The need for a system of prognostication in skeletal metastasis to decide best end-of-life care - a call to arms

Abstract

Introduction: The near terminal patient with skeletal metastasis may suitably be palliated with an intramedullary nail whereas another patient with good survival potential may benefit from a more extensive resection and reconstructive procedure. In a previous study by the senior author (Nathan et al, 2005), life expectancy in patients operated on for bone metastases correlated with clinical and haematological parameters in a normogram. We performed a cross-cultural comparison to validate this normogram.

Materials and methods: We randomly selected 73 patients who had undergone surgery for metastatic bone disease between 28 December 2000 and 11 March 2009. The time to deaths was recorded from hospital records and telephone interviews. Multiple parameters including clinical, radiological and haematological were evaluated for significant prognostic value using Kaplan-Meier survivorship analysis. Statistically significant parameters were entered into a Cox regression model for statistically independent significance. A multi-tier prediction of survival was performed by workers from various levels of seniority.

Results: At the time of analysis, there were 40 deaths (55%). Median survival was 15.8 (95% CI, 7.9 to 23.7) months. Kaplan-Meier analysis showed that low haemoglobin (P = 0.0000005), presence of lymph node metastases (P = 0.00008), multiple bone metastases (P = 0.003), presence of visceral metastases, (P = 0.007), low lymphocyte count (P = 0.02) and low serum albumin (P = 0.02) were significantly associated with poor survival. By Cox regression analysis, presence of visceral metastases (P = 0.002), presence of lymph node metastases (P = 0.0002) and low haemoglobin (P = 0.01) were shown to be independent factors in the prediction of survival. Survivorship readings were superimposed onto the previous normogram and found to be similar. Independent blinded use of the normogram allowed good prediction of survival. There was a tendency to underestimate survival when patients survived beyond 1 year of skeletal metastasis.

Conclusion: Our findings are similar to that of the previous study in showing a relationship between the above-mentioned parameters and survival. This is despite differences in patient demographic characteristics and management protocols. Use of the tools may allow better siting of most appropriate surgery in metastatic bony disease.