Impact of Stroke Therapy Academic Industry Roundtable (STAIR) Guidelines on Peri-Anesthesia Care for Rat Models of Stroke: A Meta-Analysis Comparing the Years 2005 and 2015

Eligibility and selection of the papers

Both sets of search results were screened for relevance and eligibility. All references corresponding to studies written in languages other than English were excluded. The remaining references were each attributed a number and random selection proceeded until 100 articles for each year had been chosen [random.org] according to the following criteria:

1. The full text of the manuscript was available online via the library of the University of Liège.

2. The reference corresponds to an original study. More specifically, literature reviews, books and book chapters, conference proceedings, reports, and guidelines were excluded.

3. Only studies reporting the use of adult or geriatric rats were included. Studies using neonatal rats, rats younger than eight weeks or weighing less than 220g were excluded.

4. The rats were used in an in vivo model of stroke.

5. The stroke model required surgery, and the rats were expected to regain consciousness after surgery.

Only references fulfilling all five criteria were retained for further analysis.

Data extraction

Two authors (AT and CS) independently screened both sets of included references then reached consensus for each variable. The information on variables, as detailed in Table 2, was sought from each manuscript. Where the information was not reported in the manuscript, the corresponding data was recorded as ‘not reported' (NR).

Model

Focal brain ischemia produced by various means of MCAO was the most commonly used model. Significantly less MCAO was produced by techniques including thrombosis in 2015 (n = 0) than in 2005 (n = 17) (p = 0.0001).

Anaesthetic agents

The STAIR guidelines offer the following guidance regarding the choice of anaesthetic agent:

“When designing a preclinical study for neuroprotection, the protection provided by anesthetics should be taken into account. When neurotransmitters or neuroplasticity are the main foci of a study, anesthetics such as urethane, which do not disturb the action of neurotransmitters should be used.” [13].

Anaesthetic agents used for induction and maintenance are listed in Table 3. In 2005, 79 studies did not report the induction agent, compared to 73 in 2015. Maintenance of anaesthesia (or the name of the chosen molecule) was not reported in 8 and 10 of the 100 reports analysed in 2005 and 2015, respectively. Halothane as a maintenance agent was significantly less often used in 2015.

Intubation and ventilation

The STAIR guidelines offer the following guidance regarding the choice of anaesthetic agent:

“The importance of using mechanical ventilation should be determined by the anticipated impact of the surgical/anaestheic procedure on respiratory function. The potential confounding effects from respiratory functional deficits can be minimized by the use of mechanical ventilation. Unnecessary use of mechanical ventilation should be avoided when a particular MCAO model is not likely to cause respiratory problems.

Ventilation may be needed when the operation lasts long (>1 hour) and when the ischemia affects brain stem function. A mixture of 30%:70% (O2:N2 or N2O) may be used for preclinical stroke trials combined with individualised adjustment of ventilator parameters” [13].

The majority of the studies did not report the mode of ventilation (70 in 2005, 85 in 2015) or the use of oxygen supplementation (55 in 2005, 72 in 2015). In 2005, 17 studies reported spontaneous breathing, compared to 9 in 2015. Mechanical ventilation was reported in 13 studies in 2005 and 6 in 2015. There was a significant difference for the use of O₂/N₂O mixture, which was used in 36 studies in 2005 and 14 in 2015 (p = 0.0266). Some studies reported the use of O₂-enriched air (6 in 2005, 3 in 2015), others 100% O₂ (2 in 2005, 7 in 2015) or O₂ and some, room air (1 in 2005, 4 in 2015).

Anaesthesia monitoring and recovery

The STAIR guidelines offer the following guidance regarding blood sampling, blood pressure and core temperature monitoring:

“Blood sampling is necessary for periodic measurement of arterial blood gas and frequency of measurement should be selected with reference to animal size. […] Monitoring blood pressure during experiments is needed because blood pressure fluctuation affects stroke outcomes. Blood pressure can be monitored by non-invasive and invasive methods. Use non-invasive methods for experiments that cause minimal blood pressure fluctuation and require a neurological evaluation. Use invasive methods for experiments that require constant blood pressure monitoring. […] Controlling animal body temperature in a normal range is necessary for eliminating the protective effect of hypothermia and potential harmful effect of hyperthermia” [13].

The majority of all studies did not report if and how anaesthesia was monitored (71% in 2005, 80% in 2015). Nineteen studies in 2005 reported the use of arterial blood gas analysis and invasive blood pressure monitoring. This was significantly more commonly used than in 2015, where it was only reported in four studies (p = 0.0208). Control of body temperature was reported in 64 studies in 2005, and 65 in 2015.

Almost no data were found on the use of venous (86 in 2005, 97 in 2015) or arterial (74 in 2005, 87 in 2015) accesses; the administration of intravenous fluids (99 in 2005, 99 in 2015); or the control of temperature during recovery (90 in 2005, 92 in 2015). There was no significant difference in any of these parameters between 2005 and 2015.

Country

All publications reported the country where the study had been performed. Studies had been performed in 26 different countries all over the world. In 2005, most studies were performed in the USA (n = 30). In 2015, significantly more studies had been conducted in China (n = 46) when compared to 2005 (n = 11) (p < 0.0001). Details of the number of studies performed per country in 2005 and 2015 are given in Table 4.

Strain

Most of the studies reported the use of Sprague-Dawley rats for their research. This rat strain was used in 48 studies in 2005 and 72 studies in 2015. Wistar rats were the second most commonly used strain, with 34 studies in 2005 and 20 in 2015. There was no significant difference in the proportion of these two used strains across the two sample years. Other rat strains were Spontaneous Hypertensive Rats (7 in 2005, 3 in 2015), Long-Evans (7 in 2005, 2 in 2015), Wistar-Kyoto (3 in 2005, 1 in 2015), and Lister-Hooded (1 in 2005, 0 in 2015). Two papers from 2015 did not report the strain of rats used.

Anaesthetic drug

Regarding anaesthesia protocols, the STAIR guidelines explain the neuroprotective role of most anaesthetics, their effect on neurotransmitters and receptors, and their potential effect on hyperglycemia. All these factors need to be taken into account when choosing the anaesthetic protocol for a study, and the protocol chosen can contribute to the bias of the results. The only difference in the choice of anaesthetic drug noticed between 2005 and 2015 is that halothane was less commonly used in 2015. This might be explained by the higher availability of isoflurane and isoflurane-specific equipment (i.e., vaporiser) rather than being a guideline-influenced choice of the researcher. Although not univocal, isoflurane and halothane have one-third less neuro-protective properties [22]. Some studies testing isoflurane in stroke models showed delay or improvement of the lesions [23,24,25] other studies suggest no collateral effect of isoflurane exposure during experimental stroke procedures [26]. However, available data suggests that lower doses of isoflurane in mild to moderate stroke do possess neuroprotective effects.

Similarly, results from different studies indicate conflicting results regarding the potential effect of halothane on ischemic brain injury. Some studies report that halothane anesthesia improved neurological outcomes during focal and global ischemia models [23,26,27,28]; while others failed to show an exacerbation of ischemic damage and neurological outcome [14,24,29].

The second most common drug for maintenance of anaesthesia was chloral hydrate, injected intraperitoneally (20% of the studies in 2005, and 27% in 2015). Intraperitoneal administration of chloral hydrate produces only light anaesthesia and may cause adynamic ileus, peritonitis, and gastric ulcers in rats. It is not recommended as sole anaesthetic agent and should be used in conjunction with barbiturates, opioids, alpha-2 agonists, or phenothiazine tranquillizers [30]. None of the papers reviewed in our study discussed the choice of drugs or mentioned anaesthesia as a limiting factor of the study.

The selection of the anaesthetic agent may also have an impact on cerebral blood flow (CBF), a crucial parameter during the development of the ischemic lesion. The amplitude of CBF increase depends on both the nature of the agent (halothane having a more severe effect than all other inhalants) and the dose administered [31]. Injectable agents, with the exception of ketamine, tend to decrease CBF [32]. As for the neuroprotective effect, none of the investigated studies discussed the possible consequences of the selected drug on the CBF and subsequent ischemic lesion. CBF is further determined by auto regulatory mechanisms and related to the animal's blood pressure, but only 14 and 21% of the studies (2015 and 2005, respectively) reported the use of blood pressure monitoring. Regardless of the use of such monitoring, none of the studies reported strategies to keep blood pressure in the desired range. As most studies do not mention using an intravenous access line or the administration of intravenous fluids, the pertinence of solely using of blood pressure monitoring remains questionable. However, blood pressure can be considered as a surrogate for anaesthetic depth; low blood pressure may invite the researcher to lower the concentration of inspired anaesthetic agents, when feasible [13].

Ventilation/CO2 monitoring

Regardless of all the molecular modalities, ventilation is depressed during the anaesthetic phase in humans as well as animals. This is partly explained by the loss of respiratory muscles tone, changes in alveolar gas exchange [33,34], and by the blunting of the CO_2-triggered ventilatory response [35,36].). In consequence, respiratory rhythm and tidal volume decrease and CO₂ accumulates in the blood. As initially described by Grubb and collaborators [37], there is a significant linear relationship between P_aCO₂ and CBF. As long as the mean arterial pressure remains within physiological range, each mmHg increase in the P_aCO₂ causes a 1.8ml/100g/min change in the CBF. While originally demonstrated in non-human primates [37], this relationship exists across species. As a result, the anaesthetised state should be expected to influence both the size of the cerebral ischemia and its potential reperfusion, regardless of the experimental treatment being investigated. Tracheal intubation, controlled ventilation, and monitoring of arterial blood gases for rodent models of stroke are possible [14,38,39]. P_aCO₂ should be maintained within normal limits (35–45 mmHg), and CBF as close to physiological norms as possible [14]. However, our review suggests that only 6 and 13% of peer-reviewed stroke studies (2015 and 2005, respectively) used mechanical ventilation as part of the surgical phase of the protocol. Such findings conflict with previously published findings confirming the impact of hypercapnia on the size of cerebral injury [40], and the need for mechanical ventilation to control physiological variables [14], and reduce peri-operative mortality [20] in rat models of stroke. Numerous other scientific publications document the impact of P_aO₂, blood and CSF pH on CBF [41,42,43,44]. These CBF sensitivities are potential sources of noise when collecting data, suggesting the need for a wider sample size per treatment group than necessary under controlled anaesthetic conditions.

Reported monitoring techniques also included saturation of haemoglobin with oxygen and invasive and non-invasive blood pressure monitoring. The combination of invasive blood pressure monitoring and arterial blood gases being used was less common in 2015 than 2005. However, arterial blood gas analysis was more commonly measured in other combinations, which may suggest, that overall monitoring techniques have not changed over the years. Only 29 studies in 2005 and 20 studies in 2015 report any form of monitoring used.

Temperature

Controlling body temperature within physiological range is crucial to avoiding the neuroprotective effect of hypothermia. Hypothermia reduces brain metabolism [45], while hyperthermia increases the metabolic rate of the brain and therefore the ischemic outcome [46]. In general, methods to control body temperature rely on measurement of rectal temperature, assuming a correlation between brain and rectal temperatures. However, several studies have shown considerable differences between rectal and brain temperatures [47]. Roughly, two thirds of all studies reported the use of body temperature monitoring or control during stroke induction, while the rest of the studies did not mention whether any form of temperature control was used or not. The STAIR guidelines underline the importance of body temperature control, not only during stroke induction but also during the recovery period. The most popular method is a warm chamber with a controlled temperature between 28–32°C. More sophisticated telemetric feedback systems also exist [48]. Only 8 (2005) and 7 (2015) studies report their animals recovering in a temperature-controlled environment.

Model

Numerous reviews [19,49,50,51,52] have sought to appraise the characteristics and assumptions made during the creation and use of rodent models of stroke. However, the vast majority of these reviews focus on the refinement of the animal model itself (rodents vs. rabbits vs. other species); the modalities of cerebral ischemia (total vs. partial; permanent or not; arterial occlusion vs. embolisation vs. chemical vasoconstriction); appropriate windows for target drug administration; or presence of other confounding factors such as inflammation and neural regeneration. Rats are one of the most commonly used animals for stroke models [50]) partly because of the similarities between cerebral vasculatures and physiology with humans [53], simple husbandry and ease of restraint.

Because the middle cerebral artery (MCA) and its branches are most commonly responsible for cases of primary strokes in humans [54], middle cerebral artery occlusion (MCAO) is the most commonly used model of stroke induced in rats. Occlusion of the MCA can, however, be obtained in several ways. Briefly, it can either be directly occluded distally (after sub-temporal craniectomy) or cranially (occlusion of a carotid artery and thread of an occlusive filament from the chosen carotid to the MCA) [50,51]. Other indirect models are also in use, such as thrombosis, embolism, or the use of endothelin-I as a potent vasoconstrictive agent [50].

Method/study limitations

The present review has some limitations, notably concerning the methodology. Although we tried to follow the search strategy developed by [Leenars and collaborators [21], there might have been areas open for improvement. Leenars and collaborators [21] say: “to be systematic, explicit and transparent, the scientist should always report: (1) all databases and other sources searched; (2) the dates of the last search for each database and the period searched; (3) full search strategies (including all search terms) for each database; and (4) any language or publication status restrictions used.”. The present review only exploited one single database (PubMed, indexed and non-indexed papers) and only 100 randomly selected papers were analysed were analysed for each year. However, our aim was to identify methodological differences between two given years (2005 and 2015) and not to perform a complete meta-analysis of all rat stroke model papers published in the decade.

Non-reporting

While the vast majority of papers provide statements on ethical review, the rat strain and the model used, most fail to report subsequently on anaesthetic and monitoring techniques, such as the use of arterial or venous accesses, ventilation, oxygen supplementation, or the administration of fluids or analgesics. While expected, this is a major finding. Therefore, it is possible that changes in anaesthesia protocols and techniques have occurred, but were not picked up by our study. For instance, 85% of the stroke studies included in our paper did not report on the mode of ventilation selected for their animal (i.e., spontaneous, assisted or controlled), and up to 96% failed to report whether analgesic agents were at all used peri-operatively. Such levels of non-reporting are very similar to a recent report by Carbone and Austin [55]. We did not attempt to assess whether the journal’s impact factor or commitment to publication guidelines [13,56] were associated with different levels of report since results from previous studies suggest that this is not the case [55,57,58], reinforcing the idea that guidelines may not be the golden grail to reduction of publication bias, better science and eventually more effective translational medicine [55,52,59,60,61].