Occurrence, typing and inactivation kinetics of Listeria monocytogenes isolated from commercial frozen desserts

Listeria monocytogenes is a psychrotrophic pathogen of concern in dairy frozen products due to its ability to survive post-processing and grow under refrigeration. This study aimed to assess the occurrence, thermal resistance, and refrigerated growth potential of L . monocytogenes in commercial frozen dessert samples from São Paulo, Brazil, and to model the thermal inactivation kinetics in flavored ice cream mixes. A total of 100 samples (75 dairy-based and 25 water-based frozen desserts) from 12 brands and 88 flavors were analyzed according to ISO 11290-1:1996 (ISO, 2004). Isolates were characterized by PCR and PFGE. L. monocytogenes was confirmed in 3% of the samples, all of which were dairy-based and originated from a single brand. Fourteen serotype 1/2b confirmed isolates were assessed for growth at 4 °C and thermal resistance at 70 °C for 15 s and 90 s. Three representative strains, each from a distinct PFGE pulsotype, were selected to evaluate inactivation kinetics at 60, 65, and 70 °C in vanilla (pH 7.09; 36.7°Brix) and strawberry (pH 5.78; 35.5°Brix) ice cream mixes. Isolates exhibited greater phenotypic variability in thermal resistance (γ 70 °C/15s ranging from 2.5 to 5.3 log CFU/g) than in refrigerated growth (δ 4 °C/360 h = 4.3–4.7 log CFU/g). Inactivation curves were best described by first-order kinetics (R 2 > 0.9). D-values decreased with increasing temperature, and greater thermal resistance was observed in vanilla ice cream mix. Estimated z-values were 6.7–6.8 °C for both matrices. The persistence of L . monocytogenes and the protective effect of the dairy dessert matrix highlight the pathogen's potential to survive sublethal processes and grow during cold storage. These findings underscore the need for matrix-specific validations and risk-based approaches in the processing of ready-to-eat dairy frozen desserts. The high phenotypic variability among isolates further supports the inclusion of strain diversity in predictive modeling for food safety assessments.