PATIENTS CHARACTERISTICS
=======================================
=======================================
|
Dependent: First-line
|
|
IO+V
|
V
|
IO+IO
|
IO+IO+V
|
|
Total N (%)
|
|
41 (20.4)
|
85 (42.3)
|
52 (25.9)
|
23 (11.4)
|
|
Age
|
Mean (SD)
|
62.4 (10.9)
|
68.2 (11.5)
|
62.4 (11.5)
|
62.4 (9.8)
|
|
AgeG
|
≥45
|
39 (95.1)
|
82 (96.5)
|
49 (94.2)
|
22 (95.7)
|
|
|
<45
|
2 (4.9)
|
3 (3.5)
|
3 (5.8)
|
1 (4.3)
|
|
Male1
|
No
|
9 (22.0)
|
23 (27.1)
|
11 (21.2)
|
9 (39.1)
|
|
|
Yes
|
32 (78.0)
|
62 (72.9)
|
41 (78.8)
|
14 (60.9)
|
|
Nephrectomy1
|
No
|
15 (36.6)
|
19 (22.4)
|
29 (55.8)
|
9 (39.1)
|
|
|
Yes
|
26 (63.4)
|
66 (77.6)
|
23 (44.2)
|
12 (52.2)
|
|
|
(Missing)
|
0 (0.0)
|
0 (0.0)
|
0 (0.0)
|
2 (8.7)
|
|
Sarcomatoid1
|
No
|
32 (78.0)
|
67 (78.8)
|
38 (73.1)
|
20 (87.0)
|
|
|
Yes
|
7 (17.1)
|
7 (8.2)
|
8 (15.4)
|
0 (0.0)
|
|
|
(Missing)
|
2 (4.9)
|
11 (12.9)
|
6 (11.5)
|
3 (13.0)
|
|
risk
|
favourable
|
8 (19.5)
|
27 (33.8)
|
1 (2.0)
|
2 (8.7)
|
|
|
intermediate
|
27 (65.9)
|
46 (57.5)
|
25 (49.0)
|
17 (73.9)
|
|
|
poor
|
6 (14.6)
|
7 (8.8)
|
25 (49.0)
|
4 (17.4)
|
|
BoneMet1
|
No
|
32 (78.0)
|
67 (78.8)
|
36 (69.2)
|
17 (73.9)
|
|
|
Yes
|
9 (22.0)
|
17 (20.0)
|
16 (30.8)
|
6 (26.1)
|
|
|
(Missing)
|
0 (0.0)
|
1 (1.2)
|
0 (0.0)
|
0 (0.0)
|
|
LiverMet1
|
No
|
36 (87.8)
|
74 (87.1)
|
45 (86.5)
|
22 (95.7)
|
|
|
Yes
|
5 (12.2)
|
10 (11.8)
|
7 (13.5)
|
1 (4.3)
|
|
|
(Missing)
|
0 (0.0)
|
1 (1.2)
|
0 (0.0)
|
0 (0.0)
|
|
BrainMet1
|
No
|
39 (95.1)
|
81 (95.3)
|
44 (84.6)
|
23 (100.0)
|
|
|
Yes
|
2 (4.9)
|
3 (3.5)
|
8 (15.4)
|
0 (0.0)
|
|
|
(Missing)
|
0 (0.0)
|
1 (1.2)
|
0 (0.0)
|
0 (0.0)
|
|
irAE.L1
|
No
|
34 (82.9)
|
85 (100.0)
|
40 (76.9)
|
16 (69.6)
|
|
|
Yes
|
7 (17.1)
|
0 (0.0)
|
12 (23.1)
|
7 (30.4)
|
|
Line2
|
0
|
18 (43.9)
|
24 (28.2)
|
26 (50.0)
|
7 (30.4)
|
|
|
1
|
23 (56.1)
|
61 (71.8)
|
26 (50.0)
|
16 (69.6)
|
|
Line3
|
0
|
27 (65.9)
|
38 (44.7)
|
44 (84.6)
|
15 (65.2)
|
|
|
1
|
14 (34.1)
|
47 (55.3)
|
8 (15.4)
|
8 (34.8)
|
|
Line4
|
0
|
36 (87.8)
|
66 (77.6)
|
50 (96.2)
|
21 (91.3)
|
|
|
1
|
5 (12.2)
|
19 (22.4)
|
2 (3.8)
|
2 (8.7)
|
|
Line5
|
0
|
40 (97.6)
|
79 (92.9)
|
52 (100.0)
|
23 (100.0)
|
|
|
1
|
1 (2.4)
|
6 (7.1)
|
0 (0.0)
|
0 (0.0)
|
=======================================
=======================================
ANALYSIS OF VEGF POPULATION
=======================================
=======================================
=======================================
BASELINE OUTCOMES
=======================================
|
Dependent: Line1_V
|
|
V
|
|
Total N (%)
|
|
85 (100.0)
|
|
Best.Response
|
CR
|
1 (1.3)
|
|
|
PD
|
16 (21.1)
|
|
|
PR
|
28 (36.8)
|
|
|
SD
|
31 (40.8)
|
|
ORR_12w
|
No
|
54 (63.5)
|
|
|
Yes
|
19 (22.4)
|
|
|
(Missing)
|
12 (14.1)
|
|
DCR_12w
|
No
|
17 (20.0)
|
|
|
Yes
|
56 (65.9)
|
|
|
(Missing)
|
12 (14.1)
|
|
OSL1fu2
|
OS <2y
|
27 (31.8)
|
|
|
OS 2-5y
|
41 (48.2)
|
|
|
OS ≥5y
|
16 (18.8)
|
|
|
(Missing)
|
1 (1.2)
|
|
PFSL1fu2
|
PFS <2y
|
65 (76.5)
|
|
|
PFS 2-5y
|
19 (22.4)
|
|
|
PFS ≥5y
|
1 (1.2)
|
=======================================
ASOCIATIONS WITH RESPONSE IN L1 VEGF POPULATION
=======================================
ORR No vs Yes at week 12 and all variables
The only variable associated with ORR could be NLR at week 0.
|
|
No
|
Yes
|
p
|
test
|
|
n
|
54
|
19
|
|
|
|
PltHigh_0w = Yes (%)
|
0 ( 0.0)
|
0 ( 0.0)
|
NaN
|
|
|
HbLow_0w = Yes (%)
|
23 (42.6)
|
9 (47.4)
|
0.927
|
|
|
NLR_0w.cut = ≥3 (%)
|
30 (55.6)
|
5 (26.3)
|
0.054
|
|
|
PLR_0w.cut = ≥150 (%)
|
26 (50.0)
|
7 (36.8)
|
0.474
|
|
|
PltHigh_12w = Yes (%)
|
1 ( 1.9)
|
1 ( 5.3)
|
1.000
|
|
|
HbLow_12w = Yes (%)
|
29 (55.8)
|
9 (47.4)
|
0.719
|
|
|
NLR_12w.cut = ≥3 (%)
|
14 (26.4)
|
2 (11.1)
|
0.310
|
|
|
PLR_12w.cut = ≥150 (%)
|
24 (48.0)
|
6 (35.3)
|
0.530
|
|
|
BoneMet1 = Yes (%)
|
10 (18.5)
|
3 (15.8)
|
1.000
|
|
|
LiverMet1 = Yes (%)
|
9 (16.7)
|
1 ( 5.3)
|
0.392
|
|
DCR at week 12 and all variables
It’s quite similar to the previous analysis with ORR.
Only NLR_0w, HbLow_12w, and NLR_12w could be associated with
DCR.
|
|
No
|
Yes
|
p
|
test
|
|
n
|
17
|
56
|
|
|
|
PltHigh_0w = Yes (%)
|
0 ( 0.0)
|
0 ( 0.0)
|
NaN
|
|
|
HbLow_0w = Yes (%)
|
10 (58.8)
|
22 (39.3)
|
0.253
|
|
|
NLR_0w.cut = ≥3 (%)
|
12 (70.6)
|
23 (41.1)
|
0.063
|
|
|
PLR_0w.cut = ≥150 (%)
|
10 (62.5)
|
23 (41.8)
|
0.240
|
|
|
PltHigh_12w = Yes (%)
|
1 ( 6.2)
|
1 ( 1.8)
|
0.933
|
|
|
HbLow_12w = Yes (%)
|
12 (75.0)
|
26 (47.3)
|
0.094
|
|
|
NLR_12w.cut = ≥3 (%)
|
7 (43.8)
|
9 (16.4)
|
0.049
|
|
|
PLR_12w.cut = ≥150 (%)
|
7 (50.0)
|
23 (43.4)
|
0.889
|
|
|
BoneMet1 = Yes (%)
|
4 (23.5)
|
9 (16.1)
|
0.732
|
|
|
LiverMet1 = Yes (%)
|
3 (17.6)
|
7 (12.5)
|
0.890
|
|
=======================================
PREDICTORS OF RESPONSE
=======================================
Here I’ve made a logistic regression to try to measure the
relationships found in the previous analysis.
I’ve taken a look without adjusting by IMDC risk.
As we are looking for “risk factors” (means Odds Ratio), the
analysis is made to predict “no ORR”.
We can change the language latter.
Predictor of ORR
NLR 0w
NLR ≥3 at week 0 have an OR 3.50 (IC 95%: 1.16-12.10) for not
ORR
Means: NLR <3 at week 0 have an OR 3.50 (CI 95%: 1.16-12.10) for
ORR
## # A tibble: 2 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 1.71 0.336 1.60 0.109 0.899 3.40
## 2 NLR_0w.cut≥3 3.50 0.589 2.13 0.0333 1.16 12.1
Predictor of DCR
NLR 0w
NLR ≥3 at week 0 have an ORR 3.44 (CI 95%: 1.12-12.1) for not
DCR
Means: NLR <3 at week 0 have an ORR 3.44 (CI 95%: 1.12-12.1) for
DCR
## # A tibble: 2 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 0.152 0.480 -3.93 0.0000842 0.0519 0.355
## 2 NLR_0w.cut≥3 3.44 0.598 2.07 0.0385 1.12 12.1
Hb Low 12w
Hb low at week 12 have an OR 3.35 (IC 95%: 1.02-13.2) for not
DCR
Means: Hb NOT low at week 12 have an OR 3.35 (IC 95%: 1.02-13.2) for
DCR
## # A tibble: 2 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 0.138 0.533 -3.71 0.000204 0.0409 0.350
## 2 HbLow_12wYes 3.35 0.637 1.89 0.0581 1.02 13.2
NLR 12w
NLR ≥3 at week 12 have an OR 3.98 (IC 95%: 1.16-13.7) for not
DCR
Means: NLR <3 at week 12 have an OR 3.98 (IC 95%: 1.16-13.7) for
DCR
## # A tibble: 2 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 0.196 0.364 -4.48 0.00000761 0.0896 0.381
## 2 NLR_12w.cut≥3 3.98 0.622 2.22 0.0265 1.16 13.7
=======================================
OVERALL SURVIVAL FIRST-LINE
=======================================
HR according to all variables
This is just to take a first look about what could be significant.
Focus on univariable analysis.
Remember we are not including the variable “IMDC risk” here.
ORR and DCR are significant (this is obvious). Then, HbLow, PltHigh,
PLR, and NLR are significant
at both week 0 and week 12. Bone and Liver mets are not
significant.
This is just a first look. Then we will need to calculate the real
HR adjusting by risk.
|
|
Dependent: Survival
|
|
all
|
HR (univariable)
|
HR (multivariable)
|
|
15
|
ORR_12w
|
No
|
54 (74.0)
|
|
|
|
16
|
|
Yes
|
19 (26.0)
|
0.37 (0.15-0.89, p=0.026)
|
0.35 (0.12-1.08, p=0.069)
|
|
3
|
DCR_12w
|
No
|
17 (23.3)
|
|
|
|
4
|
|
Yes
|
56 (76.7)
|
0.22 (0.11-0.42, p<0.001)
|
0.14 (0.05-0.34, p<0.001)
|
|
21
|
PltHigh_0w
|
No
|
78 (100.0)
|
|
|
|
22
|
|
Yes
|
0 (0.0)
|
NA (NA-NA, p=NA)
|
NA (NA-NA, p=NA)
|
|
5
|
HbLow_0w
|
No
|
45 (57.7)
|
|
|
|
6
|
|
Yes
|
33 (42.3)
|
1.91 (1.05-3.50, p=0.035)
|
0.85 (0.33-2.20, p=0.741)
|
|
11
|
NLR_0w.cut
|
<3
|
41 (52.6)
|
|
|
|
12
|
|
≥3
|
37 (47.4)
|
1.91 (1.04-3.50, p=0.037)
|
0.26 (0.09-0.74, p=0.011)
|
|
17
|
PLR_0w.cut
|
<150
|
42 (56.0)
|
|
|
|
18
|
|
≥150
|
33 (44.0)
|
1.96 (1.06-3.61, p=0.032)
|
2.17 (0.85-5.53, p=0.104)
|
|
23
|
PltHigh_12w
|
No
|
72 (97.3)
|
|
|
|
24
|
|
Yes
|
2 (2.7)
|
20.52 (3.70-113.65, p=0.001)
|
78.67 (4.45-1391.71, p=0.003)
|
|
7
|
HbLow_12w
|
No
|
35 (47.3)
|
|
|
|
8
|
|
Yes
|
39 (52.7)
|
2.91 (1.48-5.70, p=0.002)
|
3.29 (1.19-9.12, p=0.022)
|
|
13
|
NLR_12w.cut
|
<3
|
57 (77.0)
|
|
|
|
14
|
|
≥3
|
17 (23.0)
|
2.45 (1.14-5.26, p=0.022)
|
3.95 (1.38-11.28, p=0.010)
|
|
19
|
PLR_12w.cut
|
<150
|
40 (57.1)
|
|
|
|
20
|
|
≥150
|
30 (42.9)
|
1.90 (1.01-3.58, p=0.046)
|
2.97 (1.05-8.38, p=0.040)
|
|
1
|
BoneMet1
|
No
|
67 (79.8)
|
|
|
|
2
|
|
Yes
|
17 (20.2)
|
0.76 (0.36-1.60, p=0.470)
|
0.98 (0.37-2.63, p=0.972)
|
|
9
|
LiverMet1
|
No
|
74 (88.1)
|
|
|
|
10
|
|
Yes
|
10 (11.9)
|
1.69 (0.71-4.03, p=0.234)
|
1.07 (0.37-3.05, p=0.904)
|
Differences in OS by ORR and DCR at week 12
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ ORR_12w, data = dbV,
## type = "kaplan-meier")
##
## 13 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## ORR_12w=No 54 35 37.9 31.0 55.2
## ORR_12w=Yes 18 7 65.4 51.5 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ DCR_12w, data = dbV,
## type = "kaplan-meier")
##
## 13 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## DCR_12w=No 17 14 10.2 7.85 43.3
## DCR_12w=Yes 55 28 55.2 45.08 NA
Differences in OS by HbLow, PLR and NLR at week 0
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ HbLow_0w, data = dbV,
## type = "kaplan-meier")
##
## 8 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## HbLow_0w=No 45 24 53.6 39.8 NA
## HbLow_0w=Yes 32 20 34.9 27.4 63
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ PLR_0w.cut, data = dbV,
## type = "kaplan-meier")
##
## 11 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## PLR_0w.cut=<150 41 18 55.2 39.2 NA
## PLR_0w.cut=≥150 33 24 31.0 19.3 65.4
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_0w.cut, data = dbV,
## type = "kaplan-meier")
##
## 8 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_0w.cut=<3 40 19 53.6 41.4 NA
## NLR_0w.cut=≥3 37 25 33.2 26.6 65.4
Differences in OS by HbLow, PLR and NLR at week 12
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ HbLow_12w, data = dbV,
## type = "kaplan-meier")
##
## 12 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## HbLow_12w=No 35 17 66.0 41.4 NA
## HbLow_12w=Yes 38 24 37.9 26.6 56.6
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ PLR_12w.cut, data = dbV,
## type = "kaplan-meier")
##
## 15 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## PLR_12w.cut=<150 40 19 53.6 41.4 NA
## PLR_12w.cut=≥150 30 21 32.7 26.6 66
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_12w.cut, data = dbV,
## type = "kaplan-meier")
##
## 12 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_12w.cut=<3 57 32 53.6 39.2 66
## NLR_12w.cut=≥3 16 9 26.6 17.0 NA
Differences in OS by NLR at week 0 + week 12
This is also interesting.
Patients who achieve a NLR < 3 at week 12 have better
prognosis.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_0w.cut + NLR_12w.cut,
## data = dbV, type = "kaplan-meier")
##
## 12 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_0w.cut=<3, NLR_12w.cut=<3 37 17 55.23 45.1 NA
## NLR_0w.cut=<3, NLR_12w.cut=≥3 1 1 5.03 NA NA
## NLR_0w.cut=≥3, NLR_12w.cut=<3 20 15 34.89 28.5 NA
## NLR_0w.cut=≥3, NLR_12w.cut=≥3 15 8 27.76 17.0 NA
Differences in OS by NLR at week 12 and risk
The importance of NLR at week 12 remains in all IMDC risk
categories.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ risk + NLR_12w.cut,
## data = dbV, type = "kaplan-meier")
##
## 13 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## risk=favourable, NLR_12w.cut=<3 18 7 66.0 45.08 NA
## risk=favourable, NLR_12w.cut=≥3 5 3 39.4 26.58 NA
## risk=intermediate, NLR_12w.cut=<3 33 19 55.2 39.16 NA
## risk=intermediate, NLR_12w.cut=≥3 9 4 19.3 16.95 NA
## risk=poor, NLR_12w.cut=<3 5 5 28.5 15.74 NA
## risk=poor, NLR_12w.cut=≥3 2 2 3.5 3.25 NA
=======================================
=======================================
ANALYSIS OF IO-IO POPULATION
=======================================
=======================================
=======================================
BASELINE OUTCOMES
=======================================
|
Dependent: Line1_IOIO
|
|
IO+IO
|
|
Total N (%)
|
|
52 (100.0)
|
|
Best.Response
|
CR
|
2 (3.9)
|
|
|
PD
|
17 (33.3)
|
|
|
PR
|
15 (29.4)
|
|
|
SD
|
16 (31.4)
|
|
|
mixed
|
1 (2.0)
|
|
ORR_12w
|
No
|
36 (69.2)
|
|
|
Yes
|
14 (26.9)
|
|
|
(Missing)
|
2 (3.8)
|
|
DCR_12w
|
No
|
19 (36.5)
|
|
|
Yes
|
31 (59.6)
|
|
|
(Missing)
|
2 (3.8)
|
|
OSL1fu2
|
OS <2y
|
37 (71.2)
|
|
|
OS 2-5y
|
12 (23.1)
|
|
|
OS ≥5y
|
3 (5.8)
|
|
PFSL1fu2
|
PFS <2y
|
45 (86.5)
|
|
|
PFS 2-5y
|
5 (9.6)
|
|
|
PFS ≥5y
|
1 (1.9)
|
|
|
(Missing)
|
1 (1.9)
|
=======================================
ASOCIATIONS WITH RESPONSE IN L1 IO-IO POPULATION
=======================================
ORR No vs Yes at week 12 and all variables
No variable is associated with ORR.
|
|
No
|
Yes
|
p
|
test
|
|
n
|
36
|
14
|
|
|
|
PltHigh_0w = Yes (%)
|
10 (27.8)
|
4 (30.8)
|
1.000
|
|
|
HbLow_0w = Yes (%)
|
27 (75.0)
|
9 (69.2)
|
0.970
|
|
|
NLR_0w.cut = ≥3 (%)
|
24 (66.7)
|
12 (92.3)
|
0.153
|
|
|
PLR_0w.cut = ≥150 (%)
|
27 (79.4)
|
11 (91.7)
|
0.603
|
|
|
PltHigh_12w = Yes (%)
|
10 (27.8)
|
2 (15.4)
|
0.607
|
|
|
HbLow_12w = Yes (%)
|
26 (72.2)
|
10 (76.9)
|
1.000
|
|
|
NLR_12w.cut = ≥3 (%)
|
26 (72.2)
|
7 (53.8)
|
0.386
|
|
|
PLR_12w.cut = ≥150 (%)
|
21 (65.6)
|
10 (76.9)
|
0.699
|
|
|
BoneMet1 = Yes (%)
|
14 (38.9)
|
2 (14.3)
|
0.181
|
|
|
LiverMet1 = Yes (%)
|
5 (13.9)
|
1 ( 7.1)
|
0.861
|
|
|
irAE.L1 = Yes (%)
|
6 (16.7)
|
6 (42.9)
|
0.115
|
|
DCR at week 12 and all variables
It’s quite similar to the previous analysis with ORR.
Plt High at week 12 is associated with no DCR and irAE is associated
with DCR.
|
|
No
|
Yes
|
p
|
test
|
|
n
|
19
|
31
|
|
|
|
PltHigh_0w = Yes (%)
|
5 (26.3)
|
9 (30.0)
|
1.000
|
|
|
HbLow_0w = Yes (%)
|
13 (68.4)
|
23 (76.7)
|
0.760
|
|
|
NLR_0w.cut = ≥3 (%)
|
13 (68.4)
|
23 (76.7)
|
0.760
|
|
|
PLR_0w.cut = ≥150 (%)
|
17 (89.5)
|
21 (77.8)
|
0.525
|
|
|
PltHigh_12w = Yes (%)
|
8 (42.1)
|
4 (13.3)
|
0.052
|
|
|
HbLow_12w = Yes (%)
|
15 (78.9)
|
21 (70.0)
|
0.719
|
|
|
NLR_12w.cut = ≥3 (%)
|
15 (78.9)
|
18 (60.0)
|
0.287
|
|
|
PLR_12w.cut = ≥150 (%)
|
13 (76.5)
|
18 (64.3)
|
0.600
|
|
|
BoneMet1 = Yes (%)
|
6 (31.6)
|
10 (32.3)
|
1.000
|
|
|
LiverMet1 = Yes (%)
|
4 (21.1)
|
2 ( 6.5)
|
0.274
|
|
|
irAE.L1 = Yes (%)
|
1 ( 5.3)
|
11 (35.5)
|
0.037
|
|
=======================================
PREDICTORS OF RESPONSE
=======================================
Here I’ve made a logistic regression to try to measure the
relationships found in the previous analysis.
I’ve taken a look without adjusting by IMDC risk.
As we are looking for “risk factors” (means Odds Ratio), the
analysis is made to predict “no ORR”.
We can change the language latter.
Predictor of DCR
irAE.L1
irAE after L1 have an OR 9,90 (CI 95%: 1.67-190) for DCR
## # A tibble: 2 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 1.11 0.325 0.324 0.746 0.587 2.12
## 2 irAE.L1Yes 9.90 1.09 2.10 0.0361 1.67 190.
=======================================
OVERALL SURVIVAL FIRST-LINE
=======================================
HR according to all variables
This is just to take a first look about what could be significant.
Focus on univariable analysis.
Remember we are not including the variable “IMDC risk” here.
ORR and DCR are significant (this is obvious). Then only Hb low and
NLR at week 12 are significant
This is just a first look. Then we will need to calculate the real
HR adjusting by risk.
|
|
Dependent: Survival
|
|
all
|
HR (univariable)
|
HR (multivariable)
|
|
17
|
ORR_12w
|
No
|
36 (72.0)
|
|
|
|
18
|
|
Yes
|
14 (28.0)
|
0.37 (0.14-0.98, p=0.046)
|
0.51 (0.07-3.56, p=0.496)
|
|
3
|
DCR_12w
|
No
|
19 (38.0)
|
|
|
|
4
|
|
Yes
|
31 (62.0)
|
0.26 (0.12-0.59, p=0.001)
|
0.03 (0.01-0.16, p<0.001)
|
|
23
|
PltHigh_0w
|
No
|
36 (72.0)
|
|
|
|
24
|
|
Yes
|
14 (28.0)
|
1.00 (0.44-2.28, p=0.993)
|
1.52 (0.30-7.72, p=0.611)
|
|
5
|
HbLow_0w
|
No
|
14 (28.0)
|
|
|
|
6
|
|
Yes
|
36 (72.0)
|
1.66 (0.63-4.41, p=0.305)
|
2.56 (0.51-12.77, p=0.251)
|
|
13
|
NLR_0w.cut
|
<3
|
13 (26.0)
|
|
|
|
14
|
|
≥3
|
37 (74.0)
|
1.65 (0.67-4.10, p=0.278)
|
9.43 (1.67-53.28, p=0.011)
|
|
19
|
PLR_0w.cut
|
<150
|
8 (17.0)
|
|
|
|
20
|
|
≥150
|
39 (83.0)
|
0.90 (0.34-2.42, p=0.842)
|
0.02 (0.00-0.27, p=0.002)
|
|
25
|
PltHigh_12w
|
No
|
37 (75.5)
|
|
|
|
26
|
|
Yes
|
12 (24.5)
|
1.55 (0.66-3.65, p=0.310)
|
0.75 (0.13-4.29, p=0.743)
|
|
7
|
HbLow_12w
|
No
|
13 (26.5)
|
|
|
|
8
|
|
Yes
|
36 (73.5)
|
2.58 (0.89-7.53, p=0.082)
|
2.70 (0.48-15.09, p=0.259)
|
|
15
|
NLR_12w.cut
|
<3
|
16 (32.7)
|
|
|
|
16
|
|
≥3
|
33 (67.3)
|
2.61 (1.04-6.53, p=0.041)
|
2.26 (0.65-7.87, p=0.200)
|
|
21
|
PLR_12w.cut
|
<150
|
14 (31.1)
|
|
|
|
22
|
|
≥150
|
31 (68.9)
|
1.17 (0.46-2.96, p=0.735)
|
3.32 (0.58-19.11, p=0.180)
|
|
1
|
BoneMet1
|
No
|
36 (69.2)
|
|
|
|
2
|
|
Yes
|
16 (30.8)
|
1.24 (0.57-2.72, p=0.585)
|
4.22 (1.00-17.77, p=0.049)
|
|
11
|
LiverMet1
|
No
|
45 (86.5)
|
|
|
|
12
|
|
Yes
|
7 (13.5)
|
1.13 (0.39-3.30, p=0.818)
|
0.39 (0.05-2.95, p=0.360)
|
|
9
|
irAE.L1
|
No
|
40 (76.9)
|
|
|
|
10
|
|
Yes
|
12 (23.1)
|
0.75 (0.28-1.98, p=0.557)
|
10.26 (1.37-76.61, p=0.023)
|
Differences in OS by ORR and DCR at week 12
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ ORR_12w, data = dbIOIO,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## ORR_12w=No 36 21 12.6 9.56 NA
## ORR_12w=Yes 14 5 NA 20.30 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ DCR_12w, data = dbIOIO,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## DCR_12w=No 19 15 9.56 8.25 18.6
## DCR_12w=Yes 31 11 NA 20.30 NA
Differences in OS by HbLow, PLR and NLR at week 0
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ HbLow_0w, data = dbIOIO,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## HbLow_0w=No 14 5 NA 9.76 NA
## HbLow_0w=Yes 36 22 14.8 9.56 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ PLR_0w.cut, data = dbIOIO,
## type = "kaplan-meier")
##
## 5 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## PLR_0w.cut=<150 8 5 14.3 11.83 NA
## PLR_0w.cut=≥150 39 20 18.6 9.76 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_0w.cut, data = dbIOIO,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_0w.cut=<3 13 6 26.8 14.26 NA
## NLR_0w.cut=≥3 37 21 12.6 8.64 NA
Differences in OS by HbLow, PLR and NLR at week 12
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ HbLow_12w, data = dbIOIO,
## type = "kaplan-meier")
##
## 3 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## HbLow_12w=No 13 4 NA 14.26 NA
## HbLow_12w=Yes 36 22 12.6 8.64 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ PLR_12w.cut, data = dbIOIO,
## type = "kaplan-meier")
##
## 7 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## PLR_12w.cut=<150 14 6 27.2 11.8 NA
## PLR_12w.cut=≥150 31 18 15.0 11.4 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_12w.cut, data = dbIOIO,
## type = "kaplan-meier")
##
## 3 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_12w.cut=<3 16 6 27.2 20.30 NA
## NLR_12w.cut=≥3 33 20 11.4 8.02 NA
Differences in OS by irAE at L1
This is also interesting. There is no significant difference between
the curves.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ irAE.L1, data = dbIOIO,
## type = "kaplan-meier")
##
## n events median 0.95LCL 0.95UCL
## irAE.L1=No 40 22 14.8 11.83 NA
## irAE.L1=Yes 12 5 27.2 7.85 NA
Differences in OS by NLR at week 0 + week 12
This is also interesting.
Patients who achieve a NLR < 3 at week 12 have better
prognosis.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_0w.cut + NLR_12w.cut,
## data = dbIOIO, type = "kaplan-meier")
##
## 3 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_0w.cut=<3, NLR_12w.cut=<3 7 2 NA 14.26 NA
## NLR_0w.cut=<3, NLR_12w.cut=≥3 6 4 18.60 11.40 NA
## NLR_0w.cut=≥3, NLR_12w.cut=<3 9 4 27.20 20.30 NA
## NLR_0w.cut=≥3, NLR_12w.cut=≥3 27 16 9.56 7.85 NA
Differences in OS by NLR at week 12 and risk
The importance of NLR at week 12 remains in IMDC risk
categories.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ risk + NLR_12w.cut,
## data = dbIOIO, type = "kaplan-meier")
##
## 3 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## risk=favourable, NLR_12w.cut=≥3 1 0 NA NA NA
## risk=intermediate, NLR_12w.cut=<3 11 3 NA 27.20 NA
## risk=intermediate, NLR_12w.cut=≥3 13 5 NA 8.64 NA
## risk=poor, NLR_12w.cut=<3 5 3 20.30 14.75 NA
## risk=poor, NLR_12w.cut=≥3 19 15 9.56 6.77 18.6
=======================================
=======================================
ANALYSIS OF IO + VEGF POPULATION
=======================================
=======================================
=======================================
BASELINE OUTCOMES
=======================================
|
Dependent: Line1_IOV
|
|
IO+V
|
|
Total N (%)
|
|
41 (100.0)
|
|
Best.Response
|
CR
|
3 (7.3)
|
|
|
PD
|
1 (2.4)
|
|
|
PR
|
26 (63.4)
|
|
|
SD
|
11 (26.8)
|
|
ORR_12w
|
No
|
21 (51.2)
|
|
|
Yes
|
18 (43.9)
|
|
|
(Missing)
|
2 (4.9)
|
|
DCR_12w
|
No
|
2 (4.9)
|
|
|
Yes
|
37 (90.2)
|
|
|
(Missing)
|
2 (4.9)
|
|
OSL1fu2
|
OS <2y
|
16 (39.0)
|
|
|
OS 2-5y
|
13 (31.7)
|
|
|
OS ≥5y
|
12 (29.3)
|
|
PFSL1fu2
|
PFS <2y
|
26 (63.4)
|
|
|
PFS 2-5y
|
10 (24.4)
|
|
|
PFS ≥5y
|
5 (12.2)
|
=======================================
ASOCIATIONS WITH RESPONSE IN L1 IO + VEGF POPULATION
=======================================
ORR No vs Yes at week 12 and all variables
No variable is associated with ORR
|
|
No
|
Yes
|
p
|
test
|
|
n
|
21
|
18
|
|
|
|
PltHigh_0w = Yes (%)
|
2 ( 9.5)
|
1 ( 5.6)
|
1.000
|
|
|
HbLow_0w = Yes (%)
|
6 (28.6)
|
8 (44.4)
|
0.487
|
|
|
NLR_0w.cut = ≥3 (%)
|
9 (42.9)
|
8 (44.4)
|
1.000
|
|
|
PLR_0w.cut = ≥150 (%)
|
15 (71.4)
|
9 (50.0)
|
0.298
|
|
|
PltHigh_12w = Yes (%)
|
4 (19.0)
|
0 ( 0.0)
|
0.154
|
|
|
HbLow_12w = Yes (%)
|
7 (33.3)
|
3 (16.7)
|
0.412
|
|
|
NLR_12w.cut = ≥3 (%)
|
7 (33.3)
|
4 (22.2)
|
0.680
|
|
|
PLR_12w.cut = ≥150 (%)
|
12 (60.0)
|
7 (38.9)
|
0.330
|
|
|
BoneMet1 = Yes (%)
|
5 (23.8)
|
4 (22.2)
|
1.000
|
|
|
LiverMet1 = Yes (%)
|
2 ( 9.5)
|
3 (16.7)
|
0.853
|
|
|
irAE.L1 = Yes (%)
|
5 (23.8)
|
2 (11.1)
|
0.541
|
|
DCR at week 12 and all variables
It’s quite similar to the previous analysis with ORR.
No variable is associated with ORR, but we need to considere that
just a few patient had PD.
|
|
No
|
Yes
|
p
|
test
|
|
n
|
2
|
37
|
|
|
|
PltHigh_0w = Yes (%)
|
0 ( 0.0)
|
3 ( 8.1)
|
1.000
|
|
|
HbLow_0w = Yes (%)
|
0 ( 0.0)
|
14 (37.8)
|
0.742
|
|
|
NLR_0w.cut = ≥3 (%)
|
2 (100.0)
|
15 (40.5)
|
0.358
|
|
|
PLR_0w.cut = ≥150 (%)
|
2 (100.0)
|
22 (59.5)
|
0.688
|
|
|
PltHigh_12w = Yes (%)
|
1 ( 50.0)
|
3 ( 8.1)
|
0.480
|
|
|
HbLow_12w = Yes (%)
|
1 ( 50.0)
|
9 (24.3)
|
1.000
|
|
|
NLR_12w.cut = ≥3 (%)
|
1 ( 50.0)
|
10 (27.0)
|
1.000
|
|
|
PLR_12w.cut = ≥150 (%)
|
1 (100.0)
|
18 (48.6)
|
1.000
|
|
|
BoneMet1 = Yes (%)
|
2 (100.0)
|
7 (18.9)
|
0.074
|
|
|
LiverMet1 = Yes (%)
|
0 ( 0.0)
|
5 (13.5)
|
1.000
|
|
|
irAE.L1 = Yes (%)
|
0 ( 0.0)
|
7 (18.9)
|
1.000
|
|
=======================================
PREDICTORS OF RESPONSE
=======================================
Here I’ve made a logistic regression to try to measure the
relationships found in the previous analysis.
There is no variables associated with ORR or DCR.
=======================================
OVERALL SURVIVAL FIRST-LINE
=======================================
HR according to all variables
This is just to take a first look about what could be significant.
Focus on univariable analysis.
Remember we are not including the variable “IMDC risk” here.
ORR and DCR are not significant (???).
Hb Low, Plt High, and NLR at week 0 are significant
PLt high and BoneMet at week 12 are significant. NLR at week 12
could be significant.
This is just a first look. Then we will need to calculate the real
HR adjusting by risk.
|
|
Dependent: Survival
|
|
all
|
HR (univariable)
|
HR (multivariable)
|
|
17
|
ORR_12w
|
No
|
21 (53.8)
|
|
|
|
18
|
|
Yes
|
18 (46.2)
|
0.70 (0.29-1.68, p=0.423)
|
0.43 (0.13-1.49, p=0.186)
|
|
3
|
DCR_12w
|
No
|
2 (5.1)
|
|
|
|
4
|
|
Yes
|
37 (94.9)
|
0.40 (0.09-1.74, p=0.221)
|
7.71 (0.35-169.44, p=0.195)
|
|
23
|
PltHigh_0w
|
No
|
36 (90.0)
|
|
|
|
24
|
|
Yes
|
4 (10.0)
|
9.32 (2.36-36.80, p=0.001)
|
2.01 (0.18-21.79, p=0.567)
|
|
5
|
HbLow_0w
|
No
|
25 (62.5)
|
|
|
|
6
|
|
Yes
|
15 (37.5)
|
2.19 (0.95-5.03, p=0.065)
|
2.33 (0.63-8.61, p=0.204)
|
|
13
|
NLR_0w.cut
|
<3
|
23 (57.5)
|
|
|
|
14
|
|
≥3
|
17 (42.5)
|
2.78 (1.19-6.53, p=0.019)
|
8.43 (1.56-45.47, p=0.013)
|
|
19
|
PLR_0w.cut
|
<150
|
15 (37.5)
|
|
|
|
20
|
|
≥150
|
25 (62.5)
|
1.48 (0.62-3.49, p=0.376)
|
0.49 (0.09-2.72, p=0.412)
|
|
25
|
PltHigh_12w
|
No
|
35 (89.7)
|
|
|
|
26
|
|
Yes
|
4 (10.3)
|
3.44 (0.99-11.92, p=0.051)
|
13.54 (0.96-191.05, p=0.054)
|
|
7
|
HbLow_12w
|
No
|
29 (74.4)
|
|
|
|
8
|
|
Yes
|
10 (25.6)
|
2.04 (0.79-5.25, p=0.140)
|
0.88 (0.14-5.47, p=0.889)
|
|
15
|
NLR_12w.cut
|
<3
|
28 (71.8)
|
|
|
|
16
|
|
≥3
|
11 (28.2)
|
1.86 (0.75-4.58, p=0.179)
|
0.48 (0.09-2.62, p=0.397)
|
|
21
|
PLR_12w.cut
|
<150
|
19 (50.0)
|
|
|
|
22
|
|
≥150
|
19 (50.0)
|
0.97 (0.41-2.30, p=0.952)
|
1.09 (0.24-4.91, p=0.910)
|
|
1
|
BoneMet1
|
No
|
32 (78.0)
|
|
|
|
2
|
|
Yes
|
9 (22.0)
|
2.50 (1.00-6.25, p=0.049)
|
3.44 (0.79-14.91, p=0.099)
|
|
11
|
LiverMet1
|
No
|
36 (87.8)
|
|
|
|
12
|
|
Yes
|
5 (12.2)
|
1.88 (0.63-5.65, p=0.259)
|
1.02 (0.20-5.36, p=0.978)
|
|
9
|
irAE.L1
|
No
|
34 (82.9)
|
|
|
|
10
|
|
Yes
|
7 (17.1)
|
0.59 (0.17-1.99, p=0.393)
|
0.34 (0.06-1.96, p=0.228)
|
Differences in OS by ORR and DCR at week 12
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ ORR_12w, data = dbIOV,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## ORR_12w=No 21 14 45.2 17.5 NA
## ORR_12w=Yes 18 8 41.3 31.5 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ DCR_12w, data = dbIOV,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## DCR_12w=No 2 2 26.8 6.83 NA
## DCR_12w=Yes 37 20 43.8 31.64 NA
Differences in OS by Hb Low, Plt High and NLR at week 0
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ HbLow_0w, data = dbIOV,
## type = "kaplan-meier")
##
## 1 observation deleted due to missingness
## n events median 0.95LCL 0.95UCL
## HbLow_0w=No 25 12 57.8 40.3 NA
## HbLow_0w=Yes 15 11 31.5 17.5 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ PltHigh_0w, data = dbIOV,
## type = "kaplan-meier")
##
## 1 observation deleted due to missingness
## n events median 0.95LCL 0.95UCL
## PltHigh_0w=No 36 20 45.37 34.760 NA
## PltHigh_0w=Yes 4 3 7.39 0.953 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_0w.cut, data = dbIOV,
## type = "kaplan-meier")
##
## 1 observation deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_0w.cut=<3 23 11 85.0 31.6 NA
## NLR_0w.cut=≥3 17 12 31.5 21.8 NA
Differences in OS by Hb Low, Plt High and NLR at week 12
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ HbLow_12w, data = dbIOV,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## HbLow_12w=No 29 16 46.7 34.8 NA
## HbLow_12w=Yes 10 6 21.8 14.6 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ PltHigh_12w, data = dbIOV,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## PltHigh_12w=No 35 19 45.4 34.76 NA
## PltHigh_12w=Yes 4 3 10.2 6.83 NA
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_12w.cut, data = dbIOV,
## type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_12w.cut=<3 28 15 46.7 40.3 NA
## NLR_12w.cut=≥3 11 7 21.8 14.7 NA
Differences in OS by NLR at week 0 + week 12
This is also interesting.
Patients who achieve a NLR < 3 at week 12 have better
prognosis.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ NLR_0w.cut + NLR_12w.cut,
## data = dbIOV, type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## NLR_0w.cut=<3, NLR_12w.cut=<3 19 9 85.0 41.3 NA
## NLR_0w.cut=<3, NLR_12w.cut=≥3 3 1 NA 17.5 NA
## NLR_0w.cut=≥3, NLR_12w.cut=<3 9 6 40.3 24.0 NA
## NLR_0w.cut=≥3, NLR_12w.cut=≥3 8 6 21.8 14.6 NA
Differences in OS by NLR at week 12 and risk
The importance of NLR at week 12 remains in all IMDC risk
categories.
## Call: survfit(formula = Surv(OSL1m, Ddeathcnsr) ~ risk + NLR_12w.cut,
## data = dbIOV, type = "kaplan-meier")
##
## 2 observations deleted due to missingness
## n events median 0.95LCL 0.95UCL
## risk=favourable, NLR_12w.cut=<3 7 2 NA 29.21 NA
## risk=intermediate, NLR_12w.cut=<3 19 11 46.7 40.31 NA
## risk=intermediate, NLR_12w.cut=≥3 8 5 31.5 14.69 NA
## risk=poor, NLR_12w.cut=<3 2 2 24.3 7.39 NA
## risk=poor, NLR_12w.cut=≥3 3 2 18.2 14.55 NA