Skip to main content

Impact of Long-Term Plasma Storage on Cell-Free DNA Epigenetic Biomarker Studies.

Shao J, Nguyen T, Li Z

Publication: Biomolecules (2025) • Vol 15PMID: 40723799View on PubMed

Purpose

This paper compared the cell-free DNA (cfDNA) yield, the detection of cfDNA fragments, the presence of genomic DNA contamination, sequencing library success, and the number of uniquely mapped reads (UMR) from unmatched plasma specimens to identify potential effects of patient diagnosis (solid tumors, acute myeloid leukemia (AML) or non-cancer), patient sex, patient age, plasma storage duration (0-12 years), delays to plasma separation (0 to ≥6 years), genomic DNA contamination, and cancer clinical stage.

Conclusion

Significant differences in the median cfDNA concentration were found between plasma storage durations, plasma separation delays, specimens with and without genomic DNA, specimens from males and females, and between clinical cancer stages, but the significance of these differences depended on diagnosis and/or timepoint. Univariable quantile regression that considered patient age and sex, plasma storage time, gDNA contamination, duration of delays in plasma separation, and cancer clinical stage, identified a significant effect of plasma storage duration (P=0.001) and genomic DNA contamination (P=0.02) on cfDNA yield, but these effects were limited to specimens from AML patients. Similarly, in multivariable quantile regression analysis, there were no significant effects on cfDNA yield between specimens from patients with solid tumors or non-cancer patients; however, among AML patients, there was a significant effect at the 50’th quantile of plasma storage duration (P=0.008) and patient age (P=0.02) and of genomic DNA contamination at the 25’th quantile (P=0.04) on cfDNA yield.

Significant differences in the detection of cfDNA fragments in plasma occurred between plasma storage durations, plasma separation delays, patient ages, specimens with and without genomic DNA, and between clinical cancer stages, among specimens from one or more of the groups (patients with solid tumors, patients with AML and patients without cancer). Further, in multivariate regression analysis, only genomic DNA contamination was found to be an independent factor for the detection of the presence of cfDNA fragments and the effect was only observed in plasma from patients with solid tumors (P=0.01). Multivariable regression analysis found that genomic DNA contamination was significantly associated with plasma storage duration, delayed separation of plasma, the male sex and younger ages among specimens from one or more of the groups. Similarly, sequencing library preparation success was correlated with a high cfDNA yield, lack of genomic DNA contamination and the female sex, among specimens from one or more of the groups.

The number of uniquely mapped reads (UMRs) differed significantly between plasma storage durations, specimens with and without genomic DNA, specimens from males and females among specimens from one or more of the groups (patients with solid tumors, patients with AML and patients without cancer). However, in multivariable regression analysis, increased plasma storage duration was correlated with decreased UMRs at all quantiles in specimens from non-cancer patients and in specimens from AML patients at the 25’th quantile. Genomic DNA contamination was correlated with decreased UMRs at all quantiles in specimens from patients with solid tumors, at the 50’th and 75’th quantiles in specimens from AML patients, and at the 75’th quantile in specimens from non-cancer patients. A high cfDNA yield was associated with increased UMRs at the 25’th and 50’th quantiles in patients without cancer and at the 25’th quantile in patients with AML. Finally, in patients with solid tumors, the female sex was associated with increased UMRs at the 25’th and 50’th quantiles, but in patients with AML, the female sex was associated with decreased UMRs at the 50’th quantile.

Studies (2)

Purpose

This study compared the cfDNA yield from unmatched plasma specimens to identify potential effects of patient diagnosis (solid tumors, acute myeloid leukemia (AML) or non-cancer), patient sex, patient age, plasma storage duration (0-12 years), delays to plasma separation (0 to ≥6 years), genomic DNA contamination, and cancer clinical stage. This study retrospectively analyzed plasma specimens collected between 2006 and 2022 from 622 patients with solid tumors (bladder, breast, colorectal, kidney, lung, ovarian, prostate, or uterine), 237 patients with acute myeloid leukemia, and 211 patients without cancer (diagnosis not specified). Blood was collected in K2EDTA tubes and processed immediately or after 36 days (storage conditions not specified). Plasma was obtained by centrifugation at 1350 g for 10 min at 4°C and stored at -80°C for up to 14 years. Plasma was recentrifuged twice at 13,500 g for 10 min at 4°C before cfDNA extraction using the QIAamp Circulating Nucleic Acid Extraction Kit and the QIAvac system. cfDNA was quantified using the Qubit dsDNA HS Assay Kit.

Summary of Findings

The median cfDNA yield from plasma was 30.36 ng/mL, with higher median yields from the plasma of AML patients than patients with solid tumors (65.49 ng/mL versus 27.28 ng/mL, P<2.0x10−16) or non-cancer patients (65.49 versus 34.69 ng/mL, P=1.8 x 10−5) and from non-cancer patients than patients with solid tumors (34.69 versus 27.28 ng/mL, P=1.4x10−5). No effect of plasma storage duration was found on cfDNA yield in specimens from patients with solid tumors or patients without cancer, but among AML patients, cfDNA yield was higher from plasma stored ≥12 years than 0-1.9 years (P=0.003). While cfDNA yield from the plasma of patients with solid tumors or without cancer was not significantly affected by genomic DNA contamination, the cfDNA yield from the plasma of patients with AML was significantly higher when there was genomic DNA contamination (P=0.005). cfDNA yield was higher when plasma separation occurred after 3, 4 or 5 days rather than 1 day in patients with solid tumors (P=0.009, P=0.02 and P=0.02, respectively) and after 4 days compared to 1 day in AML patients (P=0.02). There was no effect of patient age on cfDNA yield. Differences in cfDNA yield based on patient sex were limited to slightly higher yields from females with solid tumors than males (P=0.05). The cfDNA yield was higher from the plasma of patients with stage IV cancer than Stage I-III (P=0.02). Univariable quantile regression analysis that considered patient age and sex, plasma storage time, gDNA contamination, delayed plasma separation, and cancer clinical stage, identified a significant effect of plasma storage duration (P=0.001) and genomic DNA contamination (P=0.02) on cfDNA yield but these effects were limited to specimens from AML patients. In multivariable quantile regression analysis, there were no significant effects on cfDNA yield in specimens from patients with solid tumors or non-cancer patients, but among AML patients, there was a significant effect at the 50th quantile of plasma storage(P=0.008) and patient age (P=0.02) and at the 25th quantile of genomic DNA contamination (P=0.04) on cfDNA yield.

  • Biospecimens

    FluidBlood
    FluidPlasma
  • Preservation Types

    Frozen
  • Diagnoses

    NeoplasticCarcinoma
    NeoplasticLeukemia
    Not specified
  • Platforms

    DNAFluorometry
  • Pre-analytical Factors

    ClassificationPre-Analytical FactorValue(s)
    PreaquisitionDiagnosis/ patient condition
    AML
    Non-cancer
    Solid tumor
    PreaquisitionPatient age
    50-59 years
    60-69 years
    70-79 years
    <50 years
    >80 years
    PreaquisitionPatient gender
    Female
    Male
    PreaquisitionPrognostic factor
    Stage I
    Stage II
    Stage III
    Stage IV
    Biospecimen Aliquots and ComponentsCentrifugation
    Centrifugation delays investigated
    StorageStorage duration
    Plasma storage for 0-1.9 years
    Plasma storage for 10-11.9 years
    Plasma storage for 2-3.9 years
    Plasma storage for 4-5.9 years
    Plasma storage for 6-7.9 years
    Plasma storage for 8-9.9 years
    Plasma storage for ≥ 12 years
    Processing delay of 0 days
    Processing delay of 1 day
    Processing delay of 2 days
    Processing delay of 3 days
    Processing delay of 4 days
    Processing delay of 5 days
    Processing delay of ≥6 days