We performed the literature search using the PubMed database in Sep/2020. The search criteria were: LSD (Title/Abstract) OR lysergic (Title/Abstract) OR psilocybin (Title/Abstract) OR ayahuasca (Title/Abstract) OR dimethyltryptamine (Title/Abstract) AND [fMRI (Title/Abstract) OR BOLD (Title/Abstract) OR PET (Title/Abstract)]. Figure 1 (PRISMA) summarizes the number of articles and duplicates that were found. To identify functional brain imaging studies, our inclusion criteria were: 1) the studies imaged the whole brain; 2) the results presented coordinate-based data in a standard space and were not review papers; 3) the imaging method was fMRI or PET; 4) subjects were healthy controls; 5) sample size N ≥ 8 (Eickhoff et al., 2016).

From the initial identification of 78 studies, the final study included 16 fMRI studies and four PET studies (Table 1) reporting brain imaging experiments related to those drugs. We then used the foci of brain activations extracted from each of the included studies for the ALE analysis.

We exported foci data manually from each paper to a text file containing all the coordinates of the results from the original studies that passed the inclusion criteria. All coordinates were converted to MNI standard space (using the Brett transform as implemented in the tal2mni/mni2tal function of MATLAB (R2020a, Mathworks, United States). It is important to note that all MRI studies included placebo (control) groups and the data reported are comparisons of drug vs placebo effects.

Additionally, all the available PET studies (with different tracers) are discussed in a narrative manner, given the insight they provide on molecular mechanisms of action.

ALE meta-analysis was carried out as described previously by (Turkeltaub et al., 2002). To assess the statistical significance of the results we used a permutation test (1,000 permutations) and set a threshold p value < 0.001 and a minimum cluster size of 200 mm³ (Eickhoff et al., 2016). We used GingerALE (v3.0.2), the Java version of ALE developed at the Research Imaging center and available at http://brainmap.org/ale for data processing. For visualization, the results were overlaid into a standard MNI image template (Kochunov et al., 2002).

Since we found fMRI experiments with BOLD and connectivity results, we performed an ALE including all fMRI papers and two other separate analysis: 1) using the results from the BOLD amplitude changes; 2) using the connectivity results from the fMRI papers. The resulting ALE images were converted to Z scores in order to simplify interpretation and show their significance.

Activation maps related to each of the tasks were overlaid and displayed using Mango software (http://ric.uthscsa.edu/mango/) and the Talairach Daemon (http://talairach.org/) tool was used to extract anatomical labels of results. All the input files used in our analysis and output results are freely available upon request to the corresponding author.

The profound experience induced by psychedelics like DMT, Ayahuasca, LSD and Psilocybin is characterized by changes in emotion, perception and cognition, visual imagery and differences in the sense of self (Swanson, 2018; Barrett et al., 2020b; Lowe et al., 2021; Luppi et al., 2021). Figure 3 summarizes these effects.

The most obvious finding of our analysis is the deactivation of the amygdala during the psychedelic induced states, which might underlie the emotional effects of these substances. Altered processing of facial expressions with negative valence and modulation of the amygdala activity to these stimuli has been found after the administration of serotonergic psychedelics on healthy and clinical populations (Rocha et al., 2019). The decreased reactivity of the amygdala to negative stimuli was also associated with an increase of positive mood states during the acute phase (Kraehenmann et al., 2015) and also long-term (Barrett et al., 2020). These effects may be of clinical relevance in disorders associated with difficulties in emotional processing such as depression, anxiety and addiction. Previous studies evaluating anxiety disorders have found consistent findings on the role of the amygdala in the symptoms of fear and anxiety (Holzschneider and Mulert, 2011). Meta-analytic evidence revealed consistent hyperactivation of the amygdala in post-traumatic stress disorder, social anxiety disorder and specific phobia, as well as during fear conditioning in healthy subjects, suggesting a common excessive engagement of fear circuitry (Etkin and Wager, 2007). Our results also show a greater deactivation in the right amygdala. Although there are no conclusive findings on the lateralization of amygdala in emotional processing (Kraehenmann et al., 2015), some studies point to different activations. During the presentation of emotional stimulus, right amygdala hyperactivation was observed in patients with PTSD compared with trauma-exposed non-PTSD individuals (Brohawn et al., 2010), as well as in patients with obsessive-compulsive disorder compared with healthy controls (Thorsen et al., 2018). In the latter, right amygdala hyperactivation was more evident in unmedicated patients. An increased influence from right amygdala to right middle frontal gyrus and a decreased influence from right precuneus to right amygdala was also associated to the trait neuroticism, which is the tendency to experience negative emotional states and negative self-referential information processing (Pang et al., 2016). Further research should be undertaken to elucidate the long-term impact of psychedelics on amygdala responsiveness. Recent findings from healthy populations indicated a reduced amygdala response to facial stimuli 1-week post-psilocybin, returning to baseline after 1 month (Barrett et al., 2020). Nevertheless, an increased reactivity was found in clinical populations 1 day after psilocybin session (Roseman et al., 2018).

Another important finding was the deactivation of brain areas associated with the Salience Network (SN), such as the dorsal anterior cingulate cortex. This network is involved in attributing salience and selecting relevant interoceptive, autonomic and emotional stimuli (Menon, 2015). Dysfunctions on salience-processing are relevant in many psychiatric and neurological disorders, such as schizophrenia, dementia, autism, mood and anxiety disorders, drug addiction and pain (Menon, 2015; Uddin, 2015). An aberrant salience attribution to internal stimuli is proposed as a model for psychosis (Kapur, 2003), and is also conceptualized as having an important role in the symptoms of delusions and hallucinations in schizophrenia (Palaniyappan and Liddle, 2012). These findings may help to understand the early research on psychedelics as models for psychosis. A salience network dysfunction hypothesis is also considered in autism spectrum disorder, which suggests that impaired attribution to sensory stimuli might be associated with dysfunctional cognitive processes, such as social cognition (Uddin, 2015).

We also found relevant patterns in regions involved in theory of mind such as supramarginal gyrus, medial prefrontal cortex, precuneus and posterior cingulate cortex. LSD decreased the efficiency of establishing joint attention in the PCC and the temporal gyrus, an effect attributed to 5-HT2AR stimulation (Preller et al., 2018). The authors suggested a decreased differentiation between the self and the other during social interactions. This altered sense of self characterized by a decreased differentiation between self-representations and other-representations is usually called “ego dissolution” (Nour et al., 2016). In addition, psilocybin decreased the feeling of social exclusion processing in the ACC (Preller et al., 2016). These findings point to the modulation of social cognition, which may be an important mechanism contributing to the therapeutic potential of psychedelics (Preller and Vollenweider, 2019). There is evidence for the role of the supramarginal gyrus, highlighted in our analysis, in overcoming emotional egocentricity bias in social judgements (Silani et al., 2013), which suggests a possible role in empathy. The overlap between some areas involved in theory of mind and the default mode network (DMN) has led some authors to suggest the role of the DMN in the social understanding of others (Li et al., 2014), as well as the role of the PCC in attributing mental states to others (Mars et al., 2012). Tagliazucchi et al. (2016) reported that LSD-induced states increased functional connectivity in bilateral temporo-parietal junction, a key component of theory of mind, which was correlated to subjective reports of ego dissolution. In line with this, previous studies indicated the effects of tryptamine psychedelics on dimensions related to healthy social functioning such as increased emotional empathy and prosocial behaviour (Dolder et al., 2016; Pokorny et al., 2017; Mason et al., 2019; Uthaug et al., 2021), changes in personality traits agreeableness (Netzband et al., 2020) and compassion (Apud Peláez, 2020), as well as feelings of connection to others (Watts et al., 2017). A recent study in mice reported that repeated administration of low doses of LSD promoted social behaviour by potentiating 5-HT2AR and AMPA receptor neurotransmission in the mPFC via an increasing phosphorylation of the mTORC1 (de Gregorio et al., 2021).

The activation of visual areas by psychedelics induced substances, was another outcome of our quantitative meta-analysis, namely visual areas BA19 and visual fusiform region BA37. These areas are densely populated with 5-HT2A receptors. Various studies indicated a key function for 5-HT2ARs in visual processes and the pathogenesis of visual hallucinations (Moreau et al., 2010; Seillier et al., 2017). Classical hallucinogens are used as models for studying the pathophysiology of different neuropsychiatric conditions with positive psychotic symptoms, such as schizophrenia, Parkinson’s and Alzheimer’s disease, which alter individual visual and perceptual experiences. The activation of 5-HT2ARs increases the excitability of the visual cortex in the absence of external visual stimulation (Moreau et al., 2010). In addition, the activation of 5-HT2ARs mediates the visual hallucinations that are generated by serotonergic hallucinogens, such as LSD or psilocybin (Nichols, 2004; Vollenweider and Kometer, 2010). In line with this, the hallucinogen-induced decrease in alpha oscillations might allow spontaneous self-organized activity to gain perceptual quality (Kometer et al., 2013). Recent studies demonstrated that acute LSD administration to healthy subjects not only produces elementary and complex visual (pseudo)hallucinations and perceptual illusions (Carhart-Harris et al., 2016b; Preller et al., 2017; Schmid et al., 2015; Schmidt et al., 2018), but also impaired inhibitory processes (Schmid et al., 2015) and cognitive organization (Carhart-Harris et al., 2016a). Impairments in inhibition after psilocybin administration and cognitive impairments after LSD administration were attenuated by administration of the 5-HT2AR antagonist ketanserin (Quednow et al., 2012; Preller et al., 2017). However, this does not exclude the contribution of other receptor subtypes such as 5-HT1AR (Halberstadt and Geyer, 2011). 5-HT2AR activation is indeed pivotal in inducing visual hallucinations but other receptors also contribute to cognitive impairments, and their abnormal activity can be associated with cognitive deficits in neuropsychiatric disorders such as schizophrenia and Alzheimer’s disease (Švob Štrac et al. 2016). Schmidt et al. (2018) proposes that psychedelics disrupt information processing in inhibitory cortico-striato-thalamocortical (CSTC) feedback loops that have been implicated in sensory gating of internal and external information to the cortex. This psychedelic-induced disinhibition might lead to an inability to filter and inhibit exteroceptive and interoceptive stimuli, resulting in high-level processing overload and the formation of hallucinations.

In line with our results, De Araujo et al. (2012) investigated the neuronal mechanisms underlying psychedelic-induced visual mental imagery using functional magnetic resonance imaging (fMRI). The authors found that ayahuasca increased activations in mental imagery networks, including early visual areas (BA 17, 18, 19), parahippocampal gyrus, middle temporal cortex, and frontal cortex (BA10). They also showed that ayahuasca-induced changes in primary visual cortex (BA17) were preceding activation patterns in higher-level areas, indicating that ayahuasca-induced imagery is initiated in BA17, but activity is spread to higher-level cortical areas involved with episodic memory retrieval and the processing of contextual associations, such as BA30 and BA37, which might feed memory-related content. In addition to perceptual alterations of simple and elementary visual features as color, brightness, visual contrast (Klüver, 1942; Rummele and Gnirss, 1961; Kometer et al., 2013; Kometer and Vollenweider, 2018) that might be explained by increased excitation in V1 (Kometer et al., 2013; Császár-Nagy et al., 2019), complex imagery and hallucinations has been reported (Császár-Nagy et al., 2019; Díaz, 2010; dos Santos et al., 2016; Kraehenmann, 2017; Kometer et al., 2013), with personal and profound significance, stemming from autobiographical memory (Studerus et al., 2011) to current life situations (Shanon, 2010) charged with emotional content. These complex forms of hallucinogen-induced hallucination and visions, also lead the recruitment of higher level regions in the brain, given that psychedelic imagery is usually very structured, thematic and personal (Kraehenmann, 2017). In accordance to our main results, studies have reported visual hallucinations caused by neuronal stimulation of PFC (Blanke et al., 2000), temporal areas (Mégevand et al., 2014; Aminoff et al., 2016) and increased functional connectivity between PFC and primary visual cortex (Carhart-Harris et al., 2016b). Furthermore, the review conducted by dos Santos et al. (2016), suggested that hallucinogens increase introspection and positive mood by modulating brain activity in the fronto-temporo-parieto-occipital cortices. Neuromodulatory changes induced by tryptamine psychedelics can give significant input to the study of neuropsychiatric conditions where similar patterns of activation or connectivity (Barrett et al., 2020b; Madsen et al., 2021) are found and to the implementation of new pharmacological or psychotherapeutic interventions taking advantage of this link between visual imagery, autobiographical memory and emotions (Barrett et al., 2020).

The role of amygdala in this interplay is not of less importance. The amygdala plays an important role in emotional visual processing (Vuilleumier et al., 2004). Important networks between amygdala and ventral visual pathways in primates are reported (Freese and Amaral, 2005), as well as the role of the amygdala in visual awareness (Duncan and Barrett, 2007). Furl et al., 2013 suggested that the amygdala modulates visual processing by feedback connections and that it may have a contextual role during visual coding. Deactivation of the amygdala during the psychedelic induced states has been consistently found and along with the inhibition of DMN opens a therapeutic potential for accessing and transforming autobiographical memories, emotions and maladaptive perceptions.

In addition to the mechanisms described above, changes in Default Mode Network (DMN) connectivity may be another neural basis involved in the psychologic and therapeutic effects attributed to tryptamine psychedelics. DMN areas present lower levels of activity when individuals are engaged in a task requiring externally oriented attention and activate during passive rest states or internally oriented mental processes, such as autobiographical memory, mind wandering, self-reflective thought, and future thinking (Buckner et al., 2008; Andrews-Hanna et al., 2010).

In our analysis, a decreased connectivity within PCC/Precuneus, key components of the DMN, was observed. Regarding classic hallucinogens, studies revealed that psilocybin, LSD, and ayahuasca could decrease DMN functional integrity (Carhart-Harris et al., 2012, Carhart-Harris et al., 2016c; Palhano-Fontes et al., 2015; Luppi et al., 2021; Madsen et al., 2021; Mason et al., 2021),. Barrett et al. (2020b), recently proposed that Psilocybin alters default mode network integrity and fronto-parietal network modularity by reducing Claustrum functional connectivity with these circuits. This study showed that psilocybin reduced activity of left and right claustrum during the acute effects of psilocybin, leading to alterations in claustrum connectivity with brain networks that support both sensory and high-level cognitive processes. Specifically, the authors found decreased connectivity between claustrum and the DMN during the effects of psilocybin, decreased connectivity between left claustrum and fronto-parietal task control circuits and increased connectivity between right claustrum and the same fronto-parietal networks. In sum they assigned to the claustrum (dense in 5-HT_2A receptors ) a role in the psilocybin-induced disruption in both the DMN and task-positive networks. Accordingly, Madsen et al. (2021) found negative correlations between the DMN integrity and the plasma psilocin levels and subjective drug intensity. These results support the proposed theory of action for psychedelics to decrease the control of top-down structures and increase the excitability of areas involved in sensory, emotional and cognitive appraisal processes. (Barrett et al., 2020b; Mason et al., 2021). The expression and awareness of normally repressed information would explain the novelty of the experience and the new associations would facilitate the formation of new insights (Domínguez-Clavé et al., 2016; McKenna and Riba, 2018). In line with this, Mason et al. (2021) reported psilocybin-induced decreased within-network connectivity of the DMN and increased functional connectivity between the DMN and the Frontoparietal Network (FPN) and between the DMN and the Salience Network (SN), which predicted higher scores in aspects of creative thinking and long-term increases in novelty of generated ideas. However, in contrast to these findings, there have also been findings of increased DMN activity by hallucinogens (Carhart-Harris et al., 2017; Kometer et al., 2015; Petri et al., 2014; Tagliazucchi et al., 2014). Regarding the associated therapeutic potential, DMN activity is increased in depression (Sheline et al., 2009) acute and chronic pain (Alshelh et al., 2018), schizophrenia (Garrity et al., 2007) and Parkinson’s disease (Van Eimeren et al., 2009) Aberrant patterns of connectivity are also found in drug addiction (Zhang and Volkow, 2019) and eating disorders (Stopyra et al., 2019). It seems to be reduced in autism and in Alzheimer’s disease (Broyd et al., 2009).

There were surprisingly few eligible pharmacoimaging studies using PET. Two used FDG (Vollenweider et al., 1997; Gouzoulis-Mayfrank et al., 1999) and together suggested frontal and temporal hypermetabolism, which are consistent with fMRI data. Another used the 5-HT2AR agonist radioligand (11C) Cimbi-36, and showed that intake of psilocybin leads to significantly 5-HT2AR reduced occupancy in the human brain, confirming a role for this receptor subtype. However, a specific link with this receptor system is probably an overstatement, given the evidence that multiple receptors, in particular the 5-HT1AR contribute to the behavioral effects of indoleamine hallucinogens (Halberstadt and Geyer, 2011). The neural effects of these hallucinogens seem to include regions rich in both 5-HT1AR and 5-HT2AR. These probably interact with other receptor systems such as DR2 (Vollenweider et al., 1999), whose binding is decreased probably due to endogenous dopamine release.

Taken together, our results support the plausibility of further research on the therapeutic potential of tryptamine psychedelics (Lowe et al., 2021). There is a growing number of clinical trials describing promising data on safety and efficacy of psychedelics and entactogens in several psychiatric disorders, such as posttraumatic stress disorder (Mitchell et al., 2021), treatment-resistant depression (Carhart-Harris et al., 2016a; Palhano-Fontes et al., 2019), substance addictions (Johnson et al., 2014; Bogenschutz et al., 2015); obsessive-compulsive disorder (Moreno et al., 2006); anxiety associated with life-threatening diseases (Gasser et al., 2014; Griffiths et al., 2016; Ross et al., 2016) and social anxiety in autistic adults (Danforth et al., 2018). Those preliminary findings suggest the reduction of depressant, anxiety and addiction symptoms. Patients described feelings of connection, transcendence, insights, self-awareness, alterations in the perception of the self, emotional catharsis, changes in values and life orientations, reconciliations with death, as well as psychological distress (Gasser et al., 2015; Schmid et al., 2015; Belser et al., 2017; Swift et al., 2017; Watts et al., 2017; Noorani et al., 2018; Barone et al., 2019; Lowe et al., 2021), encouraging further studies. Recently, the role of psychedelics in changing behaviours related to healthy lifestyles (Teixeira et al., 2021), as well as a treatment for neurodegenerative disorders (Vann Jones and O’Kelly, 2020) and for pain conditions (Castellanos et al., 2020) has also been hypothesized. Despite the promising results, further work is required to better understand the neurobiological and psychological mechanisms of action and the potential risks underlying the therapeutic action of tryptamine psychedelics. Several questions regarding the long-term impact of psychedelics remain unanswered at the moment. Rigorous research (possibly integrating PET with fMRI (Cumming et al., 2021)) is needed, taking into account the best clinical practices.