| Geissbűhler, P et al. 2000, Switzerland [18] |
Prospective cohort study (investigative) Assess the relationship between an elevated serum B12 and an earlier death. To assess if an elevated serum B12 predicted death independently of the PINI calculation and its derivatives: CRP, AGP, albumin, and prealbumin. |
|
|
|
Relied on blood samples taken on admission (missing data: n = 4) Does not clearly distinguish between palliative and curative patients Requires invasive procedure |
Level 2++ Hawker et al: 33/36 Maltoni et al: 6/7 |
-
•
The length of survival decreased with the increase in serum B12 levels (p = 0.0015, Cox model).
-
•
In univariate analysis, PINI and its 4 components: CRP, AGP, albumin, prealbumin were predictors of length of survival.
-
•
In multivariate analyses, CRP was the most important prognostic factor, and B12 was independent of CRP in predicting survival.
-
•
There was a strong correlation between the presence of metastasis, hepatic dysfunction and an elevated serum B12 level (p<0.001).
-
•
The BCI, calculated by B12 x CRP, was a simplified and significant marker of survival (trend test, p<0.001):
-
○
BCI <10000: 50% 3 month mortality
-
○
BCI >40000: 90% 3 month mortality
|
| Pasanisi, F et al. 2001, Italy [19] |
Prospective cohort study (investigative) Evaluation of clinical, anthropometric, hematologic, and biochemical variables immediately before starting nutritional treatment and relationship to survival in terminal cancer patients with irreversible bowel obstruction receiving home parenteral nutrition. |
|
|
|
Ratio between the number of events (death) and the number of potential predictors <10 Method for identifying date of death not described Requires invasive procedure |
Level 2+ Hawker et al: 29/36 Maltoni et al: 4/7 |
|
| Glare, P et al. 2001, Australia [20] |
|
|
|
|
|
Level 2++ Hawker et al: 33/36 Maltoni et al: 6/7 |
-
•
Pirovano et al. published details of the PaP Score, which classifies patients with very advanced cancer into homogeneous risk groups for survival based on various clinical and laboratory parameters (anorexia; dyspnoea; performance status; CPS; WBC; and lymphocyte count) [21]
-
•
The PaP Score subdivides the population into 3 groups with a different probability of survival at 30 days:
-
○
Group A: >70%
-
○
Group B: 30–70%
-
○
Group C: <30% [21]
-
•
This study externally validates the PaP Score in patients referred to the hospital palliative care team. Median survivals for the three groups were:
-
○
Group A: 60 days (95% CI 41–89)
-
○
Group B: 34 days (95% CI 25–40)
-
○
Group C: 8 days (95%CI 2–11)
-
•
The percentage survival at 30 days for the three groups were:
-
○
Group A: 66%
-
○
Group B: 54%
-
○
Group C: 5%
|
| McMillan, DC et al. 2001, UK [22] |
|
|
|
-
•
Median survival:
-
○
Bronchogenic: 60 days
|
No protocol for blood sampling Does not distinguish between palliative and curative patients Clinicopathological stage at diagnosis and subsequent treatment were not included in analysis Requires invasive procedure High risk of bias |
Level 2+ Hawker et al: 26/36 Maltoni et al: 3/7 |
On univariate analysis, log10 CRP and albumin concentrations were significant predictors of survival. On multivariate analysis, log10 CRP remained a significant independent predictor of survival. CRP concentrations were associated with a higher risk of non-cancer deaths. |
| Faris M. 2003, Oman [23] |
|
|
|
|
|
Level 2+ Hawker et al: 30/36 Maltoni et al: 4/7 |
-
•
Univariate analysis demonstrated that a number of factors had a significant effect on survival including:
-
○
Peripheral oedema
-
○
Low lymphocyte count
-
○
Low serum albumin level
-
•
Multiple regression analysis demonstrated that the following as independent predictors of survival:
-
○
Peripheral oedema (p = 0.007)
|
| Ho SY et al. 2003, Taiwan [24] |
|
|
|
|
Limitations of study not discussed Serum albumin and lymphocyte count, previously identified prognostic markers, were not significant |
Level 2++ Hawker et al: 32/36 Maltoni et al: 6/7 |
-
•
Univariate analysis demonstrated that a number of factors had a significant effect on survival including:
-
○
Prealbumin ≤100 mg/L (p = 0.03)
-
○
AST > 45 U/L (p = 0.03)
-
○
ALP > 120 U/L (p = 0.02)
-
○
Serum creatinine >1.4mg/dL (p = 0.01)
-
○
BUN >20mg/dL (p<0.01)
-
•
Multivariate analysis demonstrated that a number of factors were independent predictors of poor survival:
-
○
Prealbumin ≤100 mg/L (p<0.01)
-
○
AST > 45 U/L (p = 0.01)
-
○
BUN > 20 mg/dL (p<0.01
-
•
Statistically non significant:
-
○
Serum albumin
-
○
Transferrin
-
○
Haemoglobin
-
○
Total lymphocyte count
-
○
Serum ALT
|
| Iwase S, et al. 2004, Japan [25] |
Prospective cohort study (investigative) Quantitative serial analysis of six cytokines in the plasma: TNFα, IL-1β, IL-6, IFNγ, PTHrP and LIF and basic laboratory tests in hospice inpatients with late stages of cachexia. |
|
|
1–92 days Survival >90 days: n = 1 |
Small sample size Heterogeneity of sample No protocol for collection of blood samples Primary outcomes not specified Main findings were found incidentally Includes only patients with late stages of cachexia Ratio between the number of events (death) and the number of potential predictors <10 Requires invasive procedure Prognostic significance not assessed |
Level 2+ Hawker et al: 31/36 Maltoni et al: 4/7 |
Only IL-6 (and TNFα, n = 1) was detected in all patients. Maximum plasma IL-6 4.6–870 pg/mL Concentration of IL-6 rose gradually during the early stages of cachexia. Sharp rise in the week prior to death (n = 10). Plasma levels of IL-6 ≥100 pg/mL detected in the week prior to death (n = 6). |
| Maltoni M et al. 2005, Italy [8] |
Systematic review of non-randomised controlled trials Working group review of 38 prospective studies evaluating prognostic factors in advanced cancer. To offer evidence-based clinical recommendations concerning prognosis in advanced cancer patients. |
|
|
|
Heterogeneity of studies Meta analysis not possible Poor quality of existing published studies Some subjectivity in quality criteria |
Level 2++ Hawker et al: N/A Maltoni et al: N/A |
-
•
Level B evidence based recommendations for the prognostic correlation of a number of factors including:
-
○
Leukocytosis
-
○
Lymphopaenia
-
○
Elevated serum CRP level
-
○
CACS
-
•
Factors that proved significant in at least one multivariate analysis included:
-
○
Low pseudocholinesterase
-
○
High serum B12
-
○
High serum bilirubin
-
•
Factors for which a correlation has been indicated but not confirmed or for which a statistical significance has been identified in patient populations with less advanced disease or for which contradictory data have emerged:
-
○
Anaemia
-
○
Hypoalbuminaemia
-
○
Low prealbumin
-
○
Proteinuria
-
○
Serum calcium level
-
○
Serum sodium level
-
○
Serum LDH
|
| Shin, HS et al. 2006, Korea [26] |
|
|
|
|
|
Level 2++ Hawker et al: 35/36 Maltoni et al: 5/7 |
-
•
Significantly shorter survival times were observed in the following groups in univariate analysis:
-
○
Elevated serum urate ≥7.2mg/dL (p <0.001)
-
○
Increasing or decreasing leukocytes (p <0.001)
-
○
Hypercreatininaemia >1.5mg/dL (p = 0.006)
-
○
Low serum albumin <3.0g/dL (p = 0.026)
-
○
Hyperbilirubinaemia >1.0mg/dL (p = 0.001)
-
○
INR >1.12 (p < 0.001)
-
○
Hypocholesterolaemia <130 mg/dL (p = 0.016)
-
○
High serum LDH >378IU/L (p = 0.001)
-
•
By multivariate analysis, the following variables significantly shortened survival:
-
○
High serum uric acid concentration (HR 2.637, p = 0.001)
-
○
Prolonged INR (HR 1.686, p = 0.039)
-
○
Hypocholesterolaemia (HR 2.030, p = 0.011)
-
○
High serum LDH (HR 2.136, p = 0.001)
-
•
Patients were divided into 4 groups based on serum urate levels. Median survival for each group was:
-
○
Group 1, serum urate <3.2mg/dL: 18 days (955 CI 15–21)
-
○
Group 2, serum urate ≥3.2 - <4.9mg/dL: 23 days (95% CI 9–37)
-
○
Group 3, serum urate ≥4.9 - <7.2mg/dL: 12 days (95% CI 7–17)
-
○
Group 4, serum urate ≥7.2mg/dL: 4 days (95% CI 5–11)
-
•
In univariate analysis, survival time of the fourth highest group (serum urate ≥7.2mg/dL) was significantly shorter than that of other groups (HR 2.784, p <0.001). There was no significant difference amongst the first 3 groups.
-
•
The serum urate concentration was inversely associated with survival time (HR 2.778, p = 0.003).
-
•
Serum urate concentration was measured consecutively for 3 weeks before death (n = 39) and was significantly increased between the first and the second week before death (p = 0.006).
|
| Lam PT et al. 2007, China [27] |
Prospective cohort study (explorative) To identify potential prognostic factors affecting the survival in patients with advanced cancer who were treated in a palliative care unit at a regional hospital in Hong Kong. |
|
-
•
Advanced cancer in the last 6 months of life:
-
○
Locally advanced (n = 28)
-
○
Metastatic (n = 142)
-
•
Various malignancies
-
•
Mean age: 69 years (SD 12)
|
|
|
Level 2++ Hawker et al: 30/36 Maltoni et al: 6/7 |
-
•
By univariate analysis (Cox), a number of factors affected survival including: serum albumin concentration and WBC.
-
•
Multivariate analysis (Cox) identified a number of factors were independent predictors of survival including:
-
○
Serum albumin concentration (HR 0.95, 95% CI 0.92–0.98)
-
•
Statistically non significant factors by univariate analysis:
-
○
Serum calcium
-
○
Serum sodium
-
○
Haemoglobin
|
| Kelly, L et al. 2007, UK [28] |
|
|
|
|
Relevant blood tests not taken and excluded (n = 50) Patients censored due to missing data (n = 9) Non-consecutive series of patients No clinician estimate of survival Requires invasive procedure |
Level 2++ Hawker et al: 35/36 Maltoni et al: 6/7 |
-
•
The BCI, calculated by B12 x CRP is a simplified and significant marker of survival:
-
○
BCI < 10000: 50% 3 month mortality
-
○
BCI > 40000: 90% 3 month mortality [18]
-
•
In this study, the BCI was externally validated in hospital and hospice patients receiving palliative care. Patients were divided into 3 groups according to BCI score. The median survival for each group was:
-
○
Group 1, BCI ≤10000: 71 days (95% CI 45.9–96.2)
-
○
Group 2, BCI 10001–40000: 43 days (95% CI 28.8–57.2)
-
○
Group 3, BCI >400000: 29 days (95% CI 22.0–36.1)
-
•
Patients in Group 3 had a significantly worse survival than patients in Group 2 and patients in Group 1 (p < 0.01).
-
•
However, patients in Group 1 did not have a significantly better prognosis than those in Group 2 (p = 0.091). Thus, while a high BCI score had poor prognostic significance, an intermediate or low score was more difficult to interpret.
-
•
90-day mortality estimates for each of the three risk groups were different from the figures previously reported:
-
○
Group 1, BCI ≤10000: 58.9% vs. 47.2%
-
○
Group 2, BCI 10001–40000: 64.0% vs. 72.5%
-
○
Group 3, BCI >400000: 78.9% vs. 90.6%
|
| Suh, SY et al. 2007, South Korea [29] |
|
|
|
|
Serum LDH concentrations may fluctuate with the passage of time or may reflect complications of cancer The relationship between serum LDH and survival time is not a simple negative correlation, and requires further research Preliminary study only due to sample size Single centre Requires invasive procedures |
Level 2+ Hawker et al: 34/36 Maltoni et al: 4/7 |
-
•
Significantly shorter survival times were observed in the following groups in univariate analysis (Cox):
-
○
High serum LDH ≥313IU/L (p < 0.001)
-
○
Increased leukocytes >11000 mm3 (p = 0.020)
-
○
Increased neutrophil fraction >75% (p = 0.030)
-
○
Decreasing thrombocytes (p = 0.030)
-
○
Elevated serum CRP ≥9.5 mg/dL (p = 0.015)
-
○
Elevated serum urate ≥7.2 mg/dL (p <0.001)
-
○
Low serum albumin <3.0g/dL (p = 0.019)
-
○
Hyperbilirubinaemia >1.0mg/dL (p = 0.003)
-
○
INR >1.12 (p = 0.001)
-
○
Hypocholesterolaemia <130 mg/dL (P <0.001)
-
•
The median survival time of 27 days (95% CI 19–35) was significantly longer than that of 14 days (95% CI 10–18) in the elevated serum LDH (≥313IU/L) group (HR 2.235, p<0.001).
-
•
By multivariate analysis (Cox), the following variables were independent and significant prognostic factors of poor survival time:
-
○
ECOG PS
-
○
Medium–high level of pain
-
○
Fatigue
-
○
Hypotension
-
○
Elevated serum LDH (HR 2.087, p = 0.002)
-
○
Elevated serum CRP (HR 1.984, p = 0.002)
-
○
Elevated serum urate (HR 2.853, p <0.001)
-
•
Statistically non significant:
-
○
Haemoglobin
-
○
Pleural effusion
-
○
Pneumonia
-
•
A new scoring system was developed to estimate the survival time of terminally ill cancer patients using the 7 variables identified by multivariate analysis. The accuracy of prediction for study subjects (n = 93) is below:
-
○
Prediction <3 weeks: sensitivity 76%, specificity 67%, PPV 74%, NPV 70%
-
○
Prediction <4 weeks: sensitivity 71%, specificity 73%, PPV 85%, NPV 55%
-
•
Average serum LDH concentration was measured consecutively at 2 weeks and 1 week before death (n = 25) and was 5.04.04 +/- 347.11IU/L and 630.40 +/- 417.32IU/L respectively (p = 0.008).
|
| Alsirafy SA et al. 2009, Saudi Arabia [30] |
Retrospective cohort study To assess the predictive significance of abnormalities in serum sodium, potassium, calcium, magnesium and phosphate on the admission outcome and survival of patients with cancer referred to palliative care service. |
|
|
|
|
Level 2++ Hawker et al: 31/36 Maltoni et al: 6/7 |
-
•
Hyponatremia was the most common electrolyte abnormality (64%).
-
•
By univariate analysis, the following variables were significant prognostic factors of poor survival time:
-
○
Serum sodium (p = 0.0008)
-
○
Serum calcium (p = 0.0008)
-
○
Serum magnesium (p = 0.0001)
-
•
Abnormalities in serum potassium and phosphate had no significant impact on admission outcome or overall survival.
-
•
The 3 electrolyte abnormalities associated with the highest inpatient death rate and shortest median survival were:
-
○
Hypercalcaemia (inpatient death rate: 69%, median survival: 12 days)
-
○
Hypernatremia (68%, 8 days)
-
○
Hypermagnesemia (62%, 12 days)
-
•
Combining these abnormalities in a simple scoring system divided the patients into 3 groups with significant difference in overall survival:
-
○
Two abnormalities: 8 days (95% CI 7–9)
-
○
One abnormality: 24 days (95% CI 10–38)
-
○
No abnormalities: 60 days (95% CI 26–94)
|
| Hyodo, I et al. 2010, Japan [31] |
Prospective cohort study (explorative) To develop a new prediction tool, JPOS-PI for terminally ill cancer patients with advanced solid tumors and to compare the new tool with PaP score and PPI. |
-
•
Inpatients treated in a hospice or hospital palliative care units across multiple centres in Japan
-
•
Development sample:
-
○
n = 201
-
○
Median age: 63 years (SD 12)
-
•
Test sample:
-
○
n = 208
-
○
Median age: 65 years (SD 12)
|
|
|
The accuracy and practical value of CPS remains controversial Ratio between the number of events (death) and the number of potential predictors <10 Requires invasive procedure Requires external validation |
Level 2++ Hawker et al: 33/36 Morita et al: 5/7 |
-
•
Multivariate analysis (Cox) identified 5 significant predictors (p < 0.05) in the development sample, including:
-
○
CPS
-
○
Disturbance in consciousness
-
○
Pleural effusion
-
○
WBC count
-
○
Lymphocyte percentage
-
•
JPOS-PI was developed using these predictors and divided patients into 3 risk groups. Median survival for each group was as follows:
-
○
Group A, low risk: 51 days
-
○
Group B, intermediate risk: 35 days
-
○
Group C, high risk: 16 days
-
•
Survival probability for more than 30 days in the development sample was as follows:
-
○
Group A, low risk: 78%
-
○
Group B, intermediate risk: 61%
-
○
Group C, high risk: 16%
-
•
JPOS-PI was studied in subsequent test sample, and constant results were obtained.
-
•
The 3 risk groups judged with JPOS-PI and PaP scores showed very similar survival in the test sample.
|
| Tarumi, Y et al. 2011, Canada [32] |
|
|
|
|
Includes cancer and non-cancer patients Variations in clinicians’ experience in making the predictions Until the PaP is tested in all hospitalized patients or all cancer patients, application of the Survival Rate by PaP Score requires caution Single centre Earlier published study with similar objectives [28] |
Level 2++ Hawker et al: 35/36 Maltoni et al: 6/7 |
-
•
The PaP Score subdivides the population into 3 groups with a different probability of survival at 30 days:
-
○
Group A: >70%
-
○
Group B: 30–70%
-
○
Group C: <30% [21]
-
•
This study externally validates the PaP Score in patients referred to the hospital palliative care team. The 3 groups, divided based on different ranges of PaP Score, had significantly different survival curves with 30-day-survival rates as follows:
-
○
Group A: 78%
-
○
Group B: 55%
-
○
Group C: 11%
-
•
Median survivals for the three groups were:
-
○
Group A: 89 days (95% CI 72–106)
-
○
Group B: 35 days (95% CI 30–40)
-
○
Group C: 4 days (95%CI 3–5)
|
| Feliu, J et al. 2011, Spain [33] |
Prospective cohort study (explorative) To identify more relevant clinical and laboratory variables that predict survival in terminally ill cancer patients, and to develop and validate a nomogram that predicts survival at 15, 30, and 60 days. |
|
|
|
Some relevant survival parameters such as CPS, CRP, and comorbid conditions were not assessed in this study but might provide additional prognostic information Requires invasive procedure |
Level 2++ Hawker et al: 35/36 Maltoni et al: 6/7 |
-
•
In univariate survival analysis (Cox), a number of factors were statistically significantly associated with survival time including:
-
○
Serum creatinine (p = 0.039)
-
○
Serum corrected calcium (p = 0.007)
-
○
Haemoglobin (p = 0.006)
-
○
Lymphocyte count (p = 0.000)
-
○
Serum bilirubin (p = 0.001)
-
○
Serum LDH (p = 0.000)
-
○
Serum cholesterol (p = 0.009)
-
○
Serum albumin (p = 0.000)
-
•
Multivariate analysis (Cox) identified 5 variables independently prognostic:
-
○
ECOG PS
-
○
TTD, a subjective parameter which may relate to tumour aggressiveness
-
○
Serum albumin (HR 0.729, 95% CI 0.621–0.856, p = 0.000)
-
○
Serum LDH (HR 1.127, 95% CI 1.058–1.200, p = 0.000)
-
○
Lymphocyte number (HR 0.876, 95% CI 0.791–0.970, p = 0.011)
-
•
Statistically non significant factors by univariate analysis:
-
○
Serum potassium
-
○
Serum sodium
-
○
GGT
-
○
ALP
-
○
Neutrophil count
-
○
Uric acid
-
•
A nomogram for predicting the probability of survival at 15, 30, and 60 days was constructed using the 5 variables identified by multivariate analysis.
-
•
Cumulative survival could be grouped into quartiles according to the score generated by the nomogram, and median survival by group was:
-
○
Quartile 1, score ≤-0.45: 83 days (95% CI 64–102)
-
○
Quartile 2, score -0.45–0: 33 days (95% CI 22–44)
-
○
Quartile 3, score 0–0.34: 24 days (95%CI 20–28)
-
○
Quartile 4, score >0.34: 10 days (95%CI 8–12)
-
•
The nomogram was validated with an external dataset that included 474 patients. The nomogram provided better AUC values than the PaP Score at 15, 30 and 60 days.
|
| Gwilliam, B et al. 2011, UK [34] |
|
|
Advanced cancer Palliative |
|
|
Level 2++ Hawker et al: 36/36 Maltoni et al: 6/7 |
-
•
On multivariate analysis, a number of variables independently predicted both two week and two month survival including:
-
○
Serum CRP
-
○
WBC count
-
○
Platelet count
-
○
Urea
-
•
A number of variables had prognostic significance only for 2 week including:
-
○
Serum ALT level
-
•
A number of variables had prognostic significance only for 2 month survival including:
-
○
Lymphocyte count
-
○
Neutrophil count
-
○
Serum ALP level
-
○
Serum albumin
-
•
Separate prognostic models were created for patients without (PiPS-A) or with (PiPS-B) blood results. The AUC for all 4 models varied between 0.79–0.86.
-
•
PiPS-A scores can correctly order survival times for pairs of participants 68.9% of time.
-
•
PiPS-B scores can correctly order survival times for pairs of participants 67.5% of time.
-
•
The PiPS prognostic models can predict whether patients will survive for “days,” “weeks,” or “months”.
|
| Cui, J et al. 2014, China [35] |
|
|
|
|
Serum Na, total bilirubin, direct bilirubin, AST and ALP have not previously been included in other prognostic models. Their significance in survival prediction warrants further study Ratio between the number of events (death) and the number of potential predictors <10 Requires invasive procedure |
Level 2+ Hawker et al: 32/36 Maltoni et al: 5/7 |
-
•
By univariate and multivariate analysis, 14 variables were significantly associated with survival time including
-
○
KPS
-
○
Pain
-
○
Ascites
-
○
Hydrothorax
-
○
Oedema
-
○
Delirium
-
○
Cachexia
-
○
WBC
-
○
Hemoglobin
-
○
Serum sodium
-
○
Total bilirubin
-
○
Direct bilirubin
-
○
AST
-
○
ALP
-
•
The new-ChPS Scale developed using the above prognostic factors.
-
•
When the prognostic score of a patient was > 12, the prediction of survival appeared to be < 7 days.
-
•
When the prognostic score of a patient was < 6, the prediction of survival appeared to be > 180 days (6 months).
-
•
Statistically non significant factors by univariate analysis:
-
○
Platelet count
-
○
Serum creatinine
-
○
Serum chlorine
-
○
Serum globulin
|
| Kim, ES et al. 2014, Korea [36] |
Prospective cohort study (confirmative) To validate the PiPS model for terminal cancer patients in Korea, and evaluate its value in clinical practice. |
|
|
|
Includes only hospital inpatients on palliative care ward. Further research is needed in order to generalise the findings in various clinical settings PiPs-B model requires invasive procedure Ratio between the number of events (death) and the number of potential predictors <10 |
Level 2++ Hawker et al: 35/36 Maltoni et al: 5/7 |
Gwilliam et al. developed the PiPS prognostic models, which can be used without (PiPS-A) or with (PiPS-B) blood results, and can predict whether patients are in the last “days,” “weeks,” or “months” of life [34]. In this study, the PiPs models are externally validated in patients admitted to a palliative care ward in Korea. The overall accuracy between the PiPS-A and actual survival was 52.0%. The overall accuracy between the PiPS-B and actual survival was 49.5%. The overall accuracy between clinicians’ estimates and actual survival was 46.5%, which were lower than those of the PiPS models. The “weeks” and “months” groups showed significantly prolonged survival rates than “days” group did in both PiPS-A and PiPS-B, by the Kaplan-Meier method. The sensitivity, specificity, PPV, and NPV of predictions of survival using the PiPS-A and the PiPS-B predictor models were variable. In the “days” and “weeks” groups, the PiPS-A/PiPS-B predictions and clinicians’ estimates were consistent with actual median survival. |
| Amano, K et al. 2015, Japan [37] |
|
|
-
•
Advanced cancer:
-
○
Locally extensive: n = 278
-
○
Metastatic: n = 1233
-
•
Various malignancies including haematological
-
•
Mean age:
-
○
Low, CRP <1 mg/dL: 68.8 years (SD 13.4)
-
○
Moderate, CRP ≥1 - <5 mg/dL: 69.1 years (SD 12.1)
-
○
High, CRP ≥5 - <10 mg/dL: 68.4 years (SD 12.6)
-
○
Very high, CRP ≥10 mg/dL: 66.3 years (SD 13.4)
|
-
•
Median survival:
-
○
Low, CRP <1 mg/dL: 72.0 days (95% CI 58–86)
-
○
Moderate, CRP ≥1 - <5 mg/dL: 42.0 days (95% CI 36–48)
-
○
High, CRP ≥5 - <10 mg/dL: 25.0 days (95% CI 22–28)
-
○
Very high, CRP ≥10 mg/dL: 17.0 days (95% CI 14–20)
|
Does not distinguish between different malignancies which may influence serum CRP levels No data on acute infections, acute medical conditions and cachexia, which may influence serum CRP levels Requires invasive procedure |
Level 2++ Hawker et al: 36/36 Maltoni et al: 6/7 |
-
•
Survival rate decreased and mortality rate increased with increasing CRP level.
-
•
Patients classified into 4 groups based on CRP level. Median survival times were:
-
○
Low, CRP <1 mg/dL: 72.0 days (95% CI 58–86)
-
○
Moderate, CRP ≥1 - <5 mg/dL: 42.0 days (95% CI 36–48)
-
○
High, CRP ≥5 - <10 mg/dL: 25.0 days (95% CI 22–28)
-
○
Very high, CRP ≥10 mg/dL: 17.0 days (95% CI 14–20) (p < 0.001)
-
•
The differences in survival and 30-, 60- and 90-day mortality rates among the groups were statistically significant (p < 0.001).
-
•
In the multivariate adjusted model (Cox), a significantly higher risk of mortality was observed in the moderate-, high- and very high-CRP groups than in the low-CRP group:
-
○
Moderate, CRP ≥1 - <5 mg/dL: HR 1.47 (95% CI 1.24–1.73), p < 0.001
-
○
High, CRP ≥5 - <10 mg/dL: HR 2.09 (95% CI 1.74–2.50), p < 0.001
-
○
Very High, CRP ≥10 mg/dL: HR 2.55 (95% CI 2.13–3.05), p < 0.001
|
| Baba, M et al. 2015, Japan [38] |
|
-
•
3 palliative care settings in Japan:
-
○
Hospital inpatients
-
○
Palliative care inpatients
-
○
Community palliative care services
-
•
Modified PiPs A model: n = 2212
-
•
Modified PiPs B model: n = 1257
|
-
•
Advanced cancer
-
•
Under hospital or community palliative care services
-
•
Mean age:
-
○
Hospital inpatients: 65.5 years (SD 12.6)
-
○
Palliative care unit inpatients: 70.4 years (SD 12.3)
-
○
Community palliative care service: 73.0 years (SD 12.5)
|
-
•
Median survival:
-
○
Hospital inpatients: 45 days
-
○
Palliative care unit inpatients: 24 days
-
○
Community palliative care service: 37 days
|
Includes patients receiving chemotherapy (n = 359), however subgroup analysis of patients not receiving chemotherapy demonstrated the same conclusions Modified PiPs model not compared to CPS or another prognostic model Ratio between the number of events (death) and the number of potential predictors <10 |
Level 2++ Hawker et al: 35/36 Maltoni et al: 6/7 |
-
•
Gwilliam et al. developed the PiPS prognostic models, which can be used without (PiPS-A) or with (PiPS-B) blood results, and can predict whether patients are in the last “days,” “weeks,” or “months” of life [34].
-
•
In this study, the original PiPS models were modified so that ratings could be obtained without patient participation, and externally validated across all palliative care settings.
-
•
In all palliative care settings, both the modified PiPS-A and PiPS-B was successfully validated and identified 3 risk groups with different survival rates (p < 0.001).
-
•
The median survival for each group classified by PiPs-A was as follows:
-
○
Group A, days: 10 days (95% CI 8–12)
-
○
Group B, weeks: 39 days (95% CI 34–44)
-
○
Group C, months: 134 days (95% CI: 108–160)
-
•
The absolute agreement ranged from 56–60% in the PiPS-A model and 60–62% in the PiPS-B model, similar to previous reports.
|
| Baba, M et al. 2015, Japan [39] |
Prospective cohort study (confirmative) To investigate the feasibility and accuracy of the PaP Score, D-PaP Score, PPI and modified PiPS models across 58 palliative care services in Japan. |
-
•
4 palliative care settings across 58 centres in Japan:
-
○
Hospital inpatients (n = 554)
-
○
Palliative care units inpatients (n = 820)
-
○
Community palliative care services (n = 472)
-
○
Palliative chemotherapy (n = 515)
-
•
n = 2361
|
-
•
Advanced cancer:
-
○
Locally extensive: n = 485
-
○
Metastatic: n = 1876
-
•
Various malignancies including haematological
-
•
Under hospital or community palliative care services
-
•
Mean age:
-
○
Hospital inpatients: 68.1 years (SD 12.7)
-
○
Palliative care unit inpatients: 70.8 years (SD 12.2)
-
○
Community palliative care service: 74.3 years (SD 11.9)
-
○
Palliative chemotherapy: 62.8 years (SD 12.0)
|
-
•
Median survival:
-
○
Hospital inpatients: 37 days (95% CI 31.6–42.4)
-
○
Palliative care unit inpatients: 25 days (95% CI 22.6–27.4)
-
○
Community palliative care services: 37 days (95% CI 32.1–41.9)
-
○
Palliative chemotherapy: 48 days (95% CI 41.9–54.1)
|
|
Level 2++ Hawker et al: 35/36 Maltoni et al: 6/7 |
The D-PaP Score integrates delirium as an additional predictor to the PaP score. In this study, the feasibility of previously described predictor models: PaP Score, D-PaP Score, PPI, and modified PiPS models were assessed across all palliative care settings. The feasibility of PPI and modified PiPS-A was >90% in all groups, followed by PaP and D-PaP scores; modified PiPS-B had the lowest feasibility. The accuracy of prognostic scores was ≥69% in all groups and the difference was within 13%, while c-statistics were significantly lower with the PPI than PaP and D-PaP scores. All prognostic tools can predict short (≤13 days or ≤21 days) and long (≥30 days, ≥42 days, or ≥56 days) survival probability in all these groups using the cut-off points proposed in the original studies. |
| Malik S et al. 2015, Australia [40] |
Retrospective cohort study To determine the reversibility of hypercalcaemia amongst patients whose underlying with intravenous bisphosphonates, and assess the prognostic value of correction in serum Ca. |
|
|
-
•
Median survival:
-
○
Responders: 22 days
-
○
Non-responders: 3 days
|
|
Level 2++ Hawker et al: 33/36 Maltoni et al: 6/7 |
A reduction in serum calcium concentration was associated with a significantly prolonged survival, as well as symptomatic improvement, irrespective of whether normocalcaemia was achieved. Pre-treatment serum calcium level did not independently effect survival, in contrast to the literature. |
| Taylor, P et al. 2015. UK [41] |
|
|
|
|
No data collected on hydration status, clinically assisted hydration, oxygen therapy or presence of acute illness Retrospective Reliance of death certificate for cause of death Ratio between the number of events (death) and the number of potential predictors <10 |
Level 2++ Hawker et al: 36/36 Maltoni et al: 4/7 |
Heart rate significantly increased during the last week of life, although the model intercept value only reached 99bpm on the day of death. Changes were not observed in blood pressure, and intercept values remained clinically unremarkable All respiratory variables measured changed significantly over the final 2 weeks of life with a rise in respiratory rate and reduction in oxygen saturation. Serum urea and creatinine showed a clinically and statistically significant increase. Serum sodium showed a statistically significant but clinically insignificant increase, while serum potassium showed no significant change over time. Evidence of leucocytosis and reduced lymphocyte counts, but there was no clinically significant change in the last 2 weeks of life. Recorded values were abnormal indicating that these variables are more likely to change slowly over weeks rather than days. Serum albumin showed a significant change over time and abnormal values as death approaches, showing prognostic power over both longer and shorter timescales. Other long-term variables including haemoglobin showed a significant change over 3 months of data collection, but results were masked by red cell transfusion. |
| Yoon, J et al. 2015, Korea [42] |
|
|
|
|
Single centre Retrospective Did not assess hydration, use of diuretics or clinically assisted hydration Did not assess changes in serum Na level over time Requires invasive procedure |
Level 2+ Hawker et al: 35/36 Maltoni et al: 5/7 |
-
•
In univariate analysis (Cox), CRP showed significant association with survival (HR 1.22, p<0.001).
-
•
In univariate analysis (Cox), hyponatraemia ≤ 125mEq/L was significantly associated with poor prognosis compared with eunatraemia (HR 1.91, p<0.001).
-
•
In multivariate analysis (Cox), CRP showed a significant association with survival (HR 1.16, p< 0.001).
-
•
In multivariate analysis (Cox), hyponatraemia showed statistical significance regarding survival time compared with eunatraemia:
-
○
Serum Na ≤125mEq/L: HR 1.75, p<0.001
-
○
Serum Na 126 – 135mEq/L: HR 1.19, p = 0.048
-
○
Serum Na ≤125mEq/L showed the shortest survival time compared with eunatraemia
-
•
Hyponatraemia is an independent prognostic marker in advanced cancer patients.
|
| Miura, T et al. 2015, Japan [43] |
|
|
|
-
•
Median survival:
-
○
GPS 0: 58 days (95% CI 48–81)
-
○
GPS 1: 43 days (95% CI 37–50)
-
○
GPS 3: 21 days (95% CI 19–24)
|
|
Level 2++ Hawker et al: 35/36 Maltoni et al: 6/7 |
-
•
The GPS is a simple inflammation-based prognostic score that represents elevated CRP levels (>10 mg/L) and hypoalbuminaemia (<35 g/dL). The GPS is scored as 0, 1 or 2.
-
•
In this study, the prevalence of GPS 2 was 70.9%.
-
•
The sensitivity and specificity for 3-week prognosis of GPS 2 were 0.879 and 0.410 respectively
-
•
Median survival time based on GPS was as follows (p<0.001):
-
○
GPS 0: 58 days (95% CI 48–81)
-
○
GPS 1: 43 days (95% CI 37–50)
-
○
GPS 3: 21 days (95% CI 19–24)
-
•
The NLR is the neutrophil-lymphocyte ratio, and a score ≥ 4 has been shown to be the best cut of value for predicting prognosis.
-
•
Univariate analysis (Cox) showed a significant association with survival and the following variables:
-
○
GPS of 1 or 2
-
○
NLR ≥4
-
•
Multivariate analysis (Cox) showed that NLR ≥ 4 was an independent prognostic factor (HR 1.43, 95% CI, 1.17–1.75 p < 0.001).
-
•
The GPS is a good prognostic indicator for cancer patients across all palliative care settings.
|
| Coyle S et al. 2016, UK [44] |
Prospective cohort study (investigative) To collect urine samples from patients towards the end of life and analyse them for VOCs using GC-MS, and water-soluble metabolites using NMR. |
Hospice inpatients Advanced cancer, n = 8 Non-cancer, n = 1 Healthy controls, n = 4 |
|
|
|
Conference abstract Level: N/A Hawker et al: N/A Maltoni et al: N/A |
GCMS analysis identified over 390 VOCs. There were 279 VOCs in normal urine. The number of statistically significant VOCs increased from 12 at 4 weeks before death, to 43 at 3–5 days before death. A trend analysis for the individual VOCs showed that of these VOCs, 1 was significant at four weeks before death and 22 VOCs were significant at 3–5 days before dying. The steepest rise in significant VOCs was seen in the last week of life. These VOCs may predict the dying process. |
| Niki K et al. 2016, Japan [45] |
Retrospective cohort study (explorative) To identify short-term prognostic variables by detecting a statistical change-point in laboratory test values. To compare the WPCBAL and WPBAL scores to the GPS score. |
|
|
|
|
Conference abstract Level: N/A Hawker et al: N/A Maltoni et al: N/A |
The GPS is a simple inflammation-based prognostic score that represents elevated CRP levels (>10 mg/L) and hypoalbuminaemia (<35 g/dL). The GPS is scored as 0, 1 or 2. The WPCBAL score was derived from the WBC, platelet count, BUN, AST, LDH, and CRP levels. The WPBAL score was derived from the WBC, platelet count, BUN, AST, and LDH levels. The WPCBAL and WPBAL scores showed a higher sensitivity, specificity, negative predictive value, and relative risk for the prognosis within 2 weeks of death compared to those of the GPS considering a less than three weeks prognosis. While the WPCBAL and WPBAL scores showed an equivalent positive predictive value to the GPS, both positive and negative likelihood ratio were superior to those of GPS. |
| Paulsen O et al. 2016, Norway [46] |
Randomised controlled trial Post hoc analysis to investigate whether inflammatory markers at baseline (protein production in the serum detected by Multiplex technology) were associated with symptom intensity; or associated with response to corticosteroid therapy. |
|
Advanced cancer Opioid therapy |
|
Only pre-treatment samples available for analysis No direct comparison with corticosteroid group Small sample number 6 patients excluded from analysis due to missing data Protein and not RNA levels detected Requires invasive procedure Prognostic significance not assessed |
Conference abstract Level: N/A Hawker et al: N/A Maltoni et al: N/A |
-
•
Cytokines IL-2, IL-4, IL-8, IL-10, IL-12(p70), TNF-α, IFN-γ, and MIP-1α values were below the lower limit of detection in the serum in more than 37 of the 43 analyzed patients, and these data were excluded from the analyses.
-
•
The following cytokines were elevated above the lower limit of detection in the serum:
-
○
sTNF- r1 (n = 43, median concentration 10917 pg/mL)
-
○
MCP-1 (n = 43, median 64.1 pg/mL)
-
○
IL-18 (n = 43, median 103.2 pg/mL)
-
○
TGF-β1 (n = 43, median 45145 pg/mL)
-
○
MIF (n = 42, median 134.9 pg/mL)
-
○
IL-6 (n = 21, median 2.33 pg/mL)
-
○
IL-1ra (n = 11, median 21.7 pg/mL)
-
○
IL-1β (median 0.15 pg/mL)
-
○
CRP (median 44 pg/mL)
-
○
ESR (median 42 pg/mL)
-
•
EORTC QLQ-C30 physical and role function (markers of quality of life) were highly correlated to CRP and sTNF-r1, and IL-6 and ESR, respectively.
-
•
Medium correlations were found between:
-
○
Appetite and CRP, IL-6 and IL-1β
-
○
Fatigue and IL-1ra
-
○
Cognitive function and TGF-β1
-
○
Dyspnoea and IL-1β.
|
| Wrafter S et al. 2016, Ireland [47] |
Cohort study (confirmatory) To assess the prognostic value of CRP, albumin and CRP-based scores (CAR, mGPS with standard and low cut-offs) in cancer admissions to a specialist palliative care unit in Ireland. |
|
|
|
|
Conference abstract Level: N/A Hawker et al: N/A Maltoni et al: N/A |
85% (n = 94) had serum CRP > 10mg/L. 96% (n = 106) had serum albumin <35g/L. There was a significant mean survival difference between high and normal CRP (80 versus 29 days; p< 0.001) and low and normal albumin (35 versus 87 days; p< 0.01). CAR, the CRP/albumin ratio is an inflammatory prognostic model that has been shown to predict survival in various malignancies. The mGPS combines serum albumin and CRP into a risk stratification score, and has been shown to predict survival in various malignancies. The mGPS divides patients into 3 subgroups: 0, 1 and 2. Neither of these predictive models has been used in palliative care settings. CAR was above published cut-offs for 84% (n = 92). A significant survival difference was found between mGPS groups at 7(p = 0.01) and 180 (p = 0.02) but not 30 days. A lower CRP threshold did not improve prognostication. |
