008
- Manuscript Received: 22 SEP 2008
The Association for Health Services Research has defined its topic as a “field of enquiry using quantitative or qualitative methodology to examine the impact of the organization, financing and management of healthcare services on the access to, delivery, cost, outcomes and quality of services.”1 There can be no doubt about the importance of outcomes assessment in oncology.2 Wilke et al3 have emphasized the place of patients' self-reports in the following definition:
The term Patient-Reported Outcomes (PROs) has evolved to include any endpoint derived from patient-reports, whether calculated in a clinic, in a diary or by other means, including single item outcome measures, event logs, symptom reports, formal instruments to measure health-related quality of life (HRQoL), health status, adherence and satisfaction with treatment. The term coincides with the explicit interest from drug development researchers and regulatory authorities in the appropriate utilization and reporting of treatment impact measures.
Such has been the burgeoning interest in PROs that the Quality of Life Newsletter changed its name to the Patient Reported Outcomes Newsletter.4
Included in the armament of PROs are functional assessments and measurements of HRQoL, outcomes that are neither interchangeable nor mutually exclusive but rather complementary. These are especially relevant areas of research in young people with cancer who enjoy improving prospects for long-term survival compromised by late sequelae of their diseases and treatments that vary considerably in prevalence, severity, and combination.5
A focus on HRQoL in survivors of bone and soft tissue sarcomas is justified not only by increasing survival rates but also by the age-specific incidence rates of these diseases, notably malignant bone tumors (afflicting particularly adolescents and young adults [AYA] who have potentially lengthy life expectancies), and the reported self-assessments of HRQoL in such populations of survivors (indicative of burdens of morbidity almost as great as those experienced by survivors of brain tumors).
This review will pay particular attention to the “costs” (clinical consequences) of cure resulting from limb-salvage procedures that are undertaken in increasing proportions of patients, instead of amputation, facilitated by effective neoadjuvant chemotherapy, radiotherapy, advances in radiology, and remarkable technical accomplishments in orthopedic surgical oncology.
It must be recognized that, although bone and soft tissue sarcomas are relatively common in the AYA age group (accounting together for 6%-7% of incident cases in patients 15-29 years old),6 these diseases represent only a small fraction of the total cancer burden7: 0.16% (2380 cases estimated in 2008) and 0.72% (10,390 cases estimated in 2008), respectively, in the United States.
Sample sizes are therefore a major limitation to outcomes research involving AYA patients with sarcomas of bone and soft tissue, who have 5-year survival rates of approximately 60%8 and 75%,9 respectively (the latter excluding Kaposi sarcoma, which is relatively common in this age group). Moreover, there are numerous confounders to be addressed in measuring HRQoL in survivors of these diseases. Among these are:
Age (especially prepuberty vs postpuberty)
Sex (female survivors of cancer in early life often report poorer HRQoL than males in comparable circumstances)10
Tumor characteristics (types include >50 histologic variants, size, and location)
Presurgical (neoadjuvant) and postsurgical treatment (chemotherapy and/or radiotherapy)
The era of study, reflecting the type of surgery and technical expertise as well as the changing practices of chemotherapy and radiotherapy
Time since the surgical intervention (related to functional adaptation).
Definition and Measurement of HRQoL
Quality of life is a broad concept that encompasses economic welfare, characteristics of the community (such as crime rate, and educational and recreational amenities), characteristics of the environment (such as air and water quality), and health status.11 The narrower “within the skin” construct of health-related quality of life (HRQoL) includes the opportunities that a person's health status affords, the constraints it places on the person, and the value a person places on his/her health status.11 Patrick and Erickson have defined HRQoL as “the value assigned to duration of life as modified by impairments, functional states, perceptions and social opportunities that are influenced by disease, injury, treatment and policy.”12 Simplistically stated, (health-related) quality of life is what health (status) is worth—begging the question, “worth to whom?” This review will focus on patients' self-assessments of their HRQoL, an important form of PRO.
A detailed review of the measurement of HRQoL is beyond the scope of the current undertaking. However, an effort has been made to assess the properties of instruments that have been used to measure HRQoL after pediatric orthopedic surgery, using a selected clinical context for exemplification.13 Attention was divided between generic preference-based instruments (such as the Health Utilities Index), generic pediatric health profiles (such as the Child Health Questionnaire, not recommended for this purpose), and orthopedic-specific measures.
Generic measures of HRQoL address domains (dimensions, attributes) of health of relevance to a wide variety of disorders, as well as to the general population, thereby allowing for comparisons among a variety of groups. Specific measures address domains of particular relevance to a limited circumstance, such as a single disease. Health profiles consist of subscales, for each of which a score is obtained. Preference-based measures integrate morbidity and mortality, and generate utility scores for single attributes of health as well as overall HRQoL; the latter is essential to economic evaluation based on cost-utility analysis.
Within the category of orthopedic-specific measures, a comparison was made between Pediatrics Outcomes Data Collection Instrument (PODCI)14 (also known as POSNA15), the Activities Scale for Kids,16 and the Functional Assessment Questionnaire Working Scale.17 Problems were identified with the PODCI instrument, notably with respect to missing data. The recommendation deriving from this review was that a combination of instruments should be used for each study, and that the final choice would be dictated by the specific study objectives.
Review of the Evidence
There is no single study of sufficient size and design to address all of the confounding variables listed above, and perhaps there never will be. Consequently, the quality of the evidence is necessarily imperfect. Nevertheless, some fairly clear messages are apparent.
Some of the best information comes from the Childhood Cancer Survivor Study (CCSS); a cohort of some 15,000 subjects who were diagnosed with cancer as children and adolescents in the period 1970-1986.18, 19 These investigators recorded that survivors of bone tumors (almost 1000 subjects) were almost as likely as survivors of brain tumors to report performance limitations, restriction of routine activities, and diminished ability to attend work or school.20 Furthermore, survivors of bone and soft tissue sarcomas (n >1200) and of brain tumors had the lowest HRQoL scores derived from self-assessments using Short Form 36 (SF36).21 Similar results had been reported earlier from a smaller Canadian cohort using the Health Utilities Index Mark 3 (HUI3) instrument.22 These data were provided by a national study of >2000 survivors under the auspices of the federal Canadian Childhood Cancer Surveillance and Control Program.23
The problem of small study samples is exemplified by the report from Tabone et al, who used the Child Health Questionnaire in a highly heterogeneous cohort with respect to tumor type and treatment received.24 The study by Frances et al25 (using the PODCI instrument, in some instances completed by parental proxies) again illustrates the challenge of a small, heterogeneous study sample, as reflected in the apparent lack of differential impact of age, tumor type, Enneking stage, nature of surgery, or even metastasis and recurrence of disease on HRQoL. Another study, in patients with spinal metastases undergoing surgery for relief of pain, highlights an additional methodological challenge.26 The average (mean) HRQoL score, from instruments including HUI3, would have obscured the reality that there was in fact a bimodal distribution of scores clustering around those that were “fair” (HUI3 = 0.7) and “awful” (HUI3 = 0.1).
A focus on surgical options
As expressed by investigators in the CCSS, “There is general agreement that limb-sparing techniques are the preferred approach for patients with upper extremity tumors.”27 The evidence with respect to the lower limb, a more common site of disease,27 is much less clear. The issues relating to choice of surgical intervention for lower extremity tumors have been distilled by Simon28:
- 1Will survival be the same?
- 2How do the immediate and late morbidities (complications) compare?
- 3How does function compare?
- 4Does limb-sparing surgery impart improved psychosocial/quality of life outcomes?
Again, investigators in the CCSS have addressed these crucial issues.29 Their conclusions, from a comprehensive assessment of published studies, are that survival and local recurrence rates are similar after limb-salvage or amputation, providing that surgical margins are adequate and that effective chemotherapy is administered, and that complications are more frequent after the former procedure(s). In the categorization of surgical procedures by CCSS investigators it is unclear whether Van Nes rotationplasty is considered an amputation or a form of limb salvage.29 Even more curiously, in the CCSS survey of more than 600 survivors, no rotationplasties were noted.27 Van Nes rotationplasty is an operative procedure that can be used as an alternative to above-knee amputation. It effectively converts a higher amputation into a functional below-knee amputation. For a sarcoma of the distal femur or proximal tibia, after the resection of the tumor with the adjacent knee joint, the distal aspect of the extremity, including the foot, is rotated 180° and attached proximally to the remaining femur with internal fixation. In this way the ankle joint functions as a knee joint.
With respect to functional assessment, the instrument that has been used most commonly is that developed by the Musculo-Skeletal Tumor Society (MSTS).30 From the original cohort of 227 patients, members of the MSTS from 26 institutions reported functional assessment on 78 and HRQoL measures on 29 survivors.31 Their conclusions were that functional outcomes were better after limb salvage surgery (including rotationplasty) than after above-knee amputation or hip disarticulation, but that psychosocial outcomes (not strictly comprehensive HRQoL) were no different between the 3 groups. However, the MSTS instrument relies on the subjective ratings of clinicians, and so its validity has been called into question.32 In a CCSS report, the Toronto Extremity Salvage Score (TESS)33 was used. The TESS is an instrument for functional self-assessment, measuring physical disability or activity limitations.34 The CCSS demonstrated no difference in TESS scores between amputees and survivors who had undergone limb salvage, and good correlation between TESS scores and HRQoL.27 In their earlier report using TESS,35 the Toronto investigators observed that “the differences in disability between amputation and limb-sparing patients (sic) are smaller than anticipated.”
Recent reports from Australia36 and Scandinavia37 confirm the disparity between the scores from the MSTS and TESS questionnaires. Patients who underwent limb-salvage surgery had better functional outcomes, as determined by MSTS, than those who had an amputation, but there were no statistically significant differences in TESS (or SF36) scores. A study from St. Jude Children's Hospital (SJCRH) and the Children's Hospital of Philadelphia38 revealed no difference in MSTS, TESS, or SF36 scores between the 2 surgical groups, but the amputees had lower scores on objective functional outcome using the Functional Mobility Assessment tool.39
The matter of amputation requires closer examination. Although a single institution study (from SJCRH) reported no difference in patient satisfaction between amputation and limb salvage surgery,40 the experience from Boston is salutary, as it raises issues of referral and patient biases. Although a higher proportion of limb salvage patients than amputees were pleased with the outcome, the investigators noted that their patients were referred seeking limb salvage and, as they report “several patients … were concerned about the surgeon's investment in the limb and did not want to disappoint him.”41
A separate study from the Massachusetts General Hospital reported on more than 400 patients who had undergone surgery for tumors of the lower extremity.42 Of 66 amputees, 41 were male, whereas of 342 who had undergone limb salvage the majority (247) were female. Although there was no notable difference in functional outcomes, and no difference in employment or marital status, between the surgical groups, significantly fewer limb salvage patents had active sex lives and significantly fewer had children. The CCSS reported on almost 700 survivors of lower extremity bone tumors with respect to education, employment, health insurance, and marital status.43 Almost all had graduated from high school, and nearly half of those aged >25 years had graduated from college. There was no difference between the surgical groups. Again, almost all had been employed, although men and those with higher education were more likely to have worked in the previous year. Two thirds reported being married or living as married, with no difference between the surgical groups, although women were more likely to be in this category (and to have health insurance).
In a secondary analysis of the study from Boston reported subsequently,44 it was demonstrated that the higher the surgical resection level in the lower limb, the greater the disparity in functional outcome between limb salvage and amputation, favoring the former. Whether patients who have endoprosthetic reconstruction have poorer outcomes compared with those with vascularized autografts or allografts, as has been suggested,23 remains uncertain. However, these outcome differences may be explained, at least partially, on an anatomic basis. For example, allograft and vascularized autograft reconstructions may be used more commonly for repair of diaphyseal bone defects, which would be expected to be associated with better functional outcomes than joint arthroplasty using endoprostheses.
Investigations on smaller numbers of patients from elsewhere in the United States,45 from Australia,46 and from the Netherlands47 reveal no significant differences in psychological outcomes and HRQoL between amputees and those who have undergone limb salvage. There are 2 additional perspectives that are worthy of consideration in evaluating such information: patients who proceed to amputation are seldom candidates for limb-sparing surgery,46 and patients who have amputations for cancer may fare much better than traumatic amputees in terms of adaptation to disability.48 In this latter regard, a study of >500 patients who had experienced leg-threatening injuries, from a consortium of 8 level 1 trauma centers in the United States, revealed that the HRQoL was no different in those who had reconstructive (limb-salvage) surgery and those who had amputations.49 However, the scores from a large proportion of both groups were indicative of severe disability.
Turning to rotationplasty, the more limited information appears to be fairly consistent. Studies from Austria,50 Germany,51 and the Netherlands,52 all using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire family of instruments, report that, despite the obvious adverse cosmetic impact, survivors who had undergone Van Nes procedures at the knee experienced good functional and psychosocial outcomes,50 with HRQoL better than in subjects who had undergone other limb salvage procedures,51 even approximating that of the general population.52 A recent report from Japan,53 again on a small study sample (26 procedures in 22 patients with osteosarcoma around the knee joint), included the observation of better functioning, using the MSTS instrument, and overall HRQoL (using SF36) after rotationplasty than after both amputation and endoprosthetic replacement. Interestingly, all of those undergoing rotationplasty (n = 6) were male. In the experience of the CCSS investigators,29 survivors of rotationplasty reported HRQoL little different from normal. There appears to be very few published data on functional outcomes or HRQoL after rotationplasty at the level of the hip joint,54, 55 but again, results suggest that this is a good alternative to traditional limb salvage with an endoprosthesis, especially in young children.
A summary of surgical procedures and outcomes is provided in Table 1.
|Author, Year, Group||No. Patients, Surgery Type||Mean Age at Diagnosis, y||Mean Age at Study, y||Instruments||Main Findings|
|Weddington 1985,45 Chicago, Ill||33: 14 Amp, 19 LSS||—||38||Psychological assessment including KPS||LSS=Amp|
|Postma 1992,47 Groningen, Netherlands||33: 19 Amp, 14 LSS||16||26||HSCL; GARS; Self-Esteem Scale; HRQoL and Disability visual analog rating; semistructured interview||LSS=Amp; LSS had more physical complaints and more changes in job type; Amp had lower self esteem and more social isolation|
|Rougraff 1994,31 MSTS||78 function; 29 HRQoL: 21 Amp, 8 LSS||16||27||MSTS score; Knee Society score; Psychosocial questionnaire||LSS>AKA>hip disarticulation for better function; LSS=Amp for psychosocial outcomes|
|Marsden & Swanson 1997,46 Herston, Australia||43: 10 Amp, 33 LSS||—||—||HRQoL visual analog scale questionnaire||Trend for improved HRQoL after LSS compared with Amp|
|Hudson 1998,40 Memphis, Tenn||65: 15 Amp, 50 LSS||18||26||Self-administered questionnaire||LSS=Amp|
|Davis 1999,35 Toronto, Canada||36: 12 Amp, 24 LSS, case matched||29||32||TESS; SF36; RNL||Trend for LSS>BKA>AKA for less disability and better function based on case-matched groups; more handicap following Amp|
|Hillman 1999,51 Munster, Germany||67: 34 LSS, 33 Rot||16||22||MSTS; QLQ-C30||LSS=Amp; LSS more likely to use walking aids, more work and hobby restrictions, more daily activity restrictions due to pain; no psychosocial disadvantages for Rot|
|Veenstra 2000,52 Rotterdam, Netherlands||33 Rot||25||32||QLQ-C30; SF36; SSLI-SSLD; QLQ-BR23||HRQoL, psychosocial functioning, and social supports comparable to healthy peers; physical function less than peers but ⅔ active in sports|
|Winkelmann 2000,55 Munster, Germany||8 hip Rot||9||14||MSTS; QLQ-C30; gait analysis||Good function and HRQoL for hip Rot vs amputation or limb salvage at the hip; all patients participate in sports|
|Refaat 2002,42 Boston, Mass||408: 66 Amp, 342 LSS||47.5||55||Self-reported questionnaire for function, HRQoL, and psychological responses||LSS=Amp|
|Nagarajan 2003,43 CCSS||694: 223 Amp, 471 LSS||14||30||Psychosocial outcomes questionnaire including education, employment, health insurance, and marital status||LSS=Amp; however, compared to their siblings, Amp patients had more problems with education, employment, and obtaining health insurance|
|Nagarajan 2004,27 CCSS||528: 336 Amp, 192 LSS||14||35||TESS; QOL-CS; self-perceived general health status questionnaire||LSS=Amp|
|Hopyan 2006,36 Melbourne, Australia||45: 20 Amp, 20 LSS, 5 Rot||12||26||MSTS score; TESS; SF36; Uptime||LSS>Amp for better function; trend for LSS>Amp for less disability; Rot had the best psychosocial and Uptime results|
|Pardasaney 2006,44 Boston, Mass||408: 65 Amp, 343 LSS||40||49||Self-reported questionnaire for physical and psychological responses||With progressively higher levels of surgery, there are greater functional benefits after LSS compared with Amp|
|Ginsberg 2007,38 Memphis, Tenn and Philadelphia, Pa||91: 22 Amp, 65 LSS, 4 Rot||15||20||FMA; MSTS score; TESS; SF36||LSS=Amp for function, disability, and HRQoL; Rot>LSS and Amp due to less activity restrictions (FMA), as well as better function and less disability; no HRQoL differences|
|Akahane 2007,53 Nagano, Japan||22: 7 Amp, 8 LSS, 6 Rot||22||27||MSTS; SF36||Rot>LSS, Amp for better function; Rot=LSS=Amp for HRQoL, but Rot had the highest scores|
|Aksnes 2008,37 Scandinavia||118: 47 Amp, 71 LSS||18||31||MSTS score; TESS; SF36||LSS>Amp for better function; LSS=Amp for disability and HRQoL; 105 patients work and have good HRQoL|
Do the effects of treatment warrant the substantial monetary and nonmonetary costs (clinical consequences it takes to produce them)? The essence of economic evaluation of healthcare services is a comparison of the costs and consequences of relevant alternatives.56 In a commendable exercise more than a decade ago,57 Grimer and colleagues in the United Kingdom undertook a cost-effectiveness study comparing amputation with limb salvage. These obviously differ considerably in terms of up-front surgical costs, the duration of rehabilitation, and the frequency of the need for revisions, all of which are greater with limb salvage. The costing model included “maintenance costs” in perpetuity, an assumption evidently subject to argument. Moreover, these authors noted that “Most limb-fitting centers now offer a ‘package price’ for the provision of an exoprosthesis and maintenance costs for follow-up and attention to complications.” Furthermore, for limb salvage with an endoprosthesis, they comment that “Many centers now offer a composite cost which is made up of factors related to length of stay, the price of the implant, the complexity of the surgical procedure and rehabilitation.” These circumstances may still hold in the National Health Service, but are far from universally applicable.
Grimer and colleagues factored in some interesting clinical issues that have major cost implications, namely “Endoprosthetic replacements will eventually fail” and “Most active young people will demand and use a sophisticated artificial limb…. Most will require a spare prosthesis. Many will request and use a sports limb and also a limb for swimming. A new prosthesis will be necessary at regular intervals.” Small wonder that these authors concluded that “The surprising feature is the considerable cost of having an amputation.”
Preference-based measures of HRQoL, such as HUI3, provide utility scores that can be used to calculate quality-adjusted life years (QALYs). In turn, these contribute to a particular form of cost-effectiveness—cost-utility analyses—in economic evaluation. As an illustrative example of QALYs, we can consider the case of an 18-year-old with nonmetastatic osteosarcoma of the distal femur. On a scale of 0 = being dead to 1.0 = perfect health, his HRQoL scores are as follows:
- 1Neo-adjuvant chemotherapy for 3 months: score, 0.70.
- 2Limb reconstruction and recovery for 1 month: score, 0.75.
- 3Adjuvant chemotherapy for 8 months: score, 0.80.
- 4Survival for 10 years thereafter; well: score, 0.90.
- 5Development of anthracycline-induced cardiomyopathy with progressive deterioration to death in 4 years: score, 0.50.
With each score representing the average for the respective interval, and the total survival being 15 years, the QALYs are (3 × 0.70 plus 1 × 0.75 plus 8 × 0.80)/12 = 0.77 plus 10 × 0.90 (9.00) plus 4 × 0.50 (2.00) for a total of 11.77—equivalent to the loss of 3.23 years of perfect health.
Economic evaluation of surgical alternatives for young people with lower extremity tumors is likely to reveal a tradeoff between the up-front costs of limb salvage and the maintenance costs associated with amputation (as identified by Grimer and colleagues), while comparing the utility of HRQoL from diagnosis to death or long-term survivorship, as assessed by the patients. The QALYs generated in such an analysis would help to inform the decisions of healthcare providers and consumers alike.
An example of the interpretation of cost-utility analysis was given >15 years ago by Laupacis et al58 (so the dollar figures would need upward adjustment in the current time frame). There would be general acceptance of <$20,000 per QALY gained, by 1 intervention in comparison to another; equivocal acceptance of $20,000 to $100,000 per QALY gained; and no certainty of acceptance at more than $100,000 per QALY gained. Those unfamiliar with this territory are referred for context to the article by Tengs and colleagues,59 who describe the cost per non–quality-adjusted life year for childhood immunization as <$0, whereas at the opposite extreme the figure for sickle cell disease screening of nonblack, low-risk newborns is given as $34 billion!
A Historical Gold Standard
The first report, of which we are aware, addressing quality of life in patients with extremity sarcoma, was published >25 years ago and has never been bettered with respect to study design.60 Sugarbaker and colleagues at the National Cancer Institute (NCI) of the National Institutes of Health enrolled 21 patients aged <21 years in a randomized clinical trial of amputation plus chemotherapy versus limb-sparing surgery plus chemoradiotherapy for nonmetastatic, nonrhabdomyosarcoma soft tissue sarcomas of the lower limb. By using the Sickness Impact Profile, a well validated measure of HRQoL, these investigators demonstrated that, at 1-3 years after surgery, the amputees self-reported better emotional behavior, body care and movement, and sexual functioning than the patients who had undergone limb salvage. The group at the NCI even claimed that there were no differences in the economic costs incurred by subjects in the 2 arms of the study, an outcome perhaps peculiar to that institution, although no monetary information was provided.
Could a study of such admirable rigor be repeated today? Therein lies the “call to arms—or at least legs?”
Although it appears that the number of amputations per year has decreased with time,27, 37 it cannot be assumed that limb salvage surgery will offer a better functional outcome and HRQoL for young people with bone and soft tissue sarcomas of the lower limb. Although this challenge is clearly a moving target, formal economic evaluation would make a useful contribution to future decision making by patients and their healthcare providers, surgeons and others alike.
Conflict of Interest Disclosures
The authors made no disclosures.
- 1Association for Health Services Research. Definition of health services research. Available at: http://www.abstr.org/hsrproj/define.htm Accessed on December 28, 2008.
- 4Editorial. PRO Newsl. 2005; 34: 1., .
- 5Adolescent and young adult cancer survivors: late effects of treatment. In: BleyerWA, BarrRD, eds. Cancer in Adolescents and Young Adults. Berlin, Germany: Springer-Verlag; 2007: 411-430., , , .
- 6Introduction. In: BleyerA, O'LearyM, BarrR, RiesLAG, eds. Cancer Epidemiology in Older Adolescent and Young Adults 15-29 Years of Age, Including SEER Incidence and Survival: 1975-2000. Bethesda, MD: National Cancer Institute; 2006: 1-14., , .
- 8Malignant bone tumors. In: BleyerA, O'LearyM, BarrR, RiesLAG, eds. Cancer Epidemiology in Older Adolescents and Young Adults 15-29 Years of Age, Including SEER Incidence and Survival: 1975-2000. Bethesda, MD: National Cancer Institute; 2006: 97-109., , , , , .
- 9Soft tissue sarcomas. In: BleyerA, O'LearyM, BarrR, RiesLAG, eds. Cancer Epidemiology in Older Adolescent and Young Adults 15-29 Years of Age Including SEER Incidence and Survival: 1975-2000. Bethesda, MD National Cancer Institute; 2006: 81-95., , , .
- 12Health Status and Health Policy: Allocating Resources to Health Care. New York, NY: Oxford University Press; 1995., .
- 30Limb salvage treatment versus amputation for osteosarcoma of the distal end of the femur. J Bone Joint Surg Am. 1986; 68A: 1331-1337., , , .
- 31Limb salvage compared with amputation for osteosarcoma of the distal end of the femur. J Bone Joint Surg Am. 1994; 76A: 649-656., , , , .
- 34International Classification of Functioning, Disability and Health. Geneva, Switzerland: World Health Organization; 2001.
- 41Quality of life in osteosarcoma survivors. Oncology. 1994; 11: 19-25., , , et al.
- 46Outcomes after multi-modality treatment of musculo-skeletal tumours. Acta Orthop Scand. 1997; 68( suppl 273): 101-105., .
- 50Quality of life studies in long-term survivors of childhood bone cancer undergoing rotationplasty for local tumour control [abstract]. Qual Life Res. 1997; 6: 614., , .
- 51Malignant tumor of the distal part of the femur or the proximal part of the tibia: endoprosthetic replacement or rotationplasty. J Bone Joint Surg Am. 1999; 81A; 462-468., , , , .
- 54Hip rotationplasty for malignant tumors of the proximal part of the femur. J Bone Joint Surg Am. 1986; 68A: 362-369..
- 55Type-B-111a hip rotationplasty: an alternative operation for the treatment of malignant tumors of the femur in early childhood. J Bone Joint Surg Am. 2000; 82A: 814-828..