Tail-cuff versus radiotelemetry to measure blood pressure in mice and rats

Rats and mice are commonly used as experimental models in hypertension research. Critical to the field is information on blood pressure in these models at baseline, over time and following interventions. While there have been advancements in the approaches to measure blood pressure in mice and rats, the two methods that are commonly used include invasive radiotelemetry and the non-invasive tail-cuff technique. Both methods have advantages and disadvantages as highlighted in the 2005 American Heart Association statement paper on recommendations for blood pressure measurement in experimental animals.¹ Because assessment by implantable radiotelemetry allows for continuous and direct measurement of blood pressure, this has been considered as the “standard”. However, with this notion is the perception by some authors and reviewers that tail-cuff measurements are not acceptable in the reporting of mouse and rat blood pressures. Here we clarify that this is not the case and that the selection of methods to be used should be informed primarily by the objectives of the study. The Editors feel it is appropriate to briefly review the methods of blood pressure measurement in rats and mice, and to provide some suggestions when each method should be used.

A major advantage of radiotelemetry is that it provides continuous, or near continuous, readings of systolic, diastolic and mean blood pressures and heart rate over periods of time, from days to months. Beat-to-beat variations in blood pressure values can be employed to calculate heart rate and blood pressure variability, values that facilitate insight into sympathetic and parasympathetic tone.¹ Additionally, radiotelemetry provides an assessment of changes over time, including diurnal variation of hemodynamics that are valuable. Because of the accuracy of radiotelemetry, experimental variability is reduced, and precision is increased. Accordingly, the number of animals required for comparisons over time or between groups is usually reduced.

Despite these advantages, there are downsides to the use of radiotelemetry for blood pressure measurement. The equipment needed to establish this technique in a laboratory is costly, and the telemetry transmitters must be refurbished after 2 or 3 uses, representing an ongoing expense. Unfortunately, the market for provision of this equipment remains non-competitive, and costs have increased since 2005. The implantation of telemetry transmitters is technically difficult, and there is a failure rate for these experiments often leading to loss of animals. Special housing conditions are needed for optimal radiotelemetry recordings. Animals are generally individually housed while undergoing radiotelemetry, and this has been shown to affect food intake, obesity, gene expression, behavior, and social acclimation over the long-term.² Cages are placed on transceiver plates, which are generally separated by at least 12 inches or electronically shielded. Special rooms in animal housing are set aside, and often designed so that wiring to a data exchange matrix and computer can be routed to an adjacent room. Efforts are made to minimize the entry of staff and experimenters into the room housing the animals. Not all institutions have this infrastructure available. Telemetry transmitter implantation is also quite invasive. Commonly, two incisions are necessary especially in mice, with implantation of the telemeter via a flank incision, tunneling of the catheter subcutaneously, and insertion of the catheter into the carotid or aorta often via a second incision.^3–5 For rats, a midline abdominal incision is often required to implant the transmitter in the aorta below the kidneys. These procedures yield a chronically instrumented animal at risk of infection and the bulk of this equipment likely increases discomfort. Moreover, animals need to recover for at least 10 days following this procedure before hemodynamic measurements stabilize. Measures of inflammation, including immune cell infiltration into target organs, circulating cytokines and local levels of chemokines can be affected by this extensive instrumentation, and therefore should be performed in additional, non-instrumented animals. Furthermore, the instrumentation required for telemetry interferes with magnetic resonance imaging and can make other in vivo imaging difficult.

In contrast to radiotelemetry, the tail-cuff method is non-invasive, less expensive and does not require special surgical skills.¹ The animal housing arrangements are not as elaborate for tail-cuff measures as compared to radiotelemetry, although it is recommended that the tail-cuff units be placed in a quiet, preferably dark room, and that aside from the operator, other individuals not enter the room during pressure measurements. Ideally, this room should be in the mouse housing unit, so that the animals are not transported before these measurements. Animals need to be trained for 2 to 3 days and then measurements obtained for an additional 2 or 3 days to improve accuracy of the tail-cuff method. If pressures are followed over several weeks, a common practice is to acquire values weekly, thus requiring 2 or 3 days each week for measurement. Such experiments can thus require hours of investigator time. The tail-cuff technique also requires that mice are restrained and warmed, interventions that can induce a stress response. Obviously, the tail-cuff technique does not allow assessment of changes in hemodynamics throughout the day or while the animals are freely moving, and measurement of diastolic blood pressure is sub-optimal.

Several studies have compared radiotelemetry measures of blood pressure with tail-cuff estimates.^3–7 The consensus is that the average blood pressure obtained from a sufficient number of animals by tail-cuff correlates reasonably well with telemetry measured values, but for individual measurements, there can be enormous discrepancies between the two techniques.^3–8 Wilde, et al., used Bland-Altman analysis, which is a better approach than correlative analyses to define agreement (or disagreement) between two quantitative measurements, to compare pulse-based tail-cuff estimates of pressure to simultaneously obtained telemetry values.⁶ These values were obtained by placing the telemetry receiver under the tail-cuff device, and thus could record fluctuations in pressure caused by the restraint stress and warming required by the tail-cuff method. Using this approach, they showed that the pulse-based tail-cuff method was on average 40 mmHg lower than simultaneously obtained intra-arterial measurements both during vehicle and during Ang II infusion. In contrast, when comparing non-simultaneously obtained tail-cuff and telemetry values obtained on the same day for each mouse, there was better agreement, averaging a difference of 2.63 ± 17.1 mmHg at baseline and 15.2 ± 29.2 mmHg during Ang II infusion. Again, some individual values differed by > 60 mmHg. Similar differences in diastolic pressure were observed. There is also evidence that tail-cuff devices that detect changes in pulse by monitoring tail pressure/volume are more accurate than those that detect return in flow using an optical sensor.⁶ Feng, et al., compared the volume/pressure method of tail-cuff measures of blood pressure to radiotelemetry and found good agreement in average pressure to the pressure/volume (pulse)-based approach; but again there were substantial variations between individual values.⁶

The results of these studies can be interpreted in two ways. The first is that the non-invasive tail-cuff approach is woefully inaccurate and should be abandoned or used with extreme caution. A second conclusion is that the tail-cuff measurement reflects an average pressure over several minutes and does not reflect transient fluctuations in pressure. Given that many investigators are interested in average pressures, such estimates can be acceptable with the proviso that the tail-cuff approach may underestimate systolic pressure during interventions like Ang II infusion. Importantly, these studies do not preclude the ability of the tail-cuff technique to detect differences in the average blood pressure responses to interventions like Ang II between groups of animals, particularly if these exceed 15 to 20 mmHg. This conclusion is in keeping with a validation study by Gross and Luft 20 years ago.⁷

Given these considerations, the Editors of Hypertension recognize that many laboratories do not have ready access to radiotelemetry and that this approach is not essential for many excellent studies and manuscripts that provide important information. We therefore wish to assure the scientific community that non-invasive tail-cuff measures of blood pressure are acceptable in many instances. Some general recommendations for when and how to use this technique follow:

1. The non-invasive tail-cuff approach can compare differences in average blood pressures between groups, or document large changes in blood pressure in one group of animals in response to an intervention, if these exceed 15 to 20 mmHg. Smaller differences in blood pressure may be inaccurately estimated by the non-invasive technique. An obvious conclusion is that studies that fail to detect differences in blood pressure due to genotype, sex, or intervention using the tail-cuff method might be incorrect.

2. The number of animals required for non-invasive estimates of blood pressure is likely larger than that required for radiotelemetry, due to the variability of measures with the former. The precise number of animals will vary depending on the experiment and the strains/genotypes and interventions employed.

3. Whenever possible, the person obtaining the pressures should be unaware of the identity of the groups under study.

4. In some instances, investigators have obtained the bulk of their data with radiotelemetry and have used tail-cuff measures for supplemental or confirmatory data. The opposite may also be the case, use of tail-cuff for high throughput data and telemetry for confirmatory data in smaller groups of animals. The Editors find both approaches to be generally acceptable, but always evaluate the appropriateness of such approaches on an individual basis.

5. Immediate changes in blood pressure, for example to a drug infusion or physiological stress, are not adequately quantified by the tail-cuff method. In such cases acute measurements by an indwelling femoral or carotid artery catheter may be considered. Other experimental models, such as aortic coarctation, would also require acute and concomitant measurements from catheterized femoral and carotid arteries. Likewise, short-lived differences in tail-cuff estimates of blood pressure, for example during one time point during a 3-week infusion of Ang II or at one age but not others, are suspect.

6. Diurnal fluctuation in blood pressure is an important physiological feature and is obviously not detected by the tail-cuff method. If there is reason to believe that differences in blood pressure might be greater either at night or during the day, the tail-cuff method will not suffice.

In summary, the Editors wish to assure the scientific community that we do not mandate the use of radiotelemetry for studies to be published in Hypertension. As with all manuscripts, studies using non-invasive techniques are evaluated on a case-by-case basis but are not a priori rejected because these methods are employed. We further emphasize that the selection of blood pressure measurement should be informed in large by the objectives of the experiment and the questions that need to be addressed. We encourage reviewers to not insist on radiotelemetry when large differences in average pressures are observed and the above recommendations are followed.