Transient pre-baseline antipsychotic exposure (TPAE) is a prognostic specifier in clinical high risk for psychosis: evidence from the PSYSCAN consortium study

Abstract

Baseline exposure to antipsychotics (AP) in individuals at clinical high risk for psychosis (CHR-P) has emerged as a robust prognostic marker for transition to psychosis. Recent meta-analyses show that CHR-P individuals receiving APs at study entry exhibit significantly higher transition rates, with a temporally discernible dose-effect pattern. To test the replicability of this prognostic signal, we conducted a secondary analysis of publicly available supplementary data from the PSYSCAN Consortium. Despite applying strict inclusion criteria that limited AP exposure to ≤30 days of low-dose treatment, transition rates in PSYSCAN remained higher in the AP-exposed group (28.0% vs. 12.2%; RR 2.29). Notably, this finding was not emphasized in the primary publication, thus representing inadvertent yet compelling support for the prognostic relevance of early AP need. These results reinforce the rationale for future CHR-P studies to adopt AP-naïve condition as a stratification criterion to avoid pharmacologically induced diagnostic obscuration and improve the precision of risk modelling. Furthermore, they corroborate the PSYSCAN-derived operational notion of transient pre-baseline AP exposure (TPAE) (i.e., ≤30 cumulative days in the 3 months preceding baseline, at subtherapeutic doses) as a pragmatic and testable prognostic specifier that supplements existing CHR-P criteria.

Recent meta-analyses suggest that ongoing pharmacotherapy at inception, particularly antipsychotic exposure, represents a key prognostic modulator for psychosis transition in help-seeking individuals at clinical high risk for psychosis (CHR-P)^1–5.

In particular, with respect to AP, those cumulative evidences indicate that:

1. baseline pharmacological exposure to antipsychotics (AP) in individuals of CHR-P cohorts is associated with an increased risk of transitioning to psychosis in comparison with non-exposed individuals (32.9% vs. 20.6 Risk Ratio 1.56)³;

2. such negative prognostic effect is not due to differential enrichment strategies (i.e., pre-test risk enrichment);⁶

3. the temporal dynamics of transition to psychosis differs in AP-exposed vs. AP-naive CHR, i.e., baseline AP exposure in CHR-P is associated with a persistently higher risk of transition at follow up⁴;

4. such effect follows a dose-response pattern: among CHR-P individuals exposed to antipsychotics (APs) at baseline, those who later transitioned to psychosis had received a baseline AP prescription within the range of the mean effective chlorpromazine equivalent daily dose⁵.

Following meta-analytical evidences, the prognostic risk associated with baseline AP exposure was independently confirmed in CHR-P cohorts^7–10 and is clinically interpreted not as a paradoxical iatrogenic effect but rather as a clinical proxy of increased baseline clinical severity which escapes the standard CHR-P criteria^11,12. This suggests that AP-treated individuals, despite being classified as CHR-P, may functionally represent pharmacologically attenuated first-episode psychosis, reducing the diagnostic resolution of current CHR-P stratification frameworks¹³ (which are largely agnostic to baseline pharmacological exposures)¹⁴.

Furthermore, it must be emphasized that all available treatment guidelines for individuals at CHR-P almost unanimously do not indicate AP as first therapeutic option to prevent transition to psychosis (see ref. ¹⁵ for an updated synopsis), suggesting that exposure to AP in help-seekers undergoing CHR-P assessment is a relevant clinical-anamnestic detail. In light of such increasing empirical evidence^1,4–10, the recently launched “Largest Global Prospective Cohort Study of Clinical High Risk for Psychosis” under the Accelerating Medicines Partnership® Schizophrenia (AMP® SCZ)¹⁶ has taken an important step by listing “current antipsychotic medication at the time of baseline assessment” as an exclusion criterion, ensuring more accurate assessments of the natural course of psychosis risk.

In contrast, the PSYSCAN project¹⁷ offers a unique multinational platform to critically evaluate the prognostic role of AP treatment in CHR-P individuals. Indeed, PSYSCAN’s inclusion of participants with limited, carefully documented AP exposure ( ≤ 30 days at subtherapeutic doses) creates an ideal natural platform to directly test the negative prognostic association found in meta-analyses of CHR-P help-seekers. We therefore conducted a secondary conceptually-driven analysis of PSYSCAN data to determine whether even a circumscribed, low dose and time-limited AP exposure before inception (as per PSYSCAN specific exclusion criteria) impacts the imminent risk of transition to psychosis.

PSYSCAN consortium: operational exclusion criteria and the naturalistic subgroup CHR-P individuals with brief, low-dose AP exposure

According to Tognin et al.¹⁷, in a presumably cautious, albeit undeclared, attempt to mitigate the potential confounding influence of baseline antipsychotic (AP) exposure, the PSYSCAN study adopted relatively strict inclusion criteria, i.e., excluding individuals who had received AP treatment for more than 30 cumulative days within the 3 months prior to baseline, and only at doses deemed adequate for first-episode psychosis treatment.

Nevertheless, this protocol allowed for the inclusion of CHR-P individuals with brief, low-dose AP exposure [(Transient pre-baseline antipsychotic exposure (TPAE); See Box 1], thereby inadvertently creating a naturalistic subgroup for exploratory prognostic analysis.

Box 1 Transient Pre-Baseline Antipsychotic Exposure (TPAE): a PSYSCAN-derived operational prognostic specifier.

Definition Transient Pre-Baseline Antipsychotic Exposure (TPAE) refers to a distinct subgroup of CHR-P individuals who meet the following criteria:

• Antipsychotic treatment duration: ≤30 cumulative days

• Timing: Within the 3 months prior to baseline assessment

• Dosage: Subtherapeutic (below standard First-Episode Psychosis treatment thresholds)

• Clinical status during exposure: No evidence of full-threshold psychosis

• CHR-P status at baseline: Maintained (per CAARMS, SIPS, SPI-A, or equivalent)

This construct is derived from PSYSCAN study parameters and captures short-term, pre-enrollment AP exposure that has concluded or is clinically negligible by the time of formal CHR-P assessment.

Prognostic Relevance

TPAE functions as a prognostic specifier within the CHR-P framework. In the PSYSCAN cohort, individuals with TPAE status exhibited a transition rate of 28.0%, significantly higher than their AP-naïve counterparts (12.2%), despite the controlled, low-dose exposure window.

Comparison with General Baseline AP Exposure

Feature	TPAE	General Baseline AP Exposure
Exposure duration	≤30 days	Unrestricted
Exposure window	3 months pre-baseline	Often undefined
Dose level	Subtherapeutic	Any dose, including therapeutic
AP status at baseline	Discontinued or minimal	May be ongoing
Definition source	Derived from PSYSCAN exclusion criteria	Heterogeneous across studies
Prognostic signal (RR)	2.29	~1.56 (meta-analytic estimate)
Clinical utility	Reproducible and operationally defined; stratifies high-risk CHR-P cases.	Clinically indicative, yet with low granularity (despite meta-analytic evidence of dose-effect pattern)
Interpretation	TPAE status increases the prognostic precision of CHR-P classification, indexing a CHR-P subgroup at increased risk of transition to psychosis.	Baseline antipsychotic (AP) exposure is a broad proxy for elevated imminent risk of psychosis transition in CHR-P individuals, further substantiated by meta-analytical evidence demonstrating a dose-effect relationship

PSYSCAN consortium: supplementary data on AP exposure

Although not examined or discussed in the main text of the PSYSCAN article, the Supplementary Material (Table S4) discloses a striking discrepancy in transition rates between CHR-P individuals exposed to antipsychotics (AP) at baseline and those who were AP-naïve: 28.0% (7 of 25) versus 12.2% (26 of 213), corresponding to a relative risk of 2.29 (95% CI: 1.11–4.73)¹⁸. This difference emerged despite the study’s deliberate exclusion of individuals with prolonged or high-dose AP exposure (specifically, those treated for more than 30 cumulative days in the 3 months prior to baseline at dosages considered therapeutically adequate for a first-episode psychosis). The fact that such a prognostic signal surfaced even under these restrictive conditions underscores the significance of minimal baseline AP exposure as an indicator of heightened transition risk.

Notably, because this differential transition rate was reported only in supplementary materials and not interpreted or emphasized by the authors, it gains a distinct layer of credibility: the finding is free from confirmatory bias and thus represents a form of inadvertent evidence. This strengthens the broader empirical signal previously identified in meta-analyses, which consistently link baseline AP exposure in CHR-P individuals to a persistently elevated risk of psychosis onset, likely reflecting unmeasured clinical severity that escapes current CHR-P stratification models. Moreover, in PSYSCAN study it was not possible to directly compare symptomatic clinical severity between TPAE and AP-naïve subgroups, leaving open the possibility that unmeasured baseline psychopathology may confound the observed effect. Additionally, given the multi-centre nature of PSYSCAN, site-level variability in prescribing thresholds may have introduced selection bias, possibly weakening the generalizability of the results, deserving further empirical confirmation. Finally, the analysis did not control for covariates; however, the RR consistency with prior meta-analyses^1,4,5 supports TPAE’s utility.

Accordingly, this derivative insight from the PSYSCAN data supports the call for a new generation of CHR-P studies based on explicitly AP-naïve cohorts. Such a methodological refinement is critical to disentangle intrinsic transition risk from pharmacologically modulated clinical presentations and to avoid conflating attenuated psychosis syndromes with partially treated early psychosis (or other previous clinical conditions not transparently detailed). Within the framework of precision psychiatry, this distinction is intuitively essential for improving diagnostic clarity, enhancing prognostic accuracy, and refining individualized prevention strategies.

Conclusions

This finding suggests that it is not merely the characteristics of antipsychotic (AP) treatment, such as dose or duration, that are prognostically relevant, but rather the very need for AP intervention that serves as a significant clinical indicator of heightened risk for transition to psychosis in help-seekers referred for CHR-P assessment. Such need likely reflects an underlying severity that may not be fully captured by psychometric tools alone and is certainly invisible through current CHR-P stratification^1,4–6,11.

What makes this specific evidence about TPAE prognostic value particularly compelling is its source: it does not arise from CHR-P cohorts where AP use was unrestricted at baseline¹⁹, but rather from the PSYSCAN cohort, which applied explicit exclusion criteria to limit prior AP exposure to short-term, low-dose use. This design unintentionally created a naturalistic gradient of AP exposure, allowing an indirect yet powerful test of its prognostic meaning.

Building on this, and in the spirit of enhancing precision psychiatry, our analysis underscores the necessity for a next generation of CHR-P studies. These should incorporate transparent, systematic documentation of pharmacological exposures and their related clinical needs, before inception, at baseline and after inception (from baseline to follow ups), and prioritize the recruitment of AP-naïve individuals to unmask true risk trajectories²⁰. As converging evidence suggests, baseline AP use is a non-neutral variable: it may confound diagnostic classification and obscure the distinction between an attenuated psychosis syndrome and a pharmacologically attenuated first-episode psychosis¹¹.

Within a refined clinical and research framework, AP exposure should not be dismissed as a contingent, background variable, but rather recognized as both a prognostic stratifier (at inception) and a functional proxy for early conversion (at follow up). Consequently, stratifying CHR-P individuals based on AP-naïve condition versus AP-treatment should be adopted as a gold standard in both research and clinical monitoring and TPAE status should be systematically recorded and reported to refine prognostic models. This approach is critical to improving risk prediction, refining case conceptualization, and ultimately achieving the promise of individualized preventive psychiatry.

Author contributions

A.R., M.P. and A.P. equally conceived the paper and wrote the manuscript. All authors agreed on the final version.

Data availability

The data supporting the findings of this study are publicly accessible via the supplementary materials of the following publication: Tognin S, Vieira S, Oliver D, Cullen AE, Kempton MJ, Fusar-Poli P, Mechelli A, Dazzan P, Merritt K, Maat A, de Haan L, Lawrie SM, van Amelsvoort T, Arango C, Nelson B, Galderisi S, Bressan R, Kwon JS, Mizrahi R; PSYSCAN Consortium; Kahn RS, McGuire P. PSYSCAN multi-centre study: baseline characteristics and clinical outcomes of the clinical high risk for psychosis sample. Schizophrenia (Heidelb). 2025 Apr 17;11(1):66. 10.1038/s41537-025-00598-x. PMID: 40246889; PMCID: PMC12006469. https://static-content.springer.com/esm/art%3A10.1038%2Fs41537-025-00598-x/MediaObjects/41537_2025_598_MOESM1_ESM.docx.

Competing interests

The authors declare no competing interests.