A MYC-Driven Plasma Polyamine Signature for Early Detection of Ovarian Cancer

Abstract

MYC is an oncogenic driver in the pathogenesis of ovarian cancer. We previously demonstrated that MYC regulates polyamine metabolism in triple-negative breast cancer (TNBC) and that a plasma polyamine signature is associated with TNBC development and progression. We hypothesized that a similar plasma polyamine signature may associate with ovarian cancer (OvCa) development. Using mass spectrometry, four polyamines were quantified in plasma from 116 OvCa cases and 143 controls (71 healthy controls + 72 subjects with benign pelvic masses) (Test Set). Findings were validated in an independent plasma set from 61 early-stage OvCa cases and 71 healthy controls (Validation Set). Complementarity of polyamines with CA125 was also evaluated. Receiver operating characteristic area under the curve (AUC) of individual polyamines for distinguishing cases from healthy controls ranged from 0.74-0.88. A polyamine signature consisting of diacetylspermine + N-(3-acetamidopropyl)pyrrolidin-2-one in combination with CA125 developed in the Test Set yielded improvement in sensitivity at >99% specificity relative to CA125 alone (73.7% vs 62.2%; McNemar exact test 2-sided P: 0.019) in the validation set and captured 30.4% of cases that were missed with CA125 alone. Our findings reveal a MYC-driven plasma polyamine signature associated with OvCa that complemented CA125 in detecting early-stage ovarian cancer.

Keywords: blood-based biomarkers; early detection; ovarian cancer; polyamines.

Figure A1

Spearman correlation matrix for polyamines. Heatmaps depicting spearman correlation coefficients amongst the 4 measured polyamines in the Test Set (left panel) and Validation Set (right panel). Embedded values represent Spearman correlation coefficients. All cases and controls corresponding to the respective test set were included in the analyses. Abbrev. DAS- diacetylspermine; AcSpmd- acetylspermidine; DiAcspmd- diacetylspermidine; N3AP- N-(3-acetamidopropyl)pyrrolidin-2-one.

Figure A2

Plasma levels of polyamines in cases stratified by histology and controls in the Test Set.

Figure A3

Plasma levels of polyamines in cases stratified by histology and controls in the Validation Set.

Figure A4

Scatter plot of depicting 3-marker panel scores (Y-axis) and CA125 values (X-axis) in CA125 ‘negative’ (defined as ≤35 U/mL) subjects in the Validation Set. Dashed lines represent 100% specificity cutoff.

Figure 1

Classification performances of plasma polyamines in the Test Set. (A,B) Area under the curve (AUC) of DAS for delineating all cases (n = 116) from healthy controls (n = 71) (A) or patients with benign pelvic masses (n = 72) (B).

Figure 2

Classification performance of the 3-marker panel and CA125 in the Test Set. (A) Area under the curve for the 3-marker panel consisting of DAS + N3AP + CA125, and CA125 only. (B) Confusion matrix describing the performance of the classification model corresponding to the 3-marker panel and CA125 alone at 99% specificity. Statistical significance was determined by 1-sided McNemar exact test. (C) Scatter plot illustrating the distribution of the 3-marker panel scores (Y-axis) and log10 CA125 values (X-axis). Broken lines represent 99% specificity cutoffs.

Figure 3

Classification performances of plasma polyamines in the validation set. The area under the curve (AUC) of individual polyamines for delineating all cases (n = 61) from healthy controls (n = 71).

Figure 4

Classification performance of the 3-marker panel and CA125 in the validation set. (A) Confusion matrix describing the performance of the classification model corresponding to the 3-marker panel and CA125 alone at 99% specificity. Statistical significance was determined by 1-sided McNemar exact test. (B) Scatter plot illustrating the distribution of the 3-marker panel scores (Y-axis) and log10 CA125 values (X-axis). Broken lines represent >99% specificity cutoffs.