Delivery prediction by quantitative analysis of four steroid metabolites with liquid chromatography tandem mass spectrometry in asymptomatic pregnant women

Figures & data

Figure 1. The schematic diagram for patient enrolment and delivery prediction by quantitative analysis of four steroid metabolites with an LC-MS/MS assay.

Table 1. Linearity, LLOQ and stability of the four steroid candidates in LC-MS/MS assay.

Analytes	R^2	Linear range (ng/mL)	Linear regression		LLOQ (ng/mL)	Stability, %
			Slope (±SD)	Intercept (±SD)		4 °C	21 °C
THDOC	0.997	1–75	0.066 ± 0.005	0.020 ± 0.007	1	98.2	99.5
17-OHP	0.992	1–75	0.627 ± 0.104	0.558 ± 0.057	1	97.5	100.3
A-3,17-Diol	0.992	1–75	0.155 ± 0.022	1.011 ± 0.102	1	101.2	103.4
E3-16-Glu	0.996	4–1200	0.013 ± 0.002	0.087 ± 0.005	4	107.9	110.5

THDOC: tetrahydrodeoxycorticosterone; 17-OHP: 17-alpha-hydroxyprogesterone; A-3,17-Diol: androstane-3,17-diol; E3-16-Glu: oestriol-16-glucuronide; SD: standard deviation; stability: recovery rates for plasma samples stored at 4 or 21 °C for 6 days.

Table 2. Recoveries and imprecisions of the four steroid candidates in LC-MS/MS assay.

	E-16-Gluc	THDOC	A-3,17-Diol	17-OHP
Recovery, %
QC-L	97.5	92.6	101.0	118.6
QC-M	91.2	109.4	92.4	117.0
QC-H	107.9	91.0	88.3	102.0
Intra-assay CV, % (n = 10)
QC-L	8.4	9.1	8.6	9.5
QC-M	6.9	6.9	10.7	6.2
QC-H	9.2	7.8	7.2	4.5
Inter-assay CV, % (n = 25)
QC-L	11.8	11.1	10.8	10.9
QC-M	12.2	5.2	10	13.5
QC-H	4.9	8.4	9.2	5.2

QC-L: low level quality control; QC-M: medium level quality control, QC-H: high level quality control.

Table 3. Demographic data of enrolled subjects and plasma levels of the four steroid candidates.

	Positive group* (n = 100, 17.1%)	Negative group* (n = 485, 82.9%)	Unadjusted p	GA-adjusted p
Demographic data
Age	32 (29, 34)	31 (29, 33)	0.6	–
GA of plasma collection	38 (37,3 9)	33 (32, 36)	<0.001	–
GA of spontaneous delivery	39 (38, 40)	39 (39, 40)	0.2	–
Plasma levels (ng/ml)
E3-16-Glu	250.5 (174.0, 362.0)	133.0 (87.8, 204.0)	<0.001	<0.001
THDOC	0.7 (0.5, 1.2)	0.8 (0.5, 1.1)	0.67	0.71
A-3,17-Diol	1.7 (0.4, 4.2)	1.8 (0.5, 4.6)	0.76	0.62
17-OHP	7.9 (5.5, 11.6)	5.3 (3.4, 8.0)	<0.001	0.02

Positive group: the patients delivering within 7 days of sample collection; negative group: the patients delivering 7 days after sample collection; GA: gestational age. *Data presented as median (25th, 75th) percentile.

Figure 2. Box plots representing the plasma levels of the four steroid candidates in positive and negative groups. (A) E3-16-Glu, oestriol-16-glucuronide; (B) THDOC, tetrahydrodeoxycorticosterone; (C) A-3,17-Diol, androstane-3,17-diol；(D) 17-OHP, 17-alpha-hydroxyprogesterone. *Indicates p < .001.

Figure 3. Performance assessment of the steroid candidates E3-16-Gluc and 17-OHP in delivery prediction by ROC analysis with or without gestational age adjustment. (A–C): ROC curves for E3-16-Gluc, 17-OHP, and combined testing (E3-16-Gluc and 17-OHP); (D–F): gestational age-adjusted ROC curves for E3-16-Gluc, 17-OHP, and combined testing (E3-16-Gluc and 17-OHP) respectively. AUC: area under the curve; AAUC: gestational age-adjusted AUC.

Table 4. Predictive performance of the four steroid candidates in asymptomatic pregnant women delivery.

Predictive performance	E3-16-Glu	THDOC	A-3,17-Diol	17-OHP	Combination
Sensitivity, 95% CI (%)	78.0 (68.6–85.7)	41.0 (31.3–51.3)	55.0 (44.7–65.0)	65.0 (54.8–74.3)	87.0 (78.8–92.9)
Specificity, 95% CI (%)	65.6 (61.2–69.8)	67.2 (62.8–71.4)	38.8 (34.4–43.3)	65.2 (60.7–69.4)	60.2 (55.7–64.6)
PPV, 95% CI (%)	31.8 (28.4–35.4)	20.5 (16.5–25.2)	15.6 (13.3–18.3)	27.8 (24.2–31.7)	31.1 (28.3–34.0)
NPV, 95% CI (%)	93.5 (90.9–95.5)	84.7 (82.3–86.8)	80.7 (76.6–84.2)	90.0 (87.3–92.2)	95.7 (93.1–97.4)
Unadjusted AUC, 95% CI (%)	0.77 (0.73–0.80)	0.51 (0.47–0.56)*	0.51 (0.47–0.55)*	0.70 (0.66–0.74)*	0.80 (0.76–0.83)
GA-adjusted AUC, 95% CI (%)	0.69 (0.59–0.77)	ND	ND	0.58 (0.48–0.68)*	0.69 (0.60–0.76)

Combination: combination detection of E3-16-Glu and 17-OHP; PPV: positive predictive value; NPV: negative predictive value; GA: gestational age; CI: confidence interval. *Indicates p < .05, when compared with the AUC of the combination detection of E3-16-Glu and 17-OHP in ROC analysis.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article and its Supplemental materials.