Hallucinogen persisting perceptual disorder: a scoping review covering frequency, risk factors, prevention and treatment

The investigators conducted a scoping review following PRISMA-ScR guidance, searching bibliographic databases from inception to July 2021. Databases specifically mentioned in the extracted text include Ovid MEDLINE, Ovid EMBASE and Ovid PsycINFO; additional articles were identified through reference-list checks and a ‘‘similar articles’’ tool on MEDLINE. An initial MEDLINE search returned 188 records; EMBASE and PsycINFO searches returned 297 and 242 records respectively, and iterative screening and full-text review yielded a final set of 32 included articles. Study selection was performed by two independent reviewers (M.D. and S.L.) who screened titles/abstracts and then full texts; disagreements were resolved by discussion and, where necessary, authors of primary studies were contacted for clarification. The review included studies in which subjects had a diagnosis of HPPD according to DSM or ICD criteria; the extracted text does not supply the full list of pre-specified inclusion and exclusion criteria in one place. Because of marked heterogeneity across study designs and measures, the authors pre-specified a qualitative narrative synthesis rather than a quantitative meta-analysis. Methodological quality and risk of bias were assessed with tools appropriate to each design: the Joanna Briggs Institute checklists for case reports and cross-sectional studies, the Newcastle-Ottawa Scale (NOS) adapted for case-control studies, and the NIH quality-assessment tool for before–after (pre–post) studies without a control group. Two reviewers completed these appraisals independently and resolved disagreements by discussion. Levels of evidence were classified using Centre for Evidence-Based Medicine guidance (with case series and low-quality observational studies as Level 4 and expert/opinion/benchwork as Level 5). Outcomes of interest for the narrative synthesis included reported associations between specific hallucinogens and HPPD onset, pharmacological and non-pharmacological treatments and their reported efficacy, risk or protective factors, prevalence/incidence data by drug class, EEG and functional imaging findings, proposed pathophysiological mechanisms, and evidence regarding prevention measures. The heterogeneity in designs (case reports, case series, cross-sectional, case-control and small pre–post studies) precluded pooled statistical analysis.

Thirty-two articles met inclusion criteria: 21 case reports, three pre–post studies without control groups, six cross-sectional studies and two case–control studies. Study quality varied; case reports frequently lacked detailed patient information and diagnostic work-up, cross-sectional studies relied on self-report, and the pre–post studies were small and uncontrolled. One case–control study scored 7/9 stars on the NOS while the other scored 4/9. On drug associations, lysergic acid diethylamide (LSD) was the most frequently implicated agent. Across an aggregate of 30 cases reporting prior drug histories, 20 listed LSD as associated with symptom onset; in 13 of those cases LSD alone was implicated. An analytical cross-sectional study reported that patients with more severe HPPD type II reported significantly more LSD use than those with type I (t = 18.36, p < 0.001). Natural and synthetic cannabinoids were the second most reported class (12 cases), with synthetic cannabinoids implicated in several cases and an association with HPPD type II reported in a cross-sectional study (Lev‑Ran et al.; z2 = 40, p < 0.001). MDMA was the third most reported agent (nine cases), although no included case report attributed HPPD to MDMA alone. Psilocybin, phencyclidine, ibogaine and 5‑MeO‑DiPT were each reported in three or fewer case reports, often in combination with other drugs. Treatment reports were heterogeneous and uncontrolled. Benzodiazepines, particularly clonazepam, showed consistent positive signals in uncontrolled studies and case reports: a pre–post study (n = 14) administering 2 mg/day clonazepam for 2 months reported significant reductions on clinician-rated Clinical Global Impression (CGI), self-report scales and the Hamilton Anxiety Rating Scale (HAM‑A) from baseline to endpoint (CGI t = 16.3, p < 0.001; self-report t = 18.7, p < 0.001; HAM‑A t = 32.5, p < 0.001), and some sustained improvement at 6-month follow-up. An anticonvulsant regimen using levetiracetam (500 mg morning, 1000 mg evening) in a 1‑year uncontrolled study (Casa and Bosio) produced 20/27 patients who became ‘‘flashback‑free’’ and 14/27 with >50% symptom reduction at 15 days. Lamotrigine was reported effective in several case reports at 100–200 mg/day but findings were inconsistent and possibly influenced by concomitant medications. Adrenergic agents also showed some signal: clonidine (0.025 mg three times daily) reduced symptoms in a small pre–post study (n = 6) and in case reports; propranolol was ineffective in one case, while reboxetine reportedly reduced symptoms in another. Selective serotonin reuptake inhibitors (SSRIs) produced mixed results, with sertraline (50–100 mg/day) and fluoxetine (20 mg/day) reported as effective in some cases but ineffective in others. Atypical antipsychotics (risperidone, olanzapine) had mixed and sometimes opposing effects—some reports describe symptom exacerbation (notably risperidone) while others describe improvement. Opioid antagonists produced mixed findings: naltrexone (25–50 mg/day) reduced symptoms in two case reports, whereas naloxone plus buprenorphine did not in another case in which EMDR therapy was ultimately reported to resolve symptoms. Risk-factor data were limited and inconsistent. No established risk factors were identified. Exposure–response relationships were mixed: some reports and an excluded online survey suggested unusually strong single doses or dysphoric trips preceded symptom onset in a subset of cases, while other historical reports found no relation between number of ‘‘trips’’ and HPPD onset. Genetic predisposition was hypothesised but not demonstrated. Continued use of the precipitating agent, cannabis (natural or synthetic) use after onset, and certain medications (including some atypical antipsychotics) were reported to exacerbate or trigger relapse. Epidemiological estimates remain imprecise: type I ‘‘flashbacks’’ are commonly reported but type II prevalence estimates are uncertain due to non-random sampling and self-report reliance. Clinically, HPPD type II has been associated with panic attacks, anxiety, depressive features, suicidal ideation and suicide in some reports. Pathophysiological observations were largely hypothesis-generating. EEG studies described hyperactivity of visual pathways in HPPD, resembling EEG patterns seen during acute hallucinogen effects. The authors report two broad mechanistic hypotheses: serotonergic 5‑HT2A‑mediated toxicity causing loss of GABAergic inhibitory neurons in visual cortex or the lateral geniculate nucleus, and an adrenergic overactivation model analogous to PTSD. Clinical pharmacology data—benefit from benzodiazepines (GABA‑A modulators) and clonidine (α2‑agonist)—were cited as indirect support for these mechanisms. Comparative work suggested HPPD hallucinations differ from schizophrenia phenomenologically and in affected brain regions (psychedelic states more visual and associative-cortex centric, schizophrenia more auditory and sensory-cortex centric); two small comparative studies reported lower PANSS negative and total scores in patients with schizophrenia plus HPPD versus schizophrenia alone (t values reported: negative symptoms t = 3.45, p < .01; total PANSS t = 3.09, p < .01).

Doyle and colleagues interpret the literature as indicating LSD is the most commonly reported precipitant of HPPD, but emphasise that the etiology remains unresolved. They present competing and partly complementary mechanistic hypotheses implicating serotonergic 5‑HT2A receptor activity and downstream loss or dysfunction of GABAergic inhibitory neurons in visual pathways, as well as a model invoking adrenergic overactivation similar to post‑traumatic stress disorder; pharmacological observations (benzodiazepine responsiveness, clonidine effects) and EEG evidence of visual‑pathway hyperactivity are cited as supporting elements for these hypotheses. Treatment evidence is described as heterogeneous and of limited quality. No single drug class emerges as reliably effective across cases. Positive uncontrolled signals were noted for benzodiazepines (notably clonazepam), certain anticonvulsants (levetiracetam, lamotrigine) and select adrenergic agents; atypical antipsychotics were emphasised as having inconsistent effects and in some reports worsening symptoms. The authors highlight non‑pharmacological approaches as potentially useful, noting a case report in which eye movement desensitisation and reprocessing (EMDR) led to remission following ibogaine‑related HPPD. They further report that HPPD appears less likely in controlled therapeutic or research settings than in recreational use, possibly related to subject selection, monitoring and consistency of dose and drug quality. Key limitations acknowledged by the authors include the overall low quality of component evidence (predominance of case reports and small uncontrolled studies), reliance on self‑report and non‑random samples, small sample sizes in pre–post studies and the inclusion of grey literature. The review was also restricted to studies with DSM/ICD diagnoses of HPPD, excluding earlier reports predating formal DSM definitions. Because of these constraints, prevalence and risk‑factor estimates remain imprecise. The authors call for higher‑quality research—preclinical mechanistic studies, epidemiological work to identify risk and protective factors, and double‑blind clinical trials to assess treatments and dosing paradigms—to better characterise HPPD and inform clinical practice and safety monitoring in therapeutic psychedelic research. Finally, the expert commentary appended by the authors underscores a pragmatic stance: HPPD is a serious but uncommon adverse outcome that should be monitored as psychedelic therapies advance, and therapeutic investigations provide an opportunity to study HPPD more rigorously given controlled dosing and participant selection.

The review concludes that, while HPPD is an uncommon but clinically important consequence of hallucinogen exposure, current knowledge is limited by low‑quality and heterogeneous evidence. There is a clear need for more rigorous studies to clarify etiology, identify risk factors and establish effective, evidence‑based treatments. The authors recommend future research including well‑designed double‑blind trials of candidate treatments, preclinical investigations of pathophysiology, and epidemiological studies to determine prevalence and predictors, and they urge that ongoing and future clinical trials of hallucinogens include deliberate monitoring and characterisation of HPPD.