| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Department of Surgical Oncology, Unit 444, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas 77030
Correspondence: Address correspondence and reprint requests to: Peter W. T. Pisters, MD, FACS; E-mail: ppisters{at}mdanderson.org
In this issue of Annals of Surgical Oncology, Mack and colleagues1 from the Tom Baker Cancer Center at The University of Calgary report their results with preoperative intravenous or intra-arterial doxorubicin and short-course, high-dose-per-fraction (10 fractions of 3 Gy each) radiotherapy followed by wide local resection for selected patients with extremity soft tissue sarcomas. The authors report remarkably low treatment-related complication rates, and in the subset of patients who underwent preoperative chemoradiation followed by macroscopically and microscopically complete (R0) surgical resection, a 97% 5-year local control rate was observed. Can this regimen serve as a therapeutic maquette, or are these results best regarded as aleatoric, as are many findings in clinical research? This report raises a number of important issues that warrant further consideration and discussion. These include the important issues of patient selection, complication rates, radiation fractionation and schedule, and the relative contributions of chemotherapy, radiation, and patient selection to these results.
Historical background is relevant in understanding the rational underpinning of the treatment approach used by Mack and colleagues. This approach has its roots in the protocols developed two decades earlier at the University of California, Los Angeles (UCLA) by Eilber et al.2 The UCLA group adopted large-dose-per-fraction preoperative radiotherapy (10 fractions of 3.5 Gy each) to overcome the putative radioresistance of soft tissue sarcomas.3 The investigators were influenced in their regimen design by the observation that irradiation of some melanoma cell lines resulted in a large initial shoulder on the cell survival curve, although later evidence suggested that this may not be consistent.4 Nonetheless, the potential radioresistance of melanoma and, by extension, sarcoma was attributed to the tumor cells’ capacity for repair of sublethal damage implied by this shoulder. This preclinical speculation led to pilot studies performed in patients with soft tissue sarcoma by Eilber et al.5 and others6–8 that established the relative toxicities and local control rates associated with short-course, high-dose-per-fraction radiation and concurrent intra-arterial or intravenous anthracycline-based chemotherapy. On the basis of these pilot studies, a handful of centers, including the Tom Baker Cancer Center, adopted this approach as their standard therapy for patients with localized soft tissue sarcomas.
Patient selection is one of the many issues that should be considered in the interpretation of the Tom Baker Cancer Center experience with the modified UCLA regimen. In the abstract and introduction, Mack and colleagues report that their chemoradiation approach was used in all patients who presented with soft tissue sarcomas of the extremity and trunk between 1984 and 1996, and their results are presented as a consecutive series. However, in the Methods section, the authors clarify that they excluded patients who were believed to have tumor involvement of major nerves or bone. They do not indicate what specific clinical or radiological criteria were used to exclude patients or how many patients were excluded on this basis, but given their large referral base, it is conceivable that many patients were excluded. These patient selection criteria may account in considerable part for the relatively low amputation rate and the comparatively high rate of R0 surgical resection observed in this study.
The frequencies of chemoradiation-related toxicities reported by Mack and colleagues are remarkably low. The authors report only four patients with chemotherapy-related complications. Indeed, the toxic effects attributed to chemotherapy seemed to have arisen more from the specific route of administration (intra-arterial) than from the chemotherapy itself. Mack and colleagues observed no significant chemotherapy-related toxic effects in patients treated with intravenous doxorubicin—a therapeutic feat that many medical oncologists have yet to achieve. However, the authors point out that they retrospectively determined toxic effects; thus, direct comparison of their results with toxicity data collected in prospective clinical trials is difficult. Hence, notwithstanding the best efforts of Mack and colleagues to retrospectively record treatment-related toxicities, the vagaries of the medical record and the bias associated with physician recall are such that the toxicity profile of this regimen is likely more significant than is suggested by their article.
Mack and colleagues report an overall wound complication rate of 15%. It is interesting to note that this retrospectively evaluated and determined wound complication rate in patients treated with preoperative chemoradiation is comparable to that reported after surgery alone in prospective trials.9,10 The authors correctly point out that their relatively low wound complication rate may be a function of the retrospective nature of their assessment and, possibly, relatively frequent involvement of reconstructive surgeons when clinical circumstances indicated a high risk of wound complications. The inference from these data is that involvement of reconstructive surgeons can, in some settings, abrogate the adverse effects of preoperative radiotherapy on wound healing. However, the retrospective nature of this report does not allow us to readily interpret the relative contributions of patient selection, lower radiation dose, and involvement of reconstruction surgeons to this low complication rate. Indeed, in the randomized trial comparing preoperative with postoperative radiotherapy in extremity soft tissue sarcoma conducted by the Canadian Sarcoma Group/National Cancer Institute of Canada Clinical Trials Group, a 35% major wound complication rate was observed in patients treated with preoperative radiation (50 Gy in 25 fractions) when prospectively defined criteria were applied at predetermined intervals up to 4 months after surgery.9 In contrast, the original retrospective analysis by the same group that developed the criteria ultimately used in the randomized trial showed a complication rate nearly identical to that reported by Mack et al. (16% vs. 15%).11 This implies that caution is warranted in interpreting retrospective data on wound complication rates, especially in circumstances in which wound care, such as deep packing and drainage of wounds, may continue in the community for up to 4 months. Hence, complications may be underestimated.
Mack et al. report their local control rates stratified by microscopic margin status. They provide a reasonable rationale for this in that it is likely that tumors in patients with microscopic residual disease have a different natural history than do many tumors in patients who have undergone an R0 resection. The 97% 5-year local control rate observed in the subset of patients who were selected for multimodality treatment, completed preoperative treatment, and subsequently underwent resection with both macroscopically and microscopically negative surgical margins is certainly favorable. However, it is again not clear to what extent optimal patient selection, preoperative treatment, state-of-the-art staging and surgery, and selective analysis of only R0 cases affect these results.
Important issues for future research are raised by this report. One of the most compelling is the possibility that shorter courses of radiation may be just as effective and less toxic than conventional preoperative radiation regimens of 50 Gy in 25 fractions. The latter approach results in major wound complications in one third of patients treated in single-institution and multi-institution settings, including prospective trials.9,12 However, notwithstanding the favorable results reported by Mack and colleagues, it would be difficult to conduct a phase III trial designed to prove that a lower total radiation dose provides comparable local control with reduced treatment-related morbidity and financial costs. Physician interest in an equivalence trial would likely be relatively high among surgeons but comparatively lower among radiation oncologists, who have historically focused primarily on radiation dose intensification approaches or on dose-fractionation strategies to ameliorate tissue damage rather than on approaches that reduce or withhold radiation in subsets of patients. Shorter-duration, lower-dose regimens are more attractive to patients and less costly to administer but have not been widely accepted by radiation oncologists—particularly in North America. Evidence for this is seen when examining practice patterns for patients receiving palliative radiotherapy13,14 (where concern for late treatment effects is considerably less) and in the general lack of North American adoption of short course, low-total-dose, high-dose-per-fraction external-beam radiation treatment for patients with localized rectal cancer. It is possible that the reluctance to adopt proven (in phase III trials) regimens of short-course, higher-dose-per-fraction radiotherapy15,16 in North America relates in part to the known influence of smaller fractions in protecting against late tissue toxicities. It is also possible that the reluctance to accept shorter course treatments is influenced by pecuniary matters such as reimbursement13,14—the "third rail" of therapeutic politics. To their credit, the issue of a possible conflict of interest posed by fraction-based remuneration of radiation oncologists has recently been brought into the light by radiation oncologists themselves14 in their examination of the evidence in favor of palliative single-fraction radiation treatment, in which they raise the open question of "whether radiation oncologists in the United States will practice evidence-based or remuneration-based medicine."
Regardless of the underlying reason or reasons for a general resistance to short-course radiation treatment for patients with localized soft tissue sarcoma, Mack et al. should consider performing a more detailed analysis of late tissue effects and extremity function in patients treated with this regimen. They report no long-term complications of wound breakdown, bone fracture, second malignancy, or late neurovascular adverse effects. However, the clinically significant adverse effects of edema and fibrosis were not reported. Further studies should evaluate the rates of treatment-related fibrosis and edema and should include functional assessment with a contemporary extremity function assessment tool such as the Toronto Extremity Salvage Score.17 It would also be of interest to see a cost comparison of standard preoperative radiation versus shorter-duration chemoradiation treatment. Such economic analyses can be performed with sensitivity analyses that model not only the costs of the various treatment components, but also the costs of managing treatment-related toxicities (such as wound complications) at variable rates of frequency.
Another important issue for consideration is the relative contribution of anthracycline-based chemotherapy to these results and its use in patients with localized, nonmetastatic sarcoma. The UCLA group adopted low-dose regional doxorubicin because of the drug’s documented antitumor activity and its potential as a radiosensitizer.3 It is conceivable that the radiosensitizing properties of doxorubicin increase the biological effectiveness of lower-dose radiation and allow for the administration of a total radiation dose less than the standard dose. Without controlled trials, we really cannot determine the extent to which concurrently administered anthracycline-based chemotherapy may contribute to the observed toxicities and overall treatment results with chemoradiation, including potential late tissue effects. For the moment, it seems that anthracycline-based adjuvant chemotherapy should be offered only to very carefully selected, well-informed patients with high-risk localized disease, because the second generation of phase III adjuvant chemotherapy trials for patients with localized high-risk extremity and trunk sarcomas have been uniformly negative for overall survival,18–20 and cohort analysis of the cumulative experience of Memorial Sloan-Kettering Cancer Center and the University of Texas M. D. Anderson Cancer Center with anthracycline-based chemotherapy for 674 consecutively treated patients with stage III extremity soft tissue sarcoma revealed a statistically significant inferior sarcoma-specific survival in patients who received chemotherapy (hazard ratio for death from stage III sarcoma in chemotherapy-treated patients vs. patients not receiving chemotherapy, 1.36; 95% confidence interval, 1.02–1.81; P = .04).21
In summary, the report by Mack and colleagues is thought provoking and, indeed, hypothesis generating. First among these is the hypothesis that we can achieve comparable or even better results with less treatment in selected patients. Perhaps we can indeed achieve more with less—addition through subtraction. Alternatively, we cannot be blind to the possibility that the findings of Mack and colleagues create the potential for complicating the treatment paradigm for patients with localized sarcomas by incorporating unproven chemotherapy combined with an unusual radiation dose and fractionation regimen. Perhaps we are compounding the interest in this treatment approach with an inadequate understanding of the relative contributions of chemotherapy and radiation treatment to these results and insufficient data on late effects. In the slowly evolving world of sarcoma treatment, it is increasingly clear that only collaboration and clinical trials can answer these questions for physicians and for patients with soft tissue sarcoma.
Received for publication June 2, 2005. Accepted for publication June 3, 2005.
REFERENCES
This article has been cited by other articles:
 |
|
 |
|
  J. F. Tseng, M. T. Ballo, H. N. Langstein, J. D. Wayne, J. N. Cormier, K. K. Hunt, B. W. Feig, A. W. Yasko, V. O. Lewis, P. P. Lin, et al. The Effect of Preoperative Radiotherapy and Reconstructive Surgery on Wound Complications after Resection of Extremity Soft-Tissue Sarcomas Ann. Surg. Oncol., September 1, 2006; 13(9): 1209 - 1215. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
