Study of Preoperative Radiotherapy for Sarcomas of the Extremities with Intensity-Modulation, Image-Guidance and Small Safety-margins (PREMISS)

Treatment of soft tissue sarcomas of the extremities is a challenge for the interdisciplinary team. In general, radiation therapy (RT) is indicated in stage II and III (not T1a). There are two approaches for RT in this situation, either neoadjuvant or adjuvant. For preoperative RT usually lower doses are applied, and a dose of 50 Gy has been established; in the postoperative setting higher doses between 60-66Gy are required. Thus, preoperative RT seems to be more beneficial in terms of long-term RT-associated side effects such as edema, joint stiffness, nerve lesions or bone fractures. On the other hand, preoperative RT is associated with a higher rates of wound complications after surgery (35 % vs. 17 %). This is more or less independently of the location, i.e. shoulder, upper or lower extremity and of the histology, which can be very heterogeneous including liposarcoma, leiomyosarcoma, undifferentiated sarcoma or synovial sarcoma. Thus, due to the required expertise of all disciplines, patients with such tumors should be treated at a specialized sarcoma unit [1‐6], since it has been shown that treatment at a high volume center is associated with a significantly increased survival and better functional outcome [7].

Surgery is the mainstay of treatment in sarcomas and should be evaluated in every case. If surgery with a complete removal of the tumor is not possible, RT is a curative alternative; local control rates range between 20–45 % [8]. Generally, function-preserving treatment is a main goal, whereas in the past radical excisions with compartment resections leading to a loss of function, or amputations, were performed regularly; today, function-preserving treatment is anticipated with complete removal where possible, and combination with RT when necessary.

Few randomized studies are available which is mainly due to the low incidence of soft tissue sarcomas: An older trial compared limb amputation with a combination of extremity-conserving resection plus postoperative RT; no difference in disease-free survival and overall survival was observed [9].

Two other trials assessed postoperative RT or interstitial brachytherapy, both studies showed a clear advantage for local control compared to surgery alone [10, 11]. A large number of retrospective analyses have confirmed the positive value of postoperative RT for extremity sarcomas [12‐23].

Today, extremity-conserving surgical treatment is possible in 80–95 % of all patients [24‐28]. This requires, however, a well-functioning interdisciplinary team consisting of orthopaedic surgeons, radiation oncologists, plastic surgeons, oncologists, pathologists and radiologists [29, 30].

In detail, two main concepts exist for the application of RT: preoperative vs. postoperative. As in other indications such as esophageal, pancreatic or rectal cancer, there are clear arguments in favor of preoperative RT: The treatment volume is generally much smaller, since postoperative changes as well as intraoperatively manipulated tissue including the surgical entry channel and scar do not need to be treated [31]. Compared to postoperative RT, only doses of around 50 Gy are required. The smaller treatment volume together with the lower RT dose result in lower rates of treatment-related side effects. Moreover, tumor as well as normal tissue oxygenation is not impaired due to postoperative scarring; this leads to a higher sensitivity to radiation due to the oxygen enhancement ratio (OER; [23]).

Another factor is the possibility of sterilization around the tumor using preoperative RT, leading to an improved resectability and higher rates of R0-resections [19]. This might be explained by a thickening of the tumor capsule which has been shown in experimental settings. On the other hand, in spite of the clear advantages of preoperative RT, higher rates of surgery-related wound complications have been shown by several groups [21, 32, 33].

A number of comparative analyses between pre- and postoperative RT are currently available in patients with extremity soft tissue sarcomas. A randomized prospective trial by O’Sullivan and colleagues randomized 50 Gy preoperative RT to 66 Gy postoperative RT. In the preoperative group, which consisted of 94 patients, 10 patients received an additional boost up to 16–20 Gy in cases of R1 resections. Initial data showed a slightly improved local control and survival in the preoperative group, however wound complications were 35 % compared to 17 % in the postoperative RT group; function of the extremity was comparable in both groups [13, 18]. Long term data support the beneficial risk-benefit ratio of preoperative RT [14]. In agreement with several retrospective reports lower rates of long-term side effects such as edema, fibrosis, fracture, joint stiffness or nerve toxicity as well as better functional outcome are observed [14, 15, 17, 22, 23, 32]. A meta-analysis including 1098 patients from 5 studies confirmed a higher local control and a higher overall survival of 76 % vs. 67 % for pre-operative RT [34]. A multi-institutional matched-pairs analysis including 821 patients also reported an improved overall survival after preoperative RT [20].

However, it has to be kept in mind that wound complication rates might be higher after preoperative RT with median complication rates of 16–35 %, depending on the series [12, 17‐19, 21, 32, 35]. The rate of wound complications is dependent on RT dose, patient age, comorbidities, tumor and resection volume as well as tumor location [18, 21, 32, 36]: For example, patients with wound complications generally have a much larger resection volume than those without complications (919 cm³ vs. 456 cm³) [33].

Regarding the RT technique, 3D-conformal RT was standard over many years. Modern techniques such as intensity modulated radiotherapy (IMRT) offer improved dose conformality even for long and complex shaped volumes. Treatment planning comparisons analyzing 3D vs. IMRT could show that dose coverage as well as reduction of dose to normal tissue (bone, soft tissue) are better with IMRT [37‐39]. Thus, since the implementation of IMRT for the treatment of soft tissue sarcomas, positive results were reported [40]. With helical IMRT as Tomotherapy^© dose distributions often are even more conformal, longer volumes can be treated, and the treatment machine offers online MV-CT imaging for position verification. Early reports on Tomotherapy^© for sarcomas reported excellent results as well as improved sparing of normal tissue [41‐44]. For daily repositioning image-guidance as well as positioning devices are necessary. For extremity tumors, positioning inaccuracies of 1 cm or more have been observed [45]. This can be compensated by adequate treatment volumes as well as IGRT approaches. The improvements of RT techniques enable the radiation oncologist to reduce and adapt treatment volumes. For soft tissue sarcomas, in the past, uncertainties in positioning as well as in target volume definition depending e.g. on insufficient imaging have led to very large safety margins of ≥ 5 cm proximal/distal and 2 cm circumferentially around the visible tumor volume [31, 38, 45, 46]. Since optimized imaging including magnetic resonance imaging (MRI) as well as CT or PET-diagnostics are available, these safety margins can be reduced and IGRT approaches assure high precision of treatment delivery. Results from brachytherapy series have shown that local dose application to the tumor with small margins of 1–2 cm are excellent with local control rates of 79–87 (−100) % [10, 40, 47‐50]. The main factor, however, is exact definition of the target volume and precise dose delivery. Since complication rates are dependent on the irradiated volume, the rationale for smaller safety margins is an optimization of the risk-benefit ratio [33, 51]; initial clinical data on IMRT for soft tissue sarcomas of the extremity have shown local control of 96 % at 3 years with small margins of 2 cm [52]. Intraoperative Radiotherapy (IORT) or brachytherapy can offer an enhanced therapeutic ratio: With both techniques local dose escalations directly to the target tissue are possible, without irradiation of large areas of normal tissue.

For neoadjuvant RT, doses of 50 Gy have been established, however, in some cases incomplete tumor resection with R1 margins requires individualized approaches. It has been shown that local dose escalation as a boost treatment up to 16–20 Gy with conventional fractionation can be performed, however, series from the literature show controversial results [18, 53].

Combination of percutaneous RT (40–50 Gy) and a brachytherapy boost (15–32 Gy) has been reported to be superior to percutaneous RT or brachytherapy alone [25, 49, 54‐58]. Thus, combination treatments are considered as optimal for soft tissue sarcomas with positive resection margins [29, 59].