Наркотики и вакцины: систематический обзор литературы.

1. Introduction

Cutaneous T-cell lymphomas (CTCLs) are a type of non-Hodgkin lymphoma that usually involves the skin and rarely has extracutaneous invasion [ 1 ]. CTCL incidence increases by age, with a mean age of around 54 years, and it is not common in children [ 2 , 3 ]. CTCL has different subtypes including mycosis fungoides (MFs), Sézary syndrome (SS), primary cutaneous anaplastic large lymphoma (PC-ALCL), lymphomatoid papulosis (LyP), subcutaneous panniculitis–like T-cell lymphoma (SPTCL), and extranodal NK/T-cell lymphoma. The most common subtypes are MF, PC-ALCL, and LyP with a portion about 80% of disease frequency [ 4 ].

Clinical presentation varies from single patch or plaques to generalized lesions or blood involvement. Because of the wide spectrum of manifestations, the diagnosis is based on clinical and histopathological features [ 1 , 4 ]. Disease-specific survival depends on the subtype and varies from around 10%–100% [ 4 ]. However, it prominently decreases patients' quality of life [ 5 ]. Although there is not any curative treatment, there are several treatment options such as phototherapy, chemotherapy, systemic drugs, and stem cell transplantation [ 1 ].

CTCL pathogenesis is not completely discovered yet. Studies show that several factors including various mutations, different subsets of T-cells, chemokine receptors, and microRNA dysregulation and several molecular mechanisms such as JAK-STAT pathways are involved in disease incidence and progression [ 6 , 7 ].

There are several reports of incidence, relapse, or progression of CTCLs by using specific drugs [ 8 – 11 ]. Although the exact mechanism is not clear yet, it seems this happens due to the dysregulation of specific pathways such as JAK-STAT, which have a role in disease pathogenesis. Other probable hypotheses are as follows: PD-1 inhibitors lead T-cell proliferation [ 12 ], roles of monoclonal antibodies by affecting interleukins [ 12 ], immunosuppression [ 13 ], and some other unclear immunological mechanisms [ 14 ].

The aim of this systematic review is to identify drugs and vaccines that were associated with the incidence of CTCLs, their course of disease, and outcomes through the literature.

2. Materials and Methods

This systematic review aims to clarify the drugs and vaccines that were reported to be associated with the incidence of CTCL, associated clinical characteristics, and course of disease. It follows the 2020 guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [ 15 ].

2.1. Search Strategy

A systematic search was conducted using MeSH terms/keywords CTCL and, drug-induced or drug-associated or vaccine-associated or vaccine induced through PubMed/Medline, Scopus, Web of Science, and Embase until May 10, 2024 (Supporting file 1 ).

2.2. Eligibility Criteria and Study Selection

This systematic review includes cross-sectional studies, case series, and case reports that reported new-onset CTCL disease after using a specific drug or vaccine. The were as follows: A diagnosed CTCL after receiving a specific drug or vaccine with enough evidence of that drug or vaccine triggered the CTCL. We excluded reviews, animal studies, cases that had a history of CTCLs or Hodgkin lymphoma, post-transplant cases, and cases with a phenotypic shift of disease. We excluded the cases of atopic dermatitis that showed CTCLs after using Dupilumab because of the possibility of CTCL misdiagnosed as atopic dermatitis and previous reviews in the literature on Dupilumab association with CTCLs [ 16 ] ( Figure 1 ).

2.3. Data Extraction

Initially, eight reviewers independently screened the articles and excluded unrelated ones. In the case of disagreement, the corresponding author made the final decision. Data extracted included study characteristics, patient age, sex, underlying disease, main drug, onset of CTCL after using the drug, clinical features, histopathology, CTCL subtype, stage, management, and outcomes.

2.4. Naranjo Score

For the possibility of adverse drug reaction (ADR), we used the Naranjo score [ 17 ]. It consists of 10 questions scoring from −1 to +2, and the sum of scores (range −4 to +13) interprets as follows: Definite ≥ 9, Probable if 5–8, Possible if 1–4, and Doubtful ≤ 0.

3. Results

Out of initially 14,031 screened articles, 60 were included, comprising 59 case reports and one case series, involving 71 patients. The median (IQR) age of participants was 56 (42–69) and the mean (±SE) was 53.5 ± 2.05 years, ranging 13–84 years. Among the participants 52.1% were male. Medications were categorized into four groups: conventional treatments, biologics, small molecules, and vaccine-induced CTCLs.

3.1. Conventional Treatments

Twenty-eight articles and 30 patients were included [ 18 – 45 ]. The median (IQR) age of the patients was 56 (41–65), the mean (±SE) was 53.6 ± 2.9 years (ranging 13–84 years), and the median (IQR) time from drug intake to disease onset was 24 (7.25–108), with a mean (±SE) of 79.8 ± 26.9 months (approximately 5.8 years). The most common type of CTCL was PC-ALCL, with 11 cases (36.6%). Other frequently reported types included MF ( n  = 5, 16.6%), LyP ( n  = 5, 16.6%), SPTCL ( n  = 3, 10%), SS ( n  = 2, 6.6%), and primary cutaneous NK/T-cell lymphoma ( n  = 2, 6.6%). The most frequently reported medications were fingolimod ( n  = 8, 26.6%), methotrexate (MTX) ( n  = 7, 23%), and cyclosporine ( n  = 4, 13.3%) ( Table 1 ).

3.2. Biologics

Twenty studies involving 22 individuals were analyzed [ 12 , 46 – 64 ], with a median (IQR) age of 53 (41–69), mean (±SE) of 53.7 ± 3.6 years (21–81 years), median (IQR) onset of 18 (4.5–27.5), and an average (mean ± SE) of 23.6 ± 5.4 months (ranging from 2 weeks to 8 years). The most common type of CTCL identified was MF ( n  = 6, 27.3%), followed by SS ( n  = 4, 18.2%), SPTCL ( n  = 3, 13.6%), LyP ( n  = 2, 9%), and various other forms including nonspecified CTCL, primary aggressive cutaneous epidermotropic CD8 cytotoxic T-cell lymphoma, primary cutaneous gamma/delta T-cell lymphoma, primary cutaneous small/medium CD4+ T-cell lymphoma, cutaneous CD56+ T-cell lymphoma, cutaneous anaplastic large cell lymphoma, and cutaneous CD30+ T-cell lymphoma. Infliximab ( n  = 6, 27.3%), etanercept ( n  = 5, 22.7%), and adalimumab ( n  = 4, 18.2%) were the most commonly reported drugs in this classification ( Table 2 ).

3.3. Vaccine

Eight studies involving 15 patients with new-onset CTCL were reported in this group [ 14 , 65 – 71 ]. The median (IQR) age of the participants was 52 (30–66) and mean (±SE) was 50.6 ± 5.3 years (ranging from 18 to 80 years), and the median (IQR) and average (mean ± SE) onset of CTCL after drug initiation were 10 (4.75–22.5) and 14 ± 3.2 days (ranging from 3 to 42 days), respectively. The most common type of CTCL identified was LyP ( n  = 8, 53.3%), which included five cases of LyP Type A, one case of Type C, and one case of Type D. Other types included primary cutaneous small/medium CD4+ T-cell lymphoma ( n  = 3, 20%), SS ( n  = 1, 6.6%), primary cutaneous gamma/delta T-cell lymphoma ( n  = 1, 6.6%), SPTCL ( n  = 1, 6.6%), and one patient with PC-ALCL. Six out of the 15 patients (40%) presented with clinical manifestations following the first dose of COVID-19 vaccines. Five patients (33%) were referred after the second dose and four (27%) after the third dose. In total, 11 cases (73%) received the Pfizer–BioNTech vaccine, two received Moderna (13%), one received AstraZeneca (7%), and one patient (7%) received a Recombinant Adenovirus Type 26 vector-based vaccine ( Table 3 ).

3.4. Small Molecule

Four studies included a total of four patients [ 72 – 75 ]. The median (IQR) age of the patients was 64 (45–77), with a mean (±SE) of 62 ± 8.4 years (ranging from 42 to 78 years), and the median (IQR) and average (mean ± SE) duration of disease onset were 3 (0.87–6.25) and 3.3 ± 1.4 months (ranging from 2 weeks to 7 months), respectively. Two cases (50%) were reported with LyP (one of which was Type A), while the remaining cases were diagnosed with primary cutaneous small/medium CD4+ T-cell lymphoma (25%) and SS (25%) ( Table 4 ).

3.5. Drugs and CTCL

The most commonly associated drugs were TNFα-blockers (15/71, 21.1%), which included etanercept ( n  = 5), infliximab ( n  = 6), and adalimumab ( n  = 4). Other frequently reported medications included fingolimod (8/71, 11.2%), MTX (7/71, 9.8%), and cyclosporine (4/71, 5.6%). There were two cases each of melphalan, peginterferon, and pembrolizumab. Additionally, estradiol patch, goserelin, carbamazepine, phenytoin, glatiramer, hydrochlorothiazide, atenolol, benralizumab, nivolumab, tocilizumab, rituximab, and secukinumab were also reported. LyP was the most frequently reported type of CTCL (17/71, 23.9%) following drug initiation, comprising seven cases of LyP Type A, one case of Type B, one case of Type C, and two cases of Type D. The second and third most common types were PC-ALCL (13/71, 18.3%) and MF (11/71, 15.5%). Other common types of CTCLs included SS (8/71, 11.2%), SPTCL (8/71, 11.2%), and primary cutaneous small/medium CD4+ T-cell lymphoma (5/71, 7%).

The most common underlying conditions was rheumatoid arthritis, which accounted for 15 cases (21%) and was primarily treated with MTX in seven patients, biologics in six, and small molecules in two. Multiple sclerosis was present in 10 cases (14%), with fingolimod being the predominant treatment in eight cases. Melanoma was reported in six cases (8.5%), with five patients receiving biologic treatments, while psoriasis was noted in five cases (7%), with four individuals treated with biologics. Additionally, 14 cases (20%) did not have a documented past medical history, and nearly all of these patients, except for one, received COVID-19 vaccinations. Further details on the common underlying conditions associated with the most reported cases of CTCL are provided in Table 5 .

Naranjo ADR Probability Scale is a questionnaire comprising 10 questions that assess the causality of adverse reactions associated with the drugs used in studies. Each question is assigned a score, and the total score is calculated by summing all individual scores. The final score indicates the probability of the adverse reaction as follows: ADR < 2 Doubtful, ADR 2–4 Possible, ADR 5–8 Probable, and ADR ≥ 9 Definite. The median (IQR) and average (mean ± SE) ADR were 4 (3.5–5) and 4.4 ± 0.2 (with a range of 2–8), respectively. Among the assessed drugs and vaccines, Pfizer–BioNTech COVID-19 vaccine exhibited the highest probability of adverse reactions, with a score of (mean ± SE) 5.5 ± 0.5, while adalimumab had the lowest probability, with a score of 3.5 ± 0.2 ( Table 6 ).

Regarding the outcomes, a total of 20 patients (28%) experienced disease regression after discontinuing the drug, with a median (IQR) of 6 (1.75–13.25) and mean ± SE of 8.6 ± 2.5 weeks (ranging from 4 days to 32 weeks; six cases were not reported). However, this was associated with recurrence in one case. For 14 patients (20%), chemotherapy and/or radiotherapy were initiated. Most cases diagnosed with LyP did not receive any treatment (7/17, 41%). However, three cases (17%) were treated with MTX, and three other cases (17%) received phototherapy (two of which were treated with steroids and MTX). Additionally, two cases (12%) received topical steroids and intravenous steroids combined with antihistamines, while two cases (12%) did not have any treatments mentioned. Similarly, the majority of patients with ALCL also did not receive treatment after diagnosis (7/13, 54%), although three cases (23%) received radiotherapy. Other patients were treated with excision, topical steroids, and MTX. In the MF group, five individuals (45%) were treated with phototherapy, primarily in combination with other medications. Two cases (18%) received topical corticosteroids, and one case (9%) was treated with both radiotherapy and surgery. Two patients (18%) did not receive any treatment, and one case (9%) was not reported to have received any medications.

Six patients passed away after being diagnosed with CTCL. The cases owing to various causes of death varied and included lymphoma recurrence (in the cases of PC-ALCL and MTX-associated primary cutaneous extranodal NK/T-cell lymphoma), lung lymphoma (in the case of primary cutaneous CD8+ aggressive epidermotropic cytotoxic T-cell lymphoma), cardiovascular emboli (in case of SS), upper intestinal hemorrhage (in the cases of ALCL), and progression of underlying disease (in the case of primary cutaneous small/medium CD4+ T-cell lymphoma).

4. Discussion

This systematic review provided a comprehensive overview of the reported drug—and vaccine-associated CTCL cases. The findings contributed to the understanding of potential triggers and the clinical course of this rare but challenging disease. A total of 71 patients were reviewed with a mean age of 53.5 years, and a male predominance aligned with the known epidemiology of CTCL [ 2 ]. The most common associated drug was TNF inhibitors, accounting for 21% of cases, followed by fingolimod (11.2%). The most frequently reported subtype of CTCL was LyP, followed by PC-ALCL, MF, and SS. The shortest and longest time interval between receiving drugs/vaccine and onset of the CTCL was observed in the vaccines (mean 14 days) and conventional treatments (mean 79.8 months), respectively.

Several theories have been proposed to explain the pathophysiology of drug- and vaccine-associated CTCL. Immune system dysregulation is among the most widely recognized. The dysregulation of the JAK-STAT pathway, a key signaling mechanism in CTCL progression, is particularly evident in patients treated with immunomodulatory agents or biologics [ 6 , 7 , 9 ]. Findings from previous studies on PD-1 inhibitors, monoclonal antibodies, and other immune checkpoint modulators further support the idea that immune dysregulation underlies many cases [ 12 , 13 ]. Notably, vaccine-associated cases predominantly occurred after COVID-19 vaccination, suggesting that immunological activation and/or T-cell–mediated responses to vaccine components, including mRNA might precipitate CTCL [ 14 , 66 , 70 ]. This highlights the complex interaction between therapeutic or prophylactic interventions and immune pathways in triggering CTCL.

Among the patients with underlying diseases, autoimmune conditions accounted for nearly two-thirds (37 out of 56 patients) when the vaccine group was excluded. Notably, all patients with autoimmune diseases had been treated with immunosuppressive or immunomodulatory drugs, except for one MS patient who developed CTCL after receiving a COVID-19 vaccine. This finding further supports the hypothesis that drug-associated CTCL may result from the immunosuppressive effects of treatment, the underlying immune dysfunction, or a combination of both factors.

Among the subtypes of drug-associated CTCLs, LyP was the most frequently reported in this review. Unlike its secondary rank in primary CTCL studies, this discrepancy may reflect the distinct immunological and molecular mechanisms underlying LyP [ 4 , 76 ]. As CD30-positive lymphoproliferative disorder, LyP involves complex immunological and molecular mechanisms [ 77 ]. CD30, a tumor necrosis factor (TNF) receptor superfamily member, is expressed on activated T-cells [ 78 ]. When stimulated by antigens or mitogens, CD30 activation prevents apoptosis and promotes cell survival and proliferation, contributing to the pathogenesis of LyP [ 70 , 79 ]. This review identified the COVID-19 vaccine as the most frequently associated trigger for LyP, followed by fingolimod, TNF inhibitors, and JAK inhibitors. These agents likely induce LyP through antigenic stimulation, consistent with its pathophysiology. Understanding this potential relationship emphasizes the need for clinical attention in patients developing LyP-like lesions after immunomodulatory therapies or vaccinations.

The second most commonly reported type of drug-associated CTCL in this review was PC-ALCL, another member of the CD30-positive lymphoproliferative disorder family [ 4 , 78 ]. PC-ALCL and LyP represent the opposite ends of a spectrum of disorders with distinct prognoses, clinical presentations, and clinical courses in their classic forms. However, diagnosing borderline lesions can be challenging due to their overlapping histopathologic and immunophenotypic features [ 80 , 81 ]. As previously discussed regarding the triggering role of medications in LyP pathophysiology, the shared mechanisms of PC-ALCL and LyP may clarify the frequent observation of PC-ALCL among drug-related CTCL subtypes. Notably, drugs such as cyclosporine, fingolimod, and methotrexate were commonly implicated, highlighting the need for vigilance in patients receiving these medications.

MF, the most common subtype of CTCL overall, ranked third in drug-associated cases [ 4 , 76 ]. The pathogenesis and progression of MF are driven by a complex network of malignant mechanisms, primarily influenced by disrupted signaling pathways such as TCR/PLCγ1–NFAT, TNFR–NF-κB, and JAK–STAT [ 9 , 82 ]. TNF inhibitors and fingolimod were the most frequently implicated agents. Similarly, in a previous study, the most commonly reported CTCL subtypes associated with biological agents were reported to be MF and SS [ 83 ]. The use of biologic agents such as TNF inhibitors and the resulting development of CTCL have been controversial in recent years [ 84 , 85 ]. In our study of 71 drug-associated CTCL patients, 15 were treated with TNF-α inhibitors for a range of underlying autoimmune conditions, including psoriasis, psoriatic arthritis, Crohn's disease, ankylosing spondylitis, rheumatoid arthritis, and ulcerative colitis. Biologic therapy may be able to reveal an underlying lymphoma, particularly when taking into account the goal of biologic medicines to suppress both innate and acquired immunity, which regulate the growth of cancerous lymphocytes [ 83 ]. It remains unclear whether the development of CTCL is primarily driven by the immunosuppressive effects of the biologics or by the underlying immune disease itself; therefore, further studies are necessary to clarify this uncertainty.

Interestingly, fingolimod was the second most commonly associated drug in this study. Fingolimod, a sphingosine-1-phosphate receptor (S1PR) modulator, sequesters lymphocytes in lymph nodes and reduces their circulation in the bloodstream [ 86 ]. This altered trafficking can disrupt immune surveillance, potentially allowing malignant T-cell clones to evade detection and proliferate [ 87 ]. Furthermore, fingolimod's effects on S1PR signaling may facilitate the proliferation of malignant T-cell populations, which may also affect T-cell apoptosis and survival pathways [ 38 ]. These processes highlight a potential connection between the pathophysiology of CTCL and fingolimod.

The literature extensively documented the association between cyclosporine use and skin malignancies, particularly in organ transplant recipients [ 88 , 89 ]. Additionally, cyclosporine achieved a high Naranjo ADR score due to the strong temporal relationship, exclusion of alternative causes, dose-dependent effects, and pre-existing evidence linking it to malignancies [ 17 ]. However, only four cases of cyclosporine-associated CTCLs were identified in this study. This lower-than-expected frequency could be explained by several factors. First, transplant patients were excluded from this review, which removed a substantial subset of individuals at heightened risk of cyclosporine-induced malignancies. Second, the most frequently reported malignancy associated with cyclosporine use is nonmelanoma skin cancer, whereas recent studies have not demonstrated an increased risk of lymphoma, including CTCL, among cyclosporine recipients. Furthermore, evidence from larger series indicates that cyclosporine treatment does not elevate lymphoma risk compared to azathioprine therapy [ 90 ]. These differences likely contribute to the lower observed incidence of CTCL in nontransplant patients.

The relatively rapid regression observed in six patients treated with fingolimod suggests that stopping fingolimod might play a significant role in disease resolution, highlighting its potential causal association with CTCL. In contrast, the data on MTX revealed considerable variability in regression intervals, potentially influenced by the CTCL subtype and other individual factors. Notably, disease regression following biologics tended to take longer, likely reflecting the prolonged immunomodulatory effects of these agents.

Persistent disease following drug cessation was noted in three patients who had received phenytoin, cyclosporine, and pembrolizumab. These patients developed SS, LyP Type A, and CD56+ T-cell lymphoma. The persistence of disease in these cases appears to be more closely related to the inherent nature of the CTCL subtype and the underlying disease rather than the specific drug.

The findings from the Naranjo ADR Probability Scale assessment offer important information about the possibility of drug-induced adverse events among the evaluated therapies [ 17 ]. An average score of 4.4 ± 0.2 in this study indicates that the assessed drugs fall into the “possible” category of causality. The Pfizer–BioNTech COVID-19 vaccine demonstrated the highest ADR probability (5.5 ± 0.5), which may be attributed to the unique immunologic mechanisms underlying vaccine-induced responses [ 70 ]. In contrast, adalimumab scored the lowest (3.5 ± 0.2), reflecting a relatively lower likelihood of adverse reactions. As a TNF inhibitor, its immunomodulatory effects are well-documented, and its safety profile has been extensively characterized, likely contributing to its lower ADR probability [ 85 ].

The clinical course and outcomes of drug- and vaccine-associated CTCL were notably heterogeneous across the reviewed studies. Approximately one-third of patients experienced significant improvement and disease regression upon cessation of the suspected triggering agent. This finding highlights the potential reversibility of CTCL in some cases when the associated drug or vaccine is withdrawn. However, one instance of recurrence after drug cessation suggests that the etiology of CTCL is likely multifactorial and influenced by underlying host factors [ 35 ]. Considering the final outcomes, the majority of patients in this review experienced disease regression, in line with findings reported in primary CTCL studies [ 1 , 76 , 91 ]. However, a subset showed poorer outcomes, with three patients (4.2%) experiencing recurrences and two (2.8%) exhibiting partial or no response to treatment. Among the six patients who passed away, three died from causes unrelated to CTCL, including metastatic melanoma, cardiovascular emboli, and upper gastrointestinal bleeding. However, the remaining three patients died due to lymphoma, highlighting the disease's severity, particularly in cases involving aggressive subtypes or advanced stages.

Among the patients with LyP, all achieved disease regression except for two cases of type D LyP, who experienced relapses. This finding aligns with primary CTCL studies demonstrating an almost 100% disease-specific survival rate over 10 years for LyP, reflecting its generally favorable prognosis [ 91 ]. Notably, Type D LyP, characterized by the pagetoid infiltration of small-to-medium atypical CD8+ epidermal cells, exhibits clinical behavior similar to other LyP subtypes [ 79 ]. Regarding PC-ALCL, all patients achieved disease regression without recurrence, except for one who passed away due to the extracutaneous involvement of the disease. This observation is consistent with the favorable prognosis reported in primary CTCL studies, where systemic dissemination occurs in approximately 10% of cases [ 91 , 92 ]. Similarly, the outcomes for MF were comparable with those observed in primary cases [ 2 , 76 ]. All patients experienced disease regression following drug discontinuation and received standard treatments according to their disease stage.

Treatment for drug- and vaccine-associated CTCL mirrored primary CTCL approaches [ 91 , 93 ]. Chemoradiotherapy showed mixed outcomes [ 22 , 47 , 73 ], while phototherapy proved effective, particularly in the early stages [ 67 , 69 ]. Other treatments, including corticosteroids, methotrexate, and surgical excision, had variable success, with some cases showing spontaneous regression [ 76 , 79 , 91 , 93 ].

This review was limited by its reliance on case reports and small case series, which may introduce reporting bias and lack generalizability. However, ADRs usually are reported and even reviewing the reported cases is important to a better understanding of drug-induced CTCL. Additionally, one significant limitation of our study is the reliance on temporal associations to diagnose drug-induced CTCLs. While the onset of CTCL following drug initiation provides valuable diagnostic clues, it does not establish definitive causality, particularly in cases where the disease persists or recurs after drug cessation. This highlights the need for more robust diagnostic criteria and longitudinal studies to better understand the complex relationship between drug exposure and CTCL development. Future research should focus on larger cohort studies, mechanistic investigations, and prospective surveillance to elucidate the underlying immunological and molecular pathways. Genetic predisposition and environmental factors should also be considered potential contributors to susceptibility.

In conclusion, this systematic review emphasizes the importance of recognizing CTCL as a possible, although rare,

adverse effect of certain drugs and vaccines. Our review highlights the importance of taking a history of vaccinations, especially COVID-19 vaccines, and immunosuppressive drugs such as fingolimod, TNF-a inhibitors, and MTX in patients with a recent onset CTCL as a possible cause of disease.