Identification and characteristics of studies
We ultimately identified 22 independent studies published between March 2005 and March 2020 [9, 10, 12, 13, 16, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37]. A flow chart representing selection of studies is shown in Figure 1. Totally 11 RCTs, 10 non-randomized trials and 1 cohort study were included. The quality of 11 RCTs were all demonstrated with low risk (Grade A). The only one cohort study was also evaluated as high quality with full score. Based on the criteria of MINORS, five non-randomized trials scored zero point in sample size estimation criteria for bare report (Ibrahim NK 2005, Hurria A 2015, Blum JL 2007, Marschner N 2018, Bernardo A 2017), one trial scored one point for inadequate information (Yamamoto S 2017), and another one trial (Bernardo A 2017) also scored one point in follow-up domain for higher rate (>5%) of lost to follow-up. The details of study quality were shown in Table 1. The baseline characteristics of the included studies for metastatic breast cancer were summarized in Table 2.
Table 1 Quality assessment of the included studies
Author
|
Year
|
Cochrane
|
NOS
|
MINORS
non-control
|
MINORS
control
|
Gradishar WJ
|
2005
|
Grade A
|
|
|
|
GUAN ZZ
|
2009
|
Grade A
|
|
|
|
Gradishar WJ
|
2012
|
Grade A
|
|
|
|
Ranade AA
|
2013
|
Grade A
|
|
|
|
Andres FT
|
2015
|
Grade A
|
|
|
|
Jain MM
|
2016
|
Grade A
|
|
|
|
Tamura K
|
2017
|
Grade A
|
|
|
|
Gennari A
|
2018
|
Grade A
|
|
|
|
Ciruelos E
|
2019
|
Grade A
|
|
|
|
Hara F
|
2019
|
Grade A
|
|
|
|
Schmid P
|
2019
|
Grade A
|
|
|
|
Brezden B
|
2013
|
|
9
|
|
|
Ibrahim NK
|
2005
|
|
|
14
|
|
Mirtsching B
|
2011
|
|
|
16
|
|
Fabi A
|
2015
|
|
|
16
|
|
Hurria A
|
2015
|
|
|
16
|
|
Palumbo R
|
2015
|
|
|
16
|
|
Yamamoto S
|
2017
|
|
|
15
|
|
Hurria A
|
2019
|
|
|
14
|
|
Blum JL
|
2007
|
|
|
|
22
|
Bernardo A
|
2017
|
|
|
|
21
|
Marschner N
|
2018
|
|
|
|
22
|
Table 2 Baseline characteristics of the included studies for metastatic breast cancer
Author+Year
|
Study Design
|
Sample Size (N)
|
Treatment Line
|
Nab-paclitaxel Monotherapy Regimen
|
Treatment Related Adverse Events (TRAEs)
|
Efficacy Outcomes
|
Neutropenia (%)
|
Leukopenia
(%)
|
Neuropathy
(%)
|
Fatigue
(%)
|
ORR (%)
|
CBR
(%)
|
Median PFS
(Months)
|
Median OS
(Months)
|
All
grades
|
3/4 grade
|
All grades
|
3/4 grade
|
All
grades
|
3/4
grade
|
All grades
|
3/4 grade
|
Ibrahim NK 2005
|
Phase II
Multicenter
|
63
|
Mixed
|
300mg/m2 q3w
|
9
|
51
|
91
|
24
|
64
|
11
|
40
|
13
|
48
|
|
|
14.8
|
39
|
First
|
300mg/m2 q3w
|
|
|
|
|
|
|
|
|
64
|
|
|
|
24
|
Further
|
300mg/m2 q3w
|
|
|
|
|
|
|
|
|
21
|
|
|
|
Gradishar WJ
2005
|
Phase III
RCT
Multicenter
|
229
|
Mixed
|
260mg/m2 q3w
|
|
30
|
|
7
|
|
10
|
|
7
|
33
|
|
|
15.2
|
97
|
First
|
260mg/m2 q3w
|
|
|
|
|
|
|
|
|
42
|
|
|
|
132
|
Further
|
260mg/m2 q3w
|
|
|
|
|
|
|
|
|
27
|
|
|
13.2
|
Blum JL
2007
|
Phase II
Multicenter
|
106
|
Further
|
100mg/m2 qw 3/4
|
49
|
18
|
62
|
19
|
25
|
8
|
37
|
5
|
14
|
26
|
3
|
9.2
|
75
|
Further
|
125mg/m2 qw 3/4
|
64
|
34
|
66
|
36
|
51
|
19
|
45
|
12
|
16
|
37
|
3.5
|
9.1
|
Guan ZZ
2009
|
Phase II
RCT
Multicenter
|
104
|
Further
|
260mg/m2 q3w
|
69
|
42
|
64
|
24
|
76
|
7
|
17
|
|
54
|
71
|
7.6
|
17.8
|
Mirtsching B
2011
|
Phase II
Multicenter
|
72
|
First
|
125mg/m2 qw 3/4
|
14
|
11
|
11
|
1.4
|
54
|
8.3
|
58
|
6.9
|
42.2
|
68.8
|
14.5
|
29
|
Gradishar WJ
2009/2012
|
Phase II
RCT
Multicenter
|
76
|
First
|
100mg/m2 qw 3/4
|
80
|
25
|
|
|
58
|
9
|
34
|
0
|
63
|
83
|
7.5
|
22.2
|
74
|
First
|
150mg/m2 qw 3/4
|
92
|
44
|
|
|
68
|
22
|
47
|
4
|
74
|
91
|
14.6
|
33.8
|
76
|
First
|
300mg/m2 q3w
|
93
|
44
|
|
|
73
|
21
|
36
|
5
|
46
|
72
|
10.9
|
27.7
|
Brezden B
2013
|
Phase II
Multicenter
|
47
|
First
|
100mg/m2 qw 3/4
|
14.9
|
9
|
|
|
46.8
|
0
|
70
|
|
30
|
51
|
6
|
20.9
|
76
|
First
|
100mg/m2 qw 3/4
|
17.1
|
9
|
|
|
57.9
|
0
|
72
|
|
28
|
57
|
6.7
|
20
|
Ranade AA
2013
|
Phase II
RCT
Multicenter
|
55
|
Mixed
|
220mg/m2 q3w
|
|
28
|
|
|
|
1.5
|
|
|
40
|
67
|
|
|
53
|
Mixed
|
300mg/m2 q3w
|
|
38
|
|
|
|
17
|
|
|
40
|
72
|
|
|
Palumbo R
2015
|
Phase II
Multicenter
|
52
|
Further
|
260mg/m2 q3w
|
77
|
21.2
|
88.5
|
25
|
48.1
|
5.8
|
27
|
0
|
48.1
|
76.9
|
8.9
|
|
Fabi A
2015
|
Phase II
Single-center
|
42
|
Further
|
Mixed
|
95
|
70
|
95
|
25
|
80
|
12
|
72
|
13
|
23.8
|
50
|
4.6
|
|
32
|
Further
|
125mg/m2 qw 3/4
|
|
|
|
|
|
|
|
|
28.1
|
50
|
|
|
10
|
Further
|
260mg/m2 q3w
|
|
|
|
|
|
|
|
|
10
|
50
|
|
|
Hurria A
2015
|
Phase II
Multicenter
|
39
|
Mixed
|
100mg/m2 qw 3/4
|
26
|
3
|
36
|
5
|
8
|
0
|
31
|
5
|
31
|
69
|
5.7
|
19.4
|
Andres FT
2015
|
Phase II
RCT
Multicenter
|
21
|
Mixed
|
100mg/m2 qw 3/4
|
38
|
14
|
|
|
48
|
5
|
54
|
10
|
38
|
52
|
3.7
|
|
Jain MM
2016
|
Phase III
RCT
Multicenter
|
58
|
Mixed
|
260mg/m2 q3w
|
33
|
21
|
28
|
16
|
60
|
17
|
36
|
7
|
43
|
75
|
7.9
|
|
Yamamoto S
2017
|
Phase II
Multicenter
|
35
|
Mixed
|
180mg/m2 q3w
|
|
46
|
|
|
|
0
|
|
|
23
|
51
|
6.5
|
44.7
|
Tamura K
2017
|
Phase II
RCT
Multicenter
|
98
|
First
|
150mg/m2 qw 3/4
|
97
|
78
|
96
|
58
|
88
|
22
|
33
|
1
|
61.2
|
96.9
|
11.2
|
42.4
|
Bernardo A
2017
|
Cohort
|
209
|
Further
|
Mixed
|
9.4
|
3
|
|
|
29.9
|
2.1
|
27.4
|
1.7
|
32.1
|
57.7
|
|
18
|
68
|
Further
|
125mg/m2 qw 3/4
|
8.7
|
1.1
|
|
|
17.4
|
2.2
|
12
|
1.1
|
31.5
|
54.3
|
|
16.9
|
121
|
Further
|
260mg/m2 q3w
|
9.9
|
4.2
|
|
|
38
|
2.1
|
37.3
|
2.1
|
32.4
|
59.9
|
|
18
|
Gennari A
2018
|
Phase II
RCT
Multicenter
|
83
|
First
|
150mg/m2 q2w
|
55.4
|
21.7
|
|
|
69.5
|
8
|
|
48.2
|
47
|
65.1
|
7.9
|
25.8
|
86
|
First
|
100mg/m2 qw 3/4
|
68.6
|
27.9
|
|
|
73.2
|
5.8
|
|
46.5
|
54.7
|
68.6
|
9
|
26.2
|
86
|
First
|
75mg/m2 qw
|
72.2
|
24.4
|
|
|
70
|
5.8
|
|
10.5
|
44.7
|
60.5
|
8.5
|
25.5
|
Marschner N
2018
|
Cohort
Multicenter
|
697
|
Mixed
|
Mixed
|
|
4
|
|
7.5
|
39.6
|
4.3
|
20.8
|
1.3
|
37.2
|
68.3
|
5.9
|
15.6
|
491
|
Mixed
|
≤150mg/m2
qw
|
|
|
|
|
|
|
|
|
39.1
|
68.8
|
6
|
16.3
|
194
|
Mixed
|
220-260mg/m2
q3w
|
|
|
|
|
|
|
|
|
33
|
67.5
|
5.7
|
15.1
|
Hurria A
2019
|
Phase II
Multicenter
|
40
|
Mixed
|
100mg/m2 qw 3/4
|
44
|
11
|
18
|
3
|
10
|
5
|
55
|
5
|
35
|
75
|
6.5
|
21.2
|
Ciruelos E
2019
|
Phase II
RCT
Multicenter
|
16
|
First
|
100mg/m2 qw 3/4
|
37.5
|
0
|
50.1
|
6.3
|
81.3
|
0
|
43.9
|
6.3
|
37.5
|
|
|
|
14
|
First
|
150mg/m2 q2w
|
18.8
|
0
|
25.1
|
6.3
|
62.6
|
0
|
87.6
|
0
|
12.5
|
|
|
|
16
|
First
|
150mg/m2 qw 3/4
|
64.2
|
50
|
50
|
28.6
|
78.6
|
35.7
|
64.3
|
14.3
|
42.9
|
|
|
|
Hara F
2019
|
Phase II
RCT
Multicenter
|
48
|
Mixed
|
180mg/m2 q3w
|
50
|
14.6
|
60.4
|
14.6
|
81.3
|
8.3
|
70.8
|
0
|
37.8
|
|
6.8
|
|
45
|
Mixed
|
220mg/m2 q3w
|
73.3
|
37.7
|
77.8
|
26.6
|
84.4
|
8.9
|
77.8
|
0
|
44.1
|
|
7.3
|
|
47
|
Mixed
|
260mg/m2 q3w
|
57.4
|
25.4
|
66
|
19.1
|
91.5
|
31.9
|
80.9
|
2.1
|
48.7
|
|
6.7
|
|
Schmid P
2018/2019
|
Phase III
RCT
Multicenter
|
449
|
First
|
100mg/m2 qw 3/4
|
15
|
8.2
|
|
|
23
|
2.7
|
44
|
3
|
45.9
|
72.4
|
5.5
|
18.7
|
Nab-paclitaxel treatment patterns
Totally 3287 MBC patients treated with nab-paclitaxel monotherapy were included in the current study. In our analysis, there were 1685 (51.26%) MBC patients who received nab-paclitaxel as first-line therapy, 640 patients (19.47%) as further line therapy, and the rest 962 MBC patients (29.27%) as mixed (first or further) line therapy. Furthermore, a majority of MBC patients (n=1966, 60.40%) had nab-paclitaxel administered weekly, 1190 MBC patients (36.56%) had nab-paclitaxel administered triweekly, and 99 MBC patients (3.04%) with biweekly 150mg/m2 nab-paclitaxel administration. Among 1190 MBC patients with triweekly nab-paclitaxel schedule (q3w), despite of 194 patients (16.30%) being administered at an imprecise reported dosage of 220-260 mg/m2, 192 patients (16.13%) were administered at the dosage of 300 mg/m2, 621 patients (52.18%) were at 260 mg/m2, 100 patients (8.40%) were at 220 mg/m2, and 83 patients (6.97%) were at 180 mg/m2. Among 1966 MBC patients with weekly nab-paclitaxel schedule, 956 patients (48.62%) were administered at the dosage of 100mg/m2 qw 3/4, 247 patients (12.56%) were at 125mg/m2 qw 3/4, 186 patients (9.5%) were at 150mg/m2 qw 3/4, and 86 patients (4.4%) were at 75mg/m2 qw. Still, 491 MBC patients (24.97%) with weekly nab-paclitaxel monotherapy in the NABUCCO study were reported with administration at ≤150mg/m2 qw schedule.
Nab-paclitaxel monotherapy safety profiles
All 3287 MBC patients were included in the safety analysis. Neutropenia, leukopenia, peripheral sensory neuropathy and fatigue were the four chosen representative TRAEs of nab-paclitaxel monotherapy.
According to our study, 22 included studies with 31 individual groups reported the incidence of treatment related neutropenia after nab-paclitaxel monotherapy in 3287 MBC patients. After data integration, the overall incidence of all grades neutropenia was 52% (95% CI, 38%-66%), and the incidence of grade 3/4 neutropenia was 24% (95% CI, 16%-32%) (Figure 2A2B). 19 individual groups reported the incidence of chemotherapy related leukopenia with the overall incidence of all-grades leukopenia was 58% (95% CI, 43%-73%), and the incidence of grade 3/4 leukopenia was 17% (95% CI, 11%-24%) (Figure 2C2D). Together 3287 MBC patients in 31 individual groups reported the incidence of all grades and grade 3/4 peripheral sensory neuropathy. The overall incidence of all grades peripheral sensory neuropathy was 58% (95% CI, 48%-68%), and the incidence of grade 3/4 peripheral sensory neuropathy was only 8% (95% CI, 5%-10%) (Figure 2E2F). 27 individual groups reported the incidence of treatment related fatigue. The overall incidence of all grades fatigue was 49% (95% CI, 41%-56%), and the incidence of grade 3/4 fatigue was 5% (95% CI, 2%-8%) following nab-paclitaxel administration (Figure 2G2H).
Risk factors affecting TRAEs
According to the results of meta-regression analysis, treatment lines, nab-paclitaxel dosage, nab-paclitaxel schedule did not affect the incidence of all grades neutropenia and grade 3/4 neutropenia. However, grade 3/4 neutropenia occured less frequently in Her-2 negative patients compared with all population (Coef value=0.063, P=0.046). All grades leukopenia seemed to occur more frequently in MBC patients treated with nab-paclitaxel as further-line therapy (Coef value=0.366, P=0.056). Treatment lines, patient population, the schedule of nab-paclitaxel administration did not contribute to the development of all grades and grade 3/4 peripheral sensory neuropathy. However, the dosage of nab-paclitaxel monotherapy seemed to be a potential independent risk factor affecting the incidence of grade 3/4 peripheral sensory neuropathy (Coef value=0.201, P=0.078) as P value almost reached 0.05. Meanwhile, obvious clinical trend can be noted that grade 3/4 peripheral sensory neuropathy were more frequently recorded in patients with higher nab-paclitaxel dosage group compared with patients with lower dosage group (Figure 3A). Nab-paclitaxel related all grades fatigue occurred more frequently in MBC patients with further-line monotherapy (Coef value=-0.239, P=0.032).
Nab-paclitaxel monotherapy efficacy outcomes
All 3287 MBC patients were included in the efficacy analysis. ORR, CBR, PFS, and OS were the chosen efficacy endpoints index of nab-paclitaxel monotherapy.
33 individual groups reported the ratio of ORR as the major efficacy of nab-paclitaxel monotherapy. After data integration, the cumulative ratio of ORR was 40% (95% CI, 35%-45%). 25 individual groups reported the ratio of CBR with the cumulative ratio being 66% (95% CI, 59%-73%). Additionally, complete remission (CR) was noted in 23 individual groups and the cumulative ratio of CR was only 3% (95% CI, 2%-5%). Partial remission (PR) and stable disease (SD) were higher than CR with the cumulative ratio reaching 38% (95% CI, 32%-44%) and 28% (95% CI, 24%-31%), respectively.
23 individual groups with 2399 MBC patients reported the outcome of PFS after nab-paclitaxel monotherapy. The median PFS ranged from 3.7 to 14.6 months and the overall median PFS was 7.64 months (95% CI, 6.89-8.40 months). The outcome of OS was reported in 14 studies with 17 individual groups containing 2472 MBC patients. The median OS ranged from 15.2 to 44.7 months and the overall median OS was 24.51 months (95% CI, 21.25-27.78 months).
Risk factors affecting efficacy outcomes
According to the meta-regression analysis, we found that patients treated with nab-paclitaxel monotherapy in further line would suffer from unfavorable lower ORR (Coef value=-0.18, P=0.006) compared with the patients in other lines. In the subsequent subgroup analysis, the value of ORR was 48.2% (95% CI, 41%-45%) in patients treated with nab-paclitaxel in the first line, while 40.1% (95% CI, 35.1%-45%) in the mixed line (first or further line), and 27.2% (95% CI, 20.1%-34.3%) in the further line. Similar result and statistical significance were also obtained concerning to CBR (Coef value=-0.176, P=0.037), with CBR valued 55.3% (95% CI, 43.2%-67.4%) in the further line, 68.7% (95% CI, 62.5%-75%) in the first line, and 68.8% (95% CI, 63.5%-74%) in the mixed line. The schedule of nab-paclitaxel administration (weekly, biweekly or triweekly) did not affect the ORR (Coef value=-0.212, P=0.116). However, under the same schedule of nab-paclitaxel administration, patients appeared to have more superior ORR along with the increasing dosage of nab-paclitaxel (Coef value=0.081, P=0.044) (Figure 3B).
Patients who received first line nab-paclitaxel monotherapy was demonstrated to have longer median PFS versus mixed line therapy. The median PFS was 8.01 months (95% CI, 6.83-9.18 months) in the first line group and 6.55 months (95% CI, 5.69-7.4 months) in the mixed line group, respectively. Patients who received further line nab-paclitaxel monotherapy was demonstrated to have rather shorter median OS versus first and mixed line therapy. Median OS values were 16.18 months (95% CI, 13.41-18.94 months), 24.41 months (95% CI, 20.4-28.42 months), and 28.44 months (95% CI, 12.75-44.13 months) by further, first, and mixed line therapy, respectively. Similarly, just like the ORR, the schedule of nab-paclitaxel administration (weekly, biweekly or triweekly) did not affect the median PFS (Coef value=-2.77, P=0.162) and median OS (Coef value=6.27, P=0.623). Median PFS was 6.94 months (95% CI, 5.92-7.96 months) for weekly and 7.71 months (95% CI, 6.90-8.52 months) for q3w nab-paclitaxel monotherapy. However, under the same schedule of nab-paclitaxel administration, patients appeared to have longer PFS along with the increasing dosage of nab-paclitaxel (Coef value=2.68, P=0.03). This significance was not found with respect to the median OS (Coef value=6.27, P=0.62). Risk factors affecting nab-paclitaxel monotherapy related adverse events and efficacy outcomes were presented in Table 3.
Table 3 Risk factors affecting nab-paclitaxel monotherapy related adverse events and efficacy outcomes by meta-regression analysis
Effect Index
|
Risk Factors
|
Coef. value
|
Std. Err.
|
T value
|
P value
|
95% Confidence Interval
|
Incidence of neutropenia
3/4 grade
|
Her2 negative
|
-0.136
|
0.063
|
-2.14
|
0.046
|
-0.268 to -0.003
|
Incidence of leukopenia
all grades
|
Treatment line
|
0.366
|
0.148
|
2.48
|
0.056
|
-0.014 to 0.746
|
Incidence of neuropathy
3/4 grade
|
Nab-paclitaxel dosage
|
0.201
|
0.107
|
1.87
|
0.078
|
-0.025 to 0.427
|
Incidence of fatigue
all grades
|
Treatment line
|
-0.239
|
0.100
|
-2.40
|
0.032
|
-0.455 to -0.024
|
Overall response rate
(ORR)
|
Treatment line
|
-0.180
|
0.059
|
-3.03
|
0.006
|
-0.302 to -0.058
|
Nab-paclitaxel dosage
|
0.171
|
0.081
|
2.11
|
0.044
|
0.005 to 0.338
|
Clinical benefit rate
(CBR)
|
Treatment line
|
-0.176
|
0.077
|
-2.29
|
0.037
|
-0.340 to -0.012
|
Progression free survival
(PFS)
|
Treatment line
|
1.398
|
0.635
|
2.20
|
0.045
|
0.036 to 2.760
|
Nab-paclitaxel dosage
|
2.683
|
1.114
|
2.41
|
0.030
|
0.295 to 5.071
|
Overall survival
(OS)
|
Treatment line
|
-18.909
|
8.210
|
-2.30
|
0.040
|
-36.797 to -1.021
|