Retroperitoneal sarcomas (RPS) are rare malignancies arising within the retroperitoneal space, characterized by their infiltrative nature and propensity for local recurrence. Optimal management typically involves surgical resection; however, achieving negative margins can be challenging due to anatomical constraints and proximity to critical structures. As a result, local recurrence rates for RPS are high. Neoadjuvant radiotherapy aims to reduce tumor size and improve resectability but is associated with toxicity and delayed wound healing. While long-course radiotherapy has been employed with some success in a sub-group of RPS, it often requires prolonged treatment courses and can delay surgical resection [1]. Short-course radiotherapy (SCRT) offers the advantage of delivering a higher dose of radiation over a shorter time frame, made possible with conformal technologies with image guidance and volumetric modulation. There has been a paradigm shift in recent years on other cancer types favoring SCRT [2,3]. For sarcomas, there have been a few small prospective studies on soft tissue sarcomas of the extremities and trunk [4,5].

In this study, we sought to investigate the feasibility and utility of neoadjuvant SCRT for retroperitoneal liposarcoma, followed by immediate surgery. This feasibility study aimed to assess whether neoadjuvant SCRT targeting margins can be integrated into the treatment paradigm for RPS, and to report on early safety outcomes of ultra-fractionated radiotherapy.

The hypothesis is that SCRT to threatened margins is a safe and efficacious treatment method for retroperitoneal sarcomas. The primary objectives are to assess the safety and acute toxicities, including postoperative complications, of neoadjuvant short-course radiotherapy followed by immediate surgery in resectable retroperitoneal sarcomas, to evaluate the effects of neoadjuvant short-course radiotherapy on surgical procedures, e.g., frequency of intra-operative fibrosis, and to report short-term tumor control outcomes.

Eight patients diagnosed with retroperitoneal sarcoma were enrolled in this single-center, prospective feasibility study. Eligible patients were scheduled to receive ultra-fractionated radiotherapy before surgical resection. Safety assessments were conducted regularly throughout treatment and included monitoring of acute toxicity, wound healing complications, and surgical outcomes. Feasibility endpoints included treatment compliance, completion rates, and technical feasibility of radiotherapy delivery. This is a single-center, prospective, non-randomized feasibility study.

Patients were recruited by surgical oncologists, discussed at multi-disciplinary meetings, and referred to a radiation oncologist for discussion, communicating that neoadjuvant radiotherapy is not routine, and that SCRT is novel. Fig. 1 illustrates the flowchart for the study pathway.

SCRT was administered via image-guided-volumetric modulated arc therapy following four-dimensional computed tomography simulation. Treatment consisted of 5-fraction radiotherapy, followed by immediate surgery. Patients were prescribed 25 Gy for the first stratum. At the next level, patients were prescribed 25 Gy with simultaneous integrated boost of 30 Gy. When tumors were small, the entire tumor was treated. When tumors were deemed too large to be entirely encompassable by the radiotherapy fields, only threatened margins were irradiated. Where margins were anticipated to be “air” margins; pushing instead of infiltrative, a joint decision by radiation oncologists and surgeons was made only to include threatened margins, which typically included tumor interfaces with vertebral bodies, great vessels (aorta, inferior vena cava, portal vein), and abdominal wall musculature up to an anticipated location of the free intra-abdominal cavity, by international guidelines [6]. Dose constraints were met and were extrapolated from guidelines from Task Group 101 [7] and UK stereotactic ablative radiotherapy [8].

The study would be stopped if one or more postoperative complications of fistula or perforation happened at the site of anastomosis within the high-dose irradiation field since such events are rare. Adverse events were graded using Radiation Therapy Oncology Group score during the acute phase and National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0, up to 6 months post-operation.

Any patient diagnosed clinically, radiologically, and histology-proven to have RPS and deemed operable would be considered. The patient may have primary or locally recurrent RPS.

Clinical, radiological, and pathological assessments were conducted before and after SCRT. Surgical resection with the intent of achieving negative margins was performed as soon as logistically possible upon completion of SCRT. Surgical outcomes, including margin status and postoperative outcomes, were assessed.

Eight patients were treated with SCRT; Table 1 summarizes the included patients’ characteristics. All patients completed the planned course of ultra-fractionated radiotherapy without interruptions.

Patient 8 developed coronavirus disease 2019 (COVID-19) infection before her planned surgery. She had a recurrent multi-focal low-grade dedifferentiated liposarcoma and a joint decision was made to abandon surgical salvage for palliative chemotherapy. Patient number 8 tolerated the SCRT well and remained progression-free for 18 months after SCRT while she was maintained on palliative trabectedin. The sarcoma surgeons did not report intra-operative fibrosis for the seven patients who underwent resection. All seven patients, but one (patient 7), recovered well post-operatively. Patient 7 was an elderly patient with a large undifferentiated pleomorphic sarcoma, with pre-existing cardiac and pulmonary co-morbidities who developed an infected abscess at the distal pancreatomy bed and passed away 3 months postoperative from an unrelated sudden cardiac event. After excluding patients 7 and 8, with a median follow-up of 20.5 months, six patients were eligible for evaluation; five patients were disease-free except for one who developed an out-of-field relapse (patient 4) as she did not receive an en bloc multi-organ extended resection due to the large and central nature of her tumor, and only the central vessels were irradiated (Fig. 2).

The surgeons did not report intra-operative fibrosis for the seven patients who had resection. There were no acute toxicities greater than grade 2 during and immediately after radiotherapy. Post-operatively, patient 7 developed an abscess at the pancreatic bed and passed away from a sudden cardiac event. Patient 5 developed a chyle-leak post-operatively which was self-resolving. There were no long-term toxicities >grade 3.

RPS remains a challenging malignancy to treat due to its complex anatomical location. Currently, survival rates for low and intermediate-grade RPS are about 50% at 5 years [9]. Oncologic en bloc extended resection is required for the cure of non-metastatic retroperitoneal sarcoma [10]. Recurrent RPS also requires resection, considering the patient’s age and performance status, pace of disease, and type of initial resection. Current evidence suggests that while neoadjuvant radiotherapy was not associated with an overall benefit, the STRASS and STREXIT trials showed a consistent signal that well-differentiated (WD) and low-grade dedifferentiated liposarcomas (DDLS) demonstrated improved abdominal recurrence-free survival [1,11]. Currently, neoadjuvant radiotherapy is not routinely recommended. Recent evidence has shed light on the different behaviors of different histology subtypes of RPS, prompting STRASS 2, which is investigating the role of systemic therapies in subtypes that have a high propensity for distant metastases (high-grade DDLS, undifferentiated sarcomas, and leiomyosarcomas) [12]. Until more evidence comes to light, the use of neoadjuvant radiotherapy remains controversial and is to be discussed on an individual patient basis for WD and low-grade DDLS.

The current standard dose-fraction is about 50 Gy over 5 weeks of daily treatment (long-course radiotherapy [LCRT]) followed by delayed surgery. While LCRT may be considered for certain RPS subtypes, there are significant logistical issues. Surgeons are unwilling, understandably, to delay curative surgery for LCRT in the absence of very compelling evidence. SCRT may overcome these issues, although its toxicity may not be insignificant in patients with very large tumors. Involved margins are common in RPS and associated with worse local controls [12]. Hence, we were prompted to investigate the feasibility of a novel approach using SCRT, targeting especially the margins-at-risk in these patients. Other than logistical convenience, there could be inherent biological benefits. As the alpha/beta ratio of soft tissue sarcomas, especially low-grade, is presumably low, a high dose per fraction is more desirable, hence favoring SCRT. Boosts to high-risk margins are sometimes given as reported in some studies, with some suggesting a benefit [13,14]. There is an ongoing trial to study SCRT using proton therapy with boosts to the margins [15]. In the STRASS study, 16% of the patients had tumor progression before surgery was administered. SCRT can mitigate this problem. This sentiment is echoed by Istl and Gronchi [12].

We have provided early and small evidence that SCRT to the margins is safe, intra-, and peri-operatively, with minimal logistical difficulties for patients with RPS that are located near critical structures. One patient died from a sudden cardiac event and he did develop an abscess at the distal pancreatectomy site. We investigated and found that the radiotherapy dose at the distal pancreas was 27 Gy in 5 fractions, and our surgeon did not report increased fibrosis intra-operatively. Regarding pancreatic SCRT and extrapolating from pancreatic cancer studies, two papers studying the feasibility of photon [16] and proton [17] neoadjuvant SCRT showed that the former was associated with more toxicities than the latter due to increased fibrosis that the surgeons reported intra-operatively while performing a Whipple’s procedure. Other studies [18] have shown that neoadjuvant SCRT with chemotherapy is safe for downsizing pancreatic cancer pre-Whipple’s procedure and hence, this patient is more likely to have developed complications due to other medical factors rather than due to the SCRT given to him. Another patient’s surgery was aborted due to COVID-19 infection because this study was conducted amidst the initial pandemic. She continued to demonstrate sustained control of her recurrent tumors for at least 18 months post-SCRT. The SCRT probably contributed significantly to her extended survival on top of the palliative trabectedin.

Of the six patients who underwent resection and had longer-term follow-up, all six did not have complications peri-operatively other than one who had persistent chyle-leak, which was conservatively managed. One developed an out-of-field relapse almost two years after initial SCRT as only the tumor near the abdominal aorta, inferior vena cava and mesenteric vessels were irradiated. In the STRASS study, 39% developed at least a grade 3 complication throughout the study, and 14% developed peri-operative complications, although it is hard to ascertain how much radiotherapy contributed to these complication rates specifically [1]. Although our numbers are small, we have shown SCRT to be safe and feasible.

The current standard dose fractionation for RPS would be 50.4 or 50 Gy in 28 or 25 fractions. The equivalent dose for the short-course regimen as compared to the standard is described below (Table 2). Extrapolating from truncal and extremity soft tissue sarcomas, hypofractionated radiotherapy is gaining traction, with a few prospective trials reporting safety and equivalent tumor control ranging from 20 to 35 Gy in 5 fractions [4,5,19,20]. In terms of acute toxicities, SCRT is expected to have less (α/β=10), and in terms of sarcoma control (α/β=3), similar for 30 Gy in 5 fractions but lower for 25 Gy in 5 fractions (Table 2). As RPS is located near sensitive organs, unlike truncal and soft tissue sarcomas, we would recommend applying more caution. Hence, we propose future trials to use 30 Gy in 5 fractions as a starting point for investigations.

Many questions remained unanswered for the radiotherapy management of RPS, other than dose-fractionation. We are not sure if radiotherapy to the margins only is sufficient, and it is also difficult to predict where the at-risk margins are pre-operatively based on imaging. We await longer-term results from STRASS and STREXIT. STRASS 2 will be informative on the role of chemotherapy in high-risk retroperitoneal sarcomas, but the paradigm of low-risk retroperitoneal sarcomas remains status quo.

Our study’s strengths are that target delineation was performed with both the radiation oncologist and the primary surgeon to ensure that the areas at high risk of close or involved surgical margins were irradiated. This is also the first publication, as far as we are aware, to report on the feasibility and safety of RPS using SCRT. The surgeries were performed in a single high-volume tertiary center. The weaknesses of our study are small numbers and short follow-up. It is documented in existing literature that the risk of local recurrence does not plateau even after 15–20 years [21]. The definition of at-risk margin is also dependent on surgeons’ opinions.

This feasibility study showed that hypofractionated neoadjuvant radiotherapy to margins-at-risk is safe. We have demonstrated early results to provide insights on safety, and potential impact on surgical outcomes and early tumor control. If proven safe and efficacious in the long run with larger prospective studies, SCRT may be a valuable addition to the treatment armamentarium for this challenging disease.