Published December 8, 2025
| Version v1
Dataset
Open
SChISM: Mechanistic Modeling of cfDNA Fragmentome Dynamics Predicts Progression to Immunotherapy
Authors/Creators
-
Nguyen Phuong, Linh
(Data manager)1, 2, 3, 4
- Boutonnet, Audrey (Data collector)5
-
Greillier, Laurent
(Producer)1, 6
-
Tomasini, Pascale
(Producer)7, 6
-
Deville, Jean-Laurent
(Producer)6
-
Salas, Sebastien
(Project leader)1, 6, 3, 4
-
Ginot, Frederic
(Project member)5
-
Fina, Frédéric
(Project leader)8
-
Benzekry, Sébastien
(Supervisor)1, 2, 4
-
1.
COMPO: Méthodes computationnelles pour la prise en charge thérapeutique en oncologie
-
2.
Research Centre Inria Sophia Antipolis - Méditerranée
-
3.
Aix-Marseille University
-
4.
Centre de Recherche en Cancérologie de Marseille
- 5. Adelis Technologies
-
6.
Assistance Publique Hôpitaux de Marseille
- 7. Multidisciplinary Oncology & Therapeutic Innovations Department
- 8. ID-Solutions Oncologie
Description
Summary
This dataset contains the clinical data, longitudinal tumor size, and longitudinal size-based cfDNA concentration measurements from the SChISM clinical study (NCT05083494), which integrates advanced and/or metastatic cancer patients treated with immune checkpoint inhibitors (ICIs) in monotherapy or in combination with chemotherapy or targeted therapy. The study aims to identify size-based cfDNA concentration as potential biomarker for immunotherapy response [1].
Methods
Study design and patient population
SChISM is a prospective, multicenter, collaborative, non-interventional clinical study conducted across departments of the Assistance Publique–Hôpitaux de Marseille (AP-HM). cfDNA analysis was performed using BiaBooster™ technologies to investigate predictive and prognostic value in 128 patients with metastatic and/or recurrent carcinomas (NSCLC, HNSCC, ccRCC, and UC), initiating treatment with ICIs as standard care between April 2021 and July 2023. Follow-up extended until radiological progression under ICI therapy or 12 months of treatment. Patients with any radiological assesment or extreme cfDNA concentration (> 1,000 pg/µL at least at one timepoint) were excluded from the study. The final analytic cohort included longitudinal data from patients meeting inclusion criteria with adequate follow-up ($n=112$).
The study complied with the Declaration of Helsinki, Good Clinical Practice, and French regulations. It was classified as Category 2 research under the Jardé Law and approved by the national ethics committee, with registration through the French Ministry of Health. AP-HM, as sponsor, ensured data protection regulations (GDPR) compliance through secure anonymized electronic Case Report Forms. All participants provided written informed consent. The study is registered as NCT05083494.
Clinical data
Clinical and biological data were prospectively collected, including age, sex, Eastern Cooperative Oncology Group (ECOG) performance status, lactate dehydrogenase (LDH), PD-L1 expression—combined positive score (CPS) or tumor proportion score (TPS)—, and tumor type. Treatment response was assessed every three months by contrast-enhanced computed tomography (CT) scans, interpreted using iRECIST criteria [2]. Routine laboratory parameters collected at baseline, including neutrophil-to-lymphocyte ratio (NLR), which was included as widely studied nonspecific biomarker with prognostic and predictive relevance in immunotherapy [3,4].
Tumor size measurements
Tumor size measurements were computed from CT scans according to iRECIST guidelines. The sum of the longest diameters (SLD, denoted as TK in the dataset) of all lesions (target, non-target, and new ones) was calculated at each time point to assess tumor burden and response to treatment. Dates of radiological progression, last follow-up, or death were recorded for survival analyses.
Endpoints
Progression-free survival (PFS) was defined as the interval between first ICI infusion and progression, death, or last follow-up. Progression date corresponded to radiological confirmation or last available follow-up.
cfDNA concentration measurements
Blood collection
The certified AP-HM Biobank (CRB, NFS 96-900, ISO 9001:2015) managed plasma processing and storage. Peripheral blood (Roche cfDNA Collection Tube) was collected immediately before each ICI infusion and centrifuged at 1,600×g for 10 minutes at room temperature. Plasma supernatant was collected into a 15 mL conical tube and subjected to a second centrifugation at 4,500×g for 10 minutes to remove residual debris. Plasma was transferred into a new 15 mL tube and aliquoted in 300 to 500 µL volumes, then stored at –80°C. cfDNA analysis was performed at Adelis (Labège, France) using BIABooster™ technology [5,6].
BIABooster workflow
Experiments were carried out with a G7100A CE system (Agilent Technologies, Germany) equipped with a Zetalif fluorescence detector (Adelis, France) and a BIABooster capillary device (Adelis, # 16-BB-DNA/11, France). DNA was concentrated at the junction of two capillaries of different diameters using dual hydrodynamic and electrokinetic actuation, allowing the removal of salts and proteins, which enables in-line purification and size-dependent migration. Gel electrophoresis quantified fragments, providing cfDNA size distribution with 10 fg/µL sensitivity [6].
Plasma was pretreated with lysis buffer (56°C, 2h, 900rpm) to release cfDNA from vesicles and protein complexes. One µL was injected into the BIABooster device for cfDNA analysis. Migration curves plotted fluorescence intensity against time. Analytics software converted these into concentration and fragment size using a DNA ladder. The device gives reliable size and concentration measurements for DNA fragments between 75 and 1650 base pairs (bp).
The analytical method’s robustness was previously assessed through repeatability and reproducibility study [5]. Reproducibility was evaluated by varying the instrument, capillary device, buffer and ladder batches, time, and operator, yielding standard deviations of 1-1.8 bp for the mono- and dinucleosomal peaks and a 12% coefficient of variation for cfDNA concentration.
A cfDNA curve is defined as the concentration of fragments (pg/µL) according to the fragment size (bp). Twelve quantitative variables were derived from each curve quantified at baseline, corresponding to nucleosomal multiples:
- $P_1$ and $P_2$ (bp), denoted as
PEAK1 and PEAK2: the first and second peak’s position, corresponding to the most frequent size of the fragments originating from mono- and dinucleosomes, respectively.- $P_2-P_1$ (bp), denoted as
PEAK_DIFF: the difference between $P_2$ and $P_1$.- $HP$ (pg/µL), denoted as
HEAIGHT_PEAK1: the height of the first peak.- $HW$ (pg/µL), denoted as
HALF_WIDTH1: the left half-width of the first peak at mid-height. This variable was designed to capture the expected larger fragmentation of patients compared to healthy individuals [7].- $C_{TOT}$ (pg/µL), denoted as
CONCENTRATION: the global concentration defined as the area under the cfDNA curve between 75 and 1650 bp computed by the trapeze method.- The relative concentration $R_{[x_i,x_j]}$ in pg/µL (denoted as
SIZE_$x_i$_$x_j$) of cfDNA fragments measuring between $x_i$ and $x_j$ bp, to the total concentration $C_{TOT}$, defined as the following absolute concentration$$R_{[x_i,x_j]}=\frac{C_{[x_i,x_j]}}{C_{TOT}}$$
for $[x_i,x_j]\in{[75,111],[111,240],[240,370],[370,580],[580,1650]}$.
- $R_{<75}$ and $R_{>1650}$ (r.a.u.), denoted as
LESS_75 and GREATER_1650, respectively: the relative quantity of cfDNA fragments of less than 75 bp and greater than 1650 bp, respectively, to the total concentration C_TOT.Two quantitative variables were derived from each curve for longitudinal modeling, based on different ranges of size:
- The concentration in pg/µL of short cfDNA fragments measuring between 75 and 580 bp, denoted as SHORT in the dataset.
- The concentration in pg/µL of short cfDNA fragments measuring between 580 and 1650 bp, denoted as LONG in the dataset.
References
[1] Nguyen Phuong L, Fina F, Greillier L, et al. The SChISM study: Cell-free DNA size profiles as predictors of progression in advanced carcinoma treated with immune-checkpoint inhibitors. Published online September 3, 2025. Accessed September 16, 2025. https://inria.hal.science/hal-05238567
[2] Seymour L, Bogaerts J, Perrone A, et al. iRECIST: guidelines for response criteria for use in trials testing immunotherapeutics. Lancet Oncol. 2017;18(3):e143-e152. doi:10.1016/S1470-2045(17)30074-8
[3] Bagley SJ, Kothari S, Aggarwal C, et al. Pretreatment neutrophil-to-lymphocyte ratio as a marker of outcomes in nivolumab-treated patients with advanced non-small-cell lung cancer. Lung Cancer. 2017;106:1-7. doi:10.1016/j.lungcan.2017.01.013
[4] Mezquita L, Auclin E, Ferrara R, et al. Association of the Lung Immune Prognostic Index With Immune Checkpoint Inhibitor Outcomes in Patients With Advanced Non–Small Cell Lung Cancer. JAMA Oncology. 2018;4(3):351-357. doi:10.1001/jamaoncol.2017.4771
[5] Andriamanampisoa CL, Bancaud A, Boutonnet-Rodat A, et al. BIABooster: Online DNA Concentration and Size Profiling with a Limit of Detection of 10 fg/μL and Application to High-Sensitivity Characterization of Circulating Cell-Free DNA. Anal Chem. 2018;90(6):3766-3774. doi:10.1021/acs.analchem.7b04034
[6] Boutonnet A, Pradines A, Mano M, et al. Size and Concentration of Cell-Free DNA Measured Directly from Blood Plasma, without Prior DNA Extraction. Anal Chem. 2023;95(24):9263-9270. doi:10.1021/acs.analchem.3c00998
[7] Mouliere F, Chandrananda D, Piskorz AM, et al. Enhanced detection of circulating tumor DNA by fragment size analysis. Sci Transl Med. 2018;10(466):eaat4921. doi:10.1126/scitranslmed.aat4921
Files
df_bsl.csv
Additional details
Related works
- Is documented by
- Preprint: hal-05238567v1 (Other)
- Is required by
- Preprint: hal-05241421v1 (Other)
References
- L. Nguyen Phuong, F. Fina, L. Greillier, P. Tomasini, J.-L. Deville, A. Boutonnet, F. Ginot, J.-C. Garcia, S. Salas, S. Benzekry, Mechanistic modeling of tumor and size-dependent cell-free DNA kinetics under immune checkpoint inhibition, hal.fr, https://inria.hal.science/hal-05241421v1