Short-term risk and long-term incidence rate of infection and malignancy with IL-17 and IL-23 inhibitors in adult patients with psoriasis and psoriatic arthritis: a systematic review and meta-analysis

The risk of infection and malignancy may be a concern for patients with psoriasis receiving interleukin (IL)-17 and IL-23 inhibitors, particularly with long-term treatments. We aimed to estimate the short-term risks and long-term incidence rates of infection and malignancy with IL-17 or IL-23 antagonists in adult patients with psoriasis and psoriatic arthritis through this comprehensive meta-analysis (PROSPERO registration number: CRD42022363127). We searched PubMed, MEDLINE, Web of Science and ClinicalTrials.gov until May 17, 2023 for randomized placebo-controlled trials and long-term (≥ 52 weeks) open-label extension studies. The estimates of short-term risk ratios (RRs) and long-term exposure-adjusted incidence rates (EAIRs) were pooled using R software 4.1.1 and STATA 16.0. This review included 45 randomized placebo-controlled studies and 27 open-label extension studies. Short-term RRs of serious infection, overall infection and malignancy were 1.45 (95% confidence intervals, 95% CI: 0.81-2.59), 1.20 (95% CI: 1.06-1.35), 0.83 (95% CI: 0.41-1.71) with IL-17 inhibitors; and 0.68 (95% CI: 0.38-1.22), 1.13 (95% CI: 1.00-1.28), 0.87 (95% CI: 0.37-2.04) with IL-23 inhibitors. Increased short-term risks of nasopharyngitis and Candida infection with IL-17 inhibitors were found. Long-term EAIRs of serious infection, overall infection, nonmelanoma skin cancer (NMSC), malignancies excluding NMSC, nasopharyngitis and upper respiratory tract infection were 1.11/100 patient-years (PYs), 57.78/100PYs, 0.47/100PYs, 0.24/100PYs, 15.07/100PYs, 8.52/100PYs, 3.41/100PYs with IL-17 inhibitors; and 1.09/100PYs, 48.50/100PYs, 0.40/100PYs, 0.43/100PYs, 10.75/100PYs, 5.84/100PYs with IL-23 inhibitors. Long-term EAIR of Candida infection was 3.41/100PYs with IL-17 inhibitors. No active or reactivated tuberculosis was ever reported in all the trials, and only a few cases of latent tuberculosis, hepatitis, and herpes zoster were reported during the long-term extension periods. No evidence of increased EAIRs of infection and malignancy with longer durations was found. Our study suggested that short-term risk and long-term incidence of infections and malignancies in psoriasis patients receiving IL-17 inhibitors and IL-23 inhibitors are generally low. However, close monitoring is required for nasopharyngitis and Candida infection with IL-17 inhibitors. Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO/, identifier CRD42022363127.


Introduction
Psoriasis is a chronic, inflammatory and immune-mediated skin disorder occurring worldwide, with plaque psoriasis (PsO) being the most common phenotype (1).Psoriatic arthritis (PsA), affecting approximately 30% of psoriasis patients, is a progressive and inflammatory musculoskeletal disorder resulting in declined quality of life, disability and early mortality (2,3).PsO and PsA are associated with an increased prevalence of comorbidities, including cardiovascular disease, metabolic syndromes, and depression, which lead to substantial burdens for individual patients and society (4).
Currently, a variety of cytokines have been found to play pivotal roles in the pathogenesis of psoriasis, including interleukin (IL)-17 and IL-23 (5).A number of therapeutic agents targeting the IL-17 or IL-23 pathway have been approved in the treatment of PsO and PsA, and they have demonstrated excellent efficacy in both clinical trials and real-world settings (6).With these biologics, complete or nearly complete clearance of psoriatic lesions is now attainable.However, in addition to being implicated in the pathogenesis of psoriasis, the IL-17 family and IL-23 perform fundamental functions in innate and adaptive immunity (7,8).Hence, concerns about infection and malignancy associated with deficiency of IL-17 or IL-23 immunity have been raised, especially in patients receiving long-term therapies (9).
To date, IL-17 inhibitors and IL-23 inhibitors have been proved generally safe and well-tolerated (6).However, individual trials lack large sample sizes or long-term trial durations to detect the risk of rare adverse events (AEs) such as serious infections, rare infections and malignancies (10).Furthermore, long-term safety is one of the key considerations in the treatment strategy for psoriasis, while most meta-analyses focused on the randomized controlled trials (RCTs), whose durations were routinely limited to tens of weeks (11).It is unclear, however, whether the short-term safety data could be generalized to long-term treatments.There is growing evidence from open-label extension studies demonstrating the longterm efficacy and safety profiles, with the follow-up periods extending up to several years (11).Nevertheless, to the best of our knowledge, comprehensive studies that generalize and quantify the current safety data, particularly the long-term data, are still lacking (11,12).

Study design and systematic literature search
Studies on IL-17/IL-23 inhibitors combined with other drugs instead of monotherapy were also excluded.The titles and abstracts of potentially eligible articles were screened by two independent investigators (S Wu and Y Xu).Afterwards, the full texts of the remaining studies were reviewed to identify if they met the inclusion criteria.Discrepancies between the investigators regarding the inclusion of the studies were resolved by discussion and consultation of a third researcher (L Guo).The risk of bias was evaluated by two independent reviewers (S Wu and Y Xu), using the Cochrane risk of bias tool for randomized studies of interventions (13).The divergence of opinions was resolved by consultation of another researcher (L Guo).The results were summarized and presented using Review Manager 5.4.1.

Outcomes, definitions and interventions
following data were extracted: the number of patients in treatment and placebo groups, treatment dose, duration of the placebocontrolled periods, the number of serious infections, overall infections, malignancies, nasopharyngitis, upper respiratory tract infection and Candida infection, and the type of malignancies in both the treatment group and the placebo group.For OLE studies, we extracted the following data: treatment dose, number of patients, duration of the open-label extension periods, total patient-years (PYs) of exposure, exposure-adjusted incidence rate (EAIR) with 95% CIs of serious infection, overall infection, NMSC, malignancies excluding NMSC, nasopharyngitis, upper respiratory tract infection and Candida infection.Two reviewers (S Wu, L Yang) independently performed the data extraction.Disagreements were resolved by discussion and consultation of the third researcher (L Guo), and reviewers were unanimous in their final decisions.For AEs with limited data which were hard to be estimated by quantitative metaanalyses, qualitative descriptions, evaluations and summary tables were presented.
The numbers of subjects with infection or malignancy in the treatment and placebo groups from the eligible RCTs were recorded.Patients who received any dose of the biologics were included in the outcome measures.Risk ratios (RRs) based on the number of patients with infections or malignancies and the number of patients in each group were calculated.A pooled estimate of the RR with associated 95% confidence intervals (CIs) was produced for each outcome, with each result presented with forest plots.Heterogeneity testing was performed using the I 2 test and Q-test.A common-effects model and a random-effects model were both calculated.A random-effects model would be adapted if I 2 ≥50%.Subgroup analyses were conducted by biologic and indication.We performed funnel plots and "peters" tests (more than 10 studies) to estimate the publication bias (14).Sensitivity analyses were performed to measure the consistency and robustness of the outcomes.Statistical analyses were performed by R software 4.1.1 and the "meta" package.P-values less than 0.05 were considered significant.
The long-term incidence rates of infection and malignancy were determined by EAIRs with 95% CIs, calculated by events/100PYs.Data of 95% CIs were calculated based on Poisson distribution if they were not provided by the original literature (15).Data with durations of more than 52 weeks were included and analyzed.Patients who received any doses of the biologics were included in the final measures.We produced a pooled estimate of the EAIR of each outcome, with the results presented with forest plots.The results were stratified by years, aimed to determine EAIRs of different treatment durations.Statistical analyses were performed using STATA 16.0.P-values less than 0.05 were considered significant.

Search results and trial characteristics
A total of 4,359 potentially relevant publications were identified through database searching, of which 4,116 were excluded after screening the title or abstract, 174 were excluded after full-text review.The researchers agreed on the major part of the study inclusion and data extraction, except for 6 publications, which were solved by consultation of the third researcher (L Guo).Then three publications were included, and three were excluded (1 study on children; 1 study on psoriasis erythrodermic; 1 study not placebocontrolled).Ultimately, 71 publications were qualified for inclusion (Figure 1) .A total of 25,496 patients were included in 44 randomized placebo-controlled studies  and 17,591 patients with 42,043.3PYs of total exposure were included in 28 OLE studies (17,31,48,.The characteristics of RCTs were summarized in Table 1, and the characteristics of OLE studies were summarized in Table 2.All the included studies researched monotherapy of IL-17 or IL-23 inhibitors, with no other drug combined throughout all the trial periods.The reported cases of tuberculosis, hepatitis, and herpes zoster were summarized in Tables 3 and 4, as some of these AEs were so rare that meta-analyses were not able to be performed.Original data extracted from the included literatures were summarized in Tables S2-S4.Even when the risk of blinding was unclear in some trials, the overall risk of bias was low.The detailed risk of bias of each included study assessed by the Cochrane risk of bias tool for randomized studies of interventions was presented in Figure S1.

Upper respiratory tract infection
The outcomes of upper respiratory tract infection were reported in 25 trials, including 9,073 patients in the treatment group and 3,642 patients in the placebo group.The overall RR of upper respiratory tract infection was found to be 1.19 (95% CI: 0.99-1.43)(Figure 5B).The risk of upper respiratory tract infection was   S6A).The RR of upper respiratory tract infection was not increased in either PsO or PsA (Figure S6B).

Candida infection
The outcomes of Candida infection were reported in 13 trials, including 6,519 patients in the treatment group and 2,611 patients in the placebo group.The overall RR of Candida infection was 3.10 (95% CI: 1.83-5.24)(Figure 5C).The risk of Candida infection was increased with ixekizumab (RR = 2.59, 95% CI: 1.02-6.53),and was significantly increased with bimakizumab (RR = 6.47, 95% CI: 2.29-18.33)(Figure S7A).The RR of Candida infection was increased in both PsA and in PsO (Figure S7B).

Tuberculosis
There were 12 studies reporting the outcomes of tuberculosis, including 3,829 patients in the treatment group and 1,587 patients in the placebo group.In all the included RCTs, there was no case of latent or active tuberculosis reported in both treatment and placebo groups (Table 3).Therefore, quantitative meta-analysis was not performed.Nevertheless, we could obviously observe a quite low short-term risk of tuberculosis with IL-17 inhibitors in population without history or evidence of tuberculosis.

Hepatitis
There were 2 studies reporting the outcomes of hepatitis with IL-17 inhibitors, including 952 patients in the treatment group and 425 patients in the placebo group.There were only one case of hepatitis C reported in the treatment groups (30), and one case of hepatitis B reported in the placebo groups (28) (Table 3).Generally, the incidence of hepatitis was very low in the induction periods.Meta-analysis also did not reveal an increased risk (RR = 0.39, 95% CI: 0.05-3.06)(Figure 5D).Subgroup analysis was not performed due to limited data.

Herpes zoster
There were 2 studies reporting the outcomes of herpes zoster, including 3,447 patients in the treatment group and 1,310 patients in the placebo group.The overall RR of herpes zoster was 0.32 (95% CI: 0.07-1.36)(Figure 5E).As shown in Table 3, the overall incidence of herpes zoster was low.Two cases were reported in the treatment group, and three cases were reported in the placebo group.Subgroup analysis was not performed due to limited data.

Overall infection
The overall long-term EAIR of overall infection was 57.78/100 PYs (95% CI: 39.89-75.68)(Figure 6B).The EAIR of overall infection was 62.01/100PYs for 1 year, 84.88/100 PYs for 2 years, 24.65/100 PYs for 3 years, and 17.60/100 PYs for 5 years, respectively.There was no evidence showing an increase in the incidence rate of overall infection with the prolonged duration of anti-IL-17 treatment.

Tuberculosis
There were 12 studies reporting the long-term incidence of tuberculosis.One case of latent tuberculosis was reported in the sixyear OLE period of brodalumab (210 mg, every 2 weeks) (62); one case was reported in the three-year OLE period of Ixekizumab (160 mg at week 0, then 80mg every 2 weeks) (65); and one case was reported in the one-year OLE period of Ixekizumab (160 mg at week 0, then 80mg every 4 weeks) (66).Other included studies all reported 0/100PYs of tuberculosis.No active or reactivated tuberculosis cases were ever reported in all the included studies.No quantitative meta-analysis was performed due to the limited data.In general, the long-term EAIRs were quite low.

Hepatitis
There were 5 studies reporting the long-term incidence of hepatitis.One case of hepatitis was reported in the one-year and six-year OLE periods of brodalumab (210 mg, every 2 weeks) (17); and one case of hepatitis B was reported in the one-year OLE period of ixekizumab (160 mg at week 0, then 80mg every 4 weeks) (64).No cases of hepatitis were reported by other included studies.No quantitative meta-analysis was performed due to the limited data.

Herpes zoster
There were 6 studies reporting the long-term incidence of herpes zoster.One case of herpes zoster was reported in the oneyear OLE period of brodalumab (210 mg, every 2 weeks) (17); one case in the one-year and three-year OLE periods of ixekizumab (160 mg at week 0, then 80mg every 2 weeks) (64, 65); 2 in the one-year period of ixekizumab (160 mg at week 0, then 80mg every 2 weeks) (66).Other included studies reported 0/100PYs of EAIRs though the whole trials.No quantitative meta-analysis was performed due to the limited data.In general, the EAIRs were very low.

Nasopharyngitis
The outcomes of nasopharyngitis were reported in 14 studies, including 4,855 patients in the treatment group and 2,200 in the placebo group.The overall RR of nasopharyngitis was not increased with IL-23 inhibitors (RR = 1.15, 95% CI: 0.94-1.41)(Figure 11A).Subgroup analysis by biologic showed that none of the anti-IL-23 agents could improve the risk of nasopharyngitis (Figure S11A).Subgroup analysis by indications suggested that the risk of nasopharyngitis was not increased in either PsO or PsA (Figure S11B).Pooled risk ratio (RR) of overall infection with the treatment of IL-23 inhibitors vs placebo.Frontiers in Immunology frontiersin.org

Upper respiratory tract infection
The outcomes of upper respiratory tract infection were reported in 15 studies, including 4,832 patients in the treatment group and 2,244 patients in the placebo group.The overall RR of upper respiratory tract infection was not increased with IL-23 inhibitors (RR = 1.04, 95% CI: 0.82-1.32)(Figure 11B).Subgroup analysis by biologic showed that none of the anti-IL-23 agents could improve the risk of upper respiratory tract infection (Figure S12A).Subgroup analysis by indications suggested that the risks of upper respiratory tract infection were not increased in either PsO or PsA (Figure S12B).

Tuberculosis
There were 14 studies reporting the outcomes of tuberculosis, including 4,935 patients in the treatment group and 2,130 patients in the placebo group.Similar to IL-17 inhibitors, there was zero case of latent or active tuberculosis reported in both treatment and placebo groups with IL-23 inhibitors (Table 3).As a result, quantitative meta-analysis was not performed.A very low shortterm risk of tuberculosis with IL-23 inhibitors was reported in population without history or evidence of tuberculosis.

Hepatitis
There were 2 studies reporting the outcomes of hepatitis with IL-23 inhibitors, including 621 patients in the treatment group and 310 patients in the placebo group.There were only one case of hepatitis B reported in the treatment groups (50) (Table 3).Metaanalysis was not performed due to the limited data.Generally, the incidence of hepatitis was found very low in the induction periods.

Herpes zoster
There were 7 studies reporting the outcomes of herpes zoster, including 2,650 patients in the treatment group and 1,364 patients in the placebo group.Quantitative meta-analysis suggested that the overall RR of herpes zoster was not increased with IL-23 inhibitors (RR = 0.95, 95% CI: 0.24-3.76)(Figure 11C).Subgroup analysis was not performed due to limited data.
3.5 Long-term incidence rates with IL-23 inhibitors 23 inhibitors (Figure 12A).The EAIR of serious infection was 1.33/ 100 PYs for 1 year, 1.18/100 PYs for 2 years, 1.13/100 PYs for 3 years, 1.20/100 PYs for 4 years, and 1.05/100 PYs for 5 years.No evidence of an increased incidence rate of serious infection was found with the prolonged duration of anti-IL-23 treatment.

Overall infection
The overall long-term EAIR of overall infection was 48.50/100 PYs (95% CI: 31.76-65.24)(Figure 12B).The EAIR of overall infection was 35.50/100PYs in psoriasis patients receiving IL-17 inhibitors for 1 year, 52.73/100 PYs for 2 years, 74.03/100 PYs for 3 years.There was also no evidence showing an increased incidence rate of overall infection with the prolonged duration of anti-IL-23 treatment.

Upper respiratory tract infection
The overall long-term EAIR of upper respiratory tract infection was 5.84/100PYs (95% CI: 3.94-7.74)(Figure 13B).The EAIR of upper respiratory tract infection was 6.72/100PYs for 1 year, 4.65/ 100PYs for 3 years, 10.70/100PYs for 5 years, and 4.17/100PYs.No evidence of an increased incidence rate of upper respiratory tract infection was found with the longer duration.

Tuberculosis
There were 6 studies reporting the long-term incidence of tuberculosis.Of all the included studies, the long-term EAIRs of tuberculosis were 0/100PYs, with no case of active, reactivated or latent tuberculosis reported throughout the whole trials.Therefore, the overall EAIR of tuberculosis with IL-23 inhibitors was estimated to be 0/100PYs.No quantitative meta-analysis was performed due to the limited data.

Hepatitis
There was 1 study reporting the long-term incidence of hepatitis.Only one case of hepatitis B was reported in the twoyear OLE period of guselkumab (100 mg, every 4 weeks) (80).No case of hepatitis with other types of IL-23 inhibitors was reported.No quantitative meta-analysis was performed due to the limited data.

Herpes zoster
There were 4 studies reporting the long-term incidence of herpes zoster.One case in the two-year OLE period of guselkumab (100 mg, every 4 weeks) (80), one in the three-year and five-year OLE periods of Tildrakizumab (200 mg, at week 4 and every 12 weeks thereafter) (83, 84).Other included studies reported 0/100PYs of EAIRs though the whole trials.No quantitative meta-

A B
Pooled exposure-adjusted incidence rates (EAIRs) of (A) nasopharyngitis; (B) upper respiratory tract infection in psoriasis patients receiving long-term anti-IL-23 treatment.analysis was performed due to the limited data.In general, the EAIRs were very low.

Discussion
Therapeutic agents that specifically target the IL-17 family and IL-23 have demonstrated excellent efficacy in PsO and PsA (5,6).However, the potential for infection and malignancy has been a concern with respect to the therapies which suppress immune pathways (9,(85)(86)(87).Individual studies are not large enough in sample sizes or of long enough duration to ascertain uncommon adverse events.Hence, we performed a meta-analysis of RCTs and OLE studies to estimate the risk and incidence of infections and malignancies with treatments of IL-17 or IL-23 inhibitors.The results suggested that anti-IL-17 or anti-IL-23 treatments did not increase the short-term risk of serious infection or malignancy in adult patients with psoriatic disease.Correspondingly, the longterm incidence rates of serious infection and malignancy were quite low, in particular the incidence rate of malignancy (less than 1/ 100PYs).Despite the enhanced short-term risks of overall infection and some specific infections with both IL-17 inhibitors and IL-23 inhibitors, the risk ratios were moderately increased.The long-term incidence rates of overall infection were estimated to be 57.78/100PYs with IL-17 inhibitors and 48.5/100PYs with IL-23 inhibitors.
Subgroup analyses by biological agents suggested that brodalumab, ixekizumab, secukinumab, bimekizumab, guselkumab, risankizumab and tildrakizumab are safe in general, but still showed diversity in safety profile.In this study, ixekizumab, secukinumab, bimekizumab and risankizumab increased the shortterm risk of overall infection, whereas other drugs did not.Specifically, ixekizumab might increase the short-term risk of Candida infection; secukinumab might increase the short-term risk of nasopharyngitis and upper respiratory tract infection; bimekizumab might increase the short-term risk of nasopharyngitis and Candida infection.None of guselkumab, risankizumab and tildrakizumab was able to increase the short-term risk of nasopharyngitis or upper respiratory tract infection.In general, the risks of infection and malignancy were comparable between patients with PsO and PsA, though there were some differences in individual AEs.
Nasopharyngitis and upper respiratory tract infection are considered the most frequent AEs with IL-17 or IL-23 antagonists (10,11).It is interesting to note that IL-17 inhibitors did not increase the short-term risk of upper respiratory tract infection, and slightly increased the risk of nasopharyngitis.IL-23 inhibitors did not increase the short-term risk of nasopharyngitis or upper respiratory tract infection.But the long-term EAIRs of nasopharyngitis (15.07/100PYs with IL-17 inhibitors; 10.75/ 100PYs with IL-23 inhibitors) and upper respiratory tract infection (8.52/100PYs with IL-17 inhibitors; 5.84/100PYs with IL-23 inhibitors) were somewhat high.It is possible that nasopharyngitis and upper respiratory tract infection, albeit as two common infections in real-life settings with high incidence, are likely to have comparable incidence rates in psoriasis patients with or without using IL-17/IL-23 inhibitors.However, this finding requires further validation.
It is noteworthy that Candida infection appeared to be a prominent AE for IL-17 inhibitors.Our study suggested that the short-term risk of Candida infection was approximately 3-fold higher when administrating IL-17 antagonists compared with placebo.The long-term incidence rate was estimated to be 3.41/ 100PYs.It is well-acknowledged that IL-17 plays a role in host defense against bacterial and fungal infections, particularly in mucocutaneous microbial surveillance (86,88).The deficiency of IL-17 immunity could lead to infections with Candida species (89).Hence, continued vigilance for such infection will be required during treatment with inhibitors of this pathway.In addition, It is worth noting that the risk of Candida infection was increased almost 6-fold when bimekizumab was used, which appeared to be higher than other IL-17 inhibitors.These findings were consistent with data from existing clinical trials, which demonstrated a higher incidence of Candida infection with bimekizumab than other IL-17 antagonists (41-45, 75, 76).It might be explained by the differing targets between bimekizumab and other IL-17 inhibitors (6).In comparison to ixekizumab and secukinumab which target IL-17A, bimekizumab has been shown to have dual neutralization of IL-17A and IL-17F.These findings point to the important role of IL-17F in host defense, especially at the mucocutaneous sites (86,88).Despite higher Candida infection occurrence with IL-17 antagonists, most reported cases were mild or moderate infections, none of them were systemic, and most cases could be resolved with appropriate antifungal treatment without causing trial dropouts (41)(42)(43)(44)(45).
Compared with Candida infection, the incidence rates of tuberculosis, hepatitis, and herpes zoster were quite lower.For IL-17 inhibitors, zero case of tuberculosis in the induction periods was found, and only 3 cases of latent tuberculosis in long-term extension periods were reported.For IL-23 inhibitors, no case of tuberculosis in the whole induction periods and extension periods was reported.No case of active or reactivated tuberculosis was ever reported in all the included studies and throughout the whole trial periods, either with IL-17 inhibitors or IL-23 inhibitors.Short-term risks of hepatitis and herpes zoster seemed not increased according to the meta-analyses, despite limited studies and small sample sizes.Only several cases of hepatitis and herpes zoster were reported throughout the whole trial periods.The trial researchers tended not to attribute these cases to treatment-related reasons (17, 64-66, 80, 83, 84).These results supported that IL-17 inhibitors and IL-23 inhibitors may not increase the risk of tuberculosis, hepatitis, and herpes zoster in psoriasis patients without known diagnoses or previous histories of these infections.However, it did not mean that patients with history or evidence were totally safe to receive IL-17 or IL-23 inhibitors, as patients at high risk or with previous histories of tuberculosis, hepatitis or herpes zoster were already excluded in most studies.Therefore, it is noteworthy that patients with susceptibility to opportunistic infections and hepatitis always need overall examination and prudent evaluation before and during using any biologics (6,10).
The long-term safety analyses presented results of continuous treatments of IL-17 inhibitors and IL-23 inhibitors for up to 6 years, with > 42,000 PYs of total exposure.The cumulative long-term EAIRs of serious infection and malignancies are generally comparable to the results from the psoriasis reference population captured in the Psoriasis Longitudinal Assessment and Registry (PSOLAR) (90, 91).The long-term EAIRs of overall infection, serious infection, NMSC and malignancies excluding NMSC were also comparable between IL-17 antagonists and IL-23 antagonists.IL-23 maintains IL-17producing Th17 cells, and is essential for their differentiation, activation, and survival (8).Therefore, inhibition of IL-23 blocks downstream production of IL-17A by Th17 cells and other cell types (87).This might explain the comparable safety profiles between IL-17 inhibitors and IL-23 inhibitors in our study.The EAIRs of infection and malignancy were stratified by years to estimate the EAIRs in different durations.While the present study was not designed to quantify the differences of EAIRs between different durations, these findings seemed to suggest that the EAIRs of infection and malignancy might be broadly consistent or even diminished with the prolonged treatment duration.These findings indicated that longterm sustained treatments with IL-17 or IL-23 inhibitors with appropriate doses could be safe with continued follow-ups of the patients.
This study has several limitations.Firstly, we did not assess the dose-related trends in the incidence rate or severity of infection and malignancy.Patients are included in our analysis once they have received at least one dose of either the anti-IL-17 or anti-IL-23 antagonists.Secondly, this study might fail to evaluate the risk in patients at high risk for infections or malignancies because of the possible exclusion of patients with previous histories of infections or malignancies in the original clinical studies.Thirdly, some of the results of the meta-analyses may still present small sample sizes due to the limited data, which may decrease the reliability.Finally, we only included RCTs and OLE studies in this review.There is a need for a meta-analysis of real-world studies evaluating the risk of infection and malignancy in patients with psoriatic disease who receive IL-17 or IL-23 antagonists.
In conclusion, IL-17 inhibitors and IL-23 inhibitors are generally safe in the treatment of psoriasis and psoriatic arthritis, with both short-term and long-term treatments.Nasopharyngitis and upper respiratory tract infection are the most frequent AEs, and Candida infection seemed to be the most common opportunistic infection.It is critical to exercise caution and maintain ongoing vigilance for nasopharyngitis and Candida infection in patients receiving IL-17 inhibitors.However, to comprehensively evaluate the potential risks of cancer and infection related to prolonged use of IL-17 inhibitors and IL-23 inhibitors in individuals with psoriasis, larger, long-term, prospective cohort studies will be required.

FIGURE 1 Flow
FIGURE 1Flow diagram of selection of literature.

FIGURE 3
FIGURE 3Pooled risk ratio (RR) of overall infection with the treatment of IL-17 inhibitors vs placebo.

FIGURE 4
FIGURE 4    Pooled risk ratio (RR) of malignancy with the treatment of IL-17 inhibitors vs placebo.

FIGURE 8
FIGURE 8Pooled risk ratio (RR) of serious infection with the treatment of IL-23 inhibitors vs placebo.

FIGURE 10 11
FIGURE 10Pooled risk ratio (RR) of malignancy with the treatment of IL-23 inhibitors vs placebo.

TABLE 1
Trial characteristics of randomized placebo-controlled trials.

TABLE 2
Trial characteristics of open-label extension studies.

TABLE 3
The number of tuberculosis, hepatitis, and herpes zoster in both treatment groups and placebo groups in the included RCTs.

TABLE 4
Cases of tuberculosis, hepatitis, and herpes zoster reported in the OLE studies.

TABLE 4 Continued
FIGURE 2Pooled risk ratio (RR) of serious infection with the treatment of IL-17 inhibitors vs placebo.