The previously reported study revealed retention of the X-ray shadow of barium sulfate administrated into the rat’s stomach that was associated with acute cyclophosphamide toxicity [1] and was typical for gastrointestinal (GI) stasis. This potentially fatal complication often occurs during the patient’s preparation for hematopoietic stem cell transplantation, myeloablative cytostatic conditioning [2]; in some recipients, the X-ray gastric shadow spreads to large parts of both abdominal and thoracic cavities [3]. The GI stasis clinical significance results from its negative impact on the chemotherapy outcome. GI stasis hampers the patients’ feeding, makes it pointless to prescribe oral medications; intestinal bacterial overgrowth that is associated with GI stasis results in realization of quorum sensing [4], enhanced production of toxic substances by bacteria, as well as in endotoxemia and endotoxicosis. GI stasis contributes to the intestinal barrier damage associated with the Gram-negative bacterial lipopolysaccharide release into the bloodstream, systemic inflammation [5] and sepsis [6]. GI stasis is involved in the pathogenesis of delayed chemotherapy-induced vomiting; however, prescribing antiemetics [7, 8], prokinetics or antispasmodics [7] to such patients does not resolve the acute chemotherapy-induced gastrointestinal mucositis, a defensive response to the first stage of which is believed to be represented by GI stasis [9]. The use of enteroprotectors, i.e. medications capable of preventing acute gastrointestinal mucositis, seems to be a more promising approach to prevention of GI stasis associated with myeloablative conditioning. When using cytostatic therapy with cyclophosphamide (CP), enteroprotective effect can hypothetically be achieved by alkalinization of the chyme moving through the gastrointestinal tract. With increasing рН of biological media, spontaneous hydrolysis of the active CP metabolite produced in the liver is slowed down, accumulation of more toxic metabolites, acrolein and phosphoramide mustard, is inhibited [10], and toxicity of aldehyde dehydrogenase, the key enzyme responsible for the CP detoxification, the optimal pH value of which is within the alkaline range, is increased [11]. Higher normal pH values of the cytoplasm relative to the tumor cells contribute to the CP selective anticancer activity [12]. Acidification of the chyme by intragastric administration of weak acid solutions [13] or lactulose [14] increased the severity of acute cyclophosphamide toxicity in rats. Intragastric CP administration resulted in more severe GI stasis than intraperitoneal injection of the same dose, which could be partly due to more intense CP toxification in the acidic environment [1]. The study was aimed to test the hypothesis that NaHCO₃, the alkalinizing agent administrated by gavage in the rat model of myeloablative cytostatic therapy with cyclophosphamide, would have a protective effect against GI stasis.

METHODS

The study involved 140 male Wistar rats (161–190 g) obtained from the Rappolovo laboratory animal nursery. The animals were treated in accordance with the Principles of Good Laboratory Practice, stated in the Order № 708n of the Ministry of Health of the Russian Federation of 1 August 2010. The standard rat diet and ad libitum water access were provided. The animals were randomized into experimental groups. To deprive the rats of food, they were placed in the slatter floor cages (to avoid coprophagy and consumption of the bedding components) with access to water only for a specified time. Myeloablative cytostatic therapy was modeled by a single lateral tail vein injection of the freshly prepared aqueous solution of Endoxan (Baxter Oncology GmbH; Germany) in the amount of 10 mL/kg in a dose of 390 mg/kg (≈ 1.7 LD_{99/30 day}), which was equivalent to the daily dosage for humans of 60 mg/kg used in the myeloablative conditioning regimens [15]. Laparotomy and organ harvesting were performed under the mask halothane anaesthesia. The GI stasis severity was assessed based on the relative weight of chyme in the stomach and caecum calculated as a difference between the weight of the organ filled with chyme and the empty organ (gaster, caecum) in grams relative to the body weight in kilograms.

During the first phase of the study we assessed the dynamics of the GI stasis development after the myeloablative conditioning. For that the animals were distributed into seven groups, among which the first one was represented by intact rats (n = 10) having unlimited access to food, another tree groups were deprived of food 4, 24 or 48 h after the CP administration, and the three remaining groups were deprived of food within the same time frame but did not receive CP (n = 10 in each group). All animals were subjected to laparotomy 72 h after the CP administration to assess the GI stasis severity.

During the second phase of the study we assessed the effects of NaHCO₃ on the GI stasis severity and the growth rate of gastrointestinal microbiota. For that the rats deprived of food between 24 and 72 h after the CP administration were used. The animals were distributed into five groups, among which the first one was represented by intact rats (n = 10), and all other groups were represented by the rats administered with CP (n = 10 in each group). Rats of the second groups were administered CP only; the 4% NaHCO₃ solution (pH = 8.34) in the amount of 15 mL/kg was administrated by gavage to rats in the third group 30 min before the CP administration. In the fourth group, administration of NaHCO₃ 30 min before CP was supplemented by the repeated administration of NaHCO₃ in the same dose after the CP injection; in the fifth group, NaHCO₃ was administrated by gavage four times: 30 min before, immediately after, 60 and 120 min after the CP administration. All the rats were placed in metabolic cages for urine collection 48 h after the CP injection; 50 µL of the 10% trichloroacetic acid solution per chamber were added to the urinal chambers as a preservative. The GI stasis severity was assessed 72 h after the CP administration. To assess selectivity of the NaHCO₃ protective effect, the relative weight of the spleen was measured along with the relative weight of the gaster and caecum chyme as a measure of chemotherapy-induced damage to the hemopoietic system. Urinary excretion of indican was used as a measure of the gastrointestinal microbiota growth rate [16]. The volume of urine sampled within 24 h was measured, and indican, an intestinal endotoxicosis indicator [17], the urinary excretion of which was measured in micrograms per kilogram of body weight per hour, was quantified.

During the third phase of the study we assessed morphological changes in the small intestine associated with the myeloablative cytostatic therapy modeling by the above method, as well as the effects of double NaHCO₃ administration into the stomach (30 min before and immediately after CP) on these changes. The 10 cm long small intestine sections (duodenum proximal of pylorus; jejunum 10 cm distal of flexura duodenojejunalis; ileum proximal of caecum) were fixed in 10% formalin and embedded in paraffin. The annular slices were stained with hematoxylin and eosin and then examined with the 3DHISTECH Pannoramic MIDI scanning digital microscope (Carl Zeiss AG; Germany). We enumerated intestinal villi and measured their length in 58–73 slices of each organ obtained from three animals; the results were processed using the Case Viewer application (3DHISTECH Ltd.; Hungary).

The results were presented as the mean and standard error of the mean (M ± m). The effects of the injected substances on the quantitative parameters were estimated by analysis of variance. When the resulting models were significant, the intergroup comparison of mean values was performed using the Tukey's honestly significant difference test [18]. Correlations between traits were represented as the Spearman's rank correlation coefficients (r_s). The α-value of 0.05 was considered to be a critical significance level.

RESULTS

Three days after the CP administration to rats deprived of food within 48 h before laparotomy, the dilated stomach that was filled with chyme occupied most of the abdominal cavity; it seemed to be empty in intact animals. The increase in the volume of the caecum associated with the CP administration was lower (fig. 1). Food consumed within the first 24 h after the CP injection stayed in the stomach over the next 48 h. This resulted in the 7–13 fold gastric chyme relative weight increase. Excessive accumulation of chyme in the caecum was represented as a trend (p = 0.075 for the animals deprived of food for 4 h; fig. 2). The body weight of animals administered with CP measured on the day of laparotomy made up a smaller share of body weight measured prior to exposure, than in controls: 78.9 ± 0.8% vs. 86.2 ± 0.5% (p < 0.05).

Administration of NaHCO₃ into the stomach 30 min before and immediately after the CP administration prevented GI stasis: the gastric chyme relative weight was on average 2.6 times lower, than in unprotected animals, however, it was still three times higher than in intact rats. The decrease in the relative weight of the caecal chyme was represented as a trend (p = 0.084). The fourtime administration of NaHCO₃ had no benefit over the double administration, and a single preventive NaHCO₃ administration was ineffective. The NaHCO₃ administration had little effect on the CP-induced hypotrophy of the spleen (fig. 3).

Urinary excretion of indican after the exposure to CP only was on average 1.9 times higher than in intact rats; the double intragastric NaHCO₃ administration resulted in the 1.4-fold increase, which manifested as a trend only (p = 0.067; fig. 4). In the rats administered with CP, this indicator showed a strong negative correlation (r = –0.77; p < 0.01) with the body weight measured on a day of laparotomy as a percentage of body weight measured before the exposure. Excretion of indican in rats administered with CP only positively correlated with the relative weight of the caecal chyme (r = 0.66; p < 0.05); when NaHCO₃ was administered together with CP, the correlation was weak (r = 0.15).

Three days after the CP administration, the changes (congestion, inflammation, atrophy) were observed in the small intestine, the severity of which increased in the direction from the duodenum to the ileum. The average length of intestinal villi was reduced, and a downward trend in the number of villi was observed in the annular slices of the organ. Atrophy of villi was found in the ileum. No such alterations were found in rats administered with NaHCO₃ in addition to CP (table; fig. 5).

DISCUSSION

Modelling myeloablative cytostatic therapy in rats was associated with deep inhibition of the gastrointestinal (GI) tract propulsive function with predominance of gastric stasis developing during the first hours after the CP administration. The gastric transit time exceeded three days. It is 10–48 min in healthy people [19], that is why it can be assumed that gastric stasis persists in the recipients of hematopoietic stem cells for much of the myeloablative conditioning course lasting 3–5 days [15]. It is possible that gastric stasis is involved in general health deterioration during myeloablative cytostatic therapy, as indicated by the fact that the animals lose about a quarter of their initial body weight within three days after the CP administration.

Intragastric administration of two doses of NaHCO₃ to a total dose, which was equivalent to 350 mL of the 4% solution in humans, made it possible to a significant extent, although not completely, preserve the propulsive function of the stomach. NaHCO₃ was most effective during the period that did not exceed the CP T_½ after intravenous administration to rats, i.e. 0.5 h [20]. Consequently, inhibition of the CP toxification predominated in the mechanism underlying the NaHCO₃ protective effect. Because of the weakly alkaline nature of the NaHCO₃ solution and the salt’s capability of being absorbed by the GI tract mucosa, the protective effect involved inhibition of the acrolein and phosphoramide mustard production in epithelial cells of the stomach and/or small intestine. This is also indicated by the NaHCO₃ inability to prevent the CP-induced hypotrophy of the spleen that can be explained by the buffering properties of blood not allowing one to ensure the comparable increase in рН of cells of the stem or proliferative pool of the hematopoietic system. Thus, the NaHCO₃ protective effect of the GI tract was selective, which was conductive to its testing in disorders requiring myeloablative cytostatic therapy.

The pH values that are optimal for the enzymes responsible for DNA repair are within the range close to neutral (6.5–7.5), that is why acidosis, that results from the shift from oxidative to glycolytic phosphorylation and comes along with mitochondrial damage caused by mustard agents [21], can violate DNA repair. Acidosis also leads to another effect that contributes to enterocytopenia: the expression of pro-apoptotic proteins that activate caspases [22]. This determines the possibility of the NaHCO₃ protective effects not only in acute poisoning with CP, but in poisoning with other alkylating cytostatic agents; the hypothesis needs further testing.

The CP-induced gastric stasis was associated with the increase in urinary excretion of indican (the indoxyl sulfate potassium salt), the end product of the liver metabolism of indole (oxidation to indoxyl and its sulfonation), the main source of which in experimental rats was represented by the reaction catalyzed by the gut microbiota tryptophanase. This indicates the increase in the weight and/or metabolic activity of intestinal microbiota, that produces indole but does not metabolize it, in GI stasis; in this case, a more intense production of both indole and another toxic product of the tryptophanase reaction, ammonia, in the GI tract is inevitable. The CP-induced increase in excretion of indican was lower that the increase in the relative weight of the gastric chyme, which was probably due to lower resorption capacity of the stomach compared to the intestine. When NaHCO₃ was administered, the differences in excretion of indican between the intact animals and the rats administered with CP were minor, which characterized the endotoxemia severity in the latter.

The myeloablative cytostatic therapy modeling was associated with damage to the small intestine. The severity of damage increased distally, as determined by the fact that the abundance of bacteria in the ileal chyme exceeded that in the duodenal chyme by four orders of magnitude [23]. This is consistent with the hypothesis that gastric stasis represents the body’s protective response, the biological meaning of which is to prevent injury of the small intestine damaged by the cytostatic agent, since the small intestine is a GI tract segment that is most sensitive to cytostatic agents [24]. The use of NaHCO₃ prevented dystrophic changes in all parts of the small intestine. As for duodenum and proximal jejunum, such an effect can be explained by local alkalinization, however, the NaHCO₃ protective effect on the ileum requires further investigation. This is due to the fact that the time it takes for the NaHCO₃ solution to reach it (at least 3 h according to preliminary data acquired by monitoring methylene blue administration in the rat stomach) far exceeds T_½ of CP administered to rats by intravenous injection [20].

The problem of acute chemotherapy-induced gastrointestinal mucositis treatment is far from being resolved. Antioxidants, anti-inflammatory agents and inhibitors of apoptosis are considered as possible therapeutic agents [25]. However, delivery of these agents to the GI tract segment most sensitive to cytostatic agents, the small intestine, is only achieved when there is no gastric stasis. That is why drug treatment should be preceded by the use of medications for emergency prevention of gastric stasis. The findings indicate potential benefits from early prescription of oral alkalinizing agents, such as sodium bicarbonate, for this purpose.

CONCLUSIONS

1) Single intravenous injection of cyclophosphamide in a dose equivalent to that used for myeloablative conditioning to rats results in dystrophic changes in the small intestinal mucosa, the development of gastrointestinal stasis with predominant gastric stasis within 24 h, and the excess growth of the indole-producing gastrointestinal microbiota. 2) Intragastric administration of sodium bicarbonate in a dose equivalent to 350 mL of 4% sodium bicarbonate solution in humans to rats 30 min before and immediately after the cyclophosphamide administration to a considerable extent prevents acute chemotherapy-induced gastrointestinal mucositis, gastric stasis, and the excess growth of gastrointestinal microbiota. 3) Early oral administration of sodium bicarbonate represents a promising approach to prevention of gastric stasis when performing myeloablative chemotherapy with cyclophosphamide; the approach enables further oral administration of drugs for treatment of acute chemotherapy-induced mucositis of the small intestine to patients. 4) Testing of the above approach to prevention of gastric stasis associated with myeloablative cytostatic therapy with cyclophosphamide and other alkylating cytostatic drugs is of interest.