Baseline characteristics

Table 1 shows the baseline characteristics of the enrolled patients. Clinical factors such as sex (p = 0.628), age (p = 0.174), body mass index (p = 0.757), and American Society of Anesthesiologists score (p = 0.628) were similar between the two groups. However, patients in the ileostomy group had upper (40.0%) and mid-to-low (60.0%) rectal cancers, whereas those in the control group had left-sided colon cancers (60.0%, p = 0.002). Accordingly, the surgical method was also different between the two groups (p = 0.007). Although the clinical (p = 0.513) and pathologic (p = 0.753) tumor stages were similar between the two groups, the ileostomy group had a higher proportion of patients who underwent nCRT than the control group (80.0% vs. 0.0%, p = 0.001), and the patients in the ileostomy group tended to have received more adjuvant chemotherapy (100.0% vs. 60.0%, p = 0.087) than those in the control group. The time interval between t1 and t2 was shorter in the ileostomy group than in the control group (6.0 ± 1.9 vs. 8.0 ± 2.1 months, p = 0.038).

Table 1 Baseline characteristicsAlpha diversity analysis—t 1 vs. t 2

In the ileostomy group, the complexity within the samples significantly decreased after ileostomy repair (t2) compared with that before the initial surgery (t1) in terms of the observed operational taxonomic units (OTUs) (p = 0.010) and Shannon index (p < 0.001). However, in the control group, no significant differences were observed in the OTUs (p = 0.650) and Shannon index (p = 0.880) between t1 and t2 (Fig. 1).

Fig. 1

Alpha diversity analysis: comparison between t1 and t2 in the ileostomy (A, B) and control (C, D) groups. Within-sample diversities were measured by observed operational taxonomic units (OTUs) (A, C) and the Shannon index (B, D)

Alpha diversity analysis—ileostomy vs. control groups

Before the initial surgery (t1), the two groups showed no significant differences in the complexity within samples (OTUs, p = 0.406; Shannon index, p = 0.226). However, at t2, the ileostomy group had significantly lower complexity than the control group (OTUs, p = 0.010; Shannon index, p < 0.010) (Additional file 1: Fig. S1).

Beta diversity analysis—t 1 vs. t 2

In the ileostomy group, principal coordinate analysis (PCoA) showed significant dissimilarities of the gut microbiota between the test before the initial surgery (t1) and after ileostomy repair (t2) (Jensen-Shannon, p = 0.001; generalized UniFrac, p = 0.001). However, in the control group, the beta diversity showed no significant dissimilarities between t1 and t2 (Jensen-Shannon, p = 0.121; generalized UniFrac, p = 0.096) (Fig. 2).

Fig. 2

PCoA 2D plots of beta diversity analysis: comparison between t1 and t2 in the ileostomy (A, B) and control (C, D) groups. Between-sample dissimilarities were measured by the Jensen-Shannon divergence (A, C) and generalized UniFrac distance (B, D)

Beta diversity analysis—ileostomy vs. control groups

At baseline (t1), the beta diversity showed no dissimilarities between the two groups (Jensen-Shannon, p = 0.501; generalized UniFrac, p = 0.470). However, at t2, the two groups had significant dissimilarities of the gut microbiota (Jensen-Shannon, p = 0.001; generalized UniFrac, p = 0.001) (Additional file 2: Fig. S2).

Composition of the fecal microbiota—t 1 vs. t 2

At the phylum level, in the ileostomy group, there was a significant decrease in the relative abundance of Bacteroidetes (26.1% vs. 12.1%, p = 0.023) and a significant increase in the relative abundance of Proteobacteria (5.8% vs. 17.9%, p = 0.016) between baseline (t1) and the time of ileostomy repair (t2) (Fig. 3). The Firmicutes/Bacteroidetes (F/B) ratio was significantly higher at t2 [median 21.6, interquartile range (IQR) 3.19–915.5] than at t1 (median 2.11, IQR 1.96–2.93, p = 0.034). However, in the control group, the relative abundance of Firmicutes significantly increased between t1 and t2 (54.4% vs. 67.5%, p = 0.010), while no significant difference was observed in the relative abundance of other phyla (Fig. 3).

Fig. 3

Relative abundance of bacteria at the phylum level: comparison between t1 and t2 in the ileostomy and control groups (*p < 0.05)

At the family level, the changes in Lachnospiraceae and Ruminococcaceae were noticeable. Between t1 and t2, a significant decrease was observed in the proportions of Lachnospiraceae (29.7% vs. 3.8%, p < 0.001) and Ruminococcaceae (16.5% vs. 0.6%, p < 0.001) in the ileostomy group, whereas no significant difference was observed in the control group (Lachnospiraceae, 22.0% vs. 29.9%, p = 0.050; Ruminococcaceae, 12.6% vs. 18.4%, p = 0.082).

At the genus level, in the ileostomy group, the proportions of beneficial bacteria such as Blautia (7.4% vs. 0.1%, p < 0.001), Prevotella (6.8% vs. 0.0%, p = 0.001), Faecalibacterium (6.0% vs. 0.2%, p = 0.002), and Akkermansia (0.8% vs. 0.0%, p = 0.002) decreased, whereas those of harmful bacteria such as Clostridium (0.8% vs. 16.2%, p = 0.005), Streptococcus (1.1% vs. 17.7%, p = 0.001), Enterococcus (0.1% vs. 3.7%, p = 0.001), and Acinetobacter (0.0% vs. 3.3%, p = 0.044) increased while the ileostomy was maintained between t1 and t2 (Fig. 4). In contrast, no specific tendency was found in the control group. Some beneficial bacteria increased (Faecalibacterium, 3.8% vs. 7.5%, p = 0.019), while some other beneficial and harmful bacteria decreased (Prevotella, 7.4% vs. 2.6%, p = 0.028; Streptococcus, 4.2% vs. 1.6%, p = 0.019) (Additional file 3: Fig. S3).

Fig. 4

Relative abundance of bacteria at the genus level: comparison between t1 and t2 in the ileostomy group (*p < 0.05, **p < 0.01)

Composition of the fecal microbiota—ileostomy vs. control group

The composition of the fecal microbiota was generally similar between the two groups at t1. However, at t2, a significant difference was observed in the proportion of the Proteobacteria (17.9% vs. 5.1%, p = 0.006) phylum between the ileostomy and control groups (Additional file 4: Fig. S4). In addition, the F/B ratio was higher in the ileostomy group (median 21.6, IQR 3.19–915.5) than in the control group (median 2.75, IQR 2.48–5.39) at t2; however, this difference was not found to be statistically significant (p = 0.131).

At the family level, the ileostomy group had a lower proportion of beneficial bacteria such as Lachnospiraceae (3.8% vs. 29.9%, p < 0.001) and Ruminococcaceae (0.6% vs. 18.4%, p < 0.001) and a higher level of Streptococcaceae (18.7% vs. 1.7%, p = 0.002) and Clostridiaceae (16.2% vs. 1.1%, p = 0.013) than the control group at t2.

At the genus level, the proportions of beneficial bacteria such as Blautia (0.1% vs. 9.1%, p < 0.001), Faecalibacterium (0.2% vs. 7.5%, p < 0.001), Bifidobacterium (0.6% vs. 4.8%, p = 0.01), and Akkermansia (0.0% vs. 0.1%, p = 0.013) were significantly lower, while those of harmful bacteria such as Clostridium (16.2% vs. 1.1%, p = 0.013), Streptococcus (17.7% vs. 1.6%, p = 0.002), Enterococcus (3.7% vs. 1.5%, p = 0.049), and Fusobacterium (1.5% vs. 0.3%, p = 0.019) were higher in the ileostomy group than those in the control group at t2 (Fig. 5).

Fig. 5

Relative abundance of bacteria at the genus level: comparison between the ileostomy and control groups, at t2 (*p < 0.05, **p < 0.01)

The compositions of the bacterial community at the level of genus, phylum, and species for the control and ileostomy groups at t1 and t2 are depicted in Fig. 6 and Additional file 5: Fig. S5 and Additional file 6: Fig. S6.

Fig. 6

Composition of the bacterial community at the genus level for the ileostomy and control groups at t1 and t2. The legend on the inferior side represents the 20 most abundant genera arranged in order of frequency