Graf et al developed a prognostic index for patients with advanced colorectal carcinoma. This can help identify those patients for whom more aggressive or investigational therapies may be warranted. The authors are from the University of Uppsala in Sweden.
Patient selection: metastatic or unresectable or locally recurrent tumor
Parameters evaluated using multivariate analysis:
(1) number of symptoms referable to tumor (pain, fatigue, nausea, secretion/bleeding, other)
(2) hemoglobin level in g/L
(3) chemotherapy with methotrexate, 5-fluorouracil and leucovorin (MFL)
(4) Karnofsky performance scale
(5) disease free interval in days
(6) primary tumor surgical status
|
Parameters
|
Findings
|
Points
|
|
number of symptoms
|
|
|
|
hemoglobin in g/L
|
|
|
|
chemotherapy
|
with MFL combination
|
1
|
|
|
with 5-fluorouracil alone
|
0
|
|
Karnofsky performance scale
|
> 60
|
0
|
|
|
<= 60
|
1
|
|
disease free interval
|
>= 365 days
|
1
|
|
|
< 365 days
|
0
|
|
primary tumor
|
resected
|
0
|
|
|
unresected
|
1
|
where:
• There is only partial guidance on how to enter the number of symptoms. The patient characteristic table (Table 1) the maximum designation was >= 3 symptoms. In the implementation the most allowed will be 5, but this is an arbitrary decision.
• The study did not discuss those patients who refused chemotherapy or who were treated with other regimens.
• Unresected tumor probably indicates a patient with unresectable disease.
prognostic index =
= (0.1738 * (number of symptoms)) - (0.0177 * (hemoglobin level)) - (0.5872 * (points for chemotherapy)) - (0.3838 * (points for disease free interval)) + (0.3576 * (points for Karnofsky performance)) + (0.2795 * (points for primary tumor resection))
Interpretation:
• minimum index: around -3.626
• maximum score: around 0.44 (depends on hemoglobin level)
• The index identified 4 prognostic groups.
|
Index
|
1 Year Survival
|
2 Year Survival
|
|
< -2.74
|
66%
|
16%
|
|
-2.74 to -2.30
|
42%
|
10%
|
|
-2.29 to - 1.785
|
15%
|
3%
|
|
> - 1.785
|
12%
|
0%
|
estimated from Figure 2, page 1121
