Prospective study reveals a microbiome signature that predicts the occurrence of post-operative enterocolitis in Hirschsprung disease (HSCR) patients

Figures & data

Table 1. Characteristics of 253 Hirschsprung disease patients in the cohort study by exclusive breastfeeding status.

Characteristic	Exclusive breastfeeding
	No(n = 142)	Yes(n = 111)	P
Gestational age, weeks, mean (SD)	39.3 (1.4)	39.2 (1.3)	0.83^a
Birth weight, kg, mean (SD)	3.5 (0.4)	3.4 (0.5)	0.77^a
Gender, No. (%)			0.82^b
Female	24 (16.9)	20 (18.0)
Male	118 (83.1)	91 (82.0)
Mode of delivery, No. (%)			0.03^b
Cesarean	86 (60.6)	52 (46.8)
Vaginal	56 (39.4)	59 (53.2)
HSCR classification^c, No. (%)			0.01^b
Classical segment HSCR	78 (54.9)	78 (70.3)
Long segment HSCR or total colonic aganglionosis	64 (45.1)	33 (29.7)
Ostomy, No. (%)			0.01^b
No	113 (79.6)	101 (91.0)
Yes	29 (20.4)	10 (9.0)
Age at surgery, No. (%)			0.58^b
≤180 days	122 (85.9)	98 (88.3)
>180 days	20 (14.1)	13 (11.7)
Surgical approach, No. (%)			0.04^b
Open pull-through	29 (20.4)	10 (9.0)
Totally transanal endorectal pull-through	70 (49.3)	62 (55.9)
Laparoscopic-assisted transanal pull-through	43 (30.3)	39 (35.1)

^aCalculated using t-test.

^bCalculated using χ² test.

^cHSCR is classified as classical segment HSCR when the aganglionic segment does not extend beyond the upper sigmoid, long-segment HSCR when the aganglionic segment extends to the splenic flexure or transverse colon, and total colonic aganglionosis when the aganglionic segment extends to the colon and a short segment of the terminal ileum.

Abbreviations: HSCR, Hirschsprung disease; SD, standard deviation.

Figure 1. Flow-chart of cohort and prospective nested case–control studies.

Table 2. Characteristics of 75 Hirschsprung disease patients in the case–control study.

Characteristic	Postoperative HAEC cases (n = 25)	Controls (n = 50)	P
Gestational age, weeks, mean (SD)	39.2 (1.4)	39.4 (1.1)	0.55^a
Birth weight, kg, mean (SD)	3.5 (0.5)	3.5 (0.5)	0.87^a
Gender, No. (%)			0.07^b
Female	2 (8.0)	13 (26.0)
Male	23 (92.0)	37 (74.0)
Mode of delivery, No. (%)			0.19^b
Cesarean	16 (64.0)	24 (48.0)
Vaginal	9 (36.0)	26 (52.0)
HSCR classification^c, No. (%)			0.87^b
Classical segment HSCR	15 (60.0)	29 (58.0)
Long segment HSCR or total colonic aganglionosis	10 (40.0)	21 (42.0)
Ostomy, No. (%)			1.00^d
No	22 (88.0)	43 (86.0)
Yes	3 (12.0)	7 (14.0)
Age at pull-through procedure, No. (%)			0.47^d
≤180 days	21 (84.0)	45 (90.0)
>180 days	4 (16.0)	5 (10.0)
Surgical approach, No. (%)			0.97^b
Open pull-through	3 (12.0)	7 (14.0)
Totally transanal endorectal pull-through	13 (52.0)	25 (50.0)
Laparoscopic-assisted transanal pull-through	9 (36.0)	18 (36.0)

^aCalculated using t-test.

^bCalculated using χ² test.

^cHSCR is classified as classical segment HSCR when the aganglionic segment does not extend beyond the upper sigmoid, long-segment HSCR when the aganglionic segment extends to the splenic flexure or transverse colon, and total colonic aganglionosis when the aganglionic segment extends to the colon and a short segment of the terminal ileum.

^dCalculated using Fisher’s exact test.

Abbreviations: HAEC, Hirschsprung-associated enterocolitis; HSCR, Hirschsprung disease; SD, standard deviation.

Table 3. Association between exclusive breastfeeding and risk of Hirschsprung-associated enterocolitis in the cohort study.

	Exclusive breastfeeding
		Yes
Outcome	No	RR (95% CI)	P
Overall Hirschsprung-associated enterocolitis
No. of HSCR patients	142	111
No. of enterocolitis	73	34
Unadjusted model	Reference	0.60 (0.43–0.82)	0.002
Multivariable-adjusted model^a	Reference	0.61 (0.44–0.85)	0.003
Postoperative Hirschsprung-associated enterocolitis
No. of HSCR patients without preoperative enterocolitis	88	86
No. of postoperative enterocolitis	19	9
Unadjusted model	Reference	0.48 (0.23–1.01)	0.05
Multivariable-adjusted model^a	Reference	0.52 (0.25–1.08)	0.08
Multivariable-adjusted model^b	Reference	0.51 (0.24–1.08)	0.08

^aAdjusted for gestational age (continuous), gender (female, male), mode of delivery (cesarean, vaginal), birth weight (continuous), and classical segment HSCR (yes, no).

^bAdditionally adjusted for ostomy (yes, no), age at pull-through surgery (≤180 or > 180 days), and surgery approach (open pull-through, totally transanal endorectal pull-through, Laparoscopic-assisted transanal pull-through).

Abbreviations: CI; confidence interval; HSCR, Hirschsprung disease; RR, relative risk.

Figure 2. The enteric microbiome predicted postoperative Hirschsprung-associated enterocolitis occurrence. (a) The α diversity in po-HAEC cases vs. controls. (b) Kyoto Encyclopedia of Genes and Genomes predicted the relative frequency of lipopolysaccharide biosynthesis proteins and secretion system in po-HAEC cases vs. controls. (c) Unsupervised hierarchical clustering of 131 OTUs with significantly different abundance between po-HAEC cases and controls. (d) Selective OTUs with significantly different abundance between po-HAEC cases and controls. (e) A 21-OTU signature that predicted po-HAEC occurrence. (f) Receiver operating characteristic curve for the 21-OTU signature.

Figure 3. Exclusive breastfeeding associated with an enteric microbiome and with lower occurrence of postoperative Hirschsprung-associated enterocolitis. (a) The α diversity in EBF vs. non-EBF patients. (b) Kyoto Encyclopedia of Genes and Genomes predicted the relative frequency of lipopolysaccharide biosynthesis proteins and secretion system in EBF vs. non-EBF patients. (c) Unsupervised hierarchical clustering of 51 OTUs with significantly different abundance between EBF and non-EBF patients. (d) Selective OTUs with significantly different abundance between EBF and non-EBF patients. (e) Linear discriminant analysis of the 51 OTUs separating exclusively breastfed non-HAEC subjects from the rest.

Figure 4. Enteric lipopolysaccharide concentrations in relation to exclusive breastfeeding and postoperative Hirschsprung-associated enterocolitis status. (a) LPS concentrations in EBF vs. non-EBF patients. (b) LPS concentrations in po-HAEC cases vs. controls. (c) LPS concentrations by combined categories of EBF and po-HAEC status.