Individually tailored dosage regimen of full-spectrum Cannabis extracts for autistic core and comorbid symptoms: a real-life report of multi-symptomatic benefits

2.2. Participants

Treatment with FCE was conducted by two of the authors (PM and LS) in the clinical setting. Either during or after the end of each patient’s treatment, parents, or participants (according to the better convenience in each case) were invited to take part in this study. Twenty participants diagnosed with ASD agreed to participate and were enrolled. We are interested in evaluating the benefits of FCE treatment for any condition within the autistic spectrum, therefore the inclusion criterion was ASD diagnosis (ICD 10 = F84.0) previously received by the patients, or given after clinical evaluation by the authors PM or LS. ASD patients treated with FCE for at least 3 months were included. Participation consisted of allowing the disclosure of patient’s clinical records and answering a detailed survey (Supplementary material S1).

In Table 1 we indicate participant’s weights, genders and details of cannabinoid concentrations consumed daily and per kilo. Patients’ gender is identified as male (m) or female (f) numbers throughout the text and figures. In total, 118 families were initially invited to take part in the research, and 24 of those families answered the survey. Out of those, one participant was removed because he had been under less than 3 months of FCE treatment, one participant was removed due to faulty survey answers, and two others were removed because they used CEs that deviated from the scope of this paper. Of note, three of the enrolled patients chose to interrupt the treatment of their own accord: one (2 m) due to worsening of symptoms and two due to financial limitations (7 m, 19 m), but all treatments lasted 3 months or more, and their data was maintained in the analysis. Patient 16 m used a non-full-spectrum, pure CBD product. This was the only case treated throughout with a purified CBD formulation, which was eventually supplemented with THC-rich FCE.

2.3. Treatment

Most patients received individualized treatments based on a titration protocol, which started with low doses of a CBD-rich FCE. According to the clinical evaluation of observed effects, the dose was slowly increased and could be gradually supplemented by low doses of a THC-rich FCE to improve the results. We define “CBD-rich” FCEs as those in which CBD amount is higher than THC, and “THC-rich” FCEs as those in which THC is the more abundant cannabinoid. The CBD-rich FCEs had a proportion of either 4: 1, 21: 1, 49: 1 or 7: 1 (CBD: THC), whereas THC-rich FCEs had a proportion of either 1: 22 or 1: 4 (CBD: THC, rounded to whole numbers). One exceptional case started treatment with a purified, industrial grade CBD extract (identified in Table 1), but eventually added a FCE rich in THC. Two patients presenting refractory severe behavioral symptoms have used a blend of CBD-rich and THC-rich extracts since the beginning of the treatment. These extracts were produced by three patient organizations, and the specific extracts used by each patient depended on their choice of affiliation to a specific NPO. All extracts were administered orally via drops. Thus, daily doses were varied, changed according to symptoms and side effects presented by the patient during treatment, and eventually stabilized after reaching a consensus between physician and family.

Patients were evaluated monthly for a trimester. Upon symptoms stabilization, they were reevaluated every 6 months. In rare cases weekly reassessments/adjustments were made. Criteria employed for CBD-rich FCE dosage change, or supplementation with THC-rich FCE were: (1) severe cases of psychomotor agitation and aggression with little response to CBD; (2) persistence of complaints regarding sleep pattern and appetite despite otherwise satisfactory response to CBD (More details in Figure 1). Average concentrations at the beginning of treatment were 45 mg/day: 1.94 mg/day (CBD: THC) and 75 mg/day: 2.32 mg/day at the end of treatment, normally administered in two or three daily doses.

2.4. Cannabinoid extracts

Cannabis extracts were produced and distributed to all the patients in the study by three Brazilian patient NPOs, the Brazilian Society of Medicinal Cannabis Patients (also known as AMAME) the Cannabic Society in Defense of Life (also known as Maria Flor) and the Brazilian Support Society Cannabis Hope (also known as ABRACE). The quality and composition of cannabis extract samples were assessed by high-performance liquid chromatography (HPLC) analysis performed at certified laboratories.

2.5. Parents-reported outcome survey

The treatment reported here resulted from an adaptation of available clinical data to the Brazilian reality, but it was not originally intended to produce scientific data. Therefore, the approach used by each clinician contributing to this report to assess baseline clinical indicators of the ASD patients was not the same, which precluded us from using a before-vs-after approach. Thus, we have used a structured, caregiver-reported single-assessment retrospective outcome survey sent by e-mail to the patients and their families (Supplementary material). The questionnaire was adapted from the methodology previously used by Fleury and colleagues (76). Thus, we have employed a Likert-like scale (83) composed of six options to evaluate the perceived outcome of the treatment for 10 main symptom categories, 16 specific sub-categories for abnormal behavior and communication/social interaction impairments, and five additional non-symptomatic aspects. Parents were explicitly asked to report the difference observed since FCE treatment onset. The options for each question were: “Does Not Apply,” “Considerable Worsening,” “Moderate Worsening,” “No Change,” “Moderate Improvement” and “Considerable Improvement.”

In addition, responders were asked to describe in their own words the changes observed for each symptom category to ensure that the responder properly understood the meaning of each category and that they were scoring the Likert-like scale in a consistent way throughout the study. If discrepancies were detected, the responders were contacted by telephone or text messages to further discuss the definition of that specific symptom, and then they were asked whether they wished to reconsider their answers. These open-ended questions also helped to elucidate more practical and subjective impacts of the treatment (see discussion). The parents were invited to answer the survey from March 2022 to August 2022. Clinical charts of each patient were also assessed to obtain details of the treatment, such as weight, doses and use of other medications.

The 10 main symptom categories and five non-symptomatic aspects evaluated in the survey were chosen after reviewing their prevalence in the literature, adapting some of the available instruments (mainly the Autism Treatment Evaluation Checklist, ATEC), and considering the clinical experience of the physicians and ASD patient’s parents. Symptom categories used are as follows: (1) Attention deficit and hyperactivity disorder; (2) Abnormal behaviors; (3) Sadness, melancholy, bad moods; (4) Impaired motor development and motor coordination; (5) Lack of independence for daily activities; (6) Impaired communication and personal interactions (verbal and non-verbal); (7) Intellectual and cognitive deficits; (8) Sleep issues; (9) Seizures; (10) Avoidance and/or restrictions to food intake; (11) Positive mood states; (12) Overall patient’s quality of life; (13) Adverse effects due to treatment; (14) Use of other medications; (15) Overall family’s quality of life.

In the main group of categories, concerning Abnormal behavior, and Impaired communication and verbal interactions, each survey responder is likely to evaluate the outcome in each of these two categories considering their own reality regarding the intensity and scope of behavioral and communication problems of the patient. Further outcome details regarding these aspects were collected in two separated groups. Thus, the category “abnormal behaviors” was further subdivided into nine specific aspects as follows: (1) Stereotypies; (2) Aggressiveness Toward Others; (3) Self-Aggressiveness; (4) Autistic Meltdown Crisis/Temper Tantrum; (5) Screams and Random Sounds; (6) Obsessive Compulsive Behaviors; (7) Eating Non-Foods; (8) Discomfort in Noisy/Crowded Places; (9) Excessive Appetite.

In Communication and Social Interactions, we also analyzed specific related features, as follows: (1) Impaired Verbal Communication; (2) Impaired Visual Contact; (3) Impaired Response to Their Own Name; (4) Impaired Attention to Receptive Direct Verbal Communication; (5) Production of Sounds or Isolated Words with Communicative Function; (6) Impaired Written Communication; (7) Use of Alternative Forms of Communication (gestures, signals, cards, software applications, and other systems with images).

Adverse effects and Use of other medications were not evaluated using the above-mentioned scale, but by questions with open written answers, multiple-choice questions and by consulting patients’ medical records.

3.2. Results grouped by symptom category

There was a general perception of improvement for all 10 main symptom categories and 3 non-symptomatic aspects analyzed (Figure 2). Some features improved more frequently than others. While “avoidance and/or restrictions to food intake” improved in approximately 30% of cases, ADHD symptoms, “behavioral disorders,” “communication and social interaction deficits,” “sleep disorders,” “seizures,” “patient’s quality of life” and “family’s quality of life” improved in at least 77% of cases. General improvement in “intellectual and cognitive performance” were observed in 57% of the cases, which is very striking since there is no other pharmacological treatment known to improve this impactful and recurrent aspect of severe ASD.

Benefits from using FCE were observed in 84% of seizure cases, in general accordance with many other reports from the literature (84). Considerable improvement was equal or more frequent than 33% in 12 out of the 13 main categories (Table 2 and Figure 2). Remarkably, “patient’s quality of life” and “family’s quality of life” improved in 95 and 83% of cases, respectively. Only two patients, 2 m and 15f presented an overall worsening or no change in the main symptom categories (GOS equal to −0,3 and − 1.1, respectively).

3.3. Sub-categories of abnormal behaviors

There was improvement in at least 54% of cases in all 9 sub-categories of abnormal behaviors, as indicated in Table 3. A graphic representation of these results is presented in Figure 3. The frequency of patients presenting overall improvement in each of the specific subcategories of abnormal behavior was as follows: “stereotypies” 80%; “autistic meltdown crisis/temper tantrum” 76%; “discomfort in noisy/crowded places” 72%; “aggressiveness toward others” 67%; “eating non-foods” 63%; “excessive appetite” 62%; “obsessive compulsive behaviors” 62%; “self-aggressiveness” 56%; and “screams and random sounds” 54%. The frequency of “considerable improvement” perception was above 30% for 8 out of the 9 sub-categories, and ranged from 23% for “obsessive compulsive behavior” to 50% for “eating non-foods.”

Worsening of one or more abnormal behavior sub-categories was perceived in 7 out of the 20 patients (35%), with 3 out of 12 (25.0%) reporting worsening in “aggressiveness toward others”; 2 out of 9 (22.2%) in “self-aggressiveness,” 2 out of 17 (11.8%) in “meltdown crisis/temper tantrum,” 2 out of 13 (15.4%) in “obsessive compulsive behaviors,” 1 out of 8 (12.5%) in “eating non-foods” and 2 out of 13 (15.4%) in “excessive appetite.” “Considerable worsening” was perceived in 5 out of 20 patients (25.0%) in four sub-categories, with 1 out of 12 (83.3%) reporting considerable worsening in “aggressiveness toward others,” 1 out of 9 (11.1%) in “self-aggressiveness,” 1 out of 17 (5.8%) in “meltdown crises/temper tantrums” and 2 out of 13 (15.3%) in “excessive eating.” Nevertheless, only two patients, 2 m and 15f presented an overall worsening in abnormal behavior sub-categories (ABOS equal to −0.3 and − 0.7, respectively).

3.4. Sub-categories of communication and social interactions

There was consistent improvement in all communication and social interaction sub-categories. We highlight that “attention to receptive direct verbal communication” improved in 17 out of 20 (85%), while “visual contact” and “attention to conversation” improved in 15 out of 20 (75%) of cases. “Verbal communication” improved in 8 out of 16 (50%) of cases, “response to their own name” improved in 11 out of 20 (55%) of cases, “written communication” improved in 5 out of 14 (35.7%) of cases and “alternative forms of communication” improved in 7 out of 12 (58%) of cases. Actual worsening of cases was found only in one case for “visual contact” and in one case for “attention to receptive direct verbal communication.” Only patient 2 m presented an overall worsening in communication and interaction sub-categories (CIOS equal to −0.3).

3.5. Other medications used during FCE treatment

Medications other than the FCE used by the patients are indicated in Table 5. As FCE treatment progressed, out of the 16 patients receiving other neuropsychiatric medications, four (25%) reduced the dosage for at least one type of medication, nine (56%) discontinued at least one type of medication, and six (37%) discontinued all medication. Therefore, out of 16 patients, 13 (81%) have reduced or discontinued neuropsychiatric medications after using FCE. Only two patients increased the doses of medications other than FCE during treatment, and one patient’s medications remained unaltered.

4.2. Individual tailoring of dosage and CBD to THC proportion

Regarding the importance of using whole extracts as opposed to purified phytocannabinoids, it is worth noting that in one of the best controlled studies, published by Aran and colleagues in 2021, the treatment with full-spectrum extract containing both CBD and THC has yielded slightly better results than the mixture of pure CBD and THC in proportion equivalent to the whole extract, even though the difference was not statistically significant (73). Besides, a metanalysis study has shown that the use of full-spectrum, CBD-rich extracts has yielded better clinical results than the use of pure CBD for the treatment of infantile syndromes associated with refractory epilepsy and regressive autism (84). Finally, there is one case report showing significant outcome improvements when CBD was combined with selected terpenes for the behavioral treatment of an adolescent ASD patient (107).

Studies that followed the standard 12 weeks period of clinical evaluation and used more restrictive, homogeneous samples (73, 78), may have been too short to assess specific changes in ASD behavioral symptoms, which depend on slow neurochemical, structural and cognitive adjustments. Hopefully, our real-life study design was able to better capture such changes. These more rigorous and well controlled prospective studies have mostly used a single formulation with a CBD to THC proportion of 20: 1 (72–75). Nevertheless, published studies have altogether used a diverse range of full-spectrum cannabis extracts (72–74, 76–79, 82, 108) with comparable results.

Since these studies also varied considerably in terms of sample heterogeneity, treatment time and outcome evaluating method, it is very difficult to compare the results among them. However, most of them presented some level of improvement in both core ASD symptoms and comorbidities, with a varied range of specific benefits and proportion of patients with positive outcomes. Another trend across the studies was the occurrence of mostly mild and infrequent adverse effects. Therefore, we may say that, altogether, these studies indicate that FCE-based treatment of ASD symptoms has yielded some level of positive results, regardless of sample heterogeneity, outcome evaluating method or proportion of CBD to THC dosages. However, when high THC formulations are used, the frequency of behavioral side-effects is higher (72). Hence, it is reasonable to speculate that, even though some patients may not benefit at all from cannabinoid treatment, there might exist a CBD-to-THC formulation range and cannabinoids dosage better suited to each specific condition in the ASD spectrum. The mostly beneficial results demonstrated here support this hypothesis, since our titration protocol used combinations of extracts with different CBD to THC proportions, resulting in individually tailored CBD to THC proportions. Among the 20 patients enrolled in this study, 15 have used CBD-rich formulations throughout treatment and 3 have started with CBD-rich FCEs, and eventually supplemented it with THC-rich FCEs. Among these 18 patients, only one, who belongs to the first group, showed a negative GOS. On the other hand, among the 2 patients who have used a blend of CBD-rich FCE and THC rich FCEs since the beginning, one (8 m) showed a positive result (GOS = 1), while the other (15f) showed a slightly negative score (GOS = −0,3), which resulted from some worsening in behavioral symptoms. Patient 15f was also the only one who did not follow any kind of parallel support activities like psychotherapy or sports (see Supplementary File S1), which may have had some influence over its GOS score. Among the remaining 19 patients, which displayed predominantly positive GOS scores, the most common activities reported were Inclusive Regular Schooling (60% of cases), Psychotherapy (50% of cases), Occupation Therapy and Speech Therapy (45% of cases each) and Special Schooling and Physical Activities (30% of cases each). When proposing a tailored dose FCE treatment, physicians should also observe for the possible relevance of support activities in potentializing the effects of Cannabis treatment.

Regardless of our generally small sample size, our data strongly suggests that the effective range of cannabinoid dosage and CBD to THC proportions in ASD treatment may be wide and flexible, but it is safer to start with CBD-rich FCEs, even though some good results may be obtained by starting with a blend of CBD-rich and THC-rich FCEs. Thus, our results allow us to propose safe guidelines for a dosage scheme protocol, as illustrated in Figure 1: (1) low doses of CBD-rich FCEs should be employed at first for all patients as a standard care; (2) CBD doses should be slowly increased until improvements are observed; (3) if improvements are not observed and/or side-effects begin to occur, return to the dosage that has shown no side effects and start supplementing the first CBD-rich FCE with a second THC-rich FCE, gradually increasing in the final proportion of THC to CBD. Of course, these guidelines will need to be further tested with larger and more diverse sample sizes in order to be fully validated.

Cannabinoids are considered potential treatment option for various neurodevelopmental disorders, such as schizophrenia (109–111) and autism. ASD patients have been shown to possess abnormal neuronal activation and connectivity (61, 112, 113), probably due to deficient regulation in synapses. This is a feature reminiscent of mechanisms involved in epilepsy (61). Further, there is higher incidence of epilepsy among autism patients (114, 115) and EEG records of epileptiform activity have been recorded from non-epileptic ASD children (116). Epilepsy symptoms have been successfully treated using Cannabis extracts, especially due to CBD and THC’s ability to reduce epileptiform activity (117–119), reinforcing the use of both as a possible treatment for ASD. Indeed, people on the spectrum present modified expression of cannabinoid receptors (120) and of the endocannabinoid anandamide (121), just like in animal models of autism (122). CBD might further promote ASD symptom improvement through its anti-inflammatory effects (123), as ASD symptoms has been linked to neuroinflammation (124–126). This research background guided our choice of full spectrum Cannabis extracts instead of purified ones, and we believe it plays a part on achieving the benefits reported above.

4.3. Side effects and polypharmacy

The specific side effects observed in both patients who were receiving multiple medications and patients who only had used FCE varied considerably, but were mostly mild. Patients 15f and 2 m were the two with the lowest general outcome scores (GOS equal to −0.3 and − 1.1, respectively). Both were using many neuropsychiatric medications and perceived worsening in more symptomatic categories and subcategories than all other patients. The patient with the lowest GOS of all, 2 m, has unilaterally decided to completely withdraw at once both FCE treatment and all other medications after presenting behavioral worsening. He is the only one among all patients who has discontinued FCE treatment in consequence of worsening. He is also the only one, among the six patients who discontinued all previously used medications, who presented negative GOS. In his case, the precocious, unadvised discontinuation of all medications at once precluded us from trying to adjust the doses and have a clear understand of what may have caused the worsening. Medication interactions did not seem to have caused problems in other patients, but it is a possible cause of worsening in this case. Most of the other five patients who discontinued all other medications under clinical supervision did not present negative outcome scores in any of the 10 main symptomatic categories. The only exception was patient 17f, who presented worsening in only one main symptom category (“avoidance and/or restrictions of food intake”), but still had a relatively high outcome score (GOS equal to 1.4).

All the four patients who were not using any other medications since the beginning of the FCE treatment (namely 1 m, 4 m, 9 m and 11 m) presented GOS equal or above 1.0. Concomitant use of FCE and other psychiatric medications was mostly well tolerated across participants, supporting the use of FCE both as an adjuvant and as a single pharmacological approach for ASD treatment. There is, however, significant concern about the important side effects related to polypharmacy of psychotropic medications used by people in the spectrum (26, 27, 30, 34, 35, 38, 41, 42, 44–46, 127–139), prevalent in 19% in adults (138) going as high as 81% in children (136, 137). This may reflect the considerable range of non-core ASD symptoms still untreated by conventional clinical protocols for ASD (86).

Discontinuation and/or dose reduction of other medications, as we observed in a considerable number of cases in this work, may significantly reduce the patient’s array of side effects associated to other medicines (26, 27, 30, 34, 35, 38, 41, 42, 44–46, 127, 129–135, 139). Furthermore, in the long run, such a wide range of benefits may also significantly alleviate the family’s economic burden (22).

The authors thank the support of the nonprofit organizations AMA+ME, MARIA FLOR and ABRACE, which provided details on extract concentrations for this paper. They also provide relevant social work in Brazil by enabling access to medical Cannabis products, especially for people in a vulnerable social situation, and for that they deserved to be acknowledged.