Vehicular Ad-hoc NETworks (VANETs) are a promising technology for vehicular communications since they enable safer, more efficient, and intelligent transportation systems. However, their highly dynamic topology and intermittent connections present several challenges for reliable information dissemination. Vehicle clustering which groups vehicles into clusters, is an effective approach to improve the performance and efficiency of VANETs. Several clustering algorithms have been proposed in the past decade to reduce communication overhead, manage topology changes, reduce energy consumption, and provide differentiated quality of service (QoS) levels for different services. This paper provides a new comprehensive analysis of recent clustering-based algorithms in VANETs that appeared between 2021 and 2024. These approaches are discussed and compared considering clustering criteria, cluster-head selection metrics, scalability, performances, and security. Additionally, We deeply examine essential security measures for VANETs, tackling challenges such as multi-layer attacks and leveraging advanced technologies like blockchain to ensure data integrity, confidentiality, and availability. Furthermore, an in-depth and critical analysis of the reviewed approaches is provided to identify gaps and give some future directions.