The seminal role of Cancer in our understanding of sarcoma

Authors

  • Raphael E. Pollock MD, PhD

    Corresponding author
    1. Department of Surgical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas
    • Department of Surgical Oncology, Box 444, the University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030
    Search for more papers by this author

  • All quotations were reprinted with the permission of Wiley-Liss Inc, a subsidiary of John Wiley & Sons, Inc.

Abstract

The journal Cancer has played a critical role in advancing our understanding of sarcoma, a remarkable cluster of malignancies that arise from tissues of mesenchymal origin throughout the human body and indeed across the entire vertebrate animal spectrum. Approximately equally divided between soft tissue and bone origin, there are greater than 100 recognized histologic subtypes that collectively account for approximately 2% of adult solid tumors. Cancer 2008;113(7 suppl):1969–79. © 2008 American Cancer Society.

Because of the relative rarity of malignant disease of the connective tissue, over the past 60 years the study of sarcoma has been hampered by a lack of the patient and tumor resources needed to derive definitive conclusions. As compared with the more common malignancies of epithelial origin, there are relatively fewer centers in which basic, translational, clinical, and outcomes research focusing on sarcoma has been conducted. Nonetheless, students of sarcoma have found a receptive podium and audience under the aegis of Cancer, and it is striking that many of the classic and critical sarcoma reports of the past 60 years have seen the light of day via this journal. Moreover, many key developments over this same time period in the much larger field of oncology have been the result of groundbreaking sarcoma reports that have appeared in Cancer. Subjects as diverse as the viral and chemical etiology of solid tumors, the oncogenic impact of anticancer therapies, multimodality therapy of cancer, neoadjuvant systemic approaches to malignancy, function-preserving surgery, clinical and molecular prognosis, and the development of staging systems are all examples of critical concepts introduced by sarcoma-oriented research reports that have appeared in Cancer over the past 60 years. The contributions in 3 broad categories of emphasis will form the basis of this review article. These include: sarcoma clinical and research biology, sarcoma prognosis and staging, and the treatment of sarcoma. Greater than 150 sarcoma research reports have appeared in Cancer to date. In keeping with the historical orientation of this review, this analysis will focus on contributions that have had a significant impact in sarcoma oncology, particularly highlighting reports that are more than 30 years old and therefore may have escaped the attention of the current Cancer readership.

Sarcoma Clinical and Research Biology

The Cancer sarcoma saga begins in 1948, the first year that this journal was published. Cahan et al at the Memorial Hospital, including Fred Stewart of subsequent Stewart-Treves syndrome fame, published a report in the May 1948 issue of Cancer entitled “Sarcoma Arising in Irradiated Bone: Report of Eleven Cases.”1 One of the indications for radiotherapy in the 1920s and 1930s was the treatment of nonmalignant diseases of the bone such as tuberculosis, chondroblastoma, and benign giant cell tumor. The report by Cahan et al began by explaining the 4-part rationale for avoiding radiotherapy in the treatment of nonmalignant bone disease: 1) surgery is more effective, provides the opportunity to examine the entire specimen microscopically, and “avoids the risk of irradiating a malignant tumor under the mistaken impression that it is benign”1; 2) many benign lesions are radioresistant; 3) radiation may prematurely close epiphyseal plates in growing children; and 4) “An undesirable and extremely serious sequel is courted: there is the possibility that, after an interval of years, bone sarcoma may develop at the site of the previous irradiation and a benign bone condition, with moderate incapacitation as its most serious prognosis, may be converted into a rapidly fatal disease.”1

From a historical perspective, the elegance in language usage from that era is matched by the actual structure of the report, which included a table providing details regarding research into the experimental production of bone sarcoma dating back to studies conducted in 1910 as well as a table of the collected clinical cases of bone sarcoma after “radium or Roentgen-ray therapy,” which included the first ever report of a postmastectomy radiation-induced tumor (chondrosarcoma), observed in a patient treated in 1945.

Exhaustively detailed presentations of the authors' personal collection of 11 cases, including full‒page pathology and radiology images, and extensive information regarding every hospitalization for each of these patients points to an era in biomedical publishing when page and word limits were hardly relevant considerations. On the basis of this extensive analysis of clinical cases, the authors concluded that a latent period of at least 5 years and as long as 20 years is needed for a sarcoma to develop after radiotherapy, as was an initial treatment dose of at least 30 grays (Gy); both these observations have stood the test of time over the ensuing 60 years. Cancer was also the venue for the classic report by Stewart and Treves that identified lymphangiosarcoma in postmastectomy lymphedematous extremities, a pathophysiologic condition with which the 2 authors of this article have been eponymously immortalized (Figs. 1 and 2).2

Figure 1.

An example of a classic article that appeared in Cancer in 1948, the first year of publication. Reprinted with permission of Wiley-Liss Inc, a subsidiary of John Wiley & Sons, Inc.

Figure 2.

In 1948, when contemporary report word and space limitations were not of much concern, full‒page clinical pictures, photomicrographs, and radiology studies were frequently encountered. Shown is a photograph from the original 1948 article by Stewart and Treves depicting a patient with the syndrome for which the authors are eponymously famous. Reprinted with permission of Wiley-Liss Inc, a subsidiary of John Wiley & Sons, Inc.

The association between sarcoma and neurofibromatosis was first published in Cancer in a 1962 article entitled “Sarcomas of the Peripheral Nerves and Somatic Soft Tissues Associated with Multiple Neurofibromatosis (Von Recklinghausen's Disease).” by D'Agostino et al from the Departments of Pathology and Neurosurgery at the Mayo Clinic.3 This report was the first to examine a series of sarcomas that developed in a large cohort of neurofibromatosis patients, documenting 21 such patients from a larger group of 678 neurofibromatosis patients followed at the Mayo Clinic (time period not specified) and focused on the pathology of these tumors rather than the treatment per se. One controversy regarding the etiology of such tumors, fueled by anecdotal case reports of sarcomas in neurofibromatosis patients that had appeared up to the time of this publication in Cancer, was whether such sarcomas arose from within a nerve trunk versus perineural deep tissues. The careful pathologic analysis conducted by D'Augustino et al demonstrated that both types of tissues could give rise to sarcomas, as could plexiform neurofibromas. A commonly held contention of that era was that malignant transformation in neurofibromatosis followed surgical procedures to remove benign neurofibromas; however, 12 of 21 patients in this series had no prior surgical history, thereby laying that theory to rest. Another concept current at that time was that sarcomas arising from nerves were less malignant and metastasized less often than other types of sarcomas, a contention refuted by the results of this series. Finally, this report was the first to suggest the now universally accepted clinical constellation in neurofibromatosis patients: “the development of pain, the sudden enlargement of a pre-existing mass, or the occurrence and rapid growth of new tumors should always suggest the onset of sarcoma, particularly if located in the deep tissues.”3

Over the years Cancer has been the vehicle by which initial reports identifying many of the distinct histologic sarcoma subtypes have been disseminated. These classic publications have been authored by some of the most prominent soft tissue pathologists of the past century. By virtue of their expertise and visibility, these experts were frequently called on to consultatively render definitive diagnoses in ambiguous situations, and were therefore optimally positioned to appreciate the common histologic appearances and clinical presentations of these specific tumor clusters.

The Memorial Hospital Pathology group, in which Fred W. Stewart played a leading role, published the first report of alveolar soft part sarcoma in Cancer in 1951.4 Clinical data derived from 12 patients formed the basis of this publication. The strikingly similar microscopic appearance of the tumors from these patients was described as “characterized by the pseudoalveolar, or organoid, arrangement of the cells…present a more strictly alveolar form, with a peripheral row of cells and a central space devoid of organized structure…it was this pattern that was responsible for the term ‘alveolar soft-part sarcoma’.”4 This report also identified several other aspects of alveolar soft-part sarcoma that are now appreciated as distinctive, including that (surprisingly) the brain was the site of metastasis in 2 of the 12 cases, and that many of the patients lived for many years after surgical resection of the primary tumor, even in the presence of long‒standing metastatic progression.

A 1952 report in Cancer by Stout and Verner was the first to demonstrate the possibility that chondrosarcoma could arise from extraskeletal soft tissues.5 Reminiscent of the report by Stewart and Treves, discussed earlier, this article consisted of 7 cases in which Stout and Verner had been asked to serve as consultants. The anatomic location and distribution of these lesions precluded the possibility that they were derived from degenerating benign chondromas, and it was concluded that these tumors arose de novo in the soft tissues. In a similar manner, Fine and Stout provided the initial description of extraskeletal osteosarcoma in a report published in Cancer in 1956.6

A final example of this type of report is that from Enzinger as a single author, which appeared in 1970 in Cancer and established epithelioid sarcoma as a unique histologic subtype.7 Enzinger identified several distinctive characteristics of this clinical entity, including its typical presentation as a small and innocuous‒appearing firm nodule that developed on a distal extremity. The tumors had a tendency to grow along tendon sheaths and muscular fasciae. Microscopically, a central area of necrosis was a common finding, as was the acidophilic‒staining appearance of the tumor cells, which was the result of the presence of abundant hyalinized collagen within the tumor tissues. Enzinger also commented on the prominent pattern of lymph node and cutaneous metastases, and the high rate of regional recurrence—all of which are now commonly accepted features of this specific form of soft tissue sarcoma.

Laboratory research in sarcoma has also been the subject of several extremely important reports that have been published in Cancer over the past 60 years. One of the first such reports was published in 1948, in which Schoenbach et al examined the effects of folic acid inhibitors on experimental sarcoma being carried as a tumor transplant in recipient mice.8 Several anecdotal reports concerning the use of various folic acid inhibitors had already appeared in the biomedical literature by 1948; however, this was the first publication involving a controlled experimental system in which this new form of anticancer therapy was examined for its effect on sarcoma. Marked sarcoma growth inhibition and altered tumor histology, without detectable tumor-associated hemorrhage or necrosis, suggested that the observed effects were specifically because of the use of a folic acid antagonist (aminopterin), a tumor inhibition that could be overcome by the concomitant administration of exogenous folic acid.

The research use of carcinogens to induce the development of experimental sarcomas is another area of laboratory investigation in which Cancer played an important role. One of the first such reports was that of Ghadially and Roy, published in 1966, in which these sarcoma researchers were able to induce the development of synovial sarcoma by injecting dimethyl-benzanthracene (DMBA) into the knee joint of rats.9

One final example of groundbreaking sarcoma research that first appeared in Cancer was an initial report concerning the detection of tumor-associated antigens in soft tissue sarcoma, written by Eilber and Morton in 1970.10 The sera from patients bearing a variety of soft tissue and skeletal sarcomas could be demonstrated to contain antibodies to a sarcoma-specific antigen retrieved from a cultured human liposarcoma cell line. The sarcoma-specific antigen could be recovered from other sarcoma histologic subtype cell lines; however, the antigen could not be retrieved from either fibroblasts from sarcoma patients or from nonsarcoma malignancies. Interestingly, there also was a high titer of antibody to the sarcoma-specific antigen recovered from the family members of sarcoma patients. Cell-free extracts derived from either sarcomas or sarcoma cell lines could be used to induce the formation of a sarcoma-specific antigen by normal human cells. This report raised questions regarding the possibility of a viral etiology for human sarcoma, and helped to generate awareness about the reality of host immunologic responses to tumor-associated antigens. It is interesting that these issues are still pertinent to the yet unresolved questions of spontaneous sarcoma etiology and also to the prospect of harnessing host immunity as part of the therapeutic armamentarium for sarcoma patients.

Sarcoma Prognosis and Staging

Over the past 30 years, there has been much interest in establishing an understanding of the prognostic factors affecting outcome in sarcoma. Some of these factors have been incorporated into sarcoma staging systems, and Cancer has frequently seen service as a podium for the presentation of these efforts. If the years from 1948 through 1978 might be characterized as an era in which basic sarcoma histologic subtypes and their natural history were being established, the ensuing several decades have witnessed the emergence of methodologies to assess prognosis, including important methodologic developments and applications such as multivariate regression analysis, Kaplan-Meier survival analysis, etc.

Staging systems for solid tumors have a long history that antedates Cancer. Primacy in the effort to initially develop a staging system for solid tumors is usually credited to Lacassagne et al, who developed the first staging system (for uterine cancer) in 1929 on behalf of the League of Nations Cancer Commission. This fascinating part of oncology history, interrupted by World War II, is unfortunately beyond the scope of this report and can be found in review elsewhere.11 Developing a staging system for soft tissue sarcoma was initiated by the American Joint Committee on Cancer (AJCC) in 1968 with their formation of a Task Force for Soft Tissue Sarcomas, whose first report appeared in Cancer in 1977.12

This Committee consisted of representatives from 13 institutions, who collected clinical case records for 1215 patients with soft tissue sarcomas who were treated between 1954 and 1968. It became apparent that the traditional TNM staging criteria would be inadequate to describe the complex biology of soft tissue sarcoma, and therefore grade (G) was introduced as an additional staging parameter. Difficulties in achieving consensus grade assignments among the sarcoma pathologists serving on the AJCC Task Force were encountered in nearly one-third of the greater than 600 cases analyzed; a problem that has continued to be a source of controversy in sarcoma staging systems ever since.

In addition to the introduction of grade as a criterion of sarcoma staging, this report also was the first in which the various parameters were arranged into staging categories that could be correlated with overall survival. The basic assignment of stage criteria was established with this initial report and has persisted, with modifications, over the ensuing 31 years. Other prognostic factors such as patient age, superficial location of the tumor, and histologic subtype were recognized as having an impact, and were incorporated into the determination of grade: “For the sake of achieving simplicity in the staging these features are considered in assigning the grade of malignancy.”12

The issue of grade assignment has continued to be problematic in the pathologic assessment of soft tissue sarcomas. Because it has proven difficult to achieve consensus regarding the specific criteria for the assessment of grade, it has likewise been difficult to reach agreement concerning precisely how grade should best be incorporated into the staging of soft tissue sarcomas. Several different grading systems have been proposed, including what is currently the most widely accepted algorithm, which has been advanced by the French Federation of Cancer Centers, a system whose antecedents were initially published in Cancer in 1988.13 Coindre et al developed a soft tissue sarcoma grading system that included 3 factors: tumor differentiation, mitosis count, and percentage of tumor necrosis. Each factor was scored using a 3-tiered and rigidly defined set of criteria, and grade was determined by summing the numeric value for each of the 3 criteria. The resultant grade assignment provided a discrete 3‒grade stratification of prognosis; when assessed in a multivariate analysis, grade emerged as the strongest prognostic factor. This report established the backdrop against which alternative grading assessment algorithms were developed and assessed. Due to its demonstrable superiority compared with other grading approaches, the French grading process is now being incorporated into the AJCC staging system for soft tissue sarcoma.

Several important reports have explored other considerations relevant to prognosis in sarcoma. Tanabe et al addressed the issue of prognosis in patients with extremity soft tissue sarcoma who received preoperative radiotherapy, focusing on outcomes for patients in whom a microscopically positive margin was observed on permanent pathologic analysis.14 All patients had intermediate‒grade or high‒grade extremity soft tissue sarcomas and all were treated at a single institution using preoperative radiotherapy and limb salvage surgical resection. A subset of this group was found to have microscopically positive margins on final pathologic analysis, even though intraoperative margin assessment had been negative. It is interesting to note that although the rate of local recurrence in these margin‒positive patients was higher than those in whom negative margins were achieved, the majority of the microscopically margin‒positive patients never developed a local recurrence. Moreover, overall survival was noted to be equivalent in margin‒positive and margin‒negative patients. Tanabe et al concluded that local control of extremity sarcoma, although very important as an objective, is not a critical factor prognostic for overall survival; consequently, the goal of primary tumor therapy should be to preserve function rather than destroy function in the attempt to secure negative microscopic permanent margins.

The microscopic assessment of sarcoma is critical in ascertaining histologic subtype, assigning grade, and examining the margin status of resected tumors. These factors all have prognostic relevance in this malignant disease cluster. Much attention, especially within the last 10 years, has focused on developing an understanding regarding the molecular determinants of tumor inception, proliferation, and metastasis, both as potential targets for therapy as well as markers of prognosis. In the early 1990s, the role of mutation in the p53 tumor suppressor gene came to be appreciated as an important prognostic factor in many malignancies, and the first report of this process in sarcoma appeared in Cancer in 1994.15 Kawai et al examined a large variety of soft tissue sarcomas using immunohistochemistry to detect the presence of nuclear‒accumulated (mutated) p53 protein. The presence of nuclear‒accumulated p53 was observed in approximately one-third of the sarcomas analyzed, a percentage that has been validated in other subsequent series that have shown it to be the most commonly mutated gene in this malignancy. The presence of nuclear‒accumulated p53 protein could be demonstrated to significantly correlate with tumor grade as well as overall and disease-free survival. Restoration of wild‒type p53 remains as a possible therapeutic strategy in solid tumors, including sarcoma, and this initial report has helped to promote the inclusion of p53 mutation in our current attempts to develop molecular prognostic signatures for sarcoma.

Therapy for Sarcoma

From ancient times until the second half of the 20th century, therapy for sarcoma consisted of surgical resection when feasible, and unfortunately, frequently also when not truly feasible. Over the past 60 years, the introduction of radiotherapy and systemic chemotherapy has helped to improve primary sarcoma control and facilitate limb‒salvage surgery, and also has enhanced the control of metastases in several specific clinical contexts. Many of the seminal reports concerning these advances in sarcoma therapy have appeared in Cancer and merit consideration as part of this review.

The introduction of radiotherapy into the management of sarcoma began in the early 1960s and antedated that of chemotherapy by approximately a decade. Sarcoma applications of external beam irradiation were first conducted in a systematic manner (as opposed to anecdotal usage) in the palliative management of soft tissue sarcomas, perhaps because of concerns regarding radiation-induced sarcomagenesis as discussed earlier, concerns that were buttressed by the then current perceptions that soft tissue sarcomas were radioresistant. In 1961, Perry and Chu reported on their experience with the use of palliative radiotherapy in the management of soft tissue sarcoma patients at Memorial Hospital.16 This experience included 104 patients with recurrent or metastatic soft tissue sarcoma who were treated between 1949 and 1959. Patients were assessed for regression of objective findings (signs) and/or subjective findings (symptoms). The tumor dose-response rate varied by histologic subtype; regression could be observed in the majority of patients after 30 Gy of radiation were administered; liposarcomas were the most sensitive subtype, and the average duration of response was 8.4 months after the completion of treatment.

An additional publication from the Memorial Hospital that appeared in Cancer in 1968 is particularly informative, and deserves a detailed examination in that it was the largest cumulative series of patients treated with radiotherapy that had been reported to date.17

The introduction to this article clearly described the bias against the use of radiotherapy that was operative at that time: “Sarcomas of the soft tissues have generally been considered as radioresistant tumors for which surgical excision has been the preferred method of treatment in the curable phase.”17 Although several experts cited in the introduction were at best modestly enthusiastic about the use of radiotherapy in soft tissue sarcoma, the authors comment that “In the last 20 years there has been no significant improvement in the overall cure rate of many types of cancer, including soft tissue sarcomas…the finding that existing techniques in both surgery and radiotherapy have just about reached their maximum effectiveness has provoked…interest in combining both methods of treatment…”17

The authors identified 142 patients who received radiotherapy from a total of 653 soft tissue sarcoma patients treated at Memorial Hospital between 1935 and 1959; this subset of patients formed the basis of the report. The majority of patients (96 patients) received radiotherapy as a postoperative treatment; 25 were treated with radiotherapy only and 46 received radiotherapy before undergoing surgical resection. In patients treated before 1946, radiation doses received had to be estimated in that only “air doses” were recorded in the medical records. The majority of patients received a total of approximately 30 Gy. Management was divisible into 2 distinct treatment eras. Between 1935 and 1945 nearly all soft tissue sarcoma patients received radiotherapy in conjunction with surgery, either as a preoperative or postoperative treatment. In 1946, this policy was modified such that radiation was only used sporadically thereafter depending on the need to either cytoreduce tumors to enhance subsequent resectability or, alternatively, to provide additional treatment postoperatively in the event of narrow or positive margins of resection. A tumor was considered to be radioresponsive if it demonstrated a clinical regression of ≥25%, foreshadowing future commonly accepted definitions of treatment response.

A particularly pertinent finding was included as the lead issue in the discussion section of this report:

“Our most important observation is the apparent control of the local tumor by radiation in 24 of 72 patients. It is obvious that an ancillary method of therapy carrying such a high rate of efficacy merits wider consideration in the management of sarcomas of the soft somatic tissues. We are of the strong opinion, therefore, that combined radiation therapy and surgery is warranted in the overall management of all varieties of soft tissue sarcomas. We further believe that preoperative radiation therapy followed by surgical excision when feasible may be the treatment of choice. We hope that formal prospective randomized clinical trials will be designed to test the point. We advocate dose levels of 3000-4000 rads in 3-4 weeks, surgery to follow in approximately 2-3 weeks.”17

The authors also mention that the 5-year and 10-year survival rates of many patients treated with radiotherapy and surgery were at least equal to and sometimes better than those achieved using surgery alone, and that patients who received preoperative radiotherapy in which 100% tumor necrosis was achieved enjoyed the highest overall survival rates, further supporting the need to prospectively evaluate the combination of radiotherapy with surgery.

An additional report appeared in Cancer in 1975 that reviewed further developments in the combination of radiotherapy and surgery for soft tissue sarcoma, as demonstrated by Suit et al in writing about their experience using these approaches at The University of Texas M. D. Anderson Cancer Center.18 Seven years had elapsed after the report cited above from Memorial Hospital, and in that short interval high‒dose radiotherapy (>60 Gy/6 weeks) coupled with conservative local excision surgery became standard, with amputation reserved for the occasional case of failure of this combined approach.18 Moreover, low‒grade, nonmetastasizing lesions such as desmoid tumors were identified for the first time as being amenable to treatment using high‒dose radiotherapy alone. The authors also presented data in support of the use of combinations of radiation therapy, multidrug chemotherapy, and limited surgery as being very efficacious in the treatment of pediatric rhabdomyosarcoma. Finally, the use of the recently developed AJCC staging system for soft tissue sarcoma discussed above was first demonstrated in this report as a means of providing effective guidance for the selection of the optimal combination of stage-specific soft tissue sarcoma treatments.

Another report from The University of Texas M. D. Anderson Cancer Center that also appeared in Cancer in 1975 further extended these observations.19 The use of combined high‒dose radiotherapy with local excision was reported to achieve local control in 87 of 100 soft tissue sarcoma patients, which compared favorably to the durable local success rates achieved at that time using radical surgery alone at either The Memorial Hospital (70%) or The University of Texas M. D. Anderson Cancer Center (75%). In drawing comparisons across institutions or treatment periods, the authors appropriately asserted the need to control for the parameters of the AJCC staging system. This caveat aside, it was particularly interesting that patients with recurrent soft tissue sarcoma who were aggressively treated (for that era) using combined therapy at The University of Texas M. D. Anderson Cancer Center had a subsequent local control rate that approached that of patients who underwent radical surgery for primary soft tissue sarcoma.

Entering the modern era of multidisciplinary soft tissue sarcoma management, several subsequent reports that appeared in Cancer have further clarified the role of surgery combined with radiotherapy. For example, Leibel et al published a series of patients so treated at the University of California at San Francisco (UCSF).20 Outstanding local control rates of 90% were achieved in a series of 81 patients treated at UCSF between 1960 and 1978. Several important observations were first made in this report, including a presentation and discussion of radiotherapy complications that occurred in multimodality-treated soft tissue sarcoma patients, and an examination of the lack of a correlation between local tumor control and the subsequent emergence of metastasis, which were most strongly predicted by the grade of the primary tumor. The authors suggested that “The high incidence of metastatic disease in Grade III lesions identifies a subgroup of patients who might benefit from adjuvant chemotherapy,”20 a potential benefit that has been pursued for the ensuing 25 years since that time without the achievement of conclusively positive results to date!

Finally, several reports of then‒novel combination therapies for soft tissue sarcoma have also appeared in Cancer. These include the first definitive report on the use of interstitial radiotherapy (brachytherapy) in the management of soft tissue sarcoma, an experience published by Shiu et al from the Memorial Sloan‒Kettering Cancer Center in 1984.21 Outstanding local control rates of 100% in previously untreated patients and 62.5% in patients with recurrent soft tissue sarcoma were presented, attributed by the authors to the joint participation of the surgeon and radiotherapist in the intraoperative mapping and placement of afterloading catheters in the tumor bed, thereby facilitating more precise and targeted delivery of radiation than was possible with external beam while sparing adjacent structures from scatter radiation effects. Moreover, the possibility of delivering radiation to well‒oxygenated tissues compared with hypo-oxygenated, postoperative surgical bed tissues had appeal, as did the temporal and financial savings for patients so treated.

An additional publication in Cancer worthy of mention is one of the first reports regarding the use of intraoperative electron beam radiotherapy in conjunction with surgical resection for retroperitoneal sarcoma by Willett et al from the Massachusetts General Hospital.22 Although this approach has not been validated over the ensuing 16 years since the publication of this study, at the time it represented an innovative strategy with which to address the very difficult problem of achieving meaningful negative margins and impacting on local control and disease recurrence in this challenging sarcoma disease site; we continue to search for better answers for sarcoma occurring in the retroperitoneum.

Cancer has played an important role as the journal in which many of the fundamental reports regarding chemotherapy applications in soft tissue sarcoma have been published. One of the first such reports involving a series of sarcoma patients with a specific, discrete histologic subtype appeared in Cancer in 1967.23 Samuels and Howe from The University of Texas M. D. Anderson Cancer Center published their personal experience with 13 patients with metastatic Ewing sarcoma in which cyclophosphamide was administered intravenously as well as intrapleurally to several individuals who had pulmonary metastases. Two patients achieved a complete response to this therapy and an additional 5 patients were partial responders. This report appears to be the first to establish a successful chemotherapy strategy for this disease. A subsequent report in Cancer by Rosen et al demonstrated the efficacy of a 4‒drug chemotherapy regimen as part of the treatment program for 20 patients with Ewing sarcoma treated at the Memorial Sloan‒Kettering Cancer Center between 1970 and 1974.24 An actuarial 5‒year survival rate of 75% was obtained, the highest such recorded rate at the time that this report was published.

The development of preoperative neoadjuvant chemotherapy for the treatment of sarcoma has been a focus of clinical research in which Cancer has played a leading role. Rosen et al from the Memorial Sloan‒Kettering Cancer Center deserve the credit for the first nonempiric use of neoadjuvant chemotherapy in the treatment of solid tumors, in this instance osteosarcoma. The underlying rationale for this novel approach was the intent to avoid the amputation of weight-bearing extremities in children who could potentially become long‒term survivors given the recently demonstrated effectiveness of adjuvant chemotherapy for this disease in the 1960s.

Based on the successful use of postamputation chemotherapy in patients with osteosarcoma, Rosen et al hoped that using chemotherapy before surgery might lead to osteosarcoma cytoreduction and the amelioration of pain, and facilitate limb salvage with prosthesis implantation in an era in which amputation was the standard surgical approach to this disease and the manufacture of customized, implantable prostheses usually took 8 to 12 weeks.25

An initial experience with 20 osteogenic sarcoma patients was launched in 1973. Patients received neoadjuvant chemotherapy while their prosthesis was being fabricated; limb‒salvage surgery with implant reconstruction was then performed, and the patients received additional chemotherapy in the adjuvant setting. Radiotherapy was avoided because of the potential difficulties that might result regarding the successful implantation of the prosthesis.

This new strategy was a radical departure from the standard of care for patients with primary osteosarcoma, and therefore it was particularly gratifying that 17 of 18 measurable primary tumors decreased in size after chemotherapy, and 16 of 20 patients achieved a complete resolution of their tumor-related pain. At the time of surgery, margin‒negative limb‒salvage resections with implantable prosthesis reconstruction were performed. This treatment approach was completed for 15 patients who were then subsequently followed for 2 to 15 months postoperatively without any evidence of disease recurrence, making this a remarkably improved approach to this disease:

“These preliminary results indicate that with the use of aggressive chemotherapy, it is possible to demonstrate objective tumor regression in primary osteogenic sarcoma, allowing the surgeon to perform en bloc resection of tumor and prosthetic replacement of the involved bone. Although the limb is preserved, it is important to stress that extensive surgery yielding tumor-free margins is performed. The ultimate evaluation of this approach to the treatment of primary osteosarcoma awaits longer observation, to determine limb function and the continued disease-free status, once adjuvant chemotherapy is discontinued.”25

The importance of the novel neoadjuvant approach to osteosarcoma developed by Rosen et al cannot be overemphasized in that it paved the way for the use of systemic neoadjuvant chemotherapy in a wide variety of tumors of both epithelial and mesenchymal origin during the ensuing years.

The application of this strategy in soft tissue sarcomas has been the source of ongoing effort and controversy in that the available chemotherapy agents generally have inferior efficacy:toxicity profiles compared with the drugs available for osteosarcoma. Consequently, the use of neoadjuvant chemotherapy in soft tissue sarcoma initially received attention in the context of a primary tumor that was either unresectable or unresectable without amputation. An early report of this approach for soft tissue sarcomas appeared in Cancer in 1987, in which the results were presented for 34 patients considered to be uresectable or unresectable without amputation who were treated at the Institute Gustave-Roussy using neoadjuvant chemotherapy.26 Objective regression was observed in 13 of 34 patients. Limb‒sparing surgery was able to be performed in 21 patients, 3 patients required amputation, and 10 patients developed disseminated disease while receiving therapy. Although 19 of the 21 patients treated with limb‒salvage surgery underwent a complete macroscopic excision, only 12 had microscopically negative margins on final pathologic analysis. The majority of patients received postoperative radiotherapy; 22 of 34 patients were considered to have disease in complete remission after all therapies were delivered. Kaplan-Meier analysis demonstrated a 2‒year survival rate of 80% in these 22 patients who otherwise would have been facing mutilating surgical procedures. This represented an important advance compared with the conventional treatment of that era.

An additional classic report was the contribution of Rosenberg et al at the National Cancer Institute, who reported on an initial, prospective, randomized trial experience in which the use of adjuvant chemotherapy was evaluated in patients after definitive local therapy in the form of limb‒salvage surgery with radiotherapy versus amputation had been performed.27 This study perhaps enjoys a retrospectively applied reputation as one of the most controversial soft tissue sarcoma studies published in the literature. The potential preadjuvant chemotherapy selection bias introduced by randomization after (very different!) definitive local therapies had been completed, questions regarding histologic subtype stratification and balance between the treatment arms of the trial, the power of the accrual targets, the survival analytic methodologies used, etc. have all been the subject of intense debate for the past 25 years. Caveats notwithstanding, the significance of this study as the first positive prospective adjuvant chemotherapy trial in adult soft tissue sarcoma has stood the test of time, and remains a sentinel effort to address an as‒yet unresolved issue in this disease.

In an attempt to further enhance the efficacy:toxicity profile of chemotherapy for patients with soft tissue sarcoma, an important and currently still unresolved question has been the application of isolated limb perfusion (ILP) for primary, nonmetastatic, soft tissue sarcoma that is unresectable or unresectable without amputation. Used in these contexts, the hope is that ILP can be used to deliver chemotherapy to extremity tumors at doses that cannot be tolerated via systemic administration. Pilot studies have demonstrated that the majority of chemotherapeutic agents for systemic soft tissue sarcoma cannot be used in ILP circuits, either due to toxic effects or unfavorable drug‒ILP circuit interactions. Consequently, most ILP experiences demonstrating efficacy in soft tissue sarcoma and other malignant diseases have used melphalan administered with recombinant tumor necrosis factor-α. As this approach gained traction in the early 1990s, the largest single‒institution report of ILP for unresectable soft tissue sarcoma was published by Gutman et al from Tel Aviv University.28 The treatment cohort was comprised of 35 patients with high‒grade soft tissue sarcomas who underwent ILP with melphalan and recombinant tumor necrosis factor-α; an overall response rate of 91% was achieved, including 13 patients (37%) with complete responses and 19 (54%) with partial responses. Limb‒sparing surgery rather than amputation was possible in 29 of 34 (85%) patients, a percentage that to my knowledge remains among the highest observed conversion to limb salvage rates recorded in the literature.

In conclusion, it is clear that Cancer has played a critical role as a forum for the broad spectrum of sarcoma-relevant research over the past 60 years; many of the seminal reports in this oncology focus appeared here first. We are entering an exciting new era in sarcoma care, and it can be anticipated that surgical approaches will continue to improve, particularly with the advent of tissue-engineered replacement body parts; that radiotherapy strategies such as intensity‒modulated radiotherapy and proton-beam therapy will evolve to become ever more precise and sparing of normal tissue; and that locoregional interventions and systemic treatments will move into the era of personalized molecular therapeutics, hopefully including molecular modifiers of chemotherapy toxicity:response profiles. Many of these improvements will be based on incisive molecular-oriented laboratory research, which is already yielding new insights into sarcoma inception, proliferation, and dissemination. In turn, this new knowledge can be expected to move from the laboratory bench to application as novel care strategies for patients with sarcoma. Given the remarkable publication track record in sarcoma research, it can be anticipated that Cancer will continue to have an impact as a major platform for the presentation of research reports in this rare disease. Please join me in support of this critical effort!