Estimating the Survival of Patients With Spinal Metastasis
Key findings
- In this study, 60 orthopedic surgeons and neurosurgeons were asked to estimate life expectancy for 12 patients with spinal metastasis whose cases were presented in writing alongside CT and MR images of their lesions
- The patients varied widely in age, type of cancer, American Spinal Injury Association classification, location of bone metastasis and number and location of other metastases
- There was poor agreement between surgeons in their estimates of survival (intraclass correlation, 0.22; 95% CI, 0.12–0.48); this was true regardless of surgeon specialty, years of practice or number of patients treated per year
- Survival estimates were typically too optimistic compared with the actual survival of the 12 patients, but life expectancy was underestimated for two patients whose actual survival was >20 months
- Patients with spinal metastases have different primary tumors, which leads to large variation in median survival, and the use of a validated prognostic algorithm appears to be warranted
Life expectancy is an important factor when discussing treatment options for patients with spinal metastases, particularly when surgery is being considered. For example, a patient should be considered for "en bloc" resection only if they are expected to outlive the extended rehabilitation period long enough to benefit from the surgery.
Subscribe to the latest updates from Orthopaedics Advances in Motion
Quirina C.B.S. Thio, former research fellow in Massachusetts General Hospital's Orthopaedic Oncology Service, Joseph H. Schwab, MD, MS, chief of the Orthopaedic Spine Center and director of spine oncology at Mass General, and co-director of the Stephan L. Harris Center for Chordoma Care at Mass General Cancer Center, and colleagues found poor interrater agreement and poor accuracy when surgeons estimated survival in a sample population of patients with spinal metastases. They present their data in the Journal of Orthopaedics.
Methods
Between May 2016 and August 2017, the researchers recruited 60 members of the Dutch Spine Society and the international Skeletal Oncology Research Group to participate in an online survey. 49 were orthopedic surgeons; the others were neurosurgeons.
The surgeons were presented with written descriptions of 12 patients who had spinal metastasis. As listed in the article, the patients varied widely in age, type of cancer, American Spinal Injury Association classification, location of bone metastasis and number and location of other metastases. CT and MR images of each lesion were provided. For each case, surgeons were asked to estimate the patient's life expectancy in months.
Variability in Estimated Survival
The intraclass correlation for estimating life expectancy was 0.22 (95% CI, 0.12–0.48), reflecting poor interobserver agreement. The results were similar regardless of surgeon specialty, years since finishing residency, numbers of patients treated annually and the continent of practice.
Accuracy of Survival Estimates
Survival estimations were overly optimistic in most cases:
- Case 6—actual survival, 1 month; median difference between actual and estimated survival, 5 months
- Case 1—actual, 1.2 months; median difference, 16 months
- Case 11—actual, 1.7 months; median difference, 4 months
- Case 2—actual, 3 months; median difference, 21 months
- Case 12—actual, 4.9 months; median difference, 7 months
- Case 5—actual, 5.5 months; median difference, 0 months
- Case 3—actual, 5.6 months; median difference, 18 months
- Case 7—actual, 6.2 months; median difference, 18 months
- Case 8—actual,12.5 months; median difference, 0 months
- Case 10—actual, 13 months, median difference, 11 months
- Case 4—actual, 20.4 months; median difference, −14 months (survival was underestimated)
- Case 9—actual, 72.8 months; median difference, −49 months (survival was underestimated)
Discrepancies remained substantial when the researchers stratified the cases by the length of survival:
- Actual survival ≤3 months (n=4)—median difference between actual and estimated survival, 9 months
- Actual survival 3–12 months (n=4)—median difference, 7 months
- Actual survival ≥12 months (n=4)—median estimated survival, −6 months
Conclusion
These results clearly show the shortcomings of surgeons' survival predictions, presumably because patients with spinal metastases have different primary tumors, which leads to large variation in survival. Multiple prognostic tools have been developed to estimate life expectancy in this group of patients, and the use of a validated algorithm may aid in making shared decisions about treatment.
view original journal article Subscription may be required
Learn more about Orthopaedic Oncology Service
Refer a patient to the Orthopaedic Oncology Service