In supralethal doses, cyclophosphamide (CP) is used for myeloablation with the aim of preparing recipients to hematopoietic stem cell transplantation. Side effects of such treatment include inhibition of gastrointestinal motility, cytostatic damage to the small intestine's epithelium, and endotoxemia. These complications disrupt enteral nutrition, reduce the effectiveness of orally administered drugs, and increase the likelihood of death of patients before transplantation. Gastrostasis and acute small intestinal mucositis, prevented by introduction of sodium bicarbonate solution (NaHCO3) into the stomach, were observed in rats after administration of a myeloablation dose of CP [1]. The currently unclear matters are the nature of the relationship between gastrointestinal stasis, enterocytopenia, and endotoxemia, and how NaHCO₃ or hydrochloric acid (agents that change the pH of gastric chyme) affect them. This study aimed to identify the relationship between gastrointestinal stasis, damage to the mucous membrane of the small intestine, and acute intestinal endotoxemia in the context of modeling myeloablation cytostatic therapy in rats.

METHODS

We used male albino Wistar rats weighing 161–190 g (branch of the Kurchatov Institute ‒ PNPI – Rappolovo Laboratory Animal Nursery; Russia). The animals received standard rat feed and drinking water ad libitum. We randomized the animals into four groups: 1) intact animals; 2) those who received CP only; 3) those who received CP and NaHCO3; 4) rats that received CP and HCl. Twenty-four hours after administration of CP, rats were placed in cages with a lattice floor that prevented coprophagia and eating of litter, with access only to water. Myeloablation was triggered by a single injection (into the tail's lateral vein) of a freshly prepared aqueous solution of Endoxan, a CP drug (Baxter Oncology GMBH; Germany). The dose was 390 mg /kg (≈ 1.7 LD_{99/30 days}), the volume — 10 ml/kg. This dose of CP was 1.7 LD_{99/30 days}, which allowed the rats to survive for at least 3 days after administration. Twice, 30 minutes before and immediately after administration of CP, the animals received intragastric injections of 0.48 M NaHCO₃ solution (pH = 8.34) in a volume of 15 ml/kg, or 0.1 M HCl solution (pH = 1) in the same volume. The animals were examined 72 hours after administration of CP.

The rats were put under halothane anesthesia for blood sampling from v. portae and organ extraction. To gage the gastrointestinal tract's (GIT) propulsive function, we measured the relative mass of gastric and intestinal chyme, which was calculated as the difference (in grams) between the mass of a chyme-filled and an empty organ (gaster, caecum), related to body weight (in kilograms). In parallel, we measured the relative mass of the spleen in order to assess the selectivity of action of NaHCO₃ or HCl.

To assess the severity of enterocytopenia in the small intestine's tissues, we assessed the activity of enterocyte markers: alkaline phosphatase, alkaline phosphatase (AP) [2], and cholinesterase (CE) [3]. Cranial segments of duodenum, caudal segments of jejunum and ilum (length — 4 cm each) were homogenized in a 15-fold volume of tris-HCl buffer (50 mM, pH = 7.4) and frozen at –20 °C. After 15 hours, the homogenates were thawed at 4 °C and centrifuged at 2000 g for 10 minutes. In the supernatant, we established the activity of AP with the help of the kinetic optimized method and using a set of reagents (Olvex Diagnosticum; Russia), and assessed the activity of CE using Ellman's assay, substrate of acetylthiocholine iodide (SigmaAldrich; USA), and ChemWell 2910 biochemical analyzer (Awareness Tech.; USA). Another determined parameter was protein content, which we quantified using the Bradford method.

The assessment of intestinal endotoxemia was based on the content of endotoxin and ammonia in portal blood plasma, and urea, which is the product of neutralization of ammonia by the liver. Endotoxin was detected with a LAL reagent in the gel-thrombus test modification, the process enabled by the ALPYR Test reagent kit (Algimed Techno LLC; Russia). To increase sensitivity, we incubated the samples at 70 °C for 15 minutes before diluting them and mixing with the LAL reagent, thus unbinding the endotoxin from albumin [4]. Ammonia concentration was determined spectrophotometrically using the Ammonium Ultra reagent kit (Sentinel Diagnostics; Italy). Concentration of urea was established after its hydrolysis to ammonia, with the help of the Urea UV reagent kit (Biosystems; Spain). Concurrently, we measured blood plasma concentrations of creatinine (reaction with picric acid), albumin (reaction with bromocresol green), and total protein (reaction with acid blue 90).

Urinary excretion of indican was used as an indicator of excessive bacterial growth [5]; for this purpose, we collected urine while the animals were in metabolic chambers from 48^th to 72^nd hour after administration of CP. Indican was quantified with Obermayer reagent [6], and excretion was expressed in micrograms per kilogram of body weight per hour.

The results were given as a mean and an error of mean (M ± m). We applied ANOVA to assess the effect of the administered substances on the studied quantitative indicators. When the models reached significance, the means between groups were compared through Tukey's honest significant difference test [7]. The differences were considered significant at p = 0.05.

RESULTS

After two days of fasting, chyme was concentrated only in gaster and caecum. Animals that received CP had the relative mass of gastric chyme 10.5 times greater than intact animals, while the mass index of the caecal chyme did not differ significantly between the groups. Both NaHCO₃ and HCl partially prevented the overfilling of stomach with chyme, but did not change the volume of content of caecum (fig. 1).

Administration of CP led to enterocytopenia. The activity of AP decreased 1.6–4.9 times in all parts of the small intestine, most significantly in the ileum; the activity of ACE was reduced only in the ileum. NaHCO₃ partially prevented the decrease of activity of AP in duodenum (fig. 2). Neither NaHCO₃ nor HCl influenced the systemic cytopenic effect of CP, which made the spleen's relative weight 57% smaller (fig. 1).

In intact rats, the portal blood plasma ammonia content was 0.88 mM, three times higher in plasma after decapitation [8]. Administration of CP did not change the blood ammonia level significantly, but content of urea has grown 4.6 times. NaHCO₃, administered by gavage, made the level of urea increase further, and HCl halved it, but the said level still remained higher than in the control group. After administration of CP, the portal blood endotoxin content was 4 times higher than in intact animals, with NaHCO₃ or HCl having little effect on this parameter. The concentration of creatinine in blood of rats that received CP was twice as high as in intact animals, and remained largely the same after administration of NaHCO₃ or HCl. There were no significant intergroup differences in the content of total protein or albumin in blood plasma (Figure 3). Administration of CP intensified urinary excretion of indican two-fold, with HCl unable to change the respective dynamics.

Intragastric injection of NaHCO₃ induced hyperindicanuria as a trend (fig. 4).

DISCUSSION

After the animals are deprived of food, the mass of their gastric contents is determined by the propulsive function of the stomach. Therefore, the overfilling of stomach with chyme after administration of CP, which we registered in this and the previous study [1], reflected the development of gastrostasis. If the change in gastrointestinal motility were limited to gastrostasis, then the flow of chyme into the cecum would have happened after its emptying, and the relative mass of chyme therein would have been decreasing. However, against the background of administration of CP, the mass of the caecal chyme did not change significantly (fig. 1). This means that CP inhibited the propulsive function not only in the stomach, but also in the colon.

It can be assumed that gastrostasis was a protective response aimed at preventing chyme from injuring the small intestine, the part of the gastrointestinal tract most sensitive to CP. The damage thereto manifested as enterocytopenia, as indicated by the decreasing content of enterocyte markers in the small intestine's tissues (fig. 2). Another consequence of gastrostasis could be restricted delivery of substrates to bacteria vegetating in the lumen of the colon. In its chyme, the specific content of bacteria is eight orders of magnitude higher than in the chyme of the stomach [9], which makes colonic microflora the source of endotoxemia. Given the division frequency of 3 h⁻¹ (the mean for Escherichia coli at 37 °C), it is theoretically possible that the amount of bacteria increases eightfold within an hour after colon loses the propulsive function. Substrate limitations stemming from gastrostasis could restrain such a rapid growth of colonic microflora, as well as production of ammonia thereby. However, gastrointestinal stasis could not suppress ammonia-producing wall microflora, the substrates for which are substances diffusing to the luminal surface of the gastrointestinal mucosa from the blood. Colon stasis, which prevented discharge of the ammonia-producing microflora from the body, could have promoted production of ammonia in the caecum. Ammonia produced by the intestinal microflora becomes part of urea in the liver. Its content in the blood of animals that received CP increased fourfold three days after administration (fig. 3), which means that production of ammonia in the intestine was intensified earlier. This hypothesis is supported by an almost twofold spike of the level of blood ammonia 3 hours after administration of myeloablation dose of CP [8]. Partially, uremia could also have been associated with a delay in the excretion of urea from the body, as backed by a twofold increase of the blood level of creatinine, a marker of renal insufficiency. Therefore, the increase in blood urea levels registered in this study 3 days after administration of CP was a marker of acute intestinal endotoxemia of a mixed type, productive and retention.

In the animals that received CP, gram-negative bacteria, which are the source of endotoxin, could be concentrated in the stomach and caecum, since other parts of the gastrointestinal tract were free of chyme. The content of endotoxin in the colonic chyme is close to 2.5 g/l [10], and its release from bacteria could have been intensified by their death due to substrate limitations imposed by gastrostasis. Colonic stasis increased the duration of endotoxin contact with the sorbing surface of the mucous membrane, therefore portal endotoxinemia after administration of CP (fig. 3) could have been both productive and redistributive. The lack of effect of CP on the blood plasma protein level indicates that the content of endotoxin in biologically active free form increased in proportion to its total plasma content.

Indican is the end product of metabolism of indole, the only source of which in the experimental conditions was a reaction catalyzed by tryptophanase of the intestinal microflora. Hyperindicanuria is a valid indicator of excessive growth of indole-producing bacteria in the gastrointestinal tract [5]. Urinary excretion of indican intensified (fig. 4) despite the impairments of the renal excretory function, which were confirmed by the increased blood level of creatinine (Figure 3). In the CP group, this indicates the predominance of productive and (or) redistributive component of intestinal endotoxemia over the retention component.

Both NaHCO₃ and HCl partially prevented gastric overfilling in rats (fig. 1), and only NaHCO₃ could stop enterocytopenia to a certain degree. The alkalizing effect of NaHCO₃ was predominantly local, since it prevented enterocytopenia only in the duodenum, but not in the caudal parts of the small intestine and not in the spleen (fig. 2). This supports the hypothesis [1] that has lower cytostatic damage to the mucous membrane's epithelium a probable mechanism of prevention of gastrostasis (induced by CP) by NaHCO₃. Administration of HCl partially prevented gastric overfilling, which could have been the result of the boost it gave to pepsin [11], and accelerated digestion of feed consumed by the animals in the next 24 hours. This explanation is supported by data on the depressing effect of CP on gastric secretion [12].

The severity of intestinal endotoxemia depends on the intensity of release of toxicants by the intestinal microflora, permeability of the enterohematic barrier, and the rate of their excretion from the body. Rats that received CP had the cytostatic damage most pronounced in the small intestine [13], while the intestinal microflora was concentrated in the stomach and caecum, where mucosal epithelium is more resistant to cytostatics. Therefore, cytostatic damage to the small intestine could not directly affect the appearance of intestinal toxicants in blood. This is further evidenced by the lack of a significant effect of NaHCO₃, which reduced the severity of enterocytopenia, on the portal blood level of endotoxin (fig. 3) and severity of indicanuria (fig. 4) in rats that received CP. The difference in the effect of NaHCO₃ and HCl on the portal blood level urea, product of neutralization of ammonia (fig. 2), could have been conditioned by ammonia becoming NH₃ in the alkaline medium, a free form that easily diffuses through biomembranes [14], while in the acidic medium ammonia was in its ionized form NH4⁺, which can hardly penetrate enterohematic barrier.

Thus, in a rat model of myeloablative cytostatic therapy, prevention of gastrostasis or cytostatic damage to the small intestine — complications of the said therapy — does not suppress acute intestinal endotoxemia. It is feasible to consider intragastric administration of NaHCO₃ in combination with agents suppressing vegetation of the colonic microflora when developing measures to prevent complications of myeloablation cytostatic therapy. These measures can be supplemented by enteric detoxification (enterosorption, intestinal lavage) aimed at removing endogenous toxicants from the sites of their secretion [15].

CONCLUSIONS

Intravenous administration of a myeloablation dose of cyclophosphamide to rats causes gastrointestinal stasis, cytostatic damage to the mucous membrane of the small intestine, and development of a mixed-type acute gut-derived endotoxemia.  The latter, in a rat model of myeloablative cytostatic therapy, was predominantly caused by penetration into blood of waste products of the caecum microflora. Intragastric administration of weak solutions of sodium bicarbonate or hydrochloric acid helps prevent gastrostasis triggered in rats by the myeloablative effect of cyclophosphamide, but these solutions do not stop acute intestinal endotoxemia. Intragastric administration of weak solutions of sodium bicarbonate prevents enterocytopenia triggered in rats by the myeloablative effect of cyclophosphamide, but it cannot halt the development of acute intestinal endotoxemia. For the measures designed to counter gastrointestinal toxicity of cyclophosphamide in the context of myeloablative cytostatic therapy, it is feasible to consider administration of sodium bicarbonate in combination with agents for enteric detoxification and suppression of vegetation of the colonic microflora.