Alternaria toxins in tomato products: from analytical advances to risk assessment and regulatory needs Authors: Stefano Sdogati, Ivan Pecorelli, Serenella Orsini, Tommaso Pacini, Rita Bibi, Angela Caporali, Emanuela Verdini, Slim Smaoui, Sofia Agriopoulou, Teresa D'Amore Published: 2026 ABSTRACT Alternaria toxins (ATs) are mycotoxins of increasing relevance in food safety risk analysis, as they are widely detected in plant foods. Tomato and tomato-based products represent major contributors to ATs dietary exposure due to their high consumption and susceptibility to Alternaria spp. contamination. Matrix properties, combined with industrial processing, can further promote toxin accumulation. Despite the growing scientific interest in these fungal metabolites, an integrated assessment on tomato products is lacking. In this context, this review systematically examines twenty liquid chromatography–mass spectrometry studies published between 2011 and 2025, covering a range of tomato-based matrices. Data extraction focused on analytical performance, sample preparation, validation parameters, matrix effects and occurrence levels. Across processed matrices, tenuazonic acid (TeA) is the most prevalent toxin, with substantial enrichment in concentrated products, whereas alternariol (AOH) and alternariol monomethyl ether (AME) frequently co-occur and, in several instances, exceed the indicative levels currently set in the European Union (EU). Occurrence data were integrated with international and national consumption datasets to support risk assessment. The analysis indicates that exposures to TeA and tentoxin (TEN) generally remain below the Thresholds of Toxicological Concern (TTCs), whereas estimated intakes of AOH and AME frequently exceed the TTC, particularly among sensitive subgroups. Finally, this review examines the global regulatory framework, indicating that, apart from the EU recommendation, no formal maximum levels for ATs currently exist worldwide.This integrated assessment spotlights data gaps and the need for harmonized monitoring and refined toxicological evaluation. Tomato products are proposed as a model matrix for refining future risk assessment and management strategies for ATs. Introduction Tomato (Solanum lycopersicum) is the most produced vegetable worldwide, with a total production of 192.32 million tons. China leads the ranking with 68 million tons followed by India (20.6 million tons), Turkey (13 million tons) and the United States of America (10.2 million tons). In the European Union (EU), Italy takes the lead with 6.1 million tons, followed by Spain (3.6 million tons) and Portugal (1.5 million tons) (Food and Agriculture Organization, 2024). The fragile epidermis of tomatoes makes them susceptible to biotic and abiotic stressors, with Alternaria infection being a major contributor to fruit decay and crop losses. Alternaria is one of the most widespread saprophyte fungi genera, able to grow in disparate types of substrates ranging from water to foods and textiles. Alternaria taxonomy was first described in 1817 by Nees and, then, by Kreissl in 1912 despite remaining confusing to this day as almost 60 species have been described based on morphological and host differences (Bautista-Banos, 2014). Optimal humidity and temperature growth for Alternaria is >80% and around 20 °C, respectively. These conditions are often met outdoors in Mediterranean and tropical climates while, in indoor ambient, Alternaria spp. proliferate in poorly ventilated and humid buildings (Woudenberg et al., 2015). Among the most common species, Alternaria alternata is known for causing leaf spots and similar diseases in hundreds of host plants. Early blight (EB) and brown spots (BS) are usually visible symptoms of infested tomatoes caused by both A. alternata and A. arborescens species (Schmey et al., 2024). A. alternata is also responsible for severe asthma in sensitized patients; it is estimated that a concentration of 100 spores/m3 is enough to elicit allergic response leading, in the worst cases, to emergency admissions due to breathing difficulties and lung failure (Abel-Fernández et al., 2023). Other relevant Alternaria spp. in plant pathology are A. longipes responsible for leaf blight in carrots (Daucus carota L.), oil palm (Elaeis gunieensis), sunflower (Helianthus annuus) and most importantly tobacco (Nicotiana spp.) (Feng et al., 2021; LaMondia, 2001). A. arborescens causes stem canker disease in tomato (visible as dark brown cankers on the stems leaves necrosis) and severe early defoliation in Celtis julianae trees (Liao et al., 2023). Alternaria burnsii is reported to infest cumin species (Cuminum cyminum), pumpkin seeds (Cucurbita maxima), wheat (Triticum aestivum) and beans causing in worst case scenarios crop losses up to 70% (Feng et al., 2020); Alternaria gossypina targets Minneola tangelo fruits, apples (Malus domestica) and cotton plants (Gossypium spp.). As Alternaria spp. grow, they produce a plethora of secondary metabolites, named Alternaria toxins (ATs). ATs behave as potent phytotoxins, contributing to plant disease, spread and rotting, and as mycotoxins thus affecting human and animal health. Currently around 70 ATs are structurally characterized; these compounds can be divided in chemically related families: dibenzo-α-pyrones (DAPs): alternariol, (AOH), alternariol monomethyl ether (AME), altenuene (ALT) and altenuisol (ATS); tetramic acid derivatives (TADs): tenuazonic acid (TeA) and isotenuazonic acid (iso-TeA); perylen-quinones (PQs): altertoxin I, II, III, IV and V (ATX-I, ATX-II, ATX-III, ATX-IV and ATX-V), stemphyltoxin III (STTX-III) and alterperylenol (ALP); A. alternata f. sp. lycopersici toxins (AAL-toxins): AAL-TA (mixture of TA1 and TA2 isomers), AAL-TB, AAL-TC, AAL-TD and AAL-TE; miscellaneous structures: altenusin (ALS), altersetin (AST), tentoxin (TEN). Fig. 1 presents the chemical structures of the main ATs.Fig. 1Chemical structures of the main Alternaria toxins (ATs), including dibenzo-α-pyrones (AOH, AME, ALT, ATS), tetramic acid derivatives (TeA, iso-TeA), perylene-quinones (ATX-I, ATX-II, ATX-III, ATX-IV, ATX-V, STTX-III, ALP), A. alternata f. sp. lycopersici host-specific toxins (AAL-TA, -TB, -TC, -TD, -TE), and selected miscellaneous structures such as tentoxin (TEN). Chemical structures of the main Alternaria toxins (ATs), including dibenzo-α-pyrones (AOH, AME, ALT, ATS), tetramic acid derivatives (TeA, iso-TeA), perylene-quinones (ATX-I, ATX-II, ATX-III, ATX-IV, ATX-V, STTX-III, ALP), A. alternata f. sp. lycopersici host-specific toxins (AAL-TA, -TB, -TC, -TD, -TE), and selected miscellaneous structures such as tentoxin (TEN). DAP is a relatively common scaffold of many natural compounds. As an example, highly hydroxylated DAPs are considered polyphenols, a class of chemicals with known antioxidant and chemoprotective properties. Biosynthetic pathways of DAPs ATs are probably based on polyketide synthase (PKS) enzymes; the proposed model biosynthesis starts with acetyl-CoA consisting in six consecutive condensation reactions, in each of which a moiety of activated malonate is included in the structure, the final reaction consists in an aromatization/lactonization step that may spontaneously occur leading to AOH or AME formation (Saha et al., 2012). ALT and ATS formation may be subjected to the same synthetic route. Log P ranges from 1.03 (ALT) to 2.55 (AME) with average pKa around 7.5 given the phenolic-like structure of these substances. TADs like TeA and iso-TeA are small polar molecules (log P = 0.57) with pKa around 3.5 expected to be a product of polyketide synthase and non-ribosomal peptide synthetase (PKS–NRPS) hybrid enzyme. Recently, Yun and colleagues demonstrated that TeA biosynthesis starts from isoleucine and acetoacetyl-CoA identifying a TAS1 as the gene responsible for its production in M. oryzae (Yun et al., 2015). PQ synthetic pathway is reported to be like dihydroxynaphthalene (DHN) melanin pathway, which is responsible for the production of pigmented compounds, often observed in fungi infestation. The DHN pathway is shared by a great variety of fungi suggesting how ATX and related compounds may be produced by Alternaria. In addition, when PQ synthesis was absent, DAPs production was increased (AOH and AME) whereas, in the absence of the latter, more ATXs were produced suggesting how, probably, the two pathways compete for the same starting materials (Gao et al., 2022). AAL-toxins are part of host-specific toxins (HST) that are selectively toxic to only the host plant contributing to fungi spread and disease onset, which mostly targets tomatoes. Their synthesis is carried out by a PKS gene (ALT1) also responsible for the biosynthesis of other relevant HSTs like ACT-toxin I-II (tangerine), ACR-toxin I, II, III, IV and IV’ (lemon), AM-toxin I, II, III (apple), AS-I toxin (sunflower), AF-toxin I, II, III (strawberry) AK-toxin I, II (Japanese pear) maculosin (knapweed) and destruxin B (brassicaceae). Chemically, AALs are aminopentol esters structurally similar to fumonisins; double hydroxylation at C4 and C5 yields AAL TA1 and TA2 while only a single -OH moiety is present in AAL TB1 and TB2. Acetylation of the free amino group leads to AAL TC1, TC2, TD1 and TD2 (Meena and Samal, 2019; Wang et al., 2022). Among miscellaneous structures, the most relevant compound is tentoxin (TEN), a cyclic peptide containing glycine, L-leucine, N-methyl-L-alanine, and N-methyl-L-dehydrophenylalanine. In 2016, two genes responsible for TEN production were discovered from A. alternata cultures: TES, a non-ribosomal peptide synthetase and TES 1 a cytochrome P450 gene (Li et al., 2016). Understanding AT production, possible countermeasures for Alternaria infestation and occurrence of ATs is pivotal for food scientists and food control entities in order to fill the remaining knowledge gap regarding these mycotoxins. Looking at this context, in this systematic review we compile and critically evaluate occurrence data for ATs in one of the most widely consumed and Alternaria-susceptible food groups, tomato products. To ensure analytical robustness, only studies employing liquid chromatography–tandem mass spectrometry (LC-MS/MS) were considered for the occurrence and risk-assessment evaluation, resulting in a strict PRISMA-based selection of twenty research studies published between 2011 and 2025. In addition to summarizing contamination patterns across diverse tomato matrices, the review provides a focused appraisal of the analytical methodologies applied, together with an in-depth dietary risk assessment integrating occurrence levels with population tomato-consumption data. Regulatory framework and toxicity Regulatory framework In 2011, the European Food Safety Authority (EFSA) gathered all available literature on toxicity and occurrence of ATs in food and feed in order to give a scientific opinion on the risks for animal and public health posed by ATs. A total of 11730 analytical results were considered, including AOH (n = 2291), AME (n = 2215), ALT (n = 1747), ATX-I (n = 1279), TeA (n = 1947), TEN (n = 1388) and sum of AAL-toxins (n = 863). EFSA concluded that exposure to AOH and AME may exceed the Threshold of Toxicological Concern (TTC), whereas TTC values were not exceeded for TEN and TeA. Five years later, EFSA published an updated dietary exposure assessment based on 24,461 analytical results from 7,916 food samples. Other than oilseeds, which were the most contaminated matrix in terms of average levels and incidence, ATs were determined in relatively high mean concentrations in tomato and tomato-based products (TeA: 351.2 μg/kg). The most exposed population segments were infants between 0.007 and 1.490 μg/kg body weight per day (bw/day) and toddlers between 0.100 and 1.614 μg/kg bw/day, respectively, applying lower-bound (LB) and upper-bound (UB) approach. For other age classes and vegetarians, the main exposure source was the consumption of tomato and tomato-related products. Both EFSA evaluations highlighted the need for more sensitive analytical methods and additional occurrence data in relevant food commodities, as current evidence remains insufficient to adequately characterize exposure to ATs (EFSA Panel on Contaminants in the Food Chain (CONTAM), 2011 ; Arcella et al., 2016). As a consequence, in 2022, the European Commission issued a recommendation establishing indicative levels for TeA, AOH and AME in foodstuffs. Food categories frequently contaminated by ATs, including spices, tomato products, oilseeds, nuts, and infant foods, were identified based on previous EFSA evaluations. A summary of these levels is provided in Table 1. Indicative levels are defined as “a threshold above which investigations should be performed, to understand the factors leading to ATs presence and/or on the effect of food processing” (Arcella et al., 2016; European Commission, 2022).Table 1Indicative levels (μg/kg) for Alternaria toxins in foodstuff (Recommendation (EU) 2022/553).Table 1FoodAlternariol (AOH)Alternariol monomethyl ether (AME)Tenuazonic Acid (TeA)Processed tomato products105500Paprika powder--10000Sesame seeds3030100Sunflower seeds30301000Sunflower oil1010100Tree nuts--100Dried figs--1000Cereal based foods for infants and young children22500 Table 1Indicative levels (μg/kg) for Alternaria toxins in foodstuff (Recommendation (EU) 2022/553).Table 1FoodAlternariol (AOH)Alternariol monomethyl ether (AME)Tenuazonic Acid (TeA)Processed tomato products105500Paprika powder--10000Sesame seeds3030100Sunflower seeds30301000Sunflower oil1010100Tree nuts--100Dried figs--1000Cereal based foods for infants and young children22500 Indicative levels (μg/kg) for Alternaria toxins in foodstuff (Recommendation (EU) 2022/553). In order to provide a comprehensive overview of AT current regulatory status, we also reviewed regulatory documents and food safety standards from other regions and major trading partners — including Australia, Brazil, Canada, China, India, Indonesia, the Republic of Korea, Saudi Arabia, South Africa, Taiwan and the United States (Table 2). No national or international maximum, provisional, or indicative levels for ATs have been established outside Europe. Only the German region of Bavaria has adopted a precautionary limit of 500 μg/kg for tenuazonic acid (TeA) in sorghum- or millet-based infant foods. In some jurisdictions, ATs are not yet regulated but are recognized as emerging contaminants and included in national or regional monitoring programs (e.g. Recommendation (EU) 2022/553, Health Canada surveys, USDA/ARS research). However, as shown in Table 2 in most countries only research studies were carried out, underlining the need for official documents. In regard of risk analysis, ATs belong to the group of food contaminants for which no Codex or national maximum levels have been established worldwide. In line with Codex guideline CXG 92-2019, risk assessment and risk management are currently based on the application of the TTC approach (Codex Alimentarius Commission, 2019).Table 2Global regulatory framework for Alternaria toxins.Table 2Country/RegionRegulatedMonitoring/SurveillanceMaximum Levels/Indicative ValuesOfficial Recommendations/NotesKey References/Sources InspectedEuropeNo (EU-level); only monitoring (Recommendation (EU) 2022/553)Yes – EU Recommendation 2022/553 (AOH, AME, TeA)No official ML, Indicative level for alternariol, alternariol monomethyl ether and tenuazonic acid in certain foodsCommission Recommendation (EU) 2022/553; EFSA CONTAM opinionsEFSA Panel on Contaminants in the Food Chain (CONTAM), 2011European Commission (2022)Germany (Bavaria)Yes (regional limits)Yes – regional and national within EUTeA 500 μg/kg in sorghum/millet-based infant foodsRegional precautionary limit for infant foodsRychlik et al. (2016)United StatesNo national limitsUSDA/ARS research (Discovery of a new biofungicide for Alternaria management), occasional monitoringNoneNoAgricultural Research Service, 2023United States Department of Agriculture, 2020CanadaNo national limitsHealth Canada targeted surveys (beer, juices, fresh bell/hot peppers, grain-based foods, infant foods, nut/seed products, pomegranate products, processed fruits and vegetables, sunflower oil, and grape-based wine)NoneSurveillance by Health Canada; data collection onlyCanadian Food Inspection Agency (2020)Canadian Food Safety Agency (2023)Australia/New ZealandNo national limitsGeneral mycotoxin surveillance, limited data on AlternariaNoneNoFSANZ Food Standards Code (no regulation)ChinaNo national limitsResearch-based monitoring onlyNoneNoChinese food safety standards (GB 2761) (no regulation);Ji et al. (2024)IndiaNo national limitsResearch-based monitoring onlyNoneNoFSSAI regulations - Contaminants (no regulation)IndonesiaNo national limitsResearch-based monitoring onlyNoneNono national MLSouth KoreaNo national limitsMFDS refers to EU guidanceNoneMFDS recommends data collectionhttps://mfds.go.kr/eng/brd/m_60Saudi ArabiaNo national limitsResearch-based monitoring onlyNoneNoSaudi Food & Drug Authority (SFDA) general contaminant standards, (no regulation);Yassin et al. (2010)South AfricaNo national limitsResearch-based monitoring onlyNoneNoHickert et al. (2017)BrazilNo national limitsResearch-based monitoring onlyNoneData collection and risk assessment ongoingANVISA contaminant rules (RDC 732/2022) no regulation. da Motta and Valente Soares (2001)TaiwanNo national limitsGeneral mycotoxin surveillance, limited data on AlternariaNoneNoTFDA Sanitation Standard for Contaminants and Toxins in Food Table 2Global regulatory framework for Alternaria toxins.Table 2Country/RegionRegulatedMonitoring/SurveillanceMaximum Levels/Indicative ValuesOfficial Recommendations/NotesKey References/Sources InspectedEuropeNo (EU-level); only monitoring (Recommendation (EU) 2022/553)Yes – EU Recommendation 2022/553 (AOH, AME, TeA)No official ML, Indicative level for alternariol, alternariol monomethyl ether and tenuazonic acid in certain foodsCommission Recommendation (EU) 2022/553; EFSA CONTAM opinionsEFSA Panel on Contaminants in the Food Chain (CONTAM), 2011European Commission (2022)Germany (Bavaria)Yes (regional limits)Yes – regional and national within EUTeA 500 μg/kg in sorghum/millet-based infant foodsRegional precautionary limit for infant foodsRychlik et al. (2016)United StatesNo national limitsUSDA/ARS research (Discovery of a new biofungicide for Alternaria management), occasional monitoringNoneNoAgricultural Research Service, 2023United States Department of Agriculture, 2020CanadaNo national limitsHealth Canada targeted surveys (beer, juices, fresh bell/hot peppers, grain-based foods, infant foods, nut/seed products, pomegranate products, processed fruits and vegetables, sunflower oil, and grape-based wine)NoneSurveillance by Health Canada; data collection onlyCanadian Food Inspection Agency (2020)Canadian Food Safety Agency (2023)Australia/New ZealandNo national limitsGeneral mycotoxin surveillance, limited data on AlternariaNoneNoFSANZ Food Standards Code (no regulation)ChinaNo national limitsResearch-based monitoring onlyNoneNoChinese food safety standards (GB 2761) (no regulation);Ji et al. (2024)IndiaNo national limitsResearch-based monitoring onlyNoneNoFSSAI regulations - Contaminants (no regulation)IndonesiaNo national limitsResearch-based monitoring onlyNoneNono national MLSouth KoreaNo national limitsMFDS refers to EU guidanceNoneMFDS recommends data collectionhttps://mfds.go.kr/eng/brd/m_60Saudi ArabiaNo national limitsResearch-based monitoring onlyNoneNoSaudi Food & Drug Authority (SFDA) general contaminant standards, (no regulation);Yassin et al. (2010)South AfricaNo national limitsResearch-based monitoring onlyNoneNoHickert et al. (2017)BrazilNo national limitsResearch-based monitoring onlyNoneData collection and risk assessment ongoingANVISA contaminant rules (RDC 732/2022) no regulation. da Motta and Valente Soares (2001)TaiwanNo national limitsGeneral mycotoxin surveillance, limited data on AlternariaNoneNoTFDA Sanitation Standard for Contaminants and Toxins in Food Global regulatory framework for Alternaria toxins. At present, the EU is the only region that has translated scientific concern into a structured monitoring framework, while in most other jurisdictions ATs remain unregulated and are addressed primarily through research initiatives or targeted surveys. This asymmetry reflects both toxicological uncertainty and limited occurrence data, resulting in heterogeneous monitoring practices and analytical performances worldwide. From a risk assessment perspective, this lack of harmonization is particularly relevant for matrices such as tomato products, where industrial processing can markedly concentrate ATs and amplify consumer exposure. Tomato-based foods may represent a strategic matrix for advancing future regulatory frameworks, integrating improved occurrence data, processing-related concentration effects, and novel evidence on co-exposure. Toxicity of Alternaria toxins AOH, AME and TeA are the most studied ATs in terms of toxicity although there is limited number of toxicokinetic studies in vivo: a 1982 research details the fate of AME in adult rats, concluding how this compound is poorly absorbed in the gastrointestinal tract and rapidly metabolized (Pollock et al., 1982). More recently, AOH toxicokinetics was evaluated administrating orally 200 and 1000 mg/kg radiolabeled AOH to mice concluding how the main routes of excretion were feces (90% of the initial dose) and urine (9%) confirming previous study on gastrointestinal absorption. The structural similarity between AOH and AME suggests a similar metabolic fate: hydroxyl moieties can be subjected to phase II metabolization (e.g., glucuronation) thus explaining the low persistence of the original compound into the bloodstream (Schuchardt et al., 2014). AOH and AME exert cytotoxicity based on the generation of reactive oxygen species after alteration of mitochondrial membrane potential. In addition, carcinogenic properties were reported for AOH and AME, including concentration-dependent induction of DNA strand breaks in hamster, human colon and human liver cells. Mutagenicity mechanism is probably based on AOH/AME covalent interaction with DNA topoisomerase II during the DNA replication forks formation and elongation leading to DNA damage (Pfeiffer et al., 2007). Same behaviors were recently discovered for ATX-I and ATX-II, with similar potencies compared to AOH (Soukup et al., 2020). TeA is known to reduce protein synthesis by inhibiting release from ribosomes: in mammalian cells (3T3, CHL, L-O2) a decrease in protein production up to 80% was observed after 72 h at concentrations in the range 200-400 μg/mL. In terms of LD50 TeA has the lowest values in mice 81-225 mg/kg bw with respect to AOH and AME. Interestingly, TeA has recently captured the spotlight as an innovative pest control compound. Herbicide-like properties rely on TeA interruption of electron flow from the primary quinone acceptor to the secondary quinone acceptor at the photosynthesis II acceptor side, resulting in the inhibition of photosynthesis. As a consequence, TeA and related compounds have been also proposed as novel insecticides: a concentration in feed of 250 μg/g is sufficient to inhibit the growth of G. mellonella larvae and causes mortality of Acheta domesticus adults (LT50 7 days) (Salimova et al., 2021; Zhang et al., 2025a, Zhang et al., 2025b). Method of analysis Recommendation (EU) 2022/553 suggests a required limit of quantification for the determination of ATs in food. In detail, LOQ should not be higher than 2 μg/kg for AOH and AME in cereal-based foods for infants and young children and 4 μg/kg in other foods, while for TeA a maximum LOQ of 20 μg/kg in all foods is suggested. Although, many years before EFSA urged the EU member states to supply occurrence data, food scientists developed several analytical methods using different techniques for the determination of ATs in fungi cultures and food. A brief review is presented here. Thin layer chromatography (TLC) TLC was predominantly used in the past as a method to obtain qualitative data on the presence of ATs. TLC-UV coupling was able to detect ATs with LODs in the range 100-700 μg/kg in black rotted tomatoes and 125-500 μg/kg in cultures of Alternaria alternata, respectively (Hasan, 1995). The advent of high-performance thin layer chromatography (HPTLC) improved resolution and first attempts at ATs quantitation. TLC/HPTLC approaches were surpassed in terms of sensitivity and precision by more modern LC methods but remain a viable option for cheap and easy qualitative determination of these substances in various matrices (Fàbrega et al., 2002; Ostry et al., 2012). Gas chromatography (GC) and liquid chromatography (LC) Mycotoxin determination using GC instrumentation is generally challenging, and very few methods are reported in literature, let alone for emerging mycotoxins like ATs. The main issue is that mycotoxins are low-volatile and relatively small molecules that require a mandatory derivatization step for sensitive determination. GC-FID analysis of TeA was performed by Harvan after acetyltrimethylsilane:trimethylsilane:pyridine derivatization (LOD: 100 μg/kg); later Scott et al. firstly performed GC-MS determination of AOH, AME, ALT, ATX-I and TeA after heptafluorobutyrate:trimethylsilyl derivatization (Scott et al., 1997). LC coupled to various types of detectors became prevalent in the early 2000s with ultraviolet (UV), fluorescence (FLD), and diode-array (DAD) detectors. Sensitivities and precision were improved with respect to GC and TLC methodologies bypassing elaborate derivatization steps. Clean-up strategies were incorporated into LC methods in order to obtain cleaner final extracts with LOQs generally <10 μg/kg. Solid Phase Extraction cartridges (SPE) were used by Solfrizzo and coworkers in ATs determination in carrots by an LC-UV methodology (ATX-I, AOH, and AME: 256 nm; TeA: 279 nm) but other clean-ups were tested (e.g. hexane defatting, QuEChERS salts) by other researchers in order to obtain better performances (Solfrizzo et al., 2004). LC-FLD coupling is great in terms of selectivity and sensitivity although this could be a “double-edged sword” as only fluorescent ATs could be determined (like DAPs and PQs) while TeA, lacking the conjugated double bond system, cannot be detected using this method. LODs between 1 and 5 μg/kg were obtained by LC-FLD methods for AOH, ALT and AME in tomato paste and legumes (Králová et al., 2011). Liquid chromatography coupled to mass spectrometry (LC-MS) LC-MS is the “state-of-the-art technology” for mycotoxins determination allowing sensitive and reliable analysis in the most disparate types of matrices. LC-MS/MS using electrospray ionization (ESI) and/or atmospheric pressure chemical ionization (APCI) in both positive and negative mode opened the possibility to determine ATs, and mycotoxins in general, with higher sensitivities and better performances in terms of precision and accuracy. LOD/LOQs below 1 μg/kg were reached by many authors determining ATs in bakery products, vegetables, jams, herbs and flavedo (Magnani et al., 2007; Sulyok et al., 2007). Mass spectrometry detection overcomes LC-FLD limitations in terms of universality maintaining or even improving sensitivity and overall performances. LC-MS/MS and high-resolution mass spectrometry (LC-HRMS) paved the way to the development of multi-mycotoxin methods determining tens of compounds including ATs and other emerging mycotoxins, providing useful information of co-contamination of foods which is a very important topic in risk assessment since these studies often evaluate one contaminant at a time (Monbaliu et al., 2010; Rasmussen et al., 2010; Hickert et al., 2015a, Hickert et al., 2015b). The possibility to use isotope labelled internal standards (ILIS) is another great advantage in the determination of ATs via LC-MS/MS. ILISs both correct the measurement bias coming from poor recovery and matrix effects (signals enhancement/suppression) facilitating method development both in terms of instrumental reliability and sample preparation. Moreover, as ATs contaminate food with variable characteristics in terms of composition, ILIS allows the development of a single sample preparation/clean-up method applicable to all the desired matrices (Asam et al., 2011; Liu and Rychlik, 2013). Immunoassays Enzyme-linked immunosorbent assays (ELISA) were developed in the 70's and are currently a viable option for time-saving and selective determination of mycotoxins in food. Drawbacks, in comparison to LC-MS/MS techniques, may be lower sensitivity and lower accuracy due to cross-reactivity between antibodies and matrix interferences. In the EU, according to legislation (Regulation (EU) 2023/2782), ELISA methods could be applied only for screening purposes. In 1996, Szurdoki and coauthors firstly developed an ELISA determination method for AAL-toxins with sensitivities around ppb/low ppm ranges and a high selectivity towards AAL-toxins (with respect to structure similar to mycotoxins like fumonisins) (Szurdoki et al., 1996). TADs (TeA and iso-TeA) are the most common AT to be determined by ELISA techniques: an indirect competitive enzyme-linked immunosorbent assay (icELISA) for iso-TeA was developed by Xiao et al., in 2018 with good sensitivities (IC50: 7.8 ng/mL), acceptable recoveries (82.3-109.8%) and a positive correlation between data obtained by icELISA and an HPLC-MS/MS reference method (Xiao et al., 2018). Open sandwich enzyme-linked immunosorbent assay (OS-ELISA) showed great potential for TeA determination and a higher sensitivity (LOD: 0.08 ng/mL) with respect to icELISA as reported by Liang and coworkers (Liang et al., 2021). In addition, a fluorescence polarization immunoassay (FPIA) was also developed using fluorescein isothiocyanate isomer I (FITC) and 5-([4,6-dichlorotriazin-2-yl]amino) fluorescein hydrochloride (DTAF) tracers with detection limits between 0.13 and 0.93 μg/mL (Wang et al., 2018). Ackermann et al. first described the production of molecular antibodies (mAB) against AOH in 2011, paving the way to its possible determination in foodstuff through screening ELISA techniques. Later, Addante-Moya developed a reliable and sensitive (LOD: 0.004 ng/mL) competitive ELISA (cELISA) for AOH and a semi-quantitative nanoparticle-based competitive lateral-flow immunochromatography assay (cLFICA) which determines AOH in under 15 min in fruit juice samples (screening target concentration: 4 ng/mL). Both methods were extremely specific with cross-reaction values < 0.3% and false-suspect rate <1% (Ackermann et al., 2011; Addante-Moya et al., 2024). ATs in tomato products: methodologies and software Methodology: systematic review This study covers the period from 2011, corresponding to the publication of the first EFSA scientific opinion on ATs, to December 2025. The work was conducted following the principles of the PROMETHEUS initiative (PROMOTING Methods for Evidence Use in Scientific Assessments) and the PRISMA 2020 guidelines, ensuring a transparent and structured approach to evidence identification, appraisal, and integration. The methodological choices reflect EFSA requirements for Weight-of-Evidence (WoE) evaluations and Expert Judgement (EJ) in regulatory risk assessment. A structured Extensive Literature Search (ELS) was performed using three major databases: Web of Science, Scopus, and EFSA Search. Keywords and controlled vocabulary terms were combined to generate database-specific search strings (reported in Table S1). All retrieved records were imported into EndNote Web for organization and de-duplication, then transferred to Rayyan to support screening. Rayyan embedded machine-learning tools were used to facilitate prioritization, detect remaining duplicates, and assist in identifying potentially relevant studies, without replacing manual expert evaluation (Khabsa et al., 2015). Screening proceeded in two stages: (1) title and abstract screening, applying predefined inclusion and exclusion criteria, (2) full-text assessment of documents meeting initial relevance requirements. Only studies fulfilling all inclusion criteria were retained. Publications meeting any of the exclusion criteria, such as pre-2011 dates, reviews or conference papers, studies limited to other food matrices, method-only studies without application to naturally contaminated samples, and toxicological or morphological investigations unrelated to occurrence data, were removed. The full set of criteria is summarized in Table 3.Table 3ELS inclusion and exclusion criteria.Table 3CriterionDecisionInclusionExclusionDefault keywords exist as a whole or at least in the title, keywords, or abstract of the article×The article is published in a peer-review scientific journal×The article is written in English×Number of samples analyzed >3×Usage of LC-MS as analytical method×Duplicate records×The full text is not available×Articles published before 2011×Reviews, book chapters, conference papers, surveys×Other matrices-only studies (infant milk powder, red date, sunflower oils/seeds, cereals, wheat, barley, edible medicine herbs, apples, fruit puree, beer, yellow peach, herbal tea×Only development, validation or collaborative studies without application on naturally contaminated/commercial samples×Only other toxins×Toxicological studies; Morphological and molecular characterization studies× Table 3ELS inclusion and exclusion criteria.Table 3CriterionDecisionInclusionExclusionDefault keywords exist as a whole or at least in the title, keywords, or abstract of the article×The article is published in a peer-review scientific journal×The article is written in English×Number of samples analyzed >3×Usage of LC-MS as analytical method×Duplicate records×The full text is not available×Articles published before 2011×Reviews, book chapters, conference papers, surveys×Other matrices-only studies (infant milk powder, red date, sunflower oils/seeds, cereals, wheat, barley, edible medicine herbs, apples, fruit puree, beer, yellow peach, herbal tea×Only development, validation or collaborative studies without application on naturally contaminated/commercial samples×Only other toxins×Toxicological studies; Morphological and molecular characterization studies× ELS inclusion and exclusion criteria. The final set of eligible studies was then subjected to detailed data extraction, including information on sample characteristics, analytical methods, and reported concentrations of ATs. These selected studies were also used for risk assessment. Dietary exposure assessment to Alternaria toxins from tomato products Approach Dietary exposure to ATs from tomato products was estimated by integrating occurrence data generated in this review (20 LC-MS-based studies), and food consumption data from the most relevant international dietary surveys. In line with EFSA guidance for contaminants, analytical results below LOD/LOQ were treated using the LB approach, imputing these values as zero for exposure calculations. The proportion of left censored data was recorded for transparency. Because no toxicological reference values (human based guidance values - HBGVs) are established for ATs and only TTC values exist, a deterministic, point estimate exposure approach was applied. Exposure was calculated by multiplying the mean concentration of each ATs in the relevant tomato product by the individual consumption amounts of that product. Hypothesis and exposure duration Consistent with EFSA practice, exposure estimates were generated for consumers only, defined as individuals reporting at least one consumption event of the tomato-based product during the survey period. Exposure was then normalized to body weight (μg/kg bw/day) using individual body weight data when available, or reference values in subpopulation models. A chronic exposure assessment was conducted, reflecting EFSA definition of chronic exposure as a long term, constant or intermittent exposure that may influence health over time. Chronic exposure is the appropriate metric for ATs because their main toxicological endpoints (genotoxicity, cytotoxicity, endocrine related effects) are driven by long term accumulation rather than acute intake. Food consumption data for tomato-based products (g/day) were gathered from three complementary sources.1)FAOSTAT consumption data—providing population level estimates for the global, European, and Italian populations.2)SCAI 2005–2006 (Italy)—providing detailed individual level consumption data for mean, 95th percentile (P95), and 99th percentile (P99) consumers for tomato-based items.3)EFSA FoodEx2 classification system—ensuring harmonized consumption categories in infants, toddlers, children, adolescents, adults, and older adults across European dietary surveys. 1)FAOSTAT consumption data—providing population level estimates for the global, European, and Italian populations. FAOSTAT consumption data—providing population level estimates for the global, European, and Italian populations. 2)SCAI 2005–2006 (Italy)—providing detailed individual level consumption data for mean, 95th percentile (P95), and 99th percentile (P99) consumers for tomato-based items. SCAI 2005–2006 (Italy)—providing detailed individual level consumption data for mean, 95th percentile (P95), and 99th percentile (P99) consumers for tomato-based items. 3)EFSA FoodEx2 classification system—ensuring harmonized consumption categories in infants, toddlers, children, adolescents, adults, and older adults across European dietary surveys. EFSA FoodEx2 classification system—ensuring harmonized consumption categories in infants, toddlers, children, adolescents, adults, and older adults across European dietary surveys. Population groups followed EFSA Comprehensive Database categories: toddlers (12–35 months), other children (3–9 years), adolescents (10–17 years), adults (18–64 years), elderly (≥65 years). These age groups were included when consumption data were available for tomato products. In addition, exposure was estimated separately for males and females when data allowed. As customary in dietary exposure assessments, the analysis included only subjects who consumed at least one item from the tomato-based food category. All consumption data were converted to g/kg bw/day, applying an assumed average adult body weight of 60 kg, and EFSA reference body weights for each age class in FoodEx2. Where different datasets overlapped, consistency checks confirmed that values were comparable (SD < 1.0) and could be used jointly. Risk characterization Risk characterization was performed using two complementary approaches.(1)Hazard Quotient/Risk Characterization Ratio (HQ or RCR)HQ(orRCR)=EDIHBGVwhere EDI is the Estimated Daily Intake and HBGV is the relevant Health-Based Guidance Value. (1)Hazard Quotient/Risk Characterization Ratio (HQ or RCR) Hazard Quotient/Risk Characterization Ratio (HQ or RCR) HQ values < 1 indicate that exposure is unlikely to pose a health concern, whereas HQ > 1 suggests a potential health risk and may warrant consideration of risk-management actions.(2)Margin of Exposure (MOE) approach for genotoxic carcinogens (AOH and AME)MOE=RPEEwhere RP is the toxicological Reference Point or Point of Departure (e.g., NOAEL, BMDL10, etc.) and EE is the Estimated Exposure. For AOH and AME, which show in vitro genotoxicity and possible clastogenicity, the MOE approach recommended by EFSA for genotoxic carcinogens was applied (EFSA Scientific Committee et al., 2025). Based on the limited toxicological database, the in vivo study (28 day study, Sprague–Dawley rats), chosen by Ji and collaborators (Ji et al., 2022) for AME, was chosen. The derived BMDL10 was 0.0336 μg/kg bw/day. Due to the similar structure, mode of action, and toxicological profile of AOH and AME, this BMDL10 value was provisionally applied to both, acknowledging the uncertainty in extrapolating this value. For substances that are both genotoxic and carcinogenic, a MOE below 10,000 is generally considered to be of potential public health concern, whereas MOE values equal to or above 10,000 are regarded as of low concern. (2)Margin of Exposure (MOE) approach for genotoxic carcinogens (AOH and AME) Margin of Exposure (MOE) approach for genotoxic carcinogens (AOH and AME) ATs in tomato products: results and discussion As shown in the PRISMA flow diagram (Fig. 2), twenty research papers published from 2011 to 2025 describing the analysis of tomato-based products via LC-MS techniques were selected among the available literature to be included in the present review. A discussion on sample preparation, LC-MS methods, validation parameters and ATs occurrence in tomato products is presented. The most studied ATs were TeA, AOH, AME and TEN included in >95% of the examined papers. ALT was also extensively evaluated (13 authors out of 20; 65%) while altertoxins, (e.g ATX-I), Alternaria alternata toxins (AALs), alterperynol (ALP) altenusin (ATS) or ATs metabolites (e.g. AME-3-sulfate, AME-3-malonylglucoside, AME 3 and 9 glucoside, AOH-3-sulfate, AOH-9-diglucoside AOH 3 and 9 glucoside, were only rarely investigated.Fig. 2Flow diagram of the extensive literature search for studies on Alternaria toxins and tomato products determination by LC-MS/MS. Flow diagram of the extensive literature search for studies on Alternaria toxins and tomato products determination by LC-MS/MS. Sample preparation and clean-up The great majority of the examined studies included only a small number of the full range of available ATs in the analytical method, focusing on the most relevant in terms of toxicity and occurrence in tomato products like underlined by EFSA evaluations. ATs extraction from tomato products is mainly based on organic solvent:H2O mixtures with the organic portion often predominant over the aqueous phase. MeOH, ACN or a combination of both were often deployed while EtOAc and CH3Cl were less used. Acidification (HCOOH or CH3COOH) was carried out by most authors probably to enhance TeA (pKa ∼ 3.5) extraction, while AOH and AME, being far less acidic compounds (pKa around 7.5), may be less subjected to pH variation of the extraction mixtures (Ostry, 2008). On the other hand, good performances for ATs were also obtained by Walravens and Sanzani groups using only ACN as extraction solvent. The addition of pectinase to tomato samples was explored by Prelle and coauthors and Ji and coworkers. Pectinases are a family of enzymes that break down pectin, a structural heteropolysaccharide found in primary plant cell walls facilitating the extraction procedures from plant and fruit samples. Acid hydrolysis (HCl) and heating were applied to reduce binding-residue effects. Different HCl amounts, concentrations, heating temperatures and time were evaluated to obtain the best compromise in terms of analyte recovery. The best performances were obtained with 2 mL of 0.5M HCl at 80 °C for 30 min (sample weight: 2 g) (Tang et al., 2023). In order to reduce matrix interferences and improve sensitivity and selectivity, several clean-up strategies were explored. The use of QuEChERS salts like MgSO4, NaCl, CH3COONa or a mixture of them was reported by many authors as a good clean-up strategy based on the dispersion of salts to extract (salting-out effect), isolating target analytes from very complex matrices. Zhang and coworkers optimized QuEChERS salts mixtures using 400 mg MgSO4, +80 mg NaCl after testing different combinations from 100 to 600 mg and 60 to 200 mg, respectively (Zhang et al., 2025). Dispersive solid phase extraction (d-SPE) was evaluated as a purification by Walravens and collaborators and later by De Berardis and coworkers. Complete losses of TeA were observed by both authors, therefore d-SPE step was not included in their final optimized procedures. Among other common clean-up strategies, octadecyl silane (C18) yielded the best performances in terms of ATs recoveries (79-96%) with respect to Florisil (43-124%) while treatment with graphitized carbon black (GCB) and primary-secondary amines (PSA) were detrimental especially for AOH and AME (Zhang et al., 2025). Alternatively, solid phase extraction (SPE) is another viable option for improving ATs analysis. SPE cartridges rely on functionalized polymeric sorbent that causes strong retention for target compounds discarding unwanted and interfering compounds (e.g., phospholipids, proteins etc.). Many authors included SPE clean-up in sample preparation using reversed phase sorbents cartridges (Bond Elut PleXA from Agilent, OASIS HLB from Waters, Strata X and Strata XL from Phenomenex). Tang and collaborators compared the performances obtained by eleven different types of SPE columns: among these, C8 SPE columns yielded the best results in terms of recoveries (77.7-112.7%) for the six ATs determined (Tang et al., 2023). Coherently, Florisil and PSA clean-up presented the same issues highlighted by Pinto and coworkers (Pinto et al., 2021). Interestingly, Rodríguez-Carrasco and coworkers developed a dispersive liquid-liquid microextraction technique (DLLME) for ATs analysis in tomato products. DLLME is based on the addition of a mixture of extraction (ES) and disperser solvents (DS) to an aqueous sample creating an organic droplet suspension containing the extracted analytes. Authors tested different ES/DS combinations, obtaining the best performances for CH3Cl (ES)/ACN (DS). Despite analytical precautions, several authors reported an important matrix effect (ME) when determining ATs in tomato products. ME is caused by co-eluted matrix components that compete with the target analytes for ionization processes, resulting in suppression/enhancement of the signal compromising the accuracy and precision of quantitative data. Authors compared matrix-matched calibration (MMC) curve slopes with the corresponding ones in solvent obtaining an estimation of signal suppression/enhancement due to matrix (Rodríguez-Carrasco et al., 2016). Severe suppression of AOH signals were reported by Noser et al. (2011); López et al., 2016 (up to 70%) and ATX-I (up to 50%) while for TEN and TeA both enhancement and suppression were observed. Strong signal suppression was also reported by Rodríguez-Carrasco et al. (2016); Laika et al. (2024). Moreover, López and collaborators results highlighted that the matrix effect depended on the combination of matrix and toxin; acidic matrices (fresh citrus and tomatoes) influenced AOH signal and other ATs to a lesser extent. SPE clean-up highly reduced matrix effects in complex matrices (chili paste, eggplant, ketchup pepper and tomato) allowing more precise determination for ATs and the possibility to reach lower LOQs (Tang et al., 2023). LC-MS determination Chromatographic separation of ATs was achieved by reversed-phase columns (mainly C18 columns) using, as aqueous mobile phases, H2O, acidified H2O or ammonium acetate/bicarbonate buffers; organic phase was mainly MeOH or, alternatively, ACN or acidified ACN. ACN was reported to suppress compounds ionization with respect to MeOH as reported by Zhao et al. (2014). The same authors also chose 0.5 mM ammonium bicarbonate over aqueous ammonia or ammonium formate buffer based on improved peak intensity and shape. TeA is a relatively strong acid and chelating compound; therefore, its chromatographic determination can be challenging. López et al., 2016 obtained good peak shape and stable retention times using a Waters Atlantis HSS-T3 column and acidified mobile phases (HCOOH 1%). MS determination was achieved in both positive and negative mode, since ATs possess functional groups that permit both protonation/deprotonation mechanisms. Briefly, TeA produce m/z 196 (M-H-) precursor ion as a deprotonation of its hydroxyl moiety while m/z 198 (M + H+) can be obtained by protonation of one of the carbonyl groups; same for DAPs like AOH, AME and ALT which can be either protonated (on the pyranone ring) or deprotonated (phenolic OH) yielding intense precursor ions. ESI was preferred to APCI probably because APCI is generally used for non-polar compounds and may not be indicated for the determination of polar and low molecular weight compounds like TeA and TADs. ATs quantitation was carried out by the majority of authors via matrix-matched calibration curves compensating for matrix effect to avoid bias in terms of measurement accuracy. Other than previously discussed clean-up strategies, Walravens et al. (2016) Lattanzio et al. (2021); Pacini et al. (2024) added isotope labelled internal standards (ILISs) at the beginning of the extraction procedure to compensate for both matrix effect and analyte loss obtaining good results in terms of apparent recovery and LOQs. Despite being relatively expensive, ILISs provide more accurate measurements and allow the application of the developed methodology to the most disparate matrices. Analytical procedures and method performances are briefly summarized in Table 4.Table 4Chromatographic conditions, MS parameters and analytical performances of methods for Alternaria toxins determination in tomato products.Table 4Alternaria ToxinsExtractionClean-upColumnMobile phasesInstrumentationLOQ (μg/kg)Recoveries (%)Precision (RSD%)RefALT, AME, AOH, ATS, ATX-I, TeA, TENACN:H2O:MeOH, 45:45:10, adjusted to pH 3Bond Elut Plexa SPE cartridgeAcquity UPLC HSS T3 (100 × 2.1 mm;1.8 μm)A: H2O:ACN, 95:5 +2 mM NH4HCO3B: ACN: H2O 95/5 + 2 mM NH4HCO3QqQ (ESI+/−)2-456-1091-9iNoser et al. (2011)ALT, AME, AOH, TeA, TENPectinase + centrifugation and filtrationStrata X SPE cartridgePursuit XRs Ultra C18A: H2OB: MeOHAPCI ±0.7-32.3551.1-108.50.8-38.2Prelle et al. (2013)AME, AOH, TeA, TENACN/H2O/MeOH, 45:45:10OASIS HLB cartridgeBEH C18 (100 × 2.1 mm; 1.7 μm)A: H2O +0.5 mM NH4HCO3B: MeOHQqQ (ESI-)0.1-5 (LOD)73.4-111.24.2-14.6RZhao et al. (2014)AAL TA1, AAL TA2 ALT, AME, AOH, iso-ALT, ATX-I, TeA, TENACN:H2O:HCOOH,49:50:1NoneHalo™ RP-Amide column (100 × 2.1 mm; 2.7 μm)A: H2O + 1% HCOOHB: MeOH + 1% HCOOHQqQ (ESI+/−)1.2-11064-837.4-24r4.2-14RHickert et al. (2015)AME, AOH, TeA, TENACN:H2O:HCOOH, 84:16:1Mini-uniprep PTFE filter vialAtlantis HSS-T3 (100 × 3.0 mm;3.0 μm)A: H2OB: ACN:H2O 96:4QqQ (ESI+/−)1-5101-11210-20RLópez et al. (2016)ALT, AME, AOH, TeA, TEN + AME-3-sulfate,AME-3-malonylglucoside, AME 3 and 9 glucoside,AOH-3-sulfate,AOH-9-diglucosideAOH 3 and 9 glucoside,ACN:H2O 84:16 with 1% AANoneWaters Atlantis HSS-T3, (100 ×3 mm,3.0 μm)A: H2OB: MeOH with 4% H2OQqQ (ESI+/−)1.0-5.080-120NRLópez et al. (2016)AME, AOH, TEN1 g NaCl +ACN:CH3Cl 9:1noneGemini® C18 (150 × 2 mm; 3 μm)A: H2O + 1% HCOOH and 5 mM AFB: MeOH + 1% HCOOH and 5 mM AFQqQ (ESI+)1.75-3.581-944-9r6-15RRodríguez-Carrasco et al. (2016)ALT, AME, AOH, ATX-I, TeA, TENACN (ACN:H2O 56:44 for dried tomatoes)Na2SO4Acquity UPLC HSS T3 (100 × 2.1 mm;1.8 μm)A: H2O:CH3COOH, 99:1B: ACN:CH3COOH, 99:1QqQ (ESI+/−)9.8-61.589.3-110.61.2-15.7r2.4-17.9RWalravens et al. (2016)AME, AOH, TeA, TENACN+0.1% HCOOH:H2O80:20QuEChERS (MgSO4 4.0 g + NaCl 0.5 g)Luna, phenyl-hexyl (150 × 4.6 mm; 5 μm)A: 15 mL of (NH4)2CO3 1M diluted to 1 L with MeOHQqQ (ESI+)4-16064.2-106.53.33-9.21RDe Berardis et al. (2018)ALP, ALS, ALT, altenuic acid, AME, AOH, ATX-I, ATX-II, iso-ALT, STTX, TeA, TEN + AOH and AME metabolitesMeOH:H2O:CH3COOH,79:20:1NoneAscentis Express C18, (100 × 2.1 mm; 2.7 μm)A: H2O + 5 mM AA (pH adjusted to pH 8.7)B: MeOHQqQ (ESI-)0.2-1251-879-49RPuntscher et al. (2019)AME, AOH, TeA, TENACNExtraction salts (MgSO4 NaCl,C6H6Na2O7 · 1.5H2O, C6H5Na3O7 · 2H2O)Luna C8 (250 × 4.6 mm; 5 μm)A: H2O + 1 mM NH4HCO3:MeOH, 95:5B: MeOHQqQ (ESI-)3-580.9-113.84.63-16.9 rSanzani et al. (2019)AME, AOH, TeA, TENACN:H2O,80:20OASIS HLB cartridgeBetasil RP-18 (150 × 2.1 mm; 5 μm)A: H2O + 0.2% HCOOHB: ACN + 0.2% HCOOHQqQ (ESI+)282.3-87.4<4.1Bertuzzi et al. (2021)ALT, AME, AOH, TeA, TENMeOH:H2O:CH3COOH,85:14:1Strata-XLAscentis Express C18 (100 × 2.1 mm; 2.7 μm)A: 5 mM AAB: MeOHQqQ (APCI-)0.8-25.6100-1233-18r4-22RLattanzio et al. (2021)ALT, AME, AOH, TeA, TENACN:H2O:HCOOH,79:20:1,+ pectinaseanhydrousMgSO4 (4 g), anhydrous CH3COONa (0.5 g), and anhydrous NaCl (0.5 g), +200 mg C18BEH C18 (100 × 2.1 mm; 1.7 μm)A: 5 mM AAB: MeOHQqQ (ESI+/−)0.2-3.075.5-110.04.0-11.9rJi et al. (2022)ALT, AME, AOH, ATX-I, TeA, TENACN:EtOAc:HCOOH,60:39:1NoneNot reportedA: H2O + 0.1% FAB: MeOHQTOF (ESI+)4.2-14.662-871-5Maldonado et al. (2023)ALS, ALT, AME, AOH, TeA, TENHCl hydrolysis +ACNC8 SPE columnAcquity UPLC BEH C18 (50 ×2.1 mm,1.7 μm)A: H2O + 1% FAB: ACNQqQ (ESI+/−)2-572.0-107.42.4-15.1Tang et al. (2023)ALT, AME, AOH, TeA, TENMeOH:H2O:CH3COOH, 80:19:1Strata-XLKinetex EVO C18 (100 × 3.0 mm; 2.6 μm)A: 5 mM AAB: MeOHQqQ (ESI-)0.10-1.3078.4-101.51.6-6.4r5.0-11.8RLaika et al. (2024)ALT, AME, AOH, TeA, TENMeOH:H2O:CH3COOH, 85:14:1Strata-XLAscentis Express C18 (100 × 2.1 mm; 2.7 μm)A: 5 mM AAB: MeOHQqQ (ESI-)1.5-7589-1124.3-9.9rPacini et al. (2024)AME, AOH, TeA, TENACN:H2O:MeOH 45:45:10, pH adjusted to 3-4OASIS HLBC18 column (50 × 2.1 mm;1.7 μm)A: 1 mM ammonium bicarbonateB: MeOHQqQ (ESI-)0.1-1 (LOD)80-120<10r,RXu et al. (2024)ALT, AME, AOH, TeA, TENH2O:ACN+1% FA, 1:2400 mg MgSO4, +80 mg NaCl, +50 mg C18Acquity UPLC HSS T3 (2.1 mm × 50 mm; 1.8 μm)A: 2 mM AAB: ACNQqQ (ESI-)0.02-0.2075.0-103.20.4-8.8r1.3-8.7RZhang et al. (2025)AA: ammonium acetate AF: ammonium formate; FA: formic acid i: intra-day precision; r: in repeatability condition; R: in reproducibility conditions. Table 4Chromatographic conditions, MS parameters and analytical performances of methods for Alternaria toxins determination in tomato products.Table 4Alternaria ToxinsExtractionClean-upColumnMobile phasesInstrumentationLOQ (μg/kg)Recoveries (%)Precision (RSD%)RefALT, AME, AOH, ATS, ATX-I, TeA, TENACN:H2O:MeOH, 45:45:10, adjusted to pH 3Bond Elut Plexa SPE cartridgeAcquity UPLC HSS T3 (100 × 2.1 mm;1.8 μm)A: H2O:ACN, 95:5 +2 mM NH4HCO3B: ACN: H2O 95/5 + 2 mM NH4HCO3QqQ (ESI+/−)2-456-1091-9iNoser et al. (2011)ALT, AME, AOH, TeA, TENPectinase + centrifugation and filtrationStrata X SPE cartridgePursuit XRs Ultra C18A: H2OB: MeOHAPCI ±0.7-32.3551.1-108.50.8-38.2Prelle et al. (2013)AME, AOH, TeA, TENACN/H2O/MeOH, 45:45:10OASIS HLB cartridgeBEH C18 (100 × 2.1 mm; 1.7 μm)A: H2O +0.5 mM NH4HCO3B: MeOHQqQ (ESI-)0.1-5 (LOD)73.4-111.24.2-14.6RZhao et al. (2014)AAL TA1, AAL TA2 ALT, AME, AOH, iso-ALT, ATX-I, TeA, TENACN:H2O:HCOOH,49:50:1NoneHalo™ RP-Amide column (100 × 2.1 mm; 2.7 μm)A: H2O + 1% HCOOHB: MeOH + 1% HCOOHQqQ (ESI+/−)1.2-11064-837.4-24r4.2-14RHickert et al. (2015)AME, AOH, TeA, TENACN:H2O:HCOOH, 84:16:1Mini-uniprep PTFE filter vialAtlantis HSS-T3 (100 × 3.0 mm;3.0 μm)A: H2OB: ACN:H2O 96:4QqQ (ESI+/−)1-5101-11210-20RLópez et al. (2016)ALT, AME, AOH, TeA, TEN + AME-3-sulfate,AME-3-malonylglucoside, AME 3 and 9 glucoside,AOH-3-sulfate,AOH-9-diglucosideAOH 3 and 9 glucoside,ACN:H2O 84:16 with 1% AANoneWaters Atlantis HSS-T3, (100 ×3 mm,3.0 μm)A: H2OB: MeOH with 4% H2OQqQ (ESI+/−)1.0-5.080-120NRLópez et al. (2016)AME, AOH, TEN1 g NaCl +ACN:CH3Cl 9:1noneGemini® C18 (150 × 2 mm; 3 μm)A: H2O + 1% HCOOH and 5 mM AFB: MeOH + 1% HCOOH and 5 mM AFQqQ (ESI+)1.75-3.581-944-9r6-15RRodríguez-Carrasco et al. (2016)ALT, AME, AOH, ATX-I, TeA, TENACN (ACN:H2O 56:44 for dried tomatoes)Na2SO4Acquity UPLC HSS T3 (100 × 2.1 mm;1.8 μm)A: H2O:CH3COOH, 99:1B: ACN:CH3COOH, 99:1QqQ (ESI+/−)9.8-61.589.3-110.61.2-15.7r2.4-17.9RWalravens et al. (2016)AME, AOH, TeA, TENACN+0.1% HCOOH:H2O80:20QuEChERS (MgSO4 4.0 g + NaCl 0.5 g)Luna, phenyl-hexyl (150 × 4.6 mm; 5 μm)A: 15 mL of (NH4)2CO3 1M diluted to 1 L with MeOHQqQ (ESI+)4-16064.2-106.53.33-9.21RDe Berardis et al. (2018)ALP, ALS, ALT, altenuic acid, AME, AOH, ATX-I, ATX-II, iso-ALT, STTX, TeA, TEN + AOH and AME metabolitesMeOH:H2O:CH3COOH,79:20:1NoneAscentis Express C18, (100 × 2.1 mm; 2.7 μm)A: H2O + 5 mM AA (pH adjusted to pH 8.7)B: MeOHQqQ (ESI-)0.2-1251-879-49RPuntscher et al. (2019)AME, AOH, TeA, TENACNExtraction salts (MgSO4 NaCl,C6H6Na2O7 · 1.5H2O, C6H5Na3O7 · 2H2O)Luna C8 (250 × 4.6 mm; 5 μm)A: H2O + 1 mM NH4HCO3:MeOH, 95:5B: MeOHQqQ (ESI-)3-580.9-113.84.63-16.9 rSanzani et al. (2019)AME, AOH, TeA, TENACN:H2O,80:20OASIS HLB cartridgeBetasil RP-18 (150 × 2.1 mm; 5 μm)A: H2O + 0.2% HCOOHB: ACN + 0.2% HCOOHQqQ (ESI+)282.3-87.4<4.1Bertuzzi et al. (2021)ALT, AME, AOH, TeA, TENMeOH:H2O:CH3COOH,85:14:1Strata-XLAscentis Express C18 (100 × 2.1 mm; 2.7 μm)A: 5 mM AAB: MeOHQqQ (APCI-)0.8-25.6100-1233-18r4-22RLattanzio et al. (2021)ALT, AME, AOH, TeA, TENACN:H2O:HCOOH,79:20:1,+ pectinaseanhydrousMgSO4 (4 g), anhydrous CH3COONa (0.5 g), and anhydrous NaCl (0.5 g), +200 mg C18BEH C18 (100 × 2.1 mm; 1.7 μm)A: 5 mM AAB: MeOHQqQ (ESI+/−)0.2-3.075.5-110.04.0-11.9rJi et al. (2022)ALT, AME, AOH, ATX-I, TeA, TENACN:EtOAc:HCOOH,60:39:1NoneNot reportedA: H2O + 0.1% FAB: MeOHQTOF (ESI+)4.2-14.662-871-5Maldonado et al. (2023)ALS, ALT, AME, AOH, TeA, TENHCl hydrolysis +ACNC8 SPE columnAcquity UPLC BEH C18 (50 ×2.1 mm,1.7 μm)A: H2O + 1% FAB: ACNQqQ (ESI+/−)2-572.0-107.42.4-15.1Tang et al. (2023)ALT, AME, AOH, TeA, TENMeOH:H2O:CH3COOH, 80:19:1Strata-XLKinetex EVO C18 (100 × 3.0 mm; 2.6 μm)A: 5 mM AAB: MeOHQqQ (ESI-)0.10-1.3078.4-101.51.6-6.4r5.0-11.8RLaika et al. (2024)ALT, AME, AOH, TeA, TENMeOH:H2O:CH3COOH, 85:14:1Strata-XLAscentis Express C18 (100 × 2.1 mm; 2.7 μm)A: 5 mM AAB: MeOHQqQ (ESI-)1.5-7589-1124.3-9.9rPacini et al. (2024)AME, AOH, TeA, TENACN:H2O:MeOH 45:45:10, pH adjusted to 3-4OASIS HLBC18 column (50 × 2.1 mm;1.7 μm)A: 1 mM ammonium bicarbonateB: MeOHQqQ (ESI-)0.1-1 (LOD)80-120<10r,RXu et al. (2024)ALT, AME, AOH, TeA, TENH2O:ACN+1% FA, 1:2400 mg MgSO4, +80 mg NaCl, +50 mg C18Acquity UPLC HSS T3 (2.1 mm × 50 mm; 1.8 μm)A: 2 mM AAB: ACNQqQ (ESI-)0.02-0.2075.0-103.20.4-8.8r1.3-8.7RZhang et al. (2025)AA: ammonium acetate AF: ammonium formate; FA: formic acid i: intra-day precision; r: in repeatability condition; R: in reproducibility conditions. Chromatographic conditions, MS parameters and analytical performances of methods for Alternaria toxins determination in tomato products. AA: ammonium acetate AF: ammonium formate; FA: formic acid i: intra-day precision; r: in repeatability condition; R: in reproducibility conditions. ATs occurrence in tomato products Tomatoes are used worldwide to produce a plethora of different products, ranging from powders to sauces and juices. Authors performed ATs determination in several types of these products including fresh tomatoes, tomato sauces, pulp, puree and juices, peeled, diced and crushed tomatoes, tomato concentrates, dried tomatoes and ketchup. Some of these share a certain degree of similarity based on the production process and physicochemical properties. For this reason six subgroups were considered for discussion: ketchup, liquid tomato products (tomato sauces, puree and juices), concentrates, dried tomatoes, canned tomatoes (tomato pulp, peeled, diced and crushed tomatoes) and miscellaneous (Featherstone, 2016). “Average ± SD” values (μg/kg) were obtained by pooling each average value reported by the analyzed records; in the same manner median and max values were reported in particular for TeA, AME, AOH and TEN (Table 5).Table 5Average, median and maximum levels (μg/kg) for TeA, AME, AOH and TEN in tomato products.Table 5KetchupLiquid tomato productsConcentratesDried tomatoesCanned tomatoesMiscellaneousRecommendation (EU) 2022/553TeA average ± SD85.5±113.4102.4±80.1338.1±353.352.0124.1±126.3107.0 ± 68.5500TeA (median)41.686.0241.052.038.492.4TeA (max)1787814595552.0338.24560AME (average ± SD)3.8 ± 3.65.0 ± 8.1NCNC2.6 ± 2.63.7 ± 2.35AME (median)1.92.5NCNC2.63.8AME (max)38503.0426.638.5AOH (average ± SD)25.5±36.310.1±11.4NCNC5.0 ± 0.36.5 ± 6.810AOH (median)7.66.7NCNC5.14.4AOH (max)30057.1105.12212.825TEN (average ± SD)2.7 ± 2.22.0 ± 1.9NCNC3.7 ± 2.6NDNRTEN (median)2.81.1NCNC4.6NDTEN (max)15.88.42.0386.836NC: not calculated due to lack of data; NR: not reported; ND: not detected. Table 5Average, median and maximum levels (μg/kg) for TeA, AME, AOH and TEN in tomato products.Table 5KetchupLiquid tomato productsConcentratesDried tomatoesCanned tomatoesMiscellaneousRecommendation (EU) 2022/553TeA average ± SD85.5±113.4102.4±80.1338.1±353.352.0124.1±126.3107.0 ± 68.5500TeA (median)41.686.0241.052.038.492.4TeA (max)1787814595552.0338.24560AME (average ± SD)3.8 ± 3.65.0 ± 8.1NCNC2.6 ± 2.63.7 ± 2.35AME (median)1.92.5NCNC2.63.8AME (max)38503.0426.638.5AOH (average ± SD)25.5±36.310.1±11.4NCNC5.0 ± 0.36.5 ± 6.810AOH (median)7.66.7NCNC5.14.4AOH (max)30057.1105.12212.825TEN (average ± SD)2.7 ± 2.22.0 ± 1.9NCNC3.7 ± 2.6NDNRTEN (median)2.81.1NCNC4.6NDTEN (max)15.88.42.0386.836NC: not calculated due to lack of data; NR: not reported; ND: not detected. Average, median and maximum levels (μg/kg) for TeA, AME, AOH and TEN in tomato products. NC: not calculated due to lack of data; NR: not reported; ND: not detected. Additional information (e.g. number of samples analyzed, prevalence) together with average/median and maximum values detailed in each study were reported in supplementary materials (Table S2-S6). Processing or transformation factors (TFs) are indeed a variable that may be a source of uncertainty due to the possible incorrect identification of the category to which the product refers. However, in the current regulatory framework, as detailed in the previous section, no harmonised TFs have been established to be applied in relation to indicative levels for all types of processed products. In this context, at a national level, in some EU Member States (e.g., Italy), in order to clarify possible future maximum limits for ATs, the competent authority introduced from 2023 in the National Monitoring Program the analysis of ATs in tomato products. In the same plan, TFs for several tomato products, including diced tomatoes (TF:1.45), semi-concentrates with dry residue between 12 and 18% (TF:3), single concentrate with dry residue between 12 and 18% (TF:4.7), double concentrate with dry residue between 28 and 36% (TF:6), triple concentrate with dry residue between 36 and 38% (TF:7.5), semi-dried tomatoes (TF:2) and tomato powder and flakes (TF:20) were reported (Italian Ministry of Health, 2023). TFs may be useful to interpret compliance with the indicative levels reported in Recommendation (EU) 2022/553 for both raw and processed tomato products. TeA TeA is the most common AT in tomato products: eighteen authors out of twenty detected TeA at least once in their surveys. High prevalence was reported by most authors and in many cases, all analyzed samples contained TeA above LOQ. Maximum concentration ranged from 4.0 to 5955 μg/kg with average values between 4.5 and 901 μg/kg. The highest value was reported by Bertuzzi et al. (2021) a in concentrated tomato paste sample. TeA median values ranged between < LOD and 241.0 μg/kg. Considering all the available data, TeA was detected, on average, at 85.5 μg/kg (median: 41.6 μg/kg) in ketchup, 102.4 μg/kg (median: 86.0 μg/kg) in liquid tomato products, 338.1 μg/kg (median: 241.0 μg/kg) in tomato concentrates, 52 μg/kg (median: 52 μg/kg) in dried tomatoes and 124.1 μg/kg (median: 119.2 μg/kg) in canned tomatoes. Among miscellaneous products the highest levels were reported in fresh tomatoes by Sanzani and coworkers (up to 4560 μg/kg) although average values are comparable to the tomato products (107.0 μg/kg; median: 92.4 μg/kg). Concentrates appear to be the most contaminated tomato products: this finding can be explained by the industrial process behind tomato concentrates production. One kilogram of concentrate is obtained from 5 to 7 kg of fresh tomato depending on the type of the desired product (single, double or triple concentrate) (Food and Agriculture Organization, 2009). Consequently, the initial contamination can increase several times with respect to other tomato products like juices, sauces, ketchup or canned tomatoes. This finding stresses out the importance for food safety entities and regulatory institutions to promptly establish TFs for tomato products, to discriminate between compliant/non-compliant samples based on their production process, other than contamination levels alone. AME AME was reported by 11 authors out of 20 (55%). Average concentration of AME in ketchup was 3.8 μg/kg (median: 1.9 μg/kg; maximum 38 μg/kg); 5.0 μg/kg (median: 2.5 μg/kg; maximum: 50 μg/kg) in liquid tomato products and 2.6 μg/kg (median: 2.6 μg/kg; maximum: 6.6 μg/kg) in canned tomatoes. With respect to average values, relatively high levels of AME were reported by Zhao et al. (2014) in ketchup (up to 38 μg/kg; prevalence 90%) and by Maldonado et. 2023 and Zhang et al., 2025 in tomato purees (up to 33.1 μg/kg; prevalence: 8%) and tomato sauces (up to 50 μg/kg; prevalence: 96%), respectively. Notably, Zhao et al., 2014; Zhang et al., 2025 described diffuse AME contamination, in contrast with Maldonado et. 2023 with average and median values higher than the recommended limit for this toxin in tomato products (5 μg/kg). This finding suggests how, in this particular case, at least 50% of the analyzed samples may be non-compliant to Recommendation (EU) 2022/553. AOH AOH presence was detailed by fifteen out of twenty authors (75%) accounting for the second most common AT in tomato products. AOH was detected in liquid tomato products, on average, at 10.1 μg/kg (median: 6.7 μg/kg) in canned tomatoes at 5.0 μg/kg (median: 5.1 μg/kg) and in ketchup at 25.5 μg/kg (median: 7.6 μg/kg). More precisely, Zhao detected AOH in 45% of the ketchup samples in the range 2.5 to 300 μg/kg with a median value of 29 μg/kg. The levels detected by Zhao et al., (2014) are ten-fold higher than the average contamination reported by other authors (López et al., 2016; Tang et al., 2023; Xu et al., 2024). Nevertheless, ATs in ketchup should be consistently monitored for at least two reasons. Firstly, according to Statista.com, ketchup is one of the most consumed condiments worldwide: in the United States ketchup sales (764 million $) account for 38% of the total sauces sales (2 billion $) (Statista.com, 2017). Secondly, ketchup is mainly consumed by children and adolescents, who can be more exposed than other population segments to possible contaminants in a relatively vulnerable phase of their lives. As we previously described, AOH possess significant cancerogenic properties and its frequent detection at relatively high concentration in food mainly destined to youngsters’ consumption should be further elucidated, possibly including ketchup-based preparation like barbecue sauces and similar condiments. In addition, AOH and its methyl ether derivative (AME) are frequently co-detected, posing additional concerns for possible additional/synergistic concerns in terms of toxicity and carcinogenicity. Other ATs Other relevant ATs were evaluated by authors although rarely detected and at generally lower concentrations than TeA, AME and AOH. Briefly, tentoxin (TEN) was reported by eleven out of twenty manuscripts (55%) with average concentrations of 2.7 μg/kg in ketchup (median: 2.8 μg/kg), 2.0 μg/kg (median: 1.1 μg/kg) in liquid tomato products and 3.7 μg/kg (median: 4.6 μg/kg) in canned tomatoes. When detected, prevalence was generally <35% with the only exception of Zhao et al., (2014) which reported TEN in 86% of the analyzed ketchup samples with maximum concentrations up to 15.8 μg/kg. The highest values were reported by Sanzani et al., (2019) in fresh tomatoes (36 μg/kg) and dried tomatoes (38 μg/kg). Based on the reported data, TEN is generally less detected in tomato products with an average concentration sensibly lower than TeA, AOH and AME. Moreover, TEN is not considered a genotoxic compound and its level in food does not overcome the threshold of toxicological concern (TTC) as detailed in 2016 by EFSA's evaluation on dietary exposure assessment to ATs. Altenuene (ALT), together with AME and AOH is part of the family of dybenzopyrone ATs, sharing the same synthetic pathway. In contrast, for reasons that are not clearly elucidated right now, its detection in tomato products is sensibly lower than AME and AOH. ALT occurrence was detailed by thirteen authors but only four reported > LOD levels in tomato products. High prevalence (80%) was reported only by Prelle et al. (2013) with low concentrations (maximum level: 4.82 μg/kg). Lower amounts (up to 2.0 μg/kg) were detailed by Zhang et al., 2025 in 19 out of 43 tomato products. Conversely, Maldonado et al., 2023 highlighted ALT presence in only three samples out of thirty-eight (8%) although average levels were much higher than the ones reported for similar toxins like AME and AOH. Two hypotheses may explain the lower abundance of ALT in tomato products: i) ALT can be produced only by certain fungi isolates: Ismail et al., (2023) reported that only some Alternaria alternata strains isolates are able to produce ALT and, sometimes, other ATs were produced at higher concentrations (Ismail et al., 2023); ii) LOQs for ALT are generally higher than the ones obtained for other ATs so, in some cases, this toxin could be present in tomato samples although at levels that are below the quantitation limits of the analytical methods. Other authors focused on the evaluation of less common ATs like altertoxins, alterperylenol and AOH/AME metabolites. Altertoxins were evaluated by Walravens et al. (2016), Hickert et al. 2018, Puntscher et al. (2019); Maldonado et al. (2023). The first two never detected altertoxins in tomato products while Maldonado and Puntscher groups reported altertoxin I (ATX-I) presence. Puntscher et al., (2019) reported ATX-I at low levels (up to 3.7 μg/kg) in 23% of tomato sauces samples (average 2.2 μg/kg; median 2.3 μg/kg); comparable levels of ALP (average 4.0 μg/kg; median 4.4 μg/kg), AME-3-sulfate (average 2.5 μg/kg; median 1.0 μg/kg) and AOH-3-sulfate (average 1.2 μg/kg; median 1.0 μg/kg). In contrast, Maldonado and coworkers reported much higher concentrations of ATX-I with two tomato puree samples at 146.9 and 233.9 μg/kg. ATs metabolites (AOH and AME sulfates and glucosides) were investigated by Lopez and collaborators in 43 tomato products albeit never being detected (López et al., 2016a, López et al., 2016b). Co-occurrence of ATs Although it is very difficult to pick out contamination data of each single sample analyzed by the manuscripts included in this review, we can easily state that AT contamination is rarely caused by a single toxin. For example, AOH and AME, sharing the same synthetic route, are often present at the same time in contaminated samples. In addition, TeA was often detected in the totality of the samples analyzed, suggesting a widespread presence of this toxin in all tomato-based products. This indicates that co-occurrence of ATs is a relevant problem especially in terms of combined toxicity and possible synergistic/additive effects between two or more compounds. In vitro cytotoxicity of AOH, AME and TeA toward human intestinal epithelial cells (Caco-2) and hepatocytes (HepG2) was explored by den Hollander and colleagues. Binary and ternary combinations of ATs were tested highlighting a slight increase of cytotoxicity, with respect to single mycotoxins effects, at the highest concentrations tested and in both cell lines (den Hollander et al., 2022). Marin and coworkers evaluated the combined effects of AOH and AME on porcine epithelial cells (IPEC-1) with respect to single compounds. Co-administration of AOH and AME significantly decreased cell viability and strongly increased cytotoxicity if compared with the individual toxins, suggesting a synergistic mechanism (Marin et al., 2024). Although not conclusively demonstrated, it is likely that ATs may exert toxic synergistic effects: this should be further investigated through more in-depth studies possibly including other less evaluated, but not so irrelevant ATs like altertoxins and AAL-toxins. Risk assessment Based on the occurrence values from the 20 studies included in this review, together with consumption data from FAOSTAT, EFSA FoodEx2, and national Italian dietary surveys (Table S7), a conservative chronic dietary exposure assessment was performed for the main ATs in tomato products. Data were extracted for global, European (the only region currently establishing indicative levels), and Italian populations, including age subgroups and high consumers (P95 and P99), with Italy selected because of its high consumption of tomato products. Food consumption data were evaluated using a WoE and EJ approach to provide a comprehensive, matrix-specific assessment. Following EFSA methodology, left-censored occurrence data were treated using the LB approach, in line with the large proportion of results reported below LOQ in the available literature. The elaborations were performed for liquid tomato products (non-concentrated tomato products, canned tomato products and ketchup) based on the availability and consistency of occurrence data. For all ATs, the EDIs were compared with TTC values and HQ was derived. For carcinogenic and genotoxic AOH and AME also MOE was derived. The data obtained were shown in Table S8. Liquid tomato products-only exposure to AOH and AME frequently reached or exceeded the TTC for genotoxic substances, set at 0.0025 μg/kg bw/day, particularly in children, toddlers and in high-consumption scenarios. This pattern is consistent with EFSA previous evaluations, which likewise found that chronic exposure to AOH and AME often exceeds the corresponding TTC, indicating the need for improved toxicological understanding and improved occurrence datasets. To complement the TTC approach, exposure to AOH and AME was further compared with the BMDL10 of 0.0336 μg/kg bw/day, derived from a 28-day oral study for AME and commonly used as a provisional reference point for both toxins. Application of this value yielded very low MOE across all tomato product categories and population groups, often falling well below 100 and therefore far from the MOE ≥10,000 typically considered of low concern for genotoxic carcinogens. This result should not be interpreted as definitive evidence of high risk but rather as an indication that, given the currently limited toxicological data, a potential health concern associated with exposure to AOH and AME from tomato products cannot be excluded. Nevertheless, several uncertainties temper the interpretation of these results. The available toxicological reference point is based on a short-term study, and the use of AME values for AOH involves an inherent read-across assumption. Moreover, the true contribution of masked or conjugated toxins, as well as the role of structurally distinct Alternaria metabolites, such as PQs, remains insufficiently characterized in both occurrence and toxicological studies. Recent reviews emphasize that many relevant Alternaria metabolites still lack proper hazard characterization, making it difficult to draw firm regulatory conclusions at this time (Zhang et al., 2026). In addition, there is a notable lack of published data on processing factors and their influence on AT levels. In contrast, the estimated exposures to TeA and TEN deriving from tomato products remained well below the TTC of 1.5 μg/kg bw/day for non-genotoxic compounds, for all population groups assessed. This finding fully aligns with EFSA earlier opinion, which judged TeA and TEN exposures to be unlikely to pose a health concern based on the available evidence. Taken together, these results reinforce current regulatory perspectives. The European Commission Recommendation (EU) 2022/553 does not define legal limits but establishes indicative levels for AOH, AME, and TeA in processed tomato products, explicitly calling for improved monitoring, better analytical sensitivity, and deeper investigation into factors leading to high contamination events. The present exposure calculations support this approach: while TeA and TEN do not appear to pose appreciable risk within the tomato-only scenario, AOH and AME require continuous attention, more toxicological research, and more detailed occurrence data particularly for tomato concentrates, which show consistently higher contamination due to processing-related accumulation. The available evidence suggests that ATs in tomato products, especially AOH and AME, warrant ongoing vigilance, strengthened monitoring programs, and targeted research to address toxicological data gaps. At the same time, the results should be interpreted within the boundaries of current uncertainties. Overall, these findings are fully coherent with EFSA assessments and support a prudent, science-based approach to future risk management. Conclusions and future challenges ATs remain an underestimated group of contaminants in tomato products, despite clear evidence of their recurrent occurrence, structural diversity, and toxicological relevance. This review highlights that, while remarkable analytical progress has been made over the last decade, particularly through LC-MS/MS and multi-mycotoxin workflows, the scientific and regulatory landscape surrounding ATs is still characterized by substantial knowledge gaps. Exposure assessments based on currently available data suggest that, for some population groups, tomato-based foods may meaningfully contribute to chronic dietary intake of genotoxic ATs. Although subject to some limitations, the rigorous study selection and the stringent inclusion and exclusion criteria allow for a risk characterization based on higher-quality data, thereby reducing uncertainties related to concentration estimates. This reinforces the need for vigilance and highlights how essential high quality, well standardized data are for improving future risk characterization. Future progress in this area will depend on coordinated action at multiple levels. First, refined and standardized occurrence data are urgently needed, particularly for processed tomato products where concentration effects during production can amplify toxin presence. The near absence of validated processing or transformation factors remains a major barrier to interpreting contamination across product categories and to setting product specific guidance levels. In addition, toxicological research must advance beyond the current reliance on short term studies and structural read across approaches, expanding to long term, mechanistic and evidence-based studies, as well as to mixture toxicity investigations. At the same time, monitoring strategies must narrow it down, integrating high content methods, harmonized LOQs (as already outlined in point 3 of Recommendation (EU) 2022/553), and systematic reporting of all other ATs. As analytical sensitivity improves, the frequency of left censored data should diminish, enabling more reliable exposure modelling. The take home message of this review is clear: ATs in tomato products are neither sporadic nor negligible. They represent a persistent and globally relevant issue in a food that can be considered staple in many countries and therefore requires sustained scientific attention, stronger regulatory guidance, and practical mitigation strategies throughout the entire supply chain. Advances in detection methods are no longer sufficient on their own; meaningful progress now depends on integrating toxicology, exposure science, food technology, and regulatory policy. Only through these combined efforts can consumers be effectively protected, uncertainties reduced, and future risk management decisions robustly supported. CRediT author statement Stefano Sdogati and Teresa D'Amore conceived and designed the review. All authors contributed to the writing of the manuscript. Use of generative AI Not used. Funding This research received no external funding. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.