NEC is a devastating intestinal disease that affects 10-12% of preterm infants with a mortality between 10% and 30% [2, 3], and whose pathogenesis is still unclear. Our team has been able to reproduce this disease with an effectiveness of 73% [23] and, on this occasion, with an effectiveness of 100% and a mortality of up to 92% in Wistar rat neonates born with LPS, hypoxia and hypothermia.

This murine model has been developed based in other work such as the study by Miyake et al. [26] in 5-day-old C57BL/6 mice, where NEC is induced by feeding milk formula 3 times a day and only one LPS intake per day (dose 4mg/kg), via catheter/probe combined with 10 min hypoxia episodes 3 times a day (at 95% N₂ and 5% O₂)

Of note is the study by Zani et al. [24] using premature offspring of rats (Sprague-Dawley) obtained via caesarean section. Subsequently, NEC is induced by feeding milk formula 3 times daily, including LPS (dose 4 mg/kg) at the first feeding on day 1 and 2 of the model, via catheter/probe together with 10 min hypoxia episodes before each feeding (at 95% N₂ and 5% O₂)

Our model is very similar to the study by Zani et al. [24] with the difference that our offspring are Wistar neonates, since offspring obtained by caesarean section are excessively premature and pose many management difficulties in our environment. Even so, our model has shown adequate effectiveness [23] and would follow a procedure similar to that of other studies [27–30].

Variations have been described modifying the factors that are considered to be etiopathogenic in NEC [29, 30]: Hypoxia in % N₂ and O₂, the number of times and the time it is induced (from 50s to 10 min). Other models even Re oxygenate after hypoxia [31]; Hypothermia (if done), at 4°C in combination with hypoxia or after hypoxia with a time of 3 to 10 min [24–30, 32–37]; Administration (if any) of LPS (at doses of 1-20 mg/kg) [24, 26, 27, 32–34, 36].

Once the model and its effectiveness had been defined, the non-psychostimulant cannabinoid cannabidiol or CBD was chosen as a therapeutic option, due to its experimental use in several murine models of LPS-induced intestinal inflammation, in which it has shown multiple positive effects: reduction of histochemical damage and the promotion of positive regulation of the inflammatory response pathway (cytokines, chemokines and oxidative stress markers) [21]. Specifically, the preventive effect of CBD was observed in a Sprague-Dawley rat model of LPS-induced intestinal inflammation [21]. In this case, the protective effect of CBD was observed, reducing histopathological damage, TNF-α and Interleukin 6 (IL-6) levels along with the restoration of smooth muscle and its myoelectric activity [21].

Other studies on inflammatory bowel diseases have shown that endocannabinoids, thanks to the intestinal endocannabinoid system, are capable of inhibiting extensive pathways that release pro-inflammatory mediators such as Interleukin beta 1 (IL-ß1), multiple cytokines, the well-known TNF-α and nitric oxide, which are ultimately the main causes of these disorders [31, 38, 39]. Again, CBD specifically presents itself as a therapeutic approach to these conditions, as it not only has anti-inflammatory, antioxidant and anti-apoptotic properties for the central nervous system, but is also a major anti-inflammatory in the gut, with one of its main effects being the inhibition of neutrophilic chemotaxis [27, 38, 40, 41]. A case in point is seen in the Swiss mouse study by De Filippis et al. [27], testing the effect of CBD (at a dose of 10mg/kg) against acute infection and inflammation as a result of enteroglial protein S100B in combination with LPS (at a dose of 20mg/kg). CBD triggers inhibition against mainly neutrophil recruitment and also reduces the presence of macrophages, as represented by the intestinal macrophage-specific marker MAC3 signal. Consequently, there was a reduction in TNF-α kinase expression and intestinal histological damage observed with low caspase-3 expression.

Other studies with experimental models of NEC have assessed the protective effect of different substances such as Lactobacillus [35] vitamin B5 [42], colchicine [43] Teduglutin (glucagon-like peptide-2, GLP-2) [34] allopurinol [17] or daikenchuto (a mixture of medicinal herbs) [44]. However, none of these treatments have been successfully translated to the clinic, so there is still a need to investigate new possibilities. In this respect, cannabidiol is a very promising candidate.

Thus, the effect of CBD was studied in our NEC model, demonstrating its protective effect with an odds ratio (OR<1) value of 0.0255.

The dose of CBD used in this model is 10 mg/kg and is injected only once after induction of NEC. The dose was selected based on our group's previous studies of neonatal hypoxic-ischaemic brain damage [1–5, 25, 36, 37, 45].

The results show that, in the case of weight, the SHAM group increases moderately to almost 8 grams on the last feeding of the third day, i.e. corresponding to the weight of the usual control group [35, 37]. However, the VEH and CBD groups decrease or maintain the newborn weight, between 4 and 5 grams, with no differences between them. The same occurs in the study by Weis et al. [36], whose treatment consists of the implantation of human placental stem cells (hPSC), observing that the weight of the group with NEC and the group with NEC and treated decreases as the model progresses, worsening due to the disease itself, with no significant statistical difference to support this (p<0.03). On the other hand, in other studies using Infliximab [35] (TNF-α antagonist) and Tocilizumab [37] (Interleukin-6 signal inhibitor) as treatment, there are significant differences between the weights of the NEC group and the NEC and treatment group, with a p-value of p<0.001 and p<0.0001 respectively.

Although the calculations of volume and calories administered in the intakes of the administered milk meet the requirements established for the weight of newborns (4 Kcal/g per day) [46], the severity of the NEC manifests a clinical impact that has not been reversed by CBD treatment, which may be a sign of an underestimation of caloric requirements.

In the macroscopic study of the tissue and compared to the healthy intestine of the SHAM group, the only notable difference between the VEH and CBD groups was the difference in colour when the viscera were removed, as the consistency and dilatation corresponded to severe NEC.

The mortality of the model (92%) includes the death caused by oesophageal perforation estimated in this model (5%). In this aspect, the effect of CBD is clear, as it manages to reduce mortality by almost 24% (68%).

The analysis of the histological NEC damage grade of the groups also shows changes with respect to the model. Compared to the SHAM group, with a grade 0 for both the small and large intestine, due to its normal and healthy state, the VEH group presents, in most of the tissue, a grade 2 and 3 NEC injury. On the other hand, the CBD group has mostly grade 1 and 2 lesions.

The main difference in our method of assessment compared to other studies [33–37] is that we aim to observe the progressive change of the tissue and thus the progressive effect of the treatment (CBD). This is why we do not label as total NEC those tissues with grade 2 or higher histological involvement by NEC but quantify the percentage of grades present in the tissue based on the total present. In other studies, such as that of Feng et al. [33], where a heparin-binding epidermal growth factor (HB-EGF) is used as treatment, that of Nakame et al. [34], whose treatment consists of a trophic gut hormone, glucagon-like peptide 2 (GLP-2), and the studies by Tayman et al. [35], Weis et al. [36] and Yarci et al. [37] that we are already familiar with, the tissue with grade 2 or higher NEC damage is evaluated as NEC. All these studies also show statistically significant differences between the NEC and NEC with treatment groups. Specifically in the studies by Feng et al. [33] and Nakame et al. [34], the reduction in damage is approximately 25%, with a p-value of p<0.05. In the study by Weis et al. [36] this reduction is less, 9% with a p-value of p<0.0001, while in the studies by Tayman et al. [35] and Yarci et al. [37] this reduction is not quantified in percentage, but these data show a significant difference with p-values of p<0.002 and p<0.001 respectively.

For the immunohistochemical study, the technique for identifying or marking cell apoptosis (TUNEL) and the technique for identifying microglia and macrophages (Iba-1) have been used. On the one hand, the TUNEL technique is highly effective in marking cell death in various tissues [46, 47]. On the other hand, the Iba-1 technique, which, in addition to being used in studies of the nervous system (microglia), has been frequently used for the general identification of macrophages in intestinal tissue [48].

In the case of our results, we observed a high level of cell death or apoptosis in the intestinal tissue of the VEH group, in contrast to the SHAM and CBD groups. Specifically, there is a wide range of variation in the tissue of the CBD group, being similar to the SHAM group, which again shows the protective effect, in this case immunohistochemical, of CBD. Similar studies observe a reduction in histological damage due to a reduction in the number of apoptotic cells per area [33, 35, 37].

In other studies, such as those by Feng et al. [33], Tayman et al. [35] and Yarci et al. [37], of which we are aware, significant differences are also observed with the TUNEL technique between the groups with NEC and those with NEC and treatment. In other words, a reduction in histological damage is observed due to a reduction in the number of apoptotic cells per area. In this case the first study uses a p-value of p<0.05, like us, while the studies of Tayman et al [35] and Yarci et al [37] use p<0.001 and p<0.002 respectively.

Regarding the Iba-1 study, we observed that the SHAM group shows high fluorescence and a large amount of macrophages, on average 47.84%, with the lamina propria being the most fluorescent layer, i.e., the data correlates with the normal, healthy state of the intestine. However, the VEH and CBD groups show very low Iba-1 fluorescence, with an average of 2.85% and 6.61% respectively.

Our initial hypothesis assumed a progressive decrease in the Iba-1 signal (macrophages) as intestinal damage and inflammation sign. As such the group with the highest signal, after the results, corresponds to SHAM, which confirms our hypothesis. Thus, the VEH and CBD groups would be the ones with the least Iba-1 signal due to the damage and inflammation produced by the NEC. This is due to the displacement of macrophages from the lamina propria by polynuclear neutrophils formed as a result of the immune system's response to necrotic damage [27].

Although there are no significant differences for the Iba-1 signal between the VEH and CBD groups, in the case of VEH this may be due to the extreme necrosis of the tissue demonstrated by the grade 3 NEC damage (Fig. 3) and the apoptotic death of the tissue demonstrated by the TUNEL technique (Fig. 4A). On the other hand, the low Iba-1 signal in the CBD group may be due to the effect of CBD itself on the immune system, as the histological damage grades show that it has low levels of NEC (Fig. 3) correlated with low apoptotic cell death (Fig. 4A). In other words, the CBD group is in better histopathological condition and the low Iba-1 fluorescence may be due to the anti-inflammatory effect of CBD itself [49]. These results are in line with those shown in the study by De Filippis et al. [27].