Oxytocin (OT) is a nonapeptide characterized by a six amino acid ring structure with a three amino acid tail. This molecule is too large to cross the blood–brain barrier (BBB) [1], [2]. The delivery of drugs to the central nervous system (CNS) has been a challenging field of research for several decades, and considerable efforts have been made to enhance the delivery of therapeutic molecules across the vascular barriers of the CNS in an effective and non-invasive manner. A growing literature has suggested that intranasal delivery can be an effective alternative to invasive delivery methods for some classes of these large molecules, including OT [3], [4], [5], [6], as nasal delivery effectively bypasses the BBB.

Studies have shown that intranasal administration of OT in humans influences social information processing, with effects at both behavioral and neural levels [7]. OT has been reported to affect attachment [8], social exploration and memory [9], [10], aggression and anxiety [11], [12], emotion biases [13], and trust [7]. From neuroimaging studies, OT has been shown to modulate amygdala activity in response to threatening social stimuli [14], pain [15], and emotional faces [16].

Although many studies have shown that endogenous OT concentrations in cerebral spinal fluid (CSF) and blood can be measured reliably in humans [17], [18], [19], [20], [21], non-human primates [22], [23] and rodents [24], it is only recently that studies have measured concentration changes in CSF OT levels after intranasal administration [25], [26], [27]. Striepens and colleagues [26] have shown that both CSF and blood OT concentrations increased after OT intranasal administration in human subjects. Although the dose of 24IU resulted in modest increases after 75 minutes, and the sample of subjects used was rather small because of the invasive nature of CSF collection, this study was important as it was the first evidence of OT entering the CNS in humans following nasal delivery. Neumann and colleagues [28] measured changes in central and peripheral OT using micro-dialysis in the amygdala and hippocampus, and blood samples in rats and mice. A peak central increase was found between 30 and 60 minutes after dose delivery. Chang et al. [25] found, with two rhesus macaques, that nasal administration of OT using a nebulizer increases CSF concentrations 30 minutes after delivery; however, no plasma samples were collected in that study. A recent study that evaluated the efficacy of several OT administration routes for elevating central and peripheral OT concentration in monkeys found that only administration of OT with a nebulizer, but not with a nasal spray, significantly increased concentration of OT in CSF [27].

In the present study we build on these findings in several ways using primates. First, we used a larger number of subjects, which is necessary to validate human and non-human findings [25], [26], [27]. Second, because the delivery method most commonly used in human research is a nasal spray, we carried out a direct comparison between nasal spray and nebulizer delivery within the same group of subjects. To address these questions, we administered OT or saline to macaques using either an intranasal spray or a nebulizer, and measured OT concentration changes in the CSF as well as in blood. The CSF sampling was done by cervical puncture at the level of the cisterna magna, it being more proximal to CSF production sites than a lumbar puncture. Each subject received both delivery methods and both drug conditions on separate sessions. This design enabled us to evaluate the efficacy of the intranasal route as well as determine the optimal method for producing consistent OT increases in the CNS.

Our present findings, consistent with previous studies, show that intranasal administration of exogenous OT increases the concentration of this peptide both in CSF and plasma compared to saline. Both the intranasal spray and nebulizer administration methods resulted in elevations of OT concentration in CSF. Concurrent measurement of OT concentration in plasma showed an increase in plasma OT levels after OT administration but not after saline; however, the changes in plasma OT concentration between pre- and post-sample times differed significantly between the two delivery methods, with a greater elevation of OT after nasal spray compared to nebulizer. When we investigated the time course of OT concentration change in plasma, we found that following nebulizer delivery the OT concentration returned to baseline at the end of 40 minutes, while OT levels after intranasal spray did not return to baseline before the end of the sampling period.

Consistent with recent findings in humans [26] non-human primates [25], [27] and rodents [28], we found that intranasal administration of OT increased OT levels in CSF compared to intranasal administration of saline. Striepens et al., [26] is the first study with human participants to test CSF OT levels after OT intranasal administration. Using a between-subject design they found an OT concentration increase in CSF after 75 minutes in 3 subjects using a 24IU dose of OT intranasal administration, compared to 1 subject that underwent saline administration. This result is in line with previous human [3] and animal studies [38], [39] that have shown that some peptides (insulin, vasopressin, nerve growth factor) increase in CSF after intranasal administration. Born reported significant increases in CSF concentration following administration of vasopressin between 40 and 80 minutes [3]. The first study in non-human primates to investigate concentration changes in CSF OT levels [25] reported that following nebulizer administration of 25IU of OT, CSF OT concentrations increased after 35 minutes from 20 to 50 pg/ml in 2 rhesus monkeys. The second primate study found that CSF OT concentration increased after 60 and 120 minutes from 0.3 to 0.7 pg/ml in 4 monkeys [27]. Neumann and colleagues [28] measured changes in central and peripheral OT using a pipette to deliver exogenous OT to the rhinarium in 11 anesthetized rats and 12 mice and used microdialyis to sample from the amygdala and dorsal hippocampus. They found a peak OT increase between 30 and 60 minutes after OT administration.

Our findings add to these human [26] and primate studies [25], [27] by using a sample size of 15 male monkeys, and a within-subject design where each subject received both delivery methods and both drug conditions over 4 separate sessions, with the order of condition and delivery method randomized. However, there is some discrepancy between the time courses of OT changes in CSF across studies. While our results showed an increase 40 minutes after intranasal administration, which generally agrees with results in the Chang et al [25] and Neumann and colleagues studies [28], the time courses found in Modi et al [27] and the human study [26] differ in that the increases were not detected until 60 minutes and 75 minutes after dose administration, respectively. We did not assay these time points in the current study, so a direct comparison of level of elevation is not possible, but the differences could be explained by the fact that in our study the CSF sampling was done through cervical puncture at the cisterna magna, rather than a lumbar puncture that is much more distal to production sites. Distal sampling requires more time for changes to be detected, and also allows more opportunity for the OT to diffuse and be metabolized before reaching the lumbar area.

The present study sought to validate the human and primate results by using more subjects, and to compare the efficacy of two commonly used delivery methods: nasal spray in human studies [40], and nebulizer in primate studies [25], [27], [29]. Our current findings demonstrate that both delivery methods increased OT levels in CSF compared to saline, thus supporting the recent results in humans [26] and animals [25], [28], and indicating that both delivery methods are viable. However, contrasting findings were reported by Modi and colleagues [27] where the nebulizer, but not the nasal spray, increased CSF OT concentration after administration. This discrepancy could be explained by the number of subjects, the within- versus between-subject design, the radioimmunoassay (RIA) instead of ELISA used to analyze the CSF samples, different time points measured, different doses, different CSF sapling sites, and different OT concentrations detected at baseline (i.e. they were 10-fold smaller than our data and those reported in other studies [20], [23], [25], [36]).

Some have queried whether the elevated OT in the CNS is actually due to the intranasally administered OT crossing into the CNS [41], or if it instead triggers an increase in endogenous OT production without crossing the BBB [42]. While this question is not clinically relevant, as therapeutic elevations in CSF OT due to peripheral administration could be beneficial regardless of route of entry, learning more about the mechanism would be valuable for pharmaceutical development. It may also be possible that CNS OT is elevated via indirect effects of peripheral elevation; further studies tracking OT diffusion are needed to answer this question.

We have shown that both the nebulizer and intranasal spray delivery methods resulted in central increases of OT. The nebulizer delivery was, however, more robust across subjects, but the changes in plasma OT concentration between pre- and post-sample times differed significantly between the two delivery methods. Consistent with previous human studies that have shown increase in OT plasma concentration 30–40 minutes after intranasal administration [43], [44] we found a significant elevation of OT levels after nasal spray, compared to saline. This may be due to the formation of more or larger droplets of fluid using the spray, which would rest directly on the nasal mucosa allowing for greater opportunity for vascular absorption compared to the nebulizer, which forms a much finer mist, allowing more of the substance to be breathed in deeper towards the epithelium. Our results are in agreement with a recent study in non-human primates where only the nasal spray OT administration resulted in peripheral increases of OT, but not the nebulizer [27]. Thus, while both methods resulted in the same CNS penetration, the differences in peripheral action must be considered when choosing between them.

Furthermore, the time course of OT plasma concentration changes was found to differ between delivery methods, in that following nebulizer delivery the OT concentrations had a smaller initial increase at 10 minutes and had returned to baseline levels at the end of 40 minutes, while intranasal spray concentrations had a much sharper increase at 10 minutes [26], and did not return to baseline before the end of the sampling period. This is consistent with previous humans studies reporting that both vasopressin [3] and OT concentration was elevated compared to baseline 80–90 minutes after intranasal administration [43], and with a non-human primate study where plasma OT concentration was elevated 120 minutes after intranasal spray compared to baseline [27]. As discussed above, the use of the nasal spray may cause a greater increase in plasma OT because of the larger droplets that form on the nasal mucosa, allowing for more vascular absorption compared to the nebulizer. However, because our animals were anesthetized and in a supine position, this effect may have been even greater than in awake, erect animals, as the fluid would not run out of the nasal cavity as in an upright position, which may lead to the higher blood OT concentration.

Whether plasma OT level is indicative of central release patterns and activities within the brain and therefore associated with social behaviors remains unknown. In our data we could not investigate if there was a significant correlation between peripheral and central OT levels, because the CSF and plasma samples were not treated in the same way. We did not use extraction on the CSF samples, but we did use extraction on the plasma samples. Many studies have tried to answer this question, but most of those also assayed peripheral and central OT samples differently, making interpretation of the data difficult [45]. There are two main problems reported in the literature related to OT analysis due to methodological differences between studies. The first concerns how much sample preparation influences quantification of neuropeptides in an assay; the second is in regards to how much OT the most commonly used assays are actually measuring. The measurement of peripheral OT started with the development of RIA methods [46]. Almost 30 years later researchers began developing ELISA to avoid the use of radioactive materials. In the ELISA, there were divergent results in un-extracted compared to extracted samples: un-extracted plasma yielded values 100-fold greater than the same sample after extraction [47]. For example, humans have been reported to have basal plasma OT concentrations ranging between 0.1–23 pg/ml in extracted plasma and 99–405 pg/ml in raw plasma. The effect of extraction was also evident in samples measured by RIA, where un-extracted plasma gave values that were at least 10-fold higher than the extracted sample (reviewed in Szeto et al [48]). Although it is currently unknown whether extraction prevents cross-reactivity, or whether it eliminates the effects of binding proteins or other substances that interfere with the assay, OT-free plasma (for instance, obtained from OT knockout mice) could be used as a negative control to evaluate the issue further.

While discrepancies between results obtained with and without extraction using different assays have started to be addressed in plasma samples [47], the issue has not received the same attention for CSF samples. Most of the studies that have assayed OT in CSF used un-extracted samples [22], [23], [34], [35], [36]; only a few studies have used extracted CSF samples with RIA [20], [26], [27], [28], and only one used extracted samples with ELISA [24]. However, OT levels do not seem to differ as much between extracted and un-extracted CSF samples compared to plasma. Two studies with rhesus monkeys analyzed un-extracted CSF OT samples and reported baseline levels between 20 and 30 pg/ml [23], [25]. Human studies reported baseline CSF OT levels between 10 and 34 pg/ml without extraction [36] and around 20 pg/ml with extraction [20], [26]. Only one study reported much lower baseline CSF OT levels in monkeys that range between 0.2 and 0.4 pg/ml after sample extraction [27].

Whether peripheral measures of neuropeptides are a reliable estimate of central concentration remains an important research question and more work needs to be done to investigate the validity of measurements of OT in peripheral fluids such as plasma, saliva, and urine as a global biomarker of behaviorally relevant concentrations of central OT. Until assays are strictly validated and standardized to detect bioavailable neuropeptides, interpretation of the data remains difficult. The protocol followed when collecting and analyzing any type of biological samples for OT does have an impact on the results obtained.

Although the last decade has seen a large number of published findings adding knowledge to this topic, more effort is needed to investigate the mechanisms of intranasal administration of OT and its effect at both the central and peripheral levels. Specifically, future studies should address the pharmacokinetics of this peptide, as it is still not known how much OT reaches target sites in the brain. Second, the time course of exogenous OT should be measured with serial CSF sampling from an implanted brain or cervical port. Third, measurement of the exact amount of exogenous OT that reaches the CNS following intranasal administration must be carried out, as well as testing of different dosages in order to determine the minimum amount required to trigger central effects. Fourth, the validity of peripheral OT measurement as a biomarker for central function is still debated and there is a critical need to establish valid and reliable methods to measure OT in plasma and other biological fluids. Our findings provide evidence that intranasally administered OT does elevate CNS OT levels as well as peripheral levels. Overall, our study supports the hypothesis that delivering OT intranasally, either with a nebulizer or with a nasal spray, which has been used regularly to test social cognition in humans, results in elevations of OT in CSF.