Retroperitoneal sarcomas (RPS) are rare, anatomically diverse, and biologically heterogeneous tumours. The management of patients presenting with RPS is complex and requires an individually tailored treatment strategy informed by disease, patient, and technical factors. While patients with other cancers now benefit from advances in immunotherapy and targeted agents, surgical resection remains the only curative treatment for RPS. In their study published in this journal, Tai et al. 1 evaluate a cohort of patients with primary non-metastatic RPS referred to a tertiary sarcoma centre, with a view to defining unresectability. Their work is a welcome addition to the literature, demonstrating a low rate of technical unresectability compared to previous studies and good insight into comorbid factors that deemed patients unresectable; their work invites broader questions of how resectability in advanced gastrointestinal malignancy including RPS is defined, who should define it, and when it should be defined. The second half of the 20th century saw an evolution in many areas of complex abdominopelvic surgical oncology, where refinement of surgical technique combined with advances in imaging and perioperative medicine allowed previously established anatomically based absolute contraindications to surgery, often centred around the extent of involvement of major neurovascular, to be challenged and definitions of resectability to be revised and extended. This phase focussed on the evaluation of morbidity and mortality rates associated with increasingly radical resections, and patient selection and determination of resectability at this time depended heavily on institutional and individual surgeon technical capability. While technical constraints remain relevant, in 2025 the question for an experienced multi-disciplinary team when it comes to patient selection for many advanced abdominopelvic tumours is less commonly focussed on what can be technically resected, but what should be resected from an individual patient and biological perspective. The call to standardise resectability criteria in RPS is understandable, but it is posited that no single guideline can account for the inherent variability in biology and anatomical involvement. A logical comparator, within the field of advanced gastrointestinal surgery, when considering defining criteria for resectability is guidelines for pancreatic cancer, which the Tai et al. discuss in their article. Resection criteria for pancreatic cancer are based on a relatively consistent anatomical template, with a limited number of key vascular relationships to the pancreas 2. RPS, by contrast, may involve any structure within or adjacent to the retroperitoneum 3-5. These tumours originating in the retroperitoneum may extend into different planes, compartments, or organs, while remaining resectable 3-8. While it may seem at first reasonable to seek to apply resectability frameworks from another advanced gastrointestinal malignancy such as pancreatic cancer to RPS, particularly concerning major vascular involvement, this approach does not reflect the anatomical and biological breadth and complexity of RPS surgery. RPS are not just anatomically complex but biologically distinct. To take only the two most common RPS, the biological behaviour of well-differentiated liposarcoma (WDLPS) differs significantly from dedifferentiated liposarcoma (DDLPS) or leiomyosarcoma (LMS) 3, 5, 7. The biology of sarcoma is fundamentally different to that of pancreatic adenocarcinoma. In pancreatic adenocarcinoma, neoadjuvant therapy may downstage borderline tumours; STRASS I demonstrated no advantage to neoadjuvant radiotherapy 9, while STRASS II is currently recruiting in assessing the potential benefit of neoadjuvant chemotherapy 10. Pancreatic cancer, even when macroscopically resectable, is treated as a systemic disease 2, 11. RPS, by contrast, and particularposeLPS poses a challenge of local recurrence far more than distant recurrence with maximally invasive surgery performed with curative intent 4-7. The oncological perspective is therefore different, and this distinction shapes surgical decisions, including the radical nature of resection and reconstruction. A theoretical guideline that encompassed all anatomical considerations in RPS resection would not be generalisable as the variability lies not only in the tumour, but in institutional capability 6, 12. While standardised criteria may help in low-volume settings, the standard of care for RPS involves referral to and management at high-volume tertiary or quaternary centres 3-8, 13-15. In these settings, a guideline cannot account for the nuances of institutional experience with multivisceral resection, vascular reconstruction, disease factors and individual patient factors. Resection decisions are shaped by case-by-case subtleties that are difficult to capture by rigid criteria, where guidelines may stifle or cloud decision making 5, 7. For example, the oldest patient in the cohort was an 86 year old with DDLPS who was deemed unresectable at operation secondary to inferior vena cava involvement. Whilst not knowing the full details of the patient, were they to be 46 years old and in otherwise good health this particular tumour may well have been deemed resectable. Such an example highlights the problem with guidelines, where resectability is not solely defined by technical factors, with a complex balance of technical, biological and patient factors. For this reason, attempts to define unresectability through purely anatomical frameworks, as in pancreatic cancer, are fundamentally limited when applied to RPS. The heterogeneity in both tumour type and extent of resection in RPS is perhaps most analogous to exenterative pelvic surgery, where a variety of advanced pelvic tumour types (anorectal, genitourinary and other soft tissue tumours) require a tailored resection according to the extent of local tumour invasion. The historical list of anatomical contraindications to pelvic exenteration has been systematically challenged and revised in the last two decades and no longer applies in contemporary practice, where the decision to proceed with surgery is based on the reasonable chance of R0 resection with anticipated morbidity and functional consequences that are acceptable to an individual patient 16. This MDT framework incorporates a more nuanced approach to treatment decision-making, including incorporation of the demonstrated or anticipated biological behaviour of the individual patient's tumour (increasingly based on molecular biomarkers), as well as individual patient treatment priorities and preferences. Determining resectability in RPS lies unequivocally in the multidisciplinary team. In sarcoma centres, this decision does not rest with a single surgeon or a fixed list of contraindications. It emerges from robust pre-operative discussion that involves a review of the patient history, imaging and pathology by a group of expert clinicians who then arrive at a consensus decision for management 4-7, 12, 17. The strength of the team lies in its breadth, its collective experience in sarcoma management, and most importantly, in its interest in offering curative treatment when possible and avoiding futile surgery when not. In view of this, the MDT will often prompt a broader 'offline' discussion between surgical oncologists as to planning how surgery will be performed. These discussions must consider who needs to be involved, what the anatomical challenges are, and what infrastructure or expertise will be required to achieve a resection 4esection 4, 5, 17. Tai et al.'s study highlights that 6 of 13 patients deemed unresectable were found so intra-operatively. While this is sometimes unavoidable, a high rate of intra-operative unresectability may reflect missed opportunities for deeper interrogation of imaging and surgical planning. It reinforces the point that when MDTs are used to their full potential, with real-time imaging review and engaged multidisciplinary dialogue, many of these surprises can be anticipated and avoided 4, 12. Conversely, it should equally be recognised that where resectability is borderline, there is a place for trial dissection and operative recognition that the tumour is unresectable in the appropriately consented patient following MDT discussion. Tai and colleagues report an R2 resection rate of 8.8% in the resectable cohort. This rate is consistent with high-volume sarcoma centres, where R2 resection rates between 5% and 10% are reported 8, 14, 18. As RPS sarcoma surgery progresses in terms of technical resectability, high-volume centres need careful prospectively audit of outcomes including R2 rates. The threshold to proceed with RPS surgery should balance technical feasibility, and biological behaviour of the underlying tumour with oncologic benefit. In RPS, an R0 resection is not always technically achievable particularly in WDLPS, but true R2 resection where macroscopic disease is intentionally left behind, should be rare 3, 6, 7, 14, 17, 18. International collaboration, and the establishment of benchmarks for surgical outcomes in RPS, may further improve quality of care 14. The outcomes reported by Tai et al. help to define the anatomical boundaries of resectability in primary RPS. More importantly, their study invites reflection on how we make these complex, high-stakes decisions in contemporary RPS surgery. In our view, the key to avoiding both futile surgery and premature nihilism requires consideration of local anatomical relationships, but also an individual patient's priorities, the biology of their disease, and the technical factors that define resectability. Central to the determination of resectability is a well-resourced, skilled, and genuinely collaborative multidisciplinary team that manages a high volume of RPS. The limits of surgery in RPS should not be predefined based on anatomical criteria, but navigated in every case by an expert team invested in evidence-based best practice management. Emily Smartt: conceptualization, writing – original draft, writing – review and editing. Kilian G. M. Brown: conceptualization, writing – review and editing. David J. Coker: conceptualization, writing – original draft, writing – review and editing, supervision. Data sharing not applicable to this article as no datasets were generated or analysed during the current study.
Smartt et al. (Sun,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: