Vigano et al identified clinical factors which can help identify a patient with advanced malignancy who is likely to survive only a short period. These can help identify patients for whom terminal care is appropriate. The authors are from the University of Alberta in Edmonton, Canada.
Parameters found to be independent predictors of short survival:
(1) site of primary tumor
(2) liver metastases
(3) comorbidity
(4) weight loss in past 6 months
(5) serum albumin
(6) serum LDH (see limitations)
(7) lymphocyte count
(8) clinical estimate of survival made by the physician
Parameters |
Finding |
Points |
site of primary tumor |
lung |
1 |
|
other |
0 |
liver metastases |
absent |
0 |
|
present |
1 |
comorbidity |
none to mild |
0 |
|
moderate to severe |
1 |
weight loss in past 6 months |
<= 8.1 kg |
0 |
|
> 8.1 kg |
1 |
serum albumin |
>= 3.5 g/dL |
0 |
|
< 3.5 g/dL |
1 |
serum LDH |
<= 618 IU/L |
0 |
|
> 618 IU/L (see limitations) |
1 |
lymphocyte count |
>= 1,000 per µL |
0 |
|
< 1,000 per µL |
1 |
clinical estimate of survival by physician |
>= 2 months |
0 |
|
< 2 months |
1 |
number of parameters associated with a short survival =
= SUM(points for the 8 parameters)
Interpretation:
• minimum number of parameters: 0
• maximum number of parameters: 8
• The higher the number of parameters the greater the risk of a short survival.
• Several of the factors are measures of malnutrition while others are measures of tumor burden.
Additional items to consider including:
(1) presence of moderate to severe nausea
(2) ECOG performance scale of 4 (see limitations, below)
Limitations:
• The method, reaction temperature and reference range for LDH are not specified. This markedly limits the transportability of this parameter. For the implementation I arbitrarily used two times the upper limit of the normal reference range for this variable.
• I personally am not convinced by the arguments against using the ECOG performance scale as a predictor. It is true that the scale can be affected by acute change (page 866), but this can be compensated for. It is no more arbitrary than a physician's estimate of survival. I also am not sure that lumping ECOG scales 2, 3 and 4 together makes much sense. If your goal is predicting a short survival, I would think ECOG 4 is a useful bit of clinical information.
• An albumin in the range of 3.0 to 3.4 g/dL does not strike me as significant as < 3 g/dL.
• I wonder if grading the extent of liver metastases would improve performance. I would think extensive liver involvement would be more significant than a single small metastasis.
• As the authors say on page 862 there are many ways for a patient with advanced cancer to die. A person slowly wasting away can survive surprisingly long, while an immunocompromised patient or a patient with acute hemorrhage can die very rapidly.
Purpose: To determine if a patient with advanced cancer has risk factors associated with only a short survival based on the study of Vigano et al.
Specialty: Hematology Oncology, Surgery, general
Objective: severity, prognosis, stage
ICD-10: C00-C14, C15-C26, C30-C39, C40-C41, C43-C44, C45-C49, C50-C50, C51-C58, C60-C63, C64-C68, C69-C72, C73-C75, C76-C80, C81-C96, C97-C97,