Keywords
platinum-resistant ovarian cancer - treatment - outcomes - systemic immune-inflammation
index - neutrophil to lymphocyte ratio
Majority (70–80%) of the epithelial ovarian cancers (EOC) recur with current therapy.[1] Patients who progress within 6 months of platinum-based chemotherapy are considered
as having platinum-resistant ovarian cancer (PROC) disease and have a very poor prognosis.
The median progression-free survival (PFS) in PROC is 3 to 4 months, and the median
overall survival (OS) is 1 year.[2] Other than “platinum-refractory disease” (progression during or within 4 weeks of
platinum-based therapy [median OS: 3–5 months]), few factors have been consistently
associated with prognosis in PROC. We explored the prognostic impact of simple indices
that reflect the immunological milieu (neutrophils to lymphocyte ratio [NLR] and systemic
immune-inflammation [SII]) in patients with PROC. Inflammatory indices are prognostic
in ovarian cancer (newly diagnosed and platinum-sensitive recurrence), but there are
no reports in patients with PROC.[3]
[4] After obtaining approval from the Institutional Ethics Committee (EC Approval No:
JIP/IEC/2019/558), data of patients diagnosed with PROC between January 1, 2015 and
December 31, 2019 was collected. The diagnosis of relapse could be based on the elevation
of CA-125 or symptoms/imaging findings. PFS was defined from the start of treatment
of PROC until progression or death due to any cause. SII (platelet count × neutrophil
count)/lymphocyte count and NLR (absolute neutrophil count/absolute lymphocyte count)
were calculated. Their median values were used to divide patients into high and low
categories.
Forty-nine patients who had started treatment for PROC (n = 21 with “refractory” disease) were included in this analysis ([Fig. 1]). The median interval between the last platinum treatment to relapse was 3.2 (2.1–4.6)
months. All had undergone surgery during initial treatment, either upfront (n = 9, 18%) or interval (n = 40, 82%). At the time of diagnosis of PROC, 19 (39%) patients were symptomatic,
2 (4%) had isolated elevation of CA-125, and 28 (57%) had elevated CA-125 with abnormal
imaging. For resistant disease, the majority received only chemotherapy (n = 45 [91%]), while few underwent additional surgery (n = 4[8%]). Most received single-agent chemotherapy (n = 27), while a few received doublets (n = 22). The median number of chemotherapy cycles was 4 (range: 2–6). The overall response
rate (ORR) was 21%. After a median follow-up of 3 (range: 2–7) months, 33 (67%) patients
progressed, and 25 (57%) had died. During secondary progression, 25 patients (76%)
were symptomatic, 4 (12%) had an elevation of CA-125, and 4 (12%) had both elevations
of CA-125 and radiological imaging ([Supplementary Table S1] [online only]).
Fig. 1 The Kaplan–Meier analysis of progression-free survival (PFS) (A) and overall survival (OS) (B) in patients with platinum-resistant ovarian cancer (PROC). Comparison of PFS between
patients with high and low neutrophils to lymphocyte ratio (NLR) (C) and systemic immune-inflammation index (SII) (D). Comparison of the OS between patients with high and low SII (E).
The median PFS and OS were 4 (95% confidence interval [CI]: 2.44–4.55) and 9 months
(95% CI: 4.77–10.76), respectively ([Fig. 1A] and [1B]). Patients with a lower NLR (≤2.89) had a better PFS (5 vs. 2 months [p = 0.02]) and OS (9 vs. 5 months [p = 0.20]) when compared with patients with higher NLR (>2.89) ([Fig. 1C]). Patients with lower SII≤639 had a better PFS (9 vs. 2 months [p = 0.002]) and OS (16 vs. 8 months [p = 0.04]) in comparison to patients with higher SII ([Fig. 1D] and [1E]). On univariate analysis, the following factors were associated with a worse PFS:
NLR > 2.8 (hazard ratio [HR] =2.32, p = 0.02) and SII > 639 (HR =3.70, p = 0.002) ([Table 1]). On multivariate analysis (including NLR and SII), SII > 639 was the only factor
that predicted survival (HR =4.13, p = 0.03) for PFS.
Table 1
Univariate analysis of survival outcomes in platinum resistant/refractory patients
|
Variable
|
n
|
Median PFS
|
95% CI
|
HR
|
p-Value
|
Median OS
|
95% CI
|
HR
|
p-Value
|
|
Duration from last platinum
|
|
|
|
|
|
|
|
|
|
|
3–6 mo
|
28
|
4
|
1.47–6.59
|
1
|
0.88
|
9
|
6.67–12.05
|
1
|
0.59
|
|
< 3 mo
|
21
|
3
|
1.19–6.00
|
1.04
|
6
|
0.00–13.41
|
1.22
|
|
ECOG[a]
|
|
|
|
|
|
|
|
|
|
|
0,1
|
13
|
9
|
0.00–13.64
|
1
|
0.10
|
27
|
0.00–63.14
|
1
|
0.04
|
|
2,3
|
24
|
3
|
2.23–4.88
|
2.01
|
7
|
2.60–10.79
|
2.93
|
|
Type of therapy for PROC[b]
|
|
|
|
|
|
|
|
|
|
|
IV chemotherapy doublet
|
|
|
|
|
|
|
|
|
|
|
Yes
|
22
|
4
|
1.57–5.36
|
1
|
0.60
|
5
|
8.49–11.73
|
1
|
0.08
|
|
No
|
27
|
5
|
2.39–7.60
|
0.85
|
9
|
3.55–7.31
|
0.52
|
|
IV chemotherapy single agent
|
|
|
|
|
|
|
|
|
|
|
Yes
|
5
|
4
|
0.00–11.45
|
1
|
0.63
|
4
|
0.00–9.37
|
1
|
0.40
|
|
No
|
44
|
4
|
2.86–5.19
|
1.29
|
9
|
7.31–11.24
|
1.59
|
|
Oral etoposide
|
|
|
|
|
|
|
|
|
|
|
Yes
|
22
|
6
|
2.22–8.97
|
1
|
0.82
|
10
|
3.23–8.02
|
1
|
0.02
|
|
No
|
27
|
4
|
1.93–6.13
|
1.07
|
5
|
8.75–9.78
|
2.31
|
|
Number of previous lines of treatment
|
|
|
|
|
|
|
|
|
|
|
1
|
38
|
4
|
2.40–4.95
|
1
|
0.96
|
7
|
3.95–9.91
|
1
|
0.69
|
|
2
|
11
|
4
|
3.61–5.18
|
1.02
|
6
|
3.86–9.73
|
1.25
|
|
NLR
|
|
|
|
|
|
|
|
|
|
|
≤2.8
|
19
|
5
|
2.40–7.59
|
1
|
0.02
|
10
|
6.11–13.81
|
1
|
0.11
|
|
> 2.8
|
21
|
2
|
0.23–5.23
|
2.32
|
5
|
0.95–9.90
|
1.94
|
|
SII
|
|
|
|
|
|
|
|
|
|
|
≤639
|
12
|
9
|
2.94–15.79
|
1
|
0.002
|
16
|
3.95–27.38
|
1
|
0.04
|
|
> 639
|
28
|
2
|
0.22–4.37
|
3.70
|
8
|
3.25–12.27
|
2.49
|
|
LMR
|
|
|
|
|
|
|
|
|
|
|
> 6.7
|
20
|
4
|
2.24–6.48
|
1
|
0.09
|
9
|
6.45–12.08
|
1
|
0.75
|
|
≤6.7
|
20
|
2
|
0.12–5.51
|
1.77
|
7
|
2.62–11.24
|
1.13
|
Abbreviations: CI, confidence interval; ECOG, Eastern Cooperative Oncology Group;
LMR, lymphocyte to monocyte ratio; NLR, neutrophil to lymphocyte ratio; OS, overall
survival; PFS, progression-free survival; SII, systemic immune-inflammation index;
HR, hazard ratio.
a At the time of platinum resistance.
b The chemotherapy regimens used were paclitaxel/carboplatin (n = 7, 14%), lipodox/carbo (n = 4, 8%), single-agent lipodox (n = 5, 10%), oral etoposide (n = 22, 45%), and gemcitabine/epirubicin/carbo (n = 11, 22%).
Even though PROC has a poor prognosis, this group has recognized heterogeneity.[5] Identifying patients with PROC who may benefit from subsequent therapy is currently
based on clinical judgment (performance status, rapidity of progression, number of
previous lines, and patient wish to continue potentially toxic treatment with a low
expectation of benefit). There is a need for more objective markers to determine prognoses.
This may help us tailor more intense therapies and stratify patients included in clinical
trials in this segment. This is one of the first studies looking at the impact of
inflammatory indices in PROC. We demonstrated that SII calculated at the time of diagnosis
of PROC is a powerful independent predictor of outcomes (HR of 4.1 for PFS) among
patients with PROC undergoing second/third line of chemotherapy.
Systemic inflammation induced by cancer cells may aid tumor progression by several
mechanisms.[6] These indices have also been identified as powerful independent prognostic factors
in various cancers. In patients with newly diagnosed EOC, SII, NLR, platelet to lymphocyte
ratio, and lymphocyte to monocyte ratio have been shown to predict outcomes. Recently,
predictive abilities have been demonstrated in patients with platinum-sensitive relapsed
ovarian cancer.[3]
[7] Neutrophil infiltration of the tumor is associated with tumor growth (release of
proinvasive factors, angiogenesis)[8] while less amount of CD8+ tumor-infiltrating lymphocytes is associated with poorer
prognosis.[9]
[10] Thus, the combination of high neutrophil and low lymphocytes in the peripheral blood
reflects an immunological milieu that favors tumor growth which explains the predictive
ability of the SII. Earlier studies have also demonstrated that SII could predict
therapeutic benefit.[3] Patients with higher SII (≥730) levels did not show any benefit with the addition
of bevacizumab to chemotherapy (when compared with those with lower SII who benefited
from the addition of bevacizumab).
Other studies attempting to develop prognostic nomograms in PROC have not incorporated
SII in their models.[11]
[12] Also, there is a paucity of real-world studies on PROC; most data are from trials
or analysis of specific treatments such as bevacizumab or oral metronomic chemotherapy.
Although this study is limited by the small sample size and its retrospective nature,
ours is the first data showing that SII could be a useful prognostic predictor in
patients with platinum-refractory/resistant disease. The treatment undergone by the
patients was uniform—all our patients received chemotherapy, and there were no patients
treated with bevacizumab or other targeted agents. Though several studies in different
types of cancers have shown the usefulness of this index, it is yet to be incorporated
into practice. Future studies should study the validity of inflammatory markers in
PROC and could consider incorporating it as a biomarker in clinical trials.
