Spinal cord Compression Outcomes Of Treatment (SCOOT) Delphi study
Highlight box
Key findings
• Spinal instability and a visual analogue scale of 6/10 were considered indications for surgery. Prognosis and pain levels were major factors influencing decision-making—careful consideration in lung, myeloma and unknown primaries.
What was recommended and what is new?
• Managing patients with metastatic spinal cord compression (MSCC) is challenging; National Institute for Health and Care Excellence guidelines have been updated 2023; no guidance or agreement on tumour-specific management.
• This is the first Delphi study producing management algorithms that are tumour-specific. The algorithms are derived from the experience of national and international experts. The algorithms guide and aid in selecting appropriate management plans while allowing clinicians to make a judgment based on the individual patient.
What is the implication, and what should change now?
• Standardise care given to MSCC patients while ensuring the patient remains at the centre of the decision-making.
• Aid surgeons’ decision-making based on the experience of experts to deliver appropriate care to patients with MSCC.
• Surgeons can share the experience of experts with MSCC patients and their relatives to reach an informed management decision.
Introduction
Metastatic spinal cord compression (MSCC) is a broad term for malignant spine disease covering metastasis to the spine, myeloma and lymphoma. The spine is the commonest site for metastasis after lung and liver and is involved in up to 40% (1), while spinal cord compression may develop in 5–10% of cancer patients. Importantly, MSCC may also be the first clinical manifestation of an underlying malignancy, with around one-fifth of patients having had no cancer diagnosis in the preceding year (2).
The exact incidence of MSCC is unknown as cases are not systematically recorded in the UK. The estimated annual incidence is up to 80 cases per million, from a prospective audit in Scotland (3) and published data from Canada (4). In the United States, it is predicted that more than 20,000 cases will be diagnosed annually (5,6).
Paralysis from spinal cord compression and pain from destructive vertebral lesions are the twin threats to MSCC patients. MSCC constitutes an oncological emergency requiring urgent investigation and treatment. Whole-spine magnetic resonance imaging (MRI) is the imaging modality of choice and should be performed within 24 hours of clinical suspicion. Following radiological confirmation, corticosteroids are typically administered promptly, and definitive treatment should be initiated without delay (7). The seriousness of the risks merits invasive surgery and/or radio/chemotherapy; although often palliative in intent, both modalities have an established role in MSCC. Management, however, should be multidisciplinary and include early neuro-rehabilitation following principles similar to those applied in traumatic spinal cord injury. Rehabilitation strategies aim to optimise pain control, prevent secondary complications, support bladder and bowel function, enhance mobility and functional independence and improve overall quality of life (8).
National Institute for Health and Care Excellence (NICE) guidance National Institute for Health and Care Excellence Guideline (NG234) recommends surgery followed by radiotherapy for suitable patients offer the best outcome. The demanding and complex nature of the treatments and associated risks requires a detailed analysis of patient factors, tumour biology and prognosis for survival to formulate a patient-specific treatment.
We conducted a literature search on the thresholds for operative intervention and the outcome of surgery in patients with MSCC. While many studies focused on mobility and survival, there is a scarcity of studies on the operative threshold for pain. The complexity of the condition makes evidence generation on the treatment threshold difficult, leading us to design a high-quality Delphi study to determine current practice and the range of opinions underpinning that.
Research question
The study had three aims:
- Identify thresholds for surgical intervention in patients with MSCC;
- Identify factors that influence those thresholds;
- Identify variations in thresholds between surgeons and centres. Accomplishing these aims will help develop treatment algorithms to guide the management of these complex patients and reduce variations.
Methods
The Delphi technique is an effective way to gain and measure group consensus in health research to define diagnostic criteria, guideline development, and health outcomes. Experts provide answers to open-ended unambiguous questions in a series of rounds. Each subsequent round provides aggregated feedback on statements built on the responses. Consensus is sought through this feedback of information and iteration. The Delphi approach draws together knowledge and experience from several experts while preventing a single voice from controlling the agenda.
We used a modified Delphi technique and sought consensus on the research questions according to pre-defined criteria. Ethical approval was deemed not required by the head of research operations (Carol Maloney) following the study protocol review.
Assembling expert panel
Experts were defined as those active in MSCC research or those regularly managing patients with MSCC for more than 5 years. We identified 63 national and 27 international spinal surgeons and obtained e-mail addresses from the public domain.
Consent
Invitation letters and informed consent forms were sent electronically. All non-responders were removed from the list after 2 weeks (Appendix 1). Anonymity was maintained throughout (Appendix 2).
Survey and clinical vignettes
The authors designed six clinical vignettes following a pilot study. These were based on the common primary tumours, prostate, breast, lung, multiple myeloma (MM), renal cell carcinoma (RCC) and cancer with unknown primary (CUP). Questions included ordinal ranking, multiple-choice, checkboxes and comments boxes. The clinical vignettes and questions were sent using Google forms’ survey platform after a trial with non-experts. We decided to exclude imaging to allow a degree of generalisability and not limit the participants’ decisions to a single case.
Rounds
The rounds were planned to be 4 weeks apart. Each round remained open for 4 weeks, with reminders sent at 1 and 2 weeks. We aimed to reach a consensus, accepting that this may only be achievable on some questions and scenarios. We planned three rounds to achieve consensus in as many areas as possible and allow gradual convergence of opinion where consensus could not be reached.
Each clinical vignette had 17 questions. The experts were asked to state their management options and rationale for each clinical vignette. At the end of Round 1, areas where a threshold was reached, were reported to the experts and set aside (available at https://cdn.amegroups.cn/static/public/10.21037jss-2026-1-0027-1.pdf).
In Round 2, in areas where an agreement was not reached, Venn diagrams of the management options were presented to the expert panel. Each expert was sent an individualised list of their opinions compared to the panel’s, with overall median scores and ranges (available at https://cdn.amegroups.cn/static/public/10.21037jss-2026-1-0027-2.pdf).
In Round 3, a summary of the results and treatment algorithms, based on the responses, was sent to the experts. The experts were asked to comment on the treatment algorithms. They were also given the comments made by the rest of the panel (available at https://cdn.amegroups.cn/static/public/10.21037jss-2026-1-0027-3.pdf).
Feedback and consensus
We used RAND criteria for consensus and percentage agreement for the conclusion. Consensus is defined as agreement. Aggregated opinions offered consensus in some scenarios. We accepted a level of central tendency on an ordinal scale of 0–10, as described below.
Experts were asked to score each parameter on a scale of 0 to 10, where 0 is the least, and 10 is the most important. The parameters were then ranked based on median scores and presented to the experts in Round 2. Experts were allowed to change their answers where an 80% agreement had not been reached or comment if they disagreed with the median score. A median score of 8 to 10 was deemed very important, 5 to 7 as important and a score less than five as not important (Table 1). After the conclusion of all three rounds, the treatment algorithms developed on consensus, responses and areas not reaching an agreement were explored in a virtual meeting, including the research team.
Table 1
| Primary | Age | Duration of symptoms | Functional status | Mets at time of diagnosis | Neurological function at presentation | Number of spinal lesions | Pain | Oncologists’ predicted life expectancy | Primary tumour | Site of spinal lesions | Type of spinal lesion | Visceral mets |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Prostate | 7 [75]¶ | 8 [88]† | 9 [96]† | 7 [79]¶ | 9 [96]† | 8 [88]† | 8 [92]† | 8 [96]† | 8 [96]† | 7 [83]§ | 8 [88]† | 7 [83]§ |
| Breast | 6 [71]¶ | 8 [96]† | 9 [92]† | 7 [75]¶ | 9 [88]† | 7 [83]§ | 8 [92]† | 8 [92]† | 9 [88]† | 8 [83]† | 8 [92]† | 7 [83]§ |
| Lung | 5 [79]¶ | 7 [75]¶ | 8 [92]† | 8 [83] | 9 [88]† | 8 [88]† | 7 [88]§ | 8 [88]† | 9 [92]† | 8 [88]† | 8 [88]† | 7 [83]§ |
| RCC | 8 [79]‡ | 7 [83]§ | 9 [96]† | 8 [83] | 9 [88]† | 8 [92]† | 8 [92]† | 8 [92]† | 9 [88]† | 9 [92]† | 8 [88]† | 8 [75]‡ |
| MM | 5 [75]¶ | 7 [71]¶ | 8 [96]† | 7 [67] | 9 [96]† | 8 [79]‡ | 9 [92]† | 8 [88]† | 10 [96]† | 8 [83]† | 8 [83]† | 6 [58] |
| CUP | 6 [63]¶ | 7 [83]† | 9 [92]† | 7 [83]§ | 9 [92]† | 8 [75]‡ | 7 [83]† | 6 [75]¶ | 8 [79] | 7 [83]§ | 8 [92] | 7 [88]§ |
Data are presented as n [%]. †, agreed upon and considered very important; ‡, not agreed upon but considered very important; §, not agreed but considered it very important; ¶, agreed and considered important. CUP, cancer of unknown primary; Mets, metastases; MM, multiple myeloma; RCC, renal cell carcinoma.
Thirty-five experts initially agreed to participate; 24 completed Round 1, and 22 completed the remaining rounds. Ten were international experts, and 14 were national. Experts had a combined total of 312 years of experience in managing MSCC patients (Figure 1). Spinal Instability Neoplastic Score (SINS) was the most used scoring system (Figure 2).
The most important factors in decision-making were neurological function at presentation, functional status, and type of spinal lesion (Table 1). All three domains reached the agreement threshold of 80% and were considered important in all the clinical vignettes.
The overall rate of surgical intervention was 60%, with national experts at 57% vs. 63% for international experts. Overall, 74% of experts would not operate on patients paralysed at presentation; 33% would operate within 24 hours from the onset of symptoms (Figure 3).
A pain level of 6/10 was the threshold for surgical intervention, with 82% expecting to achieve pain relief. The management objectives varied between the clinical vignettes. Most experts expected to achieve pain relief and spinal stability more than neurological recovery (Figure 4).
Management algorithms were extrapolated from the experts’ responses for each clinical vignette and summarised into one management algorithm (Figure 5). Treatment algorithms corresponding to the six clinical vignettes (breast cancer of unknown primary, lung, multiple myeloma, prostate, and renal cell carcinoma) are presented in Figures 6-11. The experts agreed with the NICE guidelines but highlighted specific considerations in different primary tumours. In patients with lung or MM primaries, careful consideration of surgery should be made because the risks of surgery often outweigh the benefits. A diagnosis should be sought before surgical intervention or treatment in patients with CUP. All patients with RCC should undergo pre-operative embolization.
In the final online meeting, the experts concluded the following:
- The surgeon is part of the “orchestra” managing the disease to improve the quality of life and not control the disease. Good interactions with oncologists are vital to achieving this.
- The time from onset of symptoms to presentation, response to previous treatment, time from primary diagnosis to presentation and potential for future treatments or disease control dictate when and if to intervene surgically.
- Different surrogates decide when to operate, all of which address the main objective of restoring function.
- Eyeball or end-of-bed test;
- Karnofsky scale;
- The impact of visceral metastases in decision-making depends on the patient’s performance status, frailty, and prognosis.
- The aim of surgical intervention is:
- Improve spinal function by improving pain management and preventing neurological deterioration;
- Preservation of sensation is essential as a protective mechanism;
- Managing mechanical and neuropathic pain is humane, especially near the end of life.
- There is a big difference between the power grade of “0/5” and “1/5” Medical Council Research (MRC), where complete paralysis is unlikely to recover, if at all. Risks versus benefits require careful consideration.
- Significant pain and a SINS, 13 or more, warrant consideration of surgical intervention.
- There are variations in techniques; North America has a proactive approach to minimally invasive procedures; the UK has the most significant variation in management, but they have the NICE guidelines; Canada has spinal-oncology surgery in university centres.
- Avoid potentially creating more harm to the patient through surgical intervention, e.g., type of stabilisation or minimally invasive techniques.
- Follow-up:
- This is to ensure wound healing;
- Forms part of surgical checks ranging from 6 weeks to 24 months, depending on the overall status and health of the patient;
- Should avoid potentially creating more harm to the patient.
- Research:
- Partial data collection on patients is better than no data. Aim to calculate the impact on quality-adjusted life years and health economics impact.
- There are limitations to registries collecting data currently. A consensus is needed on which data is required.
- The importance of appointing well-trained research nurses.
Discussion
Following the Delphi study rounds, the experts identified age as a less important factor in all the clinical vignettes except renal cell cancer. Experts agreed that functional status, neurological function at presentation and type of spinal lesion are very important in all the clinical vignettes. Although the primary tumour was considered very important in all clinical vignettes, the experts did not reach an agreement threshold in CUP. The experts identified Visceral metastases as less important where an agreement was reached in 4 out of 6 clinical vignettes; this is contrary to the Arbeitsgemeinschaft für Osteosynthesefragen (AO) spine members’ survey, where visceral metastases were identified as an important predictor for survival by 76.48% of the 438 members surveyed (9).
The qualitative analysis revealed that the experts highly regarded previous functional status and determined how aggressive the surgical intervention would be. Experts were reluctant to operate on patients where a diagnosis was not confirmed. Although not given in the clinical vignettes, imaging carries much weight in the decision-making process, as expressed by the experts in their responses. Experts’ opinions were in keeping with NICE guidelines for surgical intervention for the unstable spine and pain; however, most experts would not operate on paralysed patients (10).
Surgery effectively decomposes neural structures and provides mechanical stability to address pain, with higher risks of morbidity from complications and depletion of the physiologic reserve. While radiotherapy and/or chemotherapy are effective in ablating tumour tissue compressing the spinal cord with lesser morbidity and complications but cannot address mechanical instability.
Experts described an acute neurological deficit as less amenable to surgery than gradually developing neurology in prostate primaries. In breast primaries, surgical intervention depends on good prognosis, responsive tumours [e.g., estrogen receptor (ER) and progesterone receptor (PR) sensitive], and performance status. The onset of symptoms and lung tumour subtype were important. Small cell lung carcinoma was the worst, and adenocarcinoma had a chemo-sensitive subtype. In non-small cell lung cancer, molecular profile and clinical status appear prognostically relevant. Patients with EGFR mutations, preserved ambulatory function, and Eastern Cooperative Oncology Group (ECOG) performance status 1–2 have been shown to exhibit improved survival, whereas rapid development of motor deficit (≤10 days) is associated with worse outcomes (11).
In RCC, anterior column reconstruction can be offered to fit patients with a good prognosis; vertebral body augmentation may be suitable for frail patients with severe pain. RCC is a very vascular tumour, and all cases undergoing surgical intervention require pre-operative embolization. In MM, the experts stated the importance of a haematologist as part of a multiple disciplinary team (MDT). Patients may require surgery if the lesions are in the mobile spine, a pathological fracture, or cord compression causing a neurological deficit. In CUP, most experts agreed that a biopsy [e.g., percutaneous biopsy, computed tomography (CT)-guided biopsy] would be required to reach a diagnosis before commencing management.
Conclusions
This Delphi study has applied the experience and knowledge acquired over 312 years (median 14 years) as national and international field experts. SINS was the most commonly used scoring system, followed by Tokuhashi.
The most important parameters influencing decision-making were the primary tumour, neurological function at presentation, functional status and type of spinal lesion. The threshold for surgery was an unstable spine and a pain level of 6 out of ten, and the experts agreed on all clinical vignettes. There was no agreement on the importance of age in decision-making. Most experts would not operate on paralysed patients.
The objectives of any intervention were mainly to control pain, prevent deterioration in neurology, and maintain function and quality of life. Careful consideration should be made in the lung, multiple myeloma and cancer of unknown origin primaries.
There remains a variation in the management of patients with MSCC; due to the complexity of these patients. Although not all regions or centres had more than one expert participating, there was more significant variation among the national experts.
We have provided up-to-date recommendations for treatment based on the most commonly occurring primary tumours. The findings from this Delphi study provide clinicians managing MSCC patients with vital information to guide management decisions and ensure patients receive the best quality care.
Acknowledgments
We would like to acknowledge all patients what we based our data on.
Footnote
Peer Review File: Available at https://jss.amegroups.com/article/view/10.21037/jss-2026-1-0027/prf
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jss.amegroups.com/article/view/10.21037/jss-2026-1-0027/coif). The authors have no conflicts of interest to declare.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Informed consent was obtained electronically from all participating experts prior to inclusion in the study.
Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.
References
- Schmidt MH, Klimo P Jr, Vrionis FD. Metastatic spinal cord compression. J Natl Compr Canc Netw 2005;3:711-9. [Crossref] [PubMed]
- Shah S, Kutka M, Lees K, et al. Management of Metastatic Spinal Cord Compression in Secondary Care: A Practice Reflection from Medway Maritime Hospital, Kent, UK. J Pers Med 2021;11:110. [Crossref] [PubMed]
- Metastatic spinal cord compression: Diagnosis and management of patients at risk of or with metastatic spinal cord compression. Cardiff (UK): National Collaborating Centre for Cancer (UK); 2008.
- Loblaw DA, Laperriere NJ, Mackillop WJ. A population-based study of malignant spinal cord compression in Ontario. Clin Oncol (R Coll Radiol) 2003;15:211-7. [Crossref] [PubMed]
- Bartels RH, van der Linden YM, van der Graaf WT. Spinal extradural metastasis: review of current treatment options. CA Cancer J Clin 2008;58:245-59. [Crossref] [PubMed]
- Claus EB. Neurosurgical management of metastases in the central nervous system. Nat Rev Clin Oncol 2011;9:79-86. [Crossref] [PubMed]
- Bachour ME, Rabbani RD, Hasan MR, et al. Spinal MRI in Patients with Suspected Metastatic Spinal Cord Compression: A Quality Improvement Audit in a District General Hospital in Kent, UK. Int J Environ Res Public Health 2025;22:401. [Crossref] [PubMed]
- Boussios S, Cooke D, Hayward C, et al. Metastatic Spinal Cord Compression: Unraveling the Diagnostic and Therapeutic Challenges. Anticancer Res 2018;38:4987-97. [Crossref] [PubMed]
- Fehlings MG, Nater A, Tetreault L. Key preoperative predictors of surgical outcomes in patients with metastatic epidural spinal cord compression: results from a survey of 438 AOSpine International members. Spine J 2016;16:S291-2.
- National Institute for Health and Care Excellence (NICE). Spinal metastases and metastatic spinal cord compression. NICE guideline NG234. Published 6 September 2023. Available online: https://www.nice.org.uk/guidance/ng234
- Vassiliou A, Bachour ME, Shah S, Boussios S. Prognostic factors in patients with metastatic spinal cord compression secondary to lung cancer: a retrospective UK single-centre study. Cancers (Basel) 2023;15:4432. [Crossref] [PubMed]
