Virtual Library
Start Your Search
Lizza Hendriks
Author of
-
+
ES09 - Biomarkers in Immunotherapy (ID 148)
- Event: WCLC 2020
- Type: Educational Session
- Track: Immuno-biology and Novel Immunotherapeutics (Phase I and Translational)
- Presentations: 1
- Moderators:
- Coordinates: 1/30/2021, 09:15 - 10:15, Scientific Program Auditorium
-
+
ES09.05 - Blood-based Biomarkers for Immunotherapy: Applications and Limitations (ID 3946)
09:15 - 10:15 | Presenting Author(s): Lizza Hendriks
- Abstract
- Presentation
Abstract
Biomarkers are needed in the treatment of lung cancer patients with immune checkpoint inhibitors (ICI), as the majority of patients will not benefit from ICI, or will even be harmed (pseudoprogression, immune related toxicity). This presentation will focus on NSCLC.
Currently, for those without a targetable oncogenic driver, only programmed-death ligand1 (PD-L1) expression on tumor cells is used for treatment selection in the ESMO guideline on metastatic NSCLC. However, PD-L1 is not a perfect biomarker, as only 32% of patients with NSCLC with high PD-L1 expression (≥50%), treated with first line monotherapy pembrolizumab are alive after five years. Importantly, PD-L1 expression levels are heterogeneous across sites of disease. Furthermore, biopsies are difficult to obtain during treatment with ICI, and easier to use biomarkers are needed both for treatment selection and monitoring.
Examples of blood-based biomarkers are circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), blood tumor mutational burden (bTMB), cytokines, soluble PD-L1, lymphocyte counts, CRP, LDH, albumin. Drawbacks of most of these biomarkers are that analyses are not standardized and that different platforms are used. Furthermore, several analyses are expensive to perform. Importantly, to interpret the results of biomarker studies, a distinction should be made between predictive and prognostic markers.
PD-L1 can be determined on CTCs. Small studies showed that there is no correlation between PD-L1 expression level on CTCs and on tumor cells on tissue. In contrast to PD-L1 expression on tissue, patients with NSCLC, treated with ICI, and baseline PD-L1+ CTCs, have a poorer progression free survival (PFS) compared with those without PD-L1+ CTCs. For soluble PD-L1, similar results have been found in small studies.
bTMB has been extensively investigated. For monotherapy anti-PD-L1, patients with a high bTMB (corrected for ctDNA concentration) have a superior PFS on ICI compared with those with a low bTMB (atezolizumab OAK and POPLAR data). In contrast, bTMB is not associated with increased benefit for patients treated with chemotherapy-ICI, as was shown for example in the KEYNOTE189. Results for ICI-ICI combinations are conflicting. For example, in an exploratory analysis of the MYSTIC study, patients with a bTMB of ≥20 mut/Mb had the best outcome with the combination of durvalumab and tremelimumab, while there was no benefit for those with a bTMB of < 20 mut/Mb. However, the NEPTUNE trial, specifically evaluating durvalumab-tremelimumab in those with a high bTMB was reported to be negative. Of note, in the b-F1RST trial, those without detectable ctDNA had the best outcome. It could be that this is a patient population with a lower disease burden and low shedding of ctDNA.
Specific ctDNA biomarkers are for example KEAP1, STK11 and ARID1A. in the MYSTIC trial, KEAP1 and STK11 seemed only prognostic, as patients with a KEAP1 or STK11 mutation had worse survival compared with the wild-type patients regardless of the treatment.
ARID1A is a negative prognostic marker, but ARID1A deficiency seems to promote antitumor immunity, increases TMB and modulates the tumor microenvironment (increase in tumor infiltrating lymphocytes). In an exploratory analysis in the MYSTIC trial, ARID1A mutations were indeed associated with improved survival in those treated with durvalumab-tremelimumab, but not durvalumab monotherapy, compared with chemotherapy.
ctDNA dynamics can also be used to select patients with the highest chance of long-term benefit, as patients with a reduction of ctDNA during ICI treatment had the best PFS on ICI. ctDNA dynamics can probably also be used to select patients for adjuvant ICI, and combination scores of ctDNA, its dynamics, PD-L1, TMB and CD8+ Tcells are in development. For example, in stage III NSCLC patients treated with chemoradiotherapy, patients with ctDNA clearance after chemoradiotherapy, survival was good regardless of adjuvant ICI. Of note, all studies are small and validation is needed.
Standard of care lab values and its dynamics during ICI can probably also be used to select patients that will, or will not, benefit from ICI. Examples are derived NLR (neutro/[leuko-neutor]) and its change during ICI treatment, or the LIPI score (dNLR and LDH combined). Of note, although large, all studies are retrospective and prospective validation is needed.
Last, T-cell receptor clonality analysis and the dynamics during ICI treatment are promising to select patients for ICI, but studies so far have been very small.
In conclusion, blood-based biomarkers are needed, to overcome the problems of tissue biopsies. Standard lab values are interesting but need prospective validation. ctDNA analysis is the most advanced, and seems especially useful for monitoring during ICI. CTC, soluble PD-L1 and Tcell receptor clonality analysis are interesting and promising but more research is needed, as is a balance between costs and usefullness.
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
-
+
FP07 - Pathology (ID 109)
- Event: WCLC 2020
- Type: Posters (Featured)
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
FP07.06 - Lung Immune Prognostic Index (LIPI) in Advanced NSCLC Patients Treated with Immunotherapy, Chemotherapy and both Combined Upfront. (ID 823)
00:00 - 00:00 | Author(s): Lizza Hendriks
- Abstract
Introduction
The Lung Immune Prognostic Index (LIPI) that combines the neutrophils/[leucocytes minus neutrophils] ratio (dNLR) and lactate dehydrogenase (LDH), is associated with outcomes in pretreated advanced NSCLC patients receiving single agent immune checkpoint inhibitors (ICI). However, its role in first line treatment of advanced NSCLC patients has not been explored yet. We assessed the value of baseline LIPI in the first line setting, for ICI-monotherapy, ICI-combination or platinum-based chemotherapy alone (CT).
Methods
We retrospectively collected data of patients treated with first-line ICI between 2016 and 2019 as single agent if PD-L1 ≥50% (ICI-cohort), or in combination with a CTLA4-inhibitor (ICI+ICI cohort), or with chemotherapy (ICI+CT cohort), from 18 centers worldwide. A control cohort of patients treated with platinum-based CT (CT-cohort) was also collected between 2011 and 2019 from 2 centers. Baseline LIPI was calculated as previously reported and correlated with overall survival (OS) and progression-free survival (PFS) in each treatment cohort.
Results
Overall, 930 patients were enrolled, 561 in the ICI-cohort, 186 in the combo ICI+CT, 55 in the ICI+ICI and 128 in the CT-cohort. Median (m) follow-up was 12.5 months. In the whole cohort, median age was 66 years, 70% male, 80% had non-squamous histology, and 84% had PS ≤1. Based on LIPI (available for 792 patients): 305 (38%) were considered good, 331 (42%) intermediate and 156 (20%) poor group.
Treatment outcomes according to LIPI scores are depicted in Table 1. The LIPI poor population had significantly worse OS compared with other LIPI groups, in the whole cohort (P<0.001) as well as in the ICI monotherapy and combo ICI+CT cohorts (both P<0.0001); and in the CT cohort (P=0.004). In term of PFS, we observed correlation between LIPI groups and outcomes in the whole cohort (P<0.001) and in the ICI- monotherapy cohort (P=0.008); however, no differences were observed in the cohorts of patients receiving chemotherapy regimens, alone (P=0.08) or combined with ICI (P=0.08). The analysis by PD-L1 expression in 756 patients with available data will be presented in the Congress.
Table 1: Median OS and PFS according to LIPI subgroups. NR = non reached.
ConclusionOutcomes
LIPI
subgroups
Overall cohort
N= 925
ICI-cohort
N=558
ICI + CT-cohort
N= 185
ICI + ICI cohort
N= 55
CT-cohort
N=127
Median OS
(95% CI)
All
16.3 (14.7-18.8)
21.0 (17.1-NR)
24.7 (16.9-27.1)
20.5 (14.1-NR)
9.79 (8.3-14.4)
LIPI good, 38.5%
19.8 (17.2-25.7)
NR (NR-NR)
25.7 (25.6-NR)
33.6 (14.7-NR)
14.42 (8.9-17.9)
LIPI interm, 41.8%
15.8 (14.3-20.3)
21.2 (17.1-NR)
20.3 (12.8-NR)
14.6 (5.5-NR)
9.30 (7.0-14.5)
LIPI poor, 19.7%
6.96 (5.6-12.5)
8.5 (3.4-13.7)
6.1 (4.9-NR)
14.1 (10.3-NR)
6.1 (5.0-NR)
Global LogRank P value
<0.0001
<0.0001
<0.0001
0.4
0.004
Overall cohort
N= 909
ICI-cohort
N=543
ICI + CT-cohort
N= 185
ICI + ICI cohort
N= 54
CT-cohort
N=127
Median PFS
(95% CI)
All
6.5 (5.9-7.1)
6.3 (5.0-7.6)
8.9 (6.80-10.9)
7.2 (5.7-30.6)
5.7 (5.3-6.4)
LIPI good, 38.7%
7.0 (5.9-8.5)
6.4 (4.5-10.8)
9.8 (7.8-13.0)
9.2 (5.7-NR)
6.0 (5.3-7.8)
LIPI interm, 41.6%
6.6 (6.1-7.6)
6.6 (5.6-8.1)
10.4 (6.4-12.4)
5.5 (2.5-NR)
6.1 (4.3-7.6)
LIPI poor, 19.7%
3.6 (3.1-5.6)
3.3 (1.9-6.7)
4.5 (2.8-8.2)
7.1 (2.56- NR)
3.7 (3.4-NR)
Global LogRank P value
<0.0001
0.008
0.08
0.4
0.08
Pretreatment LIPI was prognostic in untreated advanced NSCLC patients regardless of the type of therapy. However, LIPI was associated with PFS only in patients receiving ICI-monotherapy, with no statistically significant differences in CT-containing cohorts (alone or combined with ICI). This value of LIPI to guide treatment selection should be further explored in prospective studies.
-
+
P33 - Pathology - Immunotherapy Biomarker (ID 101)
- Event: WCLC 2020
- Type: Posters
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P33.10 - Identification of Long-Responders and Fast-Progressors under Immunotherapy Based on Early Monitoring of dNLR in Advanced NSCLC Patients (ID 2519)
00:00 - 00:00 | Author(s): Lizza Hendriks
- Abstract
Introduction
The dNLR score, based on [neutrophil/(leucocytes-neutrophils)] ratio (dNLR) at baseline (B) and before 2nd cycle (C2), has been correlated with immune checkpoint inhibitors (ICI) outcomes in NSCLC patients. We aimed to assess whether dNLR score can identify the different response phenotypes to ICI in NSCLC, particularly the fast-progressors (FP) and long-responders (LR) under ICI.
Methods
Advanced NSCLC patients receiving ICI between Nov.12 and Jan.19, were enrolled in Gustave Roussy. dNLR was retrospectively collected at B and C2. Patients were categorized as low vs. high-dNLR at each time-point (defined as ≤3 or >3), and the change between B and C2 (good = low at both time-points, poor = high at both time-points, mixed = different at each time-point). Response types were evaluated: a) “responder” (objective response (ORR) and progression-free survival (PFS) >6 months (mo.)), b) “LR” (ORR + PFS >12 mo. and median overall survival (OS) >24 mo.), c) “standard-progressor" (PD) (progressive disease as best response; not including FP) and d) “FP” (defined as early death within the 1st 12 weeks).
Results
469 patients were included: 65% males, 90% smokers, median age of 63, 75% performance status ≤1; adenocarcinoma histology in 66% and 12% harboring driver alterations. PD-L1 was ≥1% in 143/259 (55%), missing in 210 patients. The ORR was 19% (80/431); 15% were LR, 35% standard-PD and 13% FP. Overall, the median OS was 10.0 mo. [95% CI, 8.2-12.2]; 45.2 mo. [95% CI, 31.9-not reached] in LR, 4.2 mo. [95% CI, 3.2-5.7] in PD and 0.7 mo. [95% CI, 0.6-0.9] in FP population. dNLR (B) was high in 41% of FP vs. 44% of standard-PD vs. 27% of LR (P<0.001). dNLR (C2) was high in 81% of FP vs. 45% of PD vs. 14% of LR (P<0.001). Response phenotypes were strongly correlated with the dNLR score subgroups (table 1).
Overall
N=291
Long-responder
N=48
Responder
N=108
Standard progressor
N=116
Fast-progressor
N=16
P value Good 151 (52%) 32 (67%) 65 (60%) 52 (45%) 2 (13%) <0.001 Intermediate 71 (24%) 13 (27%) 24 (22%) 30 (26%) 3 (19%) <0.001 Poor 69 (24%) 3 (6%) 19 (18%) 34 (29%) 11 (69%) <0.001
Conclusion
The dNLR score, especially at C2, is an accessible and simple tool that can add information to radiological examination discriminating the different phenotypes of response under ICI, particularly aggressive patterns such as FP.
