Background: Prognosis is the challenging process of making predictions about future health outcomes, generally estimating survival time, in order to help defining the most appropriate therapeutic intervention for each patient. In patients diagnosed with solid tumors undergoing Palliative Care (PC), several prognostic tools have been validated with the aim of predicting prognosis, and are widely used by oncologists and palliativists in order to establish the more proportional therapeutic plan in each phase of the disease. However, in patients with hematological malignancies (HM) at the end of life, there is still no robust validation of these scores, and estimating the prognosis in this context is particularly challenging, resulting from the great heterogeneity of HM natural history, which range from very aggressive and rapidly evolving to pathologies with indolent behavior. Moreover, HM patients experience fluctuations in the trajectory of their disease, leading to delays in identifying and communicating the end-of-life phase, and therefore, in the transition from curative care to exclusive PC. It is known that patients with HM receive more aggressive care at the end of life and are less likely to be referred to PC services when compared to individuals with solid tumors, which may be mainly a consequence from the difficulty of estimating prognosis in HM patients.
Aims: To evaluate the capacity of prognostic, performance and symptom assessment scores commonly used for solid tumors patients under PC on predicting prognosis in patients with HM.
Methods: Retrospective data were collected from the electronic medical records of patients diagnosed with myeloma, acute leukemia and lymphoma followed at the PC outpatient clinics of the Hematology and Hemotherapy Center of the University of Campinas (Unicamp), Brazil, from 2018 to 2023. Data were evaluated at the date of the first PC consultation. CRP and albumin levels, and ECOG, PPS, CCI, Age-CCI, mGPS and ESAS scores were correlated with overall survival (OS). Statistical analysis was performed on the general population and on each subpopulation according to the underlying pathology. The significance level adopted for the study was 5%. This research was approved by the Institutional Review Board; written informed consent was obtained from all subjects.
Results: A total of 181 patients were included in the study (70 multiple myeloma, 49 acute leukemia and 62 lymphoma patients). PPS scores were positively correlated with OS (HR 0.975; 95% CI 0.965-0.986; p<0.0001). Individuals with PPS scores between 80-100% had a survival rate 3.5 times higher than those with PPS<50% (HR 3.501; 95% CI 1.977- 6.199; p=0.0001) and approximately 1.7 times higher than those with PPS between 50-70% (HR 1.695; 95% CI 1.977- 6.199; p=0.0363). Worse performance status assessed by ECOG levels were correlated with OS (HR 1.383; 95% CI 1.154-1.657; p=0.0005). Finally, mGPS scores (HR 2.446; 95% CI 1.103-5.422; p=0.0277) and albumin levels (p<0.0001; HR 0.391; 95% CI 0.289-0.528) also predicted survival in these patients. CCI, Age-CCI, number of comorbidities, CRP and ESAS scores did not predict survival in this cohort. When analyzing separately the subpopulations with myeloma, acute leukemias and lymphomas, PPS scores (HR 0.963 [0.942-0.985], HR 0.974 [0.954-0.995] and HR 0.979 [0.964-0.993], respectively, all p<0.05) and albumin levels (HR 0.272 [0.164-0.450], HR 0.497 [0.255-0.97] and HR 0.541 [0.339-0.864], respectively, all p<0.05) remained statistically related to OS. In patients with myeloma and lymphomas, ECOG was also related to OS (HR 1.816; 95% CI 1.211-2.724; p=0.0039, and HR 1.323; 95% CI 1.043-1.678; p=0.0213, respectively). CCI scores correlated with survival only in the myeloma subgroup (HR 1.306; 95% CI 1.064-1.602; p=0.0105).
Conclusions: PPS and albumin levels predicted survival in our cohort of HM patients undergoing PC. Also, ECOG and CCI predicted survival in individuals with multiple myeloma, and ECOG in lymphoma patients. These results indicate that these prognostic assessment tools could be helpful for OS prediction in patients with HM undergoing PC. Also, the routine application of prognostic tools in patients with HM could help hematology-oncologists to find the better time-points for referral to the PC team. Given the limitations of the study design, these results need to be validated in further studies.
No relevant conflicts of interest to declare.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal