A total of 157 patients who met the inclusion criteria were enrolled in the study. Forty-nine patients (31.2%) experienced FN recurrence. Baseline characteristics of the study population are shown in Table 1.
Table 1. Baseline characteristics of the patients (n=157)
|
N (%)
|
Age at initial FN
|
54.9 ± 14.1
|
Cancer types
|
|
Cervix
|
34 (21.7)
|
Ovary, tube, and primary peritoneum
|
77 (49.0)
|
Corpus
|
44 (28.0)
|
Unknown
|
1 (0.6)
|
Double primary (corpus and ovary)
|
1 (0.6)
|
Grade 4 neutropenia event before initial FN
|
64 (40.8)
|
Previous lines of chemotherapy
|
|
0
|
83 (52.9)
|
1
|
32 (20.4)
|
≥2
|
42 (26.8)
|
Chemotherapy regimen
|
|
Combination
|
118 (75.2)
|
Single
|
26 (16.6)
|
Combination + radiation
|
13 (8.3)
|
ANC at initial FN
|
174.6 ± 144.8
|
Grade 3 or 4 anemia at initial FN
|
38 (24.2)
|
Grade 3 or 4 thrombocytopenia at initial FN
|
48 (30.6)
|
Bacteremia at initial FN
|
27/143 (18.9)
|
Septic shock at initial FN
|
12 (7.6)
|
G-CSF usage at initial FN
|
|
Total dose (µg)
|
1538.9 ± 1226.8
|
Max. daily dose (µg)
|
482.5 ± 258.3
|
Duration (day)
|
3.8 ± 2.4
|
Type of G-CSF
|
|
Filgrastim
|
138 (87.9)
|
Lenograstim
|
16 (10.2)
|
Both (filgrastim, lenograstim)
|
2 (1.3)
|
No G-CSF
|
1 (0.6)
|
Recurrence of FN
|
49 (31.2)
|
Values are presented as mean±standard deviation or number (%).
ANC, absolute neutrophile count; FN, febrile neutropenia; G-CSF, granulocyte colony stimulating factor
|
The mean age at the time of initial FN was 54.9±14.1 years. In terms of cancer types of the patients, the most common cancer was ovary, tubal, and primary peritoneal cancer (n=77, 49.0%), followed by corpus (n=44, 28.0%) and cervix (n=34, 21.7%). Sixty-four patients (40.8%) experienced grade 4 neutropenia without fever before the initial FN event. Eighty-three patients (52.9%) had the first episode of FN during first-line chemotherapy without a previous history of chemotherapy. Thirty-two (20.4%) patients had a previous history of chemotherapy with one line, and 42 (26.8%) had two or more lines of chemotherapy. Regarding the chemotherapy regimen used at the time of initial FN, 118 (75.2%) and 26 (16.6%) received combination and single-regimen chemotherapy, respectively. Thirteen patients (8.3%) received radiotherapy with combination chemotherapy. The mean ANC level at diagnosis of initial FN was 174.6 ± 144.8 /mcl. Grade 3 or more anemia and thrombocytopenia at the initial FN event occurred in 38 (24.2%) and 48 women (30.6%), respectively. During the initial FN event, 27 (18.9%) patients had bacteremia and 12 (7.6%) had septic shock. Regarding the G-CSF type, most of the patients (n=138, 87.9%) used filgrastim.
Clinical and laboratory characteristics were compared between the recurrence of FN (+) and (-) (Table 2). Age≥55 years (p=0.043), previous lines of chemotherapy≤1 (p=0.002), total dose (p=0.003), and maximum daily dose (p=0.009) of G-CSF were significantly associated with recurrence of FN. The frequency of grade 3 or 4 thrombocytopenia was significantly lower in the group with recurrence of FN. However, the use of prophylactic G-CSF prior to initial FN, ANC level at the time of initial FN, grade 3 or 4 anemia, septic shock during initial FN event, CCRT, ANC after administration of G-CSF, and total duration of G-CSF usage did not differ between the two groups.
Table 2
Clinical and laboratory factors according to recurrence of febrile neutropenia
|
Recurrence of FN (-) (n=108)
|
Recurrence of FN (+) (n=49)
|
p
|
Age at initial FN ≥ 55 (year)
|
54 (50.0)
|
33 (67.3)
|
0.043
|
Previous line of chemotherapy
|
|
|
0.002
|
≤1
|
71 (65.7)
|
44 (89.8)
|
|
≥2
|
37 (34.3)
|
5 (10.2)
|
|
Concurrent chemoradiation
|
8 (7.4)
|
5(10.2)
|
0.556
|
Prophylactic G-CSF before initial FN
|
35 (32.4)
|
12 (24.5)
|
0.316
|
Blood count at initial FN
|
|
|
|
ANC (per µL)
|
177.5±145.9
|
168.3±143.5
|
0.713
|
Anemia≥ Grade 3
|
24 (22.2)
|
14 (28.6)
|
0.389
|
Thrombocytopenia≥ Grade 3
|
39 (36.1)
|
9(18.4)
|
0.025
|
Complication at initial FN
|
|
|
|
Septic shock
|
9 (8.3)
|
3 (6.1)
|
0.755
|
Bacteremia
|
21/98 (21.4)
|
6/45 (13.3)
|
0.251
|
G-CSF usage at initial FN
|
|
|
|
Total dose (µg)
|
1715.7±1290.0
|
1149.0±997.6
|
0.003
|
Duration (day)
|
4.0±2.4
|
3.2±2.3
|
0.057
|
Maximum daily dose (µg)
|
515.7±271.6
|
409.2±210.8
|
0.009
|
≤600
|
87 (80.6)
|
47 (95.9)
|
0.013
|
>600
|
21 (19.4)
|
2 (4.1)
|
|
Target ANC at (per µL)
|
|
|
|
Day of last G-CSF dose
|
2756.7±4012.6
|
3572.9±5616.5
|
0.373
|
Next day of last G-CSF dose
|
6749.1±5826.5
|
7217.9±8794.1
|
0.720
|
Values are presented as mean±standard deviation or number (%).
ANC, absolute neutrophile count; FN, febrile neutropenia; G-CSF, granulocyte colony stimulating factor
|
The results of univariate and multivariate Cox regression analyses of risk factors for recurrence of FN are shown in Table 3. The cutoff values of total dose and maximum daily dose of G-CSF used during the initial FN period were 2000 µg and 600 µg, respectively. In the univariate analysis, age≥55 years (HR, 2.03; 95% CI, 1.02-4.18; p=0.045), previous lines of chemotherapy≤1 (HR, 4.59; 1.68-12.55; p=0.003), total dose of G-CSF <2000 µg (HR, 3.29; 95% CI, 1.28-8.48; p=0.014), and maximum daily dose of G-CSF ≤600 µg (HR, 5.67; 95% CI 1.27-25.25; p=0.023) were associated with recurrent FN. However, grade 3 or 4 thrombocytopenia was associated with a decreased risk of FN recurrence (HR, 0.40; 95% CI, 0.18-0.91; p=0.028). Multivariate regression analysis showed that age≥55 years (HR, 2.42; 95% CI, 1.14-5.14; p=0.022), previous lines of chemotherapy≤1 (HR, 4.01; 95% CI, 1.40-11.55; p=0.010), and maximum daily dose of G-CSF≤ 600 µg (HR, 5.18; 95% CI, 1.12-24.02; p=0.036) were independent risk factors for recurrent FN.
Table 3
Univariate and multivariate logistic regression analysis of risk factors for recurrence of febrile neutropenia
|
N (%)
|
Univariate
|
Multivariate
|
HR
|
95% CI
|
P
|
HR
|
95% CI
|
P
|
Age at initial FN (year)
|
|
|
|
|
|
|
|
<55
|
70 (44.6)
|
1
|
|
|
1
|
|
|
≥55
|
87 (55.4)
|
2.03
|
1.02-4.18
|
0.045
|
2.42
|
1.14-5.14
|
0.022
|
Previous lines of chemotherapy
|
|
|
|
|
|
|
|
>1
|
42 (26.8)
|
1
|
|
|
1
|
|
|
≤1
|
115 (73.2)
|
4.59
|
1.68-12.55
|
0.003
|
4.01
|
1.40-11.55
|
0.010
|
Grade 3 or 4 thrombocytopenia
|
|
|
|
|
|
|
|
No
|
109 (69.4)
|
1
|
|
|
1
|
|
|
Yes
|
48 (30.6)
|
0.40
|
0.18-0.91
|
0.028
|
0.50
|
0.21-1.21
|
0.126
|
Total dose of G-CSF (µg)
|
|
|
|
|
|
|
|
≥2000
|
40 (25.5)
|
1
|
|
|
1
|
|
|
<2000
|
117 (74.5)
|
3.29
|
1.28-8.48
|
0.014
|
2.20
|
0.81-6.00
|
0.123
|
Max. daily dose of G-CSF (µg)
|
|
|
|
|
|
|
|
>600
|
23 (14.6)
|
1
|
|
|
1
|
|
|
≤600
|
134 (85.4)
|
5.67
|
1.27-25.25
|
0.023
|
5.18
|
1.12-24.02
|
0.036
|
CI, confidence interval; FN, febrile neutropenia; G-CSF, granulocyte colony stimulating factor; HR, hazard ratio
|
In the multivariate Cox regression analysis, the maximum daily dose of G-CSF was the only independent risk factor for RFS after adjusting for other relevant factors in the univariate analysis (Table 4). A maximum daily dose of G-CSF ≤ 600 µg was significantly associated with the short RFS of FN, with an HR of 4.33 (95% CI, 1.04-17-92; p=0.043).
Table 4
Univariate and multivariate Cox-regression analysis of risk factors for recurrence-free survival of febrile neutropenia
|
N (%)
|
Univariate
|
Multivariate
|
HR
|
95% CI
|
P
|
HR
|
95% CI
|
P
|
Age at initial FN (year)
|
|
|
|
|
|
|
|
<55
|
70 (44.6)
|
1
|
|
|
1
|
|
|
≥55
|
87 (55.4)
|
1.73
|
0.95-3.15
|
0.073
|
1.72
|
0.94-3.13
|
0.078
|
Previous lines of chemotherapy
|
|
|
|
|
|
|
|
>1
|
42 (26.8)
|
1
|
|
|
1
|
|
|
≤1
|
115 (73.2)
|
2.56
|
1.01-6.54
|
0.049
|
2.14
|
0.82-5.59
|
0.119
|
Grade 3 or 4 thrombocytopenia
|
|
|
|
|
|
|
|
No
|
109 (69.4)
|
1
|
|
|
1
|
|
|
Yes
|
48 (30.6)
|
0.60
|
0.29-1.25
|
0.171
|
0.69
|
0.33-1.45
|
0.322
|
Total dose of G-CSF (µg)
|
|
|
|
|
|
|
|
≥2000
|
40 (25.5)
|
1
|
|
|
1
|
|
|
<2000
|
117 (74.5)
|
2.46
|
1.05-5.78
|
0.039
|
1.82
|
0.75-4.43
|
0.185
|
Max. daily dose of G-CSF (µg)
|
|
|
|
|
|
|
|
>600
|
23 (14.6)
|
1
|
|
|
1
|
|
|
≤600
|
134 (85.4)
|
5.10
|
0.24-20.99
|
0.024
|
4.33
|
1.04-17.92
|
0.043
|
CI, confidence interval; FN, febrile neutropenia; G-CSF, granulocyte colony stimulating factor; HR, hazard ratio
|
However, there were no significant differences in both PFS (p=0.807) and OS (p=0.699) between the maximum daily dose of G-CSF ≤ 600 µg and >600 µg, as shown in Figure 1.
Additional analysis was performed to investigate the association between the maximal target level of ANC and RFS of FN. There were no significant differences in the RFS of FN according to any cutoff levels (3000/µL, 5000/µL, and 10000/µL) of the target ANC level of therapeutic G-CSF administration, as shown in Figure 2.
Table 5 shows that the side effects of G-CSF, such as bone pain, diarrhea, and thrombosis, were not different between the maximum daily dose of G-CSF ≤ 600 μg and >600 μg. For patients who had a first recurrence of FN, the recovery time from the first recurrence of FN, which was defined as the days from diagnosis of FN to ANC≥1500/μL, was not different between the two groups (4.12±2.26 vs. 3.00±2.83; p=0.500). Serious complications, such as splenic rupture or acute myeloid leukemia (AML)/myelodysplastic syndrome (MDS), were not observed during the study period.
Table 5. Recovery time from recurrence of FN and side effects of G-CSF according to maximum daily dose of G-CSF at initial FN
|
MDD ≤600 µg (n=133)
|
MDD >600 µg (n=23)
|
p
|
Recovery time from 1st recurrence of FN*
|
4.12±2.26 (n=42)
|
3.00±2.83 (n=2)
|
0.500
|
Side effects of G-CSF
|
|
|
|
Pain
|
51 (38.3)
|
5 (21.7)
|
0.125
|
Diarrhea
|
17 (12.8)
|
4 (17.4)
|
0.517
|
Thrombosis
|
2 (1.5)
|
1 (4.3)
|
0.382
|
Values are presented as mean±standard deviation or number (%).
*The days from diagnosis of FN to ANC≥1500/µL
Of 49 patients with recurrent FN, three who died at the second FN event and one who was discharged with ANC 1029 were excluded from analysis.
FN, febrile neutropenia; G-CSF, granulocyte colony stimulating factor; MDD, maximum daily dose
|