Live-attenuated vaccines of human respiratory syncytial virus for infants and young children: Progress and prospects of an old yet promising platform Authors: Jiemei Yu, Yuanhui Fu, Yanpeng Zheng, Xianglei Peng, Jinsheng He Published: 2026 ABSTRACT Highlights•No human respiratory syncytial virus (RSV) vaccine has achieved regulatory approval for RSV-seronegative children aged 6 months to 5 years.•Reverse genetics has proven instrumental in balancing the attenuation and immunogenicity of RSV live-attenuated vaccines (LAVs).•RSV LAVs show no evidence of VAERD during clinical evaluation in RSV-seronegative pediatric cohorts.•Diverse methodologies are being actively explored to further optimize and evaluate the protective efficacy and attenuation of RSV LAV candidates.•Intranasal RSV LAVs represent a promising platform for achieving a major milestone in pediatric RSV prevention. Introduction Human respiratory syncytial virus (RSV) is the most common pathogen of acute lower respiratory tract illness (LRTI) in infants and children under five years old, and represents the second most common etiology of lower respiratory infection mortality globally [1], [2]. Most children experience at least one RSV infection by age 2–3, with approximately 50 % having two infections in this period [3], [4], [5]. Annually, RSV is responsible for an estimated 33 million cases of LRTI, 3.6 million hospitalizations, 26,300 in-hospital deaths, and 101,400 overall deaths in children under five years worldwide. Notably, while infants aged 0–6 months are the most susceptible, infants and young children aged 6–60 months account for 80 % of LRTI cases, 61 % of hospitalizations, 49 % of in-hospital deaths, and 55 % of overall deaths [2]. The hospitalization burden of RSV in children under five years exceeds that of seasonal influenza [6], with RSV demonstrating higher severity in hospitalized children under 24 months, even during peak influenza season [7]. A key feature of RSV pathogenesis is its predominant replication in local mucosal tissues, especially in the superficial epithelial cells via the luminal surface and without viremia, thereby largely evading robust systemic adaptive immune responses [8], [9], [10], which likely explains why natural infection induces incomplete and short-lived immunity, leading to frequent reinfections [11], [12], [13]. Globally, over 95 % of RSV-associated LRTI episodes and 97 % of RSV-attributable deaths occur in low- and middle-income countries (LMICs) in children under 5 years [2]. In the elderly, RSV can cause severe respiratory disease with high mortality, much of which may be undiagnosed [1]. RSV is a negative-sense, single-stranded ribonucleic acid (RNA) virus that belongs to the genus Orthopneumovirus within the family Pneumoviridae [14], [15]. Its genome consists of 10 distinct genes that yield 11 proteins (NS1, NS2, N, P, M, SH, F, G, M2-1, M2-2, and L). Of note, the M2 gene contains overlapping open reading frames (ORFs) that dictate the synthesis of two individual proteins. The fusion (F) and attachment (G) glycoproteins represent the primary antigens responsible for eliciting RSV-neutralizing antibody responses. The F protein exists in two conformations: the metastable prefusion F (preF) and the stable postfusion F (postF). Notably, the preF conformation possesses a greater density of neutralizing epitopes and elicits more potent neutralizing antibody responses than postF [9]. PreF-specific antibodies account for approximately 90 % of the neutralizing activity in human sera [16], [17]. This underscores preF as the critical antigen for vaccine development. The clinical significance of this antigen is highlighted by its central role in the three recently approved preF-conformation-based vaccines. Furthermore, it is exemplified by the long-acting monoclonal antibodies (mAbs) of nirsevimab and clesrovimab; the former targets the preF-specific antigenic site Ø, while the latter binds to site IV, which is conserved across both the preF and postF conformations. This review synthesizes historical and ongoing efforts to develop live-attenuated vaccines (LAVs) against RSV tailored for seronegative infants and children, with a particular focus on the unique attributes of the LAV platform. For more than five decades, strategies for developing RSV LAVs have advanced substantially from classical empirical methods such as cold passage and chemical mutagenesis to precise, rational design using reverse genetics, which has yielded a diverse array of candidate LAVs, including biological and recombinant derivatives of wild-type RSV. Although no RSV LAV has yet attained regulatory approval, significant progress has been achieved. A pivotal advancement has been the incremental refinement of attenuation strategies, particularly those aiming at decoupling viral replication (and thus pathogenicity) from immunogenicity. This review will discuss the efforts and the critical lessons learned from the development and evaluation of RSV LAVs and offer insights to guide their future development. Challenges for pediatric RSV vaccines targeting RSV-seronegative infants and children RSV was first discovered in humans in 1957 [18], [19], thereby initiating a challenging journey of vaccine research spanning over six decades. As early as the 1960s, a formalin-inactivated whole RSV (FI-RSV) vaccine led to a tragic outcome: upon subsequent natural RSV infection, RSV-seronegative vaccinated children developed vaccine-associated enhanced respiratory disease (VAERD), resulting in two fatalities [20], [21], [22]. The underlying mechanisms of VAERD involve poor neutralizing antibodies and weak cytotoxic T lymphocyte (CTL) responses against native RSV, instead promoting a T helper 2 cell (Th2)-skewed immune response. This imbalance leads to immune complex deposition, complement activation, and pulmonary inflammation upon re-exposure to the virus, causing severe airway obstruction [20], [23]. The bottleneck persisted until the structural characterization of preF protein revitalized vaccine design, drove the research and development of structure-based vaccines and led to the approval of two subunit vaccines (Arexvy and Abrysvo) in 2023, as well as one messenger RNA (mRNA) vaccine (mRESVIA) in 2024 for older adults (≥ 60 years) and at-risk individuals (18–59 years) by the U.S. Food and Drug Administration (FDA). The U.S. Centers for Disease Control and Prevention (CDC) currently recommends RSV vaccination for all adults ≥ 75 years, as well as for individuals aged 50–74 years who are at increased risk for severe disease. Furthermore, Abrysvo was authorized in 2023 for use in pregnant women at 32–36 weeks of gestation to protect infants from birth to 6 months of age. Additionally, the long-acting mAbs nirsevimab and clesrovimab were licensed for all infants under 8 months entering their first RSV season in 2023 and in 2025, respectively [24], [25]. Nirsevimab was also recommended for high-risk infants and young children aged 8–19 months entering their second season [24]. Prior to the approval of these three vaccines and two mAbs by the FDA, the sole authorized prophylactic option against RSV disease was palivizumab (Synagis), a humanized mAb targeting antigenic site II of the F protein. However, owing to its substantial cost and the requirement for monthly administrations, the clinical utility of palivizumab has been restricted to high-risk infants with underlying lung or heart disease, primarily within developed countries [26], [27], [28]. Despite these advances, an unmet clinical need persists for RSV-seronegative pediatric populations aged 6–60 months; this vulnerable cohort bears a heavy disease burden but still lacks access to a safe and effective vaccine (Fig. 1) [2], [29], [30]. Like the historical FI-RSV vaccine, the approved subunit vaccines are derived from non-replicating platforms. They do not produce antigens de novo within host cells, predominantly stimulate major histocompatibility complex (MHC) class II pathways, and are thus poor inducers of CD8+ T cell responses. Furthermore, previous studies have already shown that CD8+ cytotoxic T lymphocytes can suppress Th2 responses and the resulting VAERD [31], [32]. Therefore, the subunit vaccines which poorly induce these T-cell responses, have historically been associated with a potential risk of Th2-mediated VAERD [9], [33], [34], [35], [36]. While preF antigens with T helper 1 cell (Th1)-skewed adjuvants can induce neutralizing antibodies and prevent VAERD at prophylactic doses in animal models, the risk re-emerges under suboptimal vaccination conditions that mimic contraction immunity [28], [36], [37].Fig. 1Active and passive RSV immunization for population at risk of severe RSV disease. Infants, young children and the elderly are at risk of severe RSV disease. Infants aged < 6 months can be protected against severe RSV disease through maternal vaccination with Abrysvo vaccine (A bivalent vaccine comprising equal amounts of recombinant RSV preF from RSV subgroups A and B). All infants under 8 months of age entering their first RSV season whose mothers did not receive the maternal vaccination should receive a single dose of either nirsevimab or clesrovimab. Furthermore, a dose of nirsevimab is recommended for selected children aged 8–19 months at elevated risk for severe RSV disease entering their second RSV season. Adults aged ≥ 75 years old are recommended to receive active immunization with either protein subunit vaccines (Abrysvo and Arexvy) or mRNA-based vaccine (mRESVIA). Currently, there are no approved vaccines to provide active, long-term immunity for RSV-seronegative pediatric populations aged 6–60 months. RSV LAV has unique advantages and is an important candidate vaccine for this age group. Illustration created using BioRender. Abbreviations: YTE (M252Y/S254T/T256E): the amino acid substitutions in Fc portion which extend the half-life of the monoclonal antibodies. RSV, human respiratory syncytial virus; preF, prefusion F; LAV, live-attenuated vaccine; VAERD, vaccine-associated enhanced respiratory disease. Active and passive RSV immunization for population at risk of severe RSV disease. Infants, young children and the elderly are at risk of severe RSV disease. Infants aged < 6 months can be protected against severe RSV disease through maternal vaccination with Abrysvo vaccine (A bivalent vaccine comprising equal amounts of recombinant RSV preF from RSV subgroups A and B). All infants under 8 months of age entering their first RSV season whose mothers did not receive the maternal vaccination should receive a single dose of either nirsevimab or clesrovimab. Furthermore, a dose of nirsevimab is recommended for selected children aged 8–19 months at elevated risk for severe RSV disease entering their second RSV season. Adults aged ≥ 75 years old are recommended to receive active immunization with either protein subunit vaccines (Abrysvo and Arexvy) or mRNA-based vaccine (mRESVIA). Currently, there are no approved vaccines to provide active, long-term immunity for RSV-seronegative pediatric populations aged 6–60 months. RSV LAV has unique advantages and is an important candidate vaccine for this age group. Illustration created using BioRender. Abbreviations: YTE (M252Y/S254T/T256E): the amino acid substitutions in Fc portion which extend the half-life of the monoclonal antibodies. RSV, human respiratory syncytial virus; preF, prefusion F; LAV, live-attenuated vaccine; VAERD, vaccine-associated enhanced respiratory disease. In December 2024, the FDA halted clinical trials of Moderna′s mRNA vaccines (mRNA-1345 and mRNA-1365) and other vaccines for RSV-seronegative pediatric populations due to the risk of VAERD. These mRNA vaccine trials failed to demonstrate protection, with some recipients aged 5 to 8 months experiencing VAERD following natural RSV infection despite robust neutralizing antibody responses. Although it is still too early to conclude on the safety and the underlying mechanism of mRNA infant vaccines against RSV infection, preliminary data revealed an elevated frequency of interleukin-5 (IL-5) responses (a Th2 cytokine) in RSV-seronegative vaccinees compared with RSV-seropositive vaccinees [38], [39]. This outcome challenges conventional paradigms, as even an mRNA-based platform, which directs intracellular translation of viral antigen to theoretically favor Th1/CTL priming, exhibits a propensity toward priming a Th2 immune response in RSV-seronegative children. This striking observation underlines the profound complexity of developing a safe and effective RSV infant vaccine. In sharp contrast, all of the RSV LAVs in clinical development are inherently devoid of VAERD risk and continue to demonstrate excellent safety profiles, positioning them as highly promising vaccine candidates for this population [38], [39], [40], [41]. The development of the pediatric RSV LAVs Although data on the safety of RSV LAVs were included in some early studies [42], [43], [44], [45], [46], [47], it was not until 2007 that the first comprehensive evidence regarding their safety was reported, demonstrating no occurrence of VAERD in seronegative infants and children upon subsequent exposure to natural RSV infection following intranasal vaccination [48]. Therefore, the core challenge currently facing the development of such vaccines lies in achieving an optimal balance: sufficient attenuation to ensure no risk of LRTI, while retaining enough immunogenicity to elicit robust protection. Furthermore, it is imperative that candidate strains for RSV LAV exhibit genetic stability to prevent reversion to a wild-type or wild-type-like phenotype. Before the early 1990s, attenuation was primarily developed through repeated low-temperature passage to generate cold-passaged (cp) mutants, or through chemical mutagenesis to produce temperature-sensitive (ts) strains, which gave rise to the first RSV LAV candidates, including cp-RSV, ts-1, ts-2, as well as ts-1 A, B, and C [46], [49], [50], [51]. These early versions were intended to exhibit restricted replication relative to wild-type RSV within the lower respiratory tract, thereby mitigating the risk of vaccine-associated LRTI. The first reported RSV LAV, developed in 1968, was a cp-RSV strain derived from the RSV A2 strain through serial passage in human and bovine embryonic kidney cells at progressively lower temperatures (final temperature 26 °C) [49]. However, in clinical trials involving seronegative children, these early cp-RSV and ts mutants proved to be either under-attenuated, genetically unstable, or over-attenuated [51], [52], [53], [54]. In the mid-1990s, the original cp-RSV underwent further chemical mutagenesis, yielding a new set of ts mutants including cpts248, cpts248/404, cpts248/955, cpts530, cpts530/1009, and cpts530/1030 et al. [55], [56], [57], [58]. When evaluated in seronegative vaccinees aged 6–36 months, the first 2 vaccines in the cpts lineage of cpts248/955 and cpts530/1009 were insufficiently attenuated [43], [59]. The cpts530/1030 mutant was not feasible for vaccinating seronegative individuals due to high viral shedding as seen with cpts248/955 in the seropositive population [57]. In contrast, cpts248/404 was more attenuated in this seronegative population and was the first RSV LAV to be evaluated in 1–2-month-old infants, where it was found to be associated with nasal congestion [44]. Critically, studies of these candidates demonstrated that intranasal RSV LAVs could replicate in the presence of maternally derived serum neutralizing antibodies [44], [56]. During that period, an RSV B subgroup LAV, cp-52B, was also developed via 52 low-temperature passages (21–32 °C) of the B1 virus [60]. While it replicated well in Vero cells, it was overattenuated in seronegative children and later found to harbor a large deletion preventing expression of the SH and G proteins, rendering it unsuitable as a vaccine candidate [61]. It is evident that conventional methods for vaccine development, such as cold passage and chemical mutagenesis, are too unpredictable and labor-intensive to be feasible for generating optimal RSV LAVs. A major advancement occurred in 1995 with the establishment of a reverse genetics system for recovering the infectious RSV A2 strain from a cloned complementary deoxyribonucleic acid (cDNA) clone. This technology not only identified mutations present in cpRSV and 6 of its biological ts derivatives as mentioned above, but also enabled, for the first time, the precise insertion or deletion of designed mutations into the viral genome, significantly accelerating the rational development of RSV LAVs. Since then, all subsequent candidates are recombinant derivatives. For example, the candidate rA2cp248/404/1030ΔSH was the first recombinant derivative, incorporating three mutations from the cpts248/404 parent strain, a deletion of the SH gene, and an additional 1030 ts mutation. It demonstrated sufficient attenuation (with virus shedding around 102.5 plaque-forming units (PFU)/mL in seronegative children), moderate immunogenicity in 1–2-month-old infants, and conferred protection against a second dose of the vaccine virus [42], [62], which is the first RSV vaccine candidate that appears to be appropriately attenuated for young infants under 6 months of age. However, a significant drawback was observed: partial loss of the ts phenotype occurred in over one-third of viral isolates, due to reversion to the wild-type sequence or other amino acid substitutions at the 248 or 1030 mutation sites [42], [63], [64]. Notably, while the 404-M2 mutation is relatively stable, a C → A nucleotide change was identified in a specimen from one infant vaccinated with cpts248/404 [44], [65]. These findings highlight that a ts phenotype reliant on one- or two-point mutations is inherently unstable and prone to reversion, rendering it unacceptable for an infant vaccine. Subsequently, owing to the elucidation of RSV gene functions, targeted deletion of the genes encoding non-essential proteins such as NS1, NS2, SH, M2-2, and G is increasingly employed to develop next-generation LAVs [42], [61], [66], [67], [68], [69], [70]. This approach often achieves a better balance between attenuation and immunogenicity by decoupling viral replication from the induction of protective immune responses, with several candidates advancing to clinical trials [44], [71], [72], [73], [74]. For instance, deleting the NS1 or the NS2 genes, which antagonize host interferon (IFN) production and apoptosis, attenuates the viral virulence by impairing its ability to evade innate immunity. This deletion may also enhance vaccine immunogenicity through an adjuvant-like effect driven by increased host IFN expression and signaling [75], [76], [77], [78]. Furthermore, the NS2 deletion may mitigate the risk of airway obstruction, a pathology linked to this protein [79]. Conversely, the M2-2 gene deletion, which regulates the transition from viral gene transcription to genome replication, reduces viral replication capacity while increasing the expression of neutralizing antigens of F and G. This shift potently enhances adaptive immune responses [80], [81]. A critical advantage of gene deletion over point mutations is the extremely high genetic barrier to reversion, as restoring an entire gene sequence is highly improbable. Furthermore, more innovative strategies are currently being employed to develop novel RSV LAVs. For example, among the candidate vaccines are MV-012–968 and RSV-MinL4.0, the former of which was constructed by codon deoptimization of the G, NS1, and NS2 genes, combined with the SH gene deletion and the removal of the secretory G protein, while the latter was created by codon-pair deoptimization of the L gene and the introduction of specific amino acid substitutions into the N, P, M2-1, and L proteins [82], [83], [84]. More recently, the “one-to-stop” approach has been adapted to develop vaccines for RNA viruses such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). This method applies RNA virus polymerase′s natural error rate to weaken the virus by introducing synonymous mutations that increase the probability of premature termination codons, thereby reducing viral fitness and pathogenicity while achieving attenuation [85], [86]. Crucially, it must be noted that RSV LAVs developed for human use must comply with the standards for human products throughout all manufacturing processes. For instance, this requires the use of qualified cell substrates, such as specific lots of Vero cells, for virus recovery and production. To enhance safety and consistency, the T7 RNA polymerase is now typically supplied via a co-transfected plasmid rather than a recombinant virus or stable cell line, and electroporation has often replaced liposome transfection [87]. A schematic of this modern, full-plasmid RSV rescue system is provided in Fig. 2.Fig. 2Schematic of RSV reverse genetics based on a full-plasmid transfection system. Cells were transfected by electroporation with four helper plasmids each encoding the RSV N, P, L, and M2-1 genes, along with another two plasmids encoding the T7 ribonucleic acid (RNA) polymerase gene and the RSV full-length antigenomic complementary deoxyribonucleic acid (cDNA), respectively. Following expression of N, P, L, M2-1, and T7 RNA polymerase, the transcription of the plasmid carrying the antigenomic cDNA is initiated by T7 RNA polymerase and generates full-length positive-sense (+) antigenomic RNA or the complementary RNA (cRNA) of the RSV genome. The RSV polymerase L then replicates cRNA to generate negative-sense (−) genomic RNA (vRNA). This vRNA subsequently serves a dual purpose: it acts as a template for the messenger RNA (mRNA) synthesis to produce viral proteins, and it functions as the template for the synthesis of cRNA to drive the amplification of progeny genomes. Both the assembly of viral proteins with genomic RNA and the subsequent budding of the progeny virions occur at the plasma membrane. Illustration created using BioRender. Abbreviations: RSV, human respiratory syncytial virus; CAG promoter, a hybrid construct consisting of the cytomegalovirus (CMV) enhancer fused to the chicken beta-actin promoter. Schematic of RSV reverse genetics based on a full-plasmid transfection system. Cells were transfected by electroporation with four helper plasmids each encoding the RSV N, P, L, and M2-1 genes, along with another two plasmids encoding the T7 ribonucleic acid (RNA) polymerase gene and the RSV full-length antigenomic complementary deoxyribonucleic acid (cDNA), respectively. Following expression of N, P, L, M2-1, and T7 RNA polymerase, the transcription of the plasmid carrying the antigenomic cDNA is initiated by T7 RNA polymerase and generates full-length positive-sense (+) antigenomic RNA or the complementary RNA (cRNA) of the RSV genome. The RSV polymerase L then replicates cRNA to generate negative-sense (−) genomic RNA (vRNA). This vRNA subsequently serves a dual purpose: it acts as a template for the messenger RNA (mRNA) synthesis to produce viral proteins, and it functions as the template for the synthesis of cRNA to drive the amplification of progeny genomes. Both the assembly of viral proteins with genomic RNA and the subsequent budding of the progeny virions occur at the plasma membrane. Illustration created using BioRender. Abbreviations: RSV, human respiratory syncytial virus; CAG promoter, a hybrid construct consisting of the cytomegalovirus (CMV) enhancer fused to the chicken beta-actin promoter. Altogether, RSV LAVs present a highly promising strategy for protecting infants and young children, offering several key advantages (Fig. 1), including: (1) Acceptable safety profile. The use of reverse genetics effectively eliminates the reversion risk of RSV LAVs, however their replication competence enables de novo synthesis of viral proteins within host cells. This process simulates the natural infection. Hence, RSV LAVs can avoid VAERD when directly administered intranasally to seronegative pediatric populations, which is the most important advantage of RSV LAVs [46], [48], [72], [82], [88], [89]. This safety profile is substantiated by clinical trial data involving thousands of seronegative infants and children. These data span several decades, from 1997 until late 2025, marked by the early termination of the Sanofi Phase 3 PEARL trials within the Respiratory Syncytial Virus Toddler (RSVt) program [42], [43], [44], [45], [48], [71], [72], [73], [74], [90], [91], [92], [93], [94], [95], [96]. Furthermore, even infants whose levels of nirsevimab had begun to decline did not have evidence of VAERD upon their exposures to natural RSV infection [97], which suggests it is also safe if RSV LAVs are inoculated following passive immunization; (2) Convenient administration and induction of robust immune responses. The nasal drop method, which delivers RSV LAV directly to the respiratory mucosa, avoids needle puncture and improves convenience and compliance. Meanwhile, this vaccine formulation, which retains a near-complete antigenic profile, combined with its intranasal delivery route, can elicit durable, full-spectrum immune responses—encompassing systemic and local mucosal immunity. Furthermore, it is highly effective in overcoming interference from maternally derived antibodies [42], [44], [73], [94], [98], [99], [100], [101]. Critically, F protein is largely presented in its prefusion conformation in RSV LAV, evokes high titers of neutralizing antibodies that can persist for 6–12 months, and primes strong anamnestic responses upon natural virus exposure [71], [102]; (3) The defined evaluation criteria for the efficacy. A ≥ 4-fold increase in the complement-enhanced 60 % plaque reduction RSV-neutralizing antibody titer (PRNT60) from pre-vaccination levels—rather than the absolute titer—has been shown to predict efficacy against RSV disease, particularly RSV-associated medically attended lower respiratory illness (RSV-MALRI). Notably, none of the recipients who achieved a ≥ 4-fold rise developed RSV-MALRI [71]. In these studies, RSV-seronegative participants are defined as children older than 6 months with a baseline PRNT60 ≤ 1:40 (< 5.3 log2), whereas seropositive participants are those with a PRNT60 > 1:40 [42], [44], [59], [93]. Alternatively, the serum IgA specific for RSV F protein is also used as a biomarker to categorize participants as RSV-naïve or RSV-experienced, which is produced only in response to RSV infection, other than transferred transplacentally from mother to child [103]; (4) The relatively small Phase 3 trial is capable of evaluating significant efficacy. RSV infection rates are much higher among infants and young children, with a basic reproductive number (R0) estimated at approximately 3.0 [104], [105]. The morbidity can vary markedly yearly; however, the 80 % of LRTI and over half of RSV deaths occur in children aged 6 months and older, the target population of RSV LAVs. Furthermore, the established correlation between a ≥ 4-fold increase in RSV PRNT60 and clinical protection in RSV-naïve children enables precise estimates of efficacy from even relatively small clinical trials [71]. As a result, the global Phase 3 PEARL trial (NCT06252285) of RSVt sponsored by Sanofi enrolled approximately 6,300 children aged 6 to 22 months across 19 countries. Representative pediatric RSV LAVs in clinical development As previously noted, establishing an optimal balance between attenuation and immunogenicity has been a major challenge in the development of RSV LAV. Early candidates such as cpts530/1009, cpts248/955, and cpts248/404 showed insufficient attenuation [43], [44], [59], while some later constructs incorporating gene deletions, including rA2cpΔNS2 and rA2cp248/404ΔNS2, were also either under- or over-attenuated [45]. Currently, approximately 15 RSV LAV candidates are undergoing clinical evaluation in RSV-seronegative infants and young children [38], [39], many of which demonstrate considerable translational potential. Below are concise introductions to the representative RSV LAV candidates, including 5 “more promising” vaccine candidates [71], which are RSV/ΔNS2/Δ1313/I1314L, MEDI/ΔM2-2, LID/ΔM2-2, LID/ΔM2-2/1030s, and D46/NS2/N/ΔM2-2-HindIII, capable of inducing a four-fold or greater rise in PRNT60 in 80 % of the vaccinees in Phase 1 clinical trials. The remaining candidates are selected to demonstrate diverse strategies for balancing attenuation and immunogenicity. Their modification strategies and key attributes are summarized in Table 1.Table 1Clinical characteristics and development stage of representative RSV LAVs in RSV-seronegative infants and children.RSV LAV candidatesRecipient age (months)Inoculation No.Dosage (PFU)Peak viral titer in NW, median (Log10 PFU/mL)Serum RSV PRNT60 ≥ 4 fold rise (%)Serum IgG RSV F antibody titer* ≥ 4 fold rise (%)Stage of clinical development(registration ID)Clinical statusRSV/ΔNS2/Δ1313/I1314L6–24Single1051060.61.8, 3.1Δ5360, 805360, 853 (NCT06252285)Terminated6–18Two105106−−67.961.983.779.4RSV/6120/ΔNS2/1030s6–24Single1053.071, 9076, 851 (NCT03387137);2 (NCT03916185)CompletedTerminatedMEDI/ΔM2-26–24Single1051.595901 (NCT01459198)CompletedRSV/2766–24Single1053.271, 9292, 1001 (NCT03227029 /NCT03422237);2 (NCT03916185)CompletedTerminatedLID/ΔM2-26–24Single1053.890901 (NCT02237209, NCT02040831)CompletedLID/cpΔM2-26–24Single1051.645451 (NCT02890381, NCT02948127)TerminatedLID/ΔM2-2/1030s6–24Single1053.185851 (NCT02952339, NCT02794870,NCT04520659)NCT02952339, NCT02794870, NCT04520659: Completed;D46/NS2/N/ΔM2-2-HindIII6–24Single1053.5951001 (NCT03102034, NCT03099291)NCT03102034, NCT03099291: Completed;RSV 6120/ΔNS1 &RSV 6120/F1/G2/ΔNS16–24Single105101.4, 101.254, 7962, 711 (NCT03596801)NCT03596801: Completed;MV-012–9686–36SingleTwoNA107NANANANANANA1 (NCT04909021)NCT04909021: Unknown statusRSV-MinL4.06–24Single103–106NANANA1 (NCT04919109)NCT04919109: Terminated*Serum IgG titers to RSV F were determined with ELISA across all the vaccinees, except the vaccinees of 2 doses of intranasal RSV/ΔNS2/Δ1313/I1314L, the preF-specific IgG titers instead.Δ The values from two independent clinical studies or two distinct recombinant viruses were marked with “,”.Abbreviations: RSV, human respiratory syncytial virus; LAVs, live-attenuated vaccines; NW, nasal wash; PFU, plaque-forming units; PRNT60, the complement-enhanced 60 % plaque-reduction RSV-neutralizing antibody titers; NA, not available. Table 1Clinical characteristics and development stage of representative RSV LAVs in RSV-seronegative infants and children.RSV LAV candidatesRecipient age (months)Inoculation No.Dosage (PFU)Peak viral titer in NW, median (Log10 PFU/mL)Serum RSV PRNT60 ≥ 4 fold rise (%)Serum IgG RSV F antibody titer* ≥ 4 fold rise (%)Stage of clinical development(registration ID)Clinical statusRSV/ΔNS2/Δ1313/I1314L6–24Single1051060.61.8, 3.1Δ5360, 805360, 853 (NCT06252285)Terminated6–18Two105106−−67.961.983.779.4RSV/6120/ΔNS2/1030s6–24Single1053.071, 9076, 851 (NCT03387137);2 (NCT03916185)CompletedTerminatedMEDI/ΔM2-26–24Single1051.595901 (NCT01459198)CompletedRSV/2766–24Single1053.271, 9292, 1001 (NCT03227029 /NCT03422237);2 (NCT03916185)CompletedTerminatedLID/ΔM2-26–24Single1053.890901 (NCT02237209, NCT02040831)CompletedLID/cpΔM2-26–24Single1051.645451 (NCT02890381, NCT02948127)TerminatedLID/ΔM2-2/1030s6–24Single1053.185851 (NCT02952339, NCT02794870,NCT04520659)NCT02952339, NCT02794870, NCT04520659: Completed;D46/NS2/N/ΔM2-2-HindIII6–24Single1053.5951001 (NCT03102034, NCT03099291)NCT03102034, NCT03099291: Completed;RSV 6120/ΔNS1 &RSV 6120/F1/G2/ΔNS16–24Single105101.4, 101.254, 7962, 711 (NCT03596801)NCT03596801: Completed;MV-012–9686–36SingleTwoNA107NANANANANANA1 (NCT04909021)NCT04909021: Unknown statusRSV-MinL4.06–24Single103–106NANANA1 (NCT04919109)NCT04919109: Terminated*Serum IgG titers to RSV F were determined with ELISA across all the vaccinees, except the vaccinees of 2 doses of intranasal RSV/ΔNS2/Δ1313/I1314L, the preF-specific IgG titers instead.Δ The values from two independent clinical studies or two distinct recombinant viruses were marked with “,”.Abbreviations: RSV, human respiratory syncytial virus; LAVs, live-attenuated vaccines; NW, nasal wash; PFU, plaque-forming units; PRNT60, the complement-enhanced 60 % plaque-reduction RSV-neutralizing antibody titers; NA, not available. Clinical characteristics and development stage of representative RSV LAVs in RSV-seronegative infants and children. *Serum IgG titers to RSV F were determined with ELISA across all the vaccinees, except the vaccinees of 2 doses of intranasal RSV/ΔNS2/Δ1313/I1314L, the preF-specific IgG titers instead. Δ The values from two independent clinical studies or two distinct recombinant viruses were marked with “,”. Abbreviations: RSV, human respiratory syncytial virus; LAVs, live-attenuated vaccines; NW, nasal wash; PFU, plaque-forming units; PRNT60, the complement-enhanced 60 % plaque-reduction RSV-neutralizing antibody titers; NA, not available. RSV/ΔNS2/Δ1313/I1314L This candidate was constructed by combining the deletion mutation with the ts mutation and was also designated SP0125, VAD00001, VAD00004, or RSVt, and was the first RSV LAV to proceed to the Phase 3 clinical trial. It was engineered by deleting the NS2 gene and the 1313 codon in the L protein. The 1313 codon deletion confers a moderate ts phenotype and is stabilized by I1314L substitution. Additional modifications include a 112-nucleotide deletion in the downstream non-translated region of the SH gene and five synonymous mutations in the terminal region of the SH open reading frame, collectively termed “6120” [72], [75], [106]. In clinical trials (NCT01893554, NCT03227029/NCT03422237), the candidate proved genetically stable, immunogenic, and well-tolerated, with viral shedding titers between 101.8 and 103.1 PFU/mL in RSV-seronegative vaccinees aged 6–24 months. It also primed strong anamnestic immune responses upon natural RSV exposure [72], [73]. Phase 2 data (NCT04491877) further indicated a favorable safety profile and a ≥ 4-fold PRNT60 rise in 62 %-68 % of RSV-seronegative vaccinees aged 6–18 months, with a marked reduction in viral shedding after a second dose, indicating effective immune priming [103], [107]. Based on these encouraging outcomes, Sanofi Pasteur initiated a global Phase 3 PEARL trial (NCT06252285) in February 2024 to evaluate two intranasal doses in children aged 6–22 months, with an expected completion date at the end of 2027. RSV/6120/ΔNS2/1030s This candidate was constructed by introducing the “6120” mutation, deleting the NS2 gene, and incorporating the “1030s” (a stable version of 1030) ts mutation comprising K (AAA) and S (TCA) at codons 1321 and 1313 in the L polymerase gene [63], [92], which was expected to be less restricted in replication and more immunogenic than RSV/ΔNS2/Δ1313/I1314L in seronegative infants and young children [108]. In a Phase 1 trial (NCT03387137), a single intranasal dose proved immunogenic and genetically stable in RSV-seronegative vaccinees aged 6–24 months, with vaccine virus shed at 103.0 PFU/mL and ≥ 4-fold increases in PRNT60 in 90 % of vaccinees. However, rhinorrhea occurred more frequently among vaccine recipients than in the placebo group, and a single instance of LRTI was associated with vaccine virus shedding, which was not observed with RSV/ΔNS2/Δ1313/I1314L [108]. In the meantime, RSV/6120/ΔNS2/1030s was also in a head-to-head comparison with RSV/ΔNS2/Δ1313/I1314L in a larger multisite Phase 1/2 study (NCT03916185) and displayed a similar safety profile but no superior immunogenicity, as indicated by the posted outcomes from the clinical trial. The study closed to enrollment prior to reaching full accrual, resulting in less precise outcome estimates than originally planned. MEDI/ΔM2-2 and RSV/276 These two closely related candidates were both derived from the M2-2 deletion. MEDI/ΔM2-2, formerly known as rA2ΔM2-2, was generated by deleting a 234-nucleotide segment from the C-terminal region of the M2-2 ORF via introduction of two HindIII sites [74], [81], [109], [110]. RSV/276 contains the same 234-nucleotide M2-2 deletion as and differs from MEDI/ΔM2-2 by only four nucleotides in the noncoding region: A1937G, G8198C, C8199G, and T13900C [73], [111]. In a trial of RSV-seronegative vaccinees aged 6–24 months (NCT01459198), MEDI/ΔM2-2 (105.0 PFU dose) was shed in 101.5 PFU/mL and induced ≥ 4-fold increases in PRNT60 in 95 % of recipients, with responses sustained for 5–12 months and potent anamnestic immunity upon natural infection [71], [74]. Despite its promising results, further evaluation was terminated under a Collaborative Research and Development Agreement (CRADA) between the National Institute of Allergy and Infectious Diseases (NIAID) and MedImmune, LLC, MA, USA [74], [111]. In a head-to-head Phase 1 trial with RSV/ΔNS2/Δ1313/I1314L (NCT03227029/NCT03422237) in RSV-seronegative vaccinees aged 6–24 months, RSV/276 also showed excellent infectivity, tolerability, and immunogenicity, priming strong anamnestic responses, displaying similar shedding (103.1 vs 103.2 PFU/mL) and similar frequency (24 % vs 21 %) of peak titer ≥ 4 log10 PFU/mL. Individuals who had been vaccinated with either vaccine candidate rarely experienced RSV-MALRI upon community exposure. RSV/276 seemed associated with more frequent mild cough in the setting of the small sample size [73]. Additionally, it was also compared with RSV/ΔNS2/Δ1313/I1314L and RSV/6120/ΔNS2/1030s in the Phase 1/2 study (NCT03916185) mentioned above. LID/ΔM2-2 and LIDcpΔM2-2 LID/ΔM2-2 was derived from an RSV A2 strain (GenBank KT992094) and incorporated a 241- nucleotide deletion from the M2-2 ORF along with the “6120” mutation. In two companion studies (NCT02040831, NCT02237209), it induced ≥ 4-fold PRNT60 rises in 90 % of RSV-seronegative vaccinees aged 6–24 months and anamnestic responses post-RSV season, with no RSV-MALRI reported. However, one case of mild and objectionable LRTI and relatively high shedding (median peak titer of 103.8 PFU/mL) with 16 % of peak titer ≥ 5 log10 PFU/mL raised potential safety concerns following vaccination [91], [94]. LIDcpΔM2-2, a further attenuated version carrying five additional cp-derived mutations, was evaluated in two Phase 1 studies (NCT02890381 and NCT02948127) [49], [112]. It showed reduced shedding (101.6 PFU/mL) but was over-attenuated in RSV-seronegative vaccinees aged 6–24 months [95]. LID/ΔM2-2/1030s This candidate built upon LID/ΔM2-2 through the addition of the “1030s” ts mutation. In three Phase 1 clinical trials (NCT02952339, NCT02794870, and NCT04520659), a single 105 PFU intranasal dose in RSV-seronegative vaccinees aged 6–24 months infected 90 % of recipients (based on the shedding vaccine virus and/or a ≥ 4-fold rise in serum RSV antibodies). Among them, 85 % of recipients shed the vaccine virus, with a median peak titer of 103.1 PFU/mL but no peak titer ≥ 5 log10 PFU/mL, and displayed ≥ 4-fold increases in PRNT60. Following surveillance, 6 of 19 vaccinees had ≥ 4-fold increases in PRNT60, indicating infection with wild-type RSV. Five of these individuals remained free of RSV-MALRI, whereas the only participant who developed RSV-MALRI was the recipient who had neither a post-vaccination serologic response nor detectable vaccine virus shedding. Adverse events included respiratory symptoms and fever, but VAERD or other concerning safety signals were not reported [92]. This vaccine demonstrated good tolerance in RSV-seronegative children, primed for anamnestic antibody responses, and was highly attractive for further evaluation. D46/NS2/N/ΔM2-2-HindIII This candidate was engineered on the LID/ΔM2-2 backbone but retained the complete 112-nucleotide 3′ noncoding region of the SH gene, while incorporating a 234-nucleotide deletion in the M2-2 gene, akin to the strategy used for MEDI/ΔM2-2. It was further distinguished by two amino acid substitutions: K51R in NS2 and T24A in the N protein [94]. In two companion clinical trials (NCT03102034 and NCT03099291) involving RSV-seronegative recipients aged 6–24 months, a single intranasal dose of 105.0 PFU led to vaccine shedding in 95 % of recipients, with median peak titers of 103.5 PFU/mL. Additionally, 95 % and 100 % of vaccinees showed ≥ 4-fold rises in PRNT60 and anti-F IgG, respectively. Compared with the LID/ΔM2-2 strain, D46/NS2/N/ΔM2-2-HindIII exhibited a narrower range and lower variability in peak viral titers with only 5 % ≥ 5 log10 PFU/mL versus 16 %. Furthermore, it demonstrated a reduced mean shedding level of 3.5 versus 3.8 log10 PFU/mL. Taken together, these findings indicate enhanced attenuation and warrant its further clinical development [94]. Notably, LID/cp/ΔM2-2, LID/ΔM2-2/1030s, and D46/NS2/N/ΔM2-2-HindIII represent three derivatives of LID/ΔM2-2, collectively constituting the LID/ΔM2-2 family. Each derivative incorporates distinct attenuating mutations, and they are ranked by attenuation level as follows: LID/ΔM2-2 < D46/NS2/N/ΔM2-2-HindIII < LID/ΔM2-2/1030s < LID/cp/ΔM2-2 [94]. RSV 6120/ΔNS1 and RSV 6120/F1/G2/ΔNS1 RSV 6120/ΔNS1 was constructed by introducing the “6120” mutation and deleting the NS1 gene. Its modified variant, RSV 6120/F1/G2/ΔNS1, incorporated additional optimization strategies: codon optimization of the F gene to enhance translation efficiency, and the repositioning of the F and G genes to the first and second genomic positions, respectively, to maximize protein expression [113]. A Phase 1, head-to-head trial (NCT03596801) was initiated in June 2018 to evaluate both candidates in RSV-seropositive children aged 12–59 months and RSV-seronegative children aged 6–24 months. According to data posted in February 2026, these two RSV LAVs displayed a good attenuation phenotype, with median peak titers of 101.4 and 101.2 PFU/mL, respectively. However, their immunogenicity was lower than expected, with 54 % and 79 % of vaccinees showing ≥ 4-fold rises in PRNT60, respectively [114]. MV-012–968 Developed by Meissa Vaccines, this LAV incorporated several modifications into its RSV A2 backbone: expression of a chimeric line19 F protein, deletion of the SH gene, ablation of the secreted form of G, and codon deoptimization of the NS1, NS2, and G genes [82], [115]. In a Phase 1 clinical trial (NCT04909021), MV-012–968 elicited strong serum neutralizing antibody responses in RSV-seronegative vaccinees aged 6–36 months, with favorable tolerability and significant attenuation. No serious adverse events were reported [116]. Despite these encouraging preliminary data, further development of the vaccine candidate has recently been halted due to strategic and budgetary reevaluations [117], [118]. RSV-MinL4.0 This candidate, developed by Codagenix, was initially attenuated through codon-pair deoptimization of the L gene, resulting in the ts strain RSV-MinL. The subsequent RSV-MinL4.0 variant was stabilized by introducing four amino acid substitutions, each in the N, P, M2-1, and L proteins [83], [84], [119], [120], [121]. Following the completion of a Phase 1 trial in healthy adults (NCT04295070), a second Phase 1 trial (NCT04919109) has been initiated and terminated. In summary, while the development of RSV LAVs is encouraging, challenges remain in balancing attenuation and immunogenicity, and in identifying the most viable candidates for advanced clinical trials. Owing to insufficient efficacy in the Phase 3 PEARL trial (NCT06252285), Sanofi discontinued the clinical development of this candidate in October 2025—a program that nevertheless marked a historical milestone as the first RSV LAV to enter Phase 3 evaluation. Nevertheless, the safety profile remained acceptable, and no signals related to VAERD were observed [122]. Because detailed data and official commentaries regarding this discontinuation have not yet been released, the underlying mechanisms of this failure remain open to speculation. For instance, one potential contributing factor is that, unlike Phase 1 and 2 trials, Phase 3 trials typically forgo baseline serologic screening prior to participant recruitment. Consequently, some infants harboring undetected residual maternal IgG, or young children with pre-existing neutralizing antibodies from prior natural RSV exposure, may also have been enrolled. This could lead to the rapid neutralization of the RSV vaccine candidate upon intranasal entry and the subsequent suppression of vaccine replication. Second, the vaccine itself may possess the insufficient immunogenicity because of over-attenuation, compromising its in vivo replication. Lastly, the immature humoral immunity of the target infant population likely plays a role, as their low frequency of somatic hypermutation typically results in reduced antibody avidity and restricted diversity [108], [123]. Consequently, developing alternative RSV LAVs with enhanced immunogenicity—such as MEDI/ΔM2-2, RSV/276, and LID/ΔM2-2/1030s (a prominent member of the LID/ΔM2-2 family)— represents a promising strategy to enhance overall vaccine efficacy. These candidates generally exhibit an acceptable safety profile and superior immunogenicity in RSV-seronegative vaccinees compared with RSVt, even at 105 PFU dose [71], [72], [74], [91], [92], [94]. Furthermore, the licensure of long-acting mAbs and the maternal Abrysvo vaccine has largely obviated the need to protect young infants using RSV LAVs. Accordingly, provided that RSV LAVs demonstrate an acceptable safety profile in RSV-seronegative pediatric populations aged ≥ 6 months, their current degree of attenuation may be deemed sufficient, rendering further attenuation unnecessary. Crucially, data generated by all individual RSV LAVs entering clinical trials remain invaluable and warrant rigorous evaluation and interpretation; these findings may well hold the clues for further enhancing vaccine immunogenicity. Discussion The recent success of RSV preF-based subunit vaccines in protecting the elderly and conferring immunity to infants via maternal immunization marks a major milestone. Currently, it underscores a critical gap: the lack of effective preventive and therapeutic strategies for RSV-seronegative pediatric populations aged 6 months and older. As comprehensively reviewed herein, intranasally administered RSV LAVs are increasingly emerging as highly viable platforms for protecting this vulnerable population. The following sections synthesize decades of research insights and key experiences in RSV LAV development, while exploring critical pathways for future development. The approach and criteria for evaluating attenuation phenotype of RSV LAVs Regarding the evaluation of attenuation phenotype of RSV LAVs in preclinical stage, the transition over the model has occurred early from chimpanzees—the only fully permissive model for RSV replication and disease [43], [56], [67], [70], [75], [112], [124], [125], [126]—to semi-permissive models like cotton rats and African green monkeys (AGMs) due to ethical constraints [127], [128]. Although phylogenetically closer to humans than rodents [129], [130], the same semi-permissive nature of AGM as in cotton rats makes it unfeasible to achieve precise attenuation assessment [82], [131], [132], [133], [134], [135], [136]. Now, this translational gap has been bridged by the development of well-differentiated human airway epithelial (HAE) cell cultures, which are the only significant indicators of the RSV attenuation phenotype in young infants and RSV-seronegative children [44], [137]. The ex vivo systems recapitulate the human airway′s structure and function, including a polarized epithelium, ciliary activity, mucus production, and relevant receptor expression [138], [139], [140]. Critically, RSV LAV growth in HAE correlates strongly with replication levels in RSV-seronegative children, providing a reliable predictor of clinical attenuation phenotypes and reducing preclinical reliance on animal models [134], [135], [137]. Evidence suggests that a restriction of RSV LAV growth in HAE by approximately 2.0 and 3.0 logs is predictive of viral attenuation in RSV-seronegative children, respectively [137], thereby rendering early-stage RSV LAV candidate screening independent of animal models and significantly enhancing development efficiency. This degree of restriction is crucial because, apart from being free of LRTI, the ideal LAVs are also sufficiently attenuated to minimize post-vaccination upper respiratory tract illness (URTI). To rigorously validate this safety profile in vivo, Phase I clinical evaluations employ an age-deescalation approach, progressing sequentially from seropositive adults to RSV-seropositive children (12–59 months), and finally to the ultimate target population of RSV-seronegative infants and children (6–24 months). Concurrently, safety endpoints including fever, otitis media, URTI such as rhinorrhea, LRTI and non-LRTI cough must be closely monitored [108]. Crucially, based on an expanding body of clinical data, a fundamental safety checkpoint for RSV LAVs is the observation of minimal or no vaccine virus shedding in RSV-experienced groups; failure to meet this criterion underlines under-attenuation, thereby halting further development of the candidate in RSV-seronegative pediatric populations [42], [43], [45], [54], [59]. Therefore, quantifying shedding magnitude remains vital for characterizing the attenuation level or the vaccine safety in early clinical trials [42], [71], [94]. However, the precise criteria for viral shedding used to assess the safety of RSV LAVs in RSV-seronegative infants and children remain unclear and are currently determined primarily based on empirical evidence. For example, the previous studies indicated if the mean peak shedding titers were ≥ 102.4 PFU/mL in RSV-seronegative infants and children, they would be 104.0-104.9 PFU/mL in 1–2-month-old young infants and associated with adverse events like nasal congestion that interferes with feeding and sleep [44], [94], and if the shedding titers were ≥ 104.4 PFU/mL, they would be insufficiently attenuated in the RSV-seronegative infants and children themselves and even triggered wheezing [43], [44], [59]. However, shedding titers of 102.5 PFU/mL in RSV-seronegative children have also been reported to be well tolerated both in seronegative children and in 1–2-month-old young infants [42]. It has been documented that RSV/ΔNS2/Δ1313/I1314L (in 106 dose) and the other promising candidates, such as MEDI ΔM2-2 and LID/ΔM2-2/1030s, exhibit mean peak shedding titers around 101.8 to 103.1 PFU/mL, 101.5 PFU/mL, and 103.1 PFU/mL, respectively [72], [73], [74], [92], [94]. It seems that 103.1 PFU/mL is the upper limit of mean peak shedding titers for an attenuated RSV LAV in RSV-seronegative infants and children. Additionally, the clinical development of intranasal RSV LAVs faces several limitations that warrant attention in future studies. In particular, the small sample sizes inherent to early-stage trials constrain the statistical power required to definitively assess vaccine-associated adverse events (e.g., cough) and protective efficacy against community-acquired RSV-MALRI [73]. The multifaceted immunological mechanisms underlying the protective efficacy of intranasal RSV LAVs Many studies have shown that intranasal, but not systemic immunization, elicits mucosal immune responses and provides superior protection against respiratory virus infections and transmission [141], [142]. The mechanisms underlying the protective efficacy of intranasal RSV LAVs have yet to be fully elucidated. Although a ≥ 4-fold rise in serum neutralizing antibody titer (PRNT60) is predictive of efficacy against RSV-MALRI following intranasal RSV LAVs [71], [102], it may not represent the spectrum of protection. Crucially, systemic neutralization alone is not invariably correlated with protection [71], [143], [144], [145], primarily due to the inefficient transudation of serum antibody into the nasal mucosa [71], [146]. For example, a recent observation is that intranasal LAVs can confer complete protection in both the upper and lower respiratory tracts of nonhuman primates, despite inducing lower serum PRNT60 titers than intramuscular subunit vaccines. This phenomenon highlights the critical contribution of immunity beyond systemic neutralization [143], as well as the multiple coordinated mechanisms required for such robust protection. Furthermore, high levels of nasal SIgA produced by local plasma cells provide early defense at the mucosal entry point and have been shown to predict protection against RSV infection [145], [147]. Beyond humoral immunity, systemic CD8+ T cells in humans and CD8+ TRM cells in the lungs of humans and mice are critically linked to optimized viral control. By facilitating rapid viral recognition and accelerated clearance, these cellular components effectively curtail viral load and dampen symptom severity upon re-exposure [148], [149], [150], [151], [152], [153]. In contrast to the protective profile observed in humans, murine systemic CD8+ T cells frequently contribute to a dual phenotype, facilitating viral clearance while simultaneously driving severe immunopathology [31], [32], [148], [149], [154], [155], [156], [157], [158]. Additionally, antibody Fc-effector functions play a significant role in fighting against RSV disease [16], [143], [144], [159], [160]. Collectively, these findings support the interpretation that a ≥ 4-fold rise in PRNT60 serves as a direct mediator of protection and a marker for the successful induction of broader, non-neutralizing immune mechanisms [71]. The elucidation of these defensive pathways, particularly the role of lung CD8+ TRM, not only deepens our understanding of vaccine-induced immunity but also provides new clues for designing and evaluating next-generation RSV LAVs with superior protection compared to the natural infection. Another emerging concept is trained innate immunity, which has been shown to contribute to heterologous and broad-spectrum protection against diverse respiratory threats following intranasal or intramuscular vaccination with adenovirus-vectored COVID-19 vaccine, influenza vaccines, or even the “universal vaccine” [161], [162], [163], [164]. This phenomenon has not yet been rigorously documented for RSV LAVs, but it presents an intriguing question: could RSV LAVs also elicit trained innate responses, RSV-specific or non-specific? Collectively, such mechanisms would be particularly advantageous against a fast-replicating pathogen like RSV, which employs a “hit-and-run” infection strategy and can be better controlled by the rapidly triggered innate immune responses and the local humoral and CD8+ TRM responses at the site of RSV entry [8], [165]. Future studies are needed to dissect the relative contributions of diverse immune components to the overall efficacy of intranasal RSV LAVs. Additionally, current immunogenicity assessments predominantly rely on established correlates of protection such as serum RSV-neutralizing antibodies and anti-F IgG titers. Expanding these to include measurements such as RSV-neutralizing antibodies in nasal washes (NW), preF-specific antibodies in sera and NW, and the cellular immunity would provide a more comprehensive evaluation [73], [92]. Ongoing efforts aiming at enhancing the immunogenicity of RSV LAV candidate strains Despite compelling theoretical advantages of intranasal immunization, the clinical development of intranasal vaccines for the “Big 3” respiratory viruses (RSV, SARS-CoV-2, and influenza virus) has not been as successful as anticipated. At present, intranasal vaccine development centers on two main platforms. The first is virus-vectored vaccines such as those based on modified vaccinia Ankara virus (MVA), bovine/human parainfluenza virus (PIV) type 3 (rB/HPIV3), adenovirus, and influenza virus. However, very few have reached the market. To date, dNS1-RBD (a live-attenuated influenza virus-vectored vaccine) and BBV154 (an adenovirus-vectored vaccine) received emergency use authorization in China and India, respectively [166], [167], [168], [169], [170]. Notably, due to failure to meet the predefined efficacy threshold (≥ 30 %) following homologous prime-boost in Phase 3 trial, dNS1-RBD was suggested to serve as a primary immunization or heterologous booster [171]. The intranasal BBV154 candidate appeared superior in inducing serum neutralizing antibody titers compared to its intramuscular counterpart in Phase 3 study [169]. The second platform comprises LAVs, specifically targeting RSV and influenza viruses. Currently, FluMist® and Ganwu® are the only two licensed intranasal live-attenuated influenza vaccines (LAIVs) available for respiratory viral infections. FluMist® has shown superior efficacy compared with intramuscular inactivated influenza vaccines (IIV) in children aged two years and older. For this age group, the mucosal IgA, T-cell responses, and innate immunity elicited by intranasal LAIV appear as crucial for protection as hemagglutinin-inhibiting (HAI) antibodies [99], [172], [173], [174]. Ganwu®, a product developed by China′s BCHT Biotech, is approved for children and adolescents aged 3 to 17. Regarding the intranasal RSV LAVs, although none are currently approved, candidates engineered by deleting the M2-2 gene or its modified variants warrant further clinical evaluation due to their enhanced immunogenicity. Notably, the genetic background of the parent strain significantly influences the balance between the attenuation phenotype and immunological phenotype in the M2-2-deleted RSV LAVs. For example, an M2-2-deleted LAV derived from the MSA1 clinical isolate was less attenuated than its A2-derived counterpart, suggesting that the A2 genome may harbor strain-dependent attenuating sequences not found in clinical isolates [134]. Furthermore, even subtle nucleotide differences in the A2 backbone can affect the attenuation phenotype of different RSV LAVs deriving from M2-2 deletion, as described above. Therefore, exploring alternative viral backbones represents a rational approach to optimizing vaccine performance. For example, the Long strain, isolated in 1957 and sharing the GA1 genotype as A2, possesses unique attributes that render it an attractive parental backbone. In addition to priming a Th1-biased response, the Long strain can elicit robust IFN-α production in human dendritic cells, underscoring an immunological profile distinct from that of A2 strain [175], [176], [177]. Recent findings have established proof-of-concept by incorporating well-characterized cp and ts mutations or introducing an M2-2 deletion, into the Long strain backbone. In a murine model, these modifications yielded attenuation and immunogenicity profiles similar to those observed for A2-based candidates, although validation through direct head-to-head comparisons remains warranted [110], [178]. Meanwhile, the emergence of currently dominant circulating strains such as ON1 and BA9, which harbor duplications in their G protein′s mucin-like domains, adds another layer of consideration for RSV LAV development. These genotypic changes likely provide a population-level fitness advantage. Furthermore, neutralizing antibody responses in convalescent sera are most potent against homotypic viruses, underscoring the potential benefit of optimizing the spectrum of the induced protective immune response by matching candidate LAVs to currently circulating strains [51], [179], [180]. Moreover, the protection mediated by non-neutralizing antibodies via Fc-effector functions, as noted above, coupled with the antigenic divergence in G protein between the circulating strains and historical prototype strains (e.g., A2), warrants the development of RSV LAVs based on circulating strain backbones or their corresponding neutralizing antigens. Other strategies involve N6-Methyladenosine (m6A) modification of RSV RNA during RSV replication and the resulting escape from host IFN-mediated antiviral responses, which may serve as targets for improving immunogenicity of RSV LAVs [181], [182], [183]. Furthermore, minor variations in F and L proteins can profoundly modulate the in vitro and in vivo characteristics of RSV LAVs, including viral replication kinetics, genetic stability, preF expression, pathogenicity, immunogenicity, and immune polarization [135], [184], [185], [186]. In summary, the next generation of RSV LAVs poised for clinical development hold the potential for significantly enhanced immunogenicity by integrating M2-2 deletion, alternative backbones (such as Long or ON1 strains), and nucleotide or protein modifications, either alone or in combination. Combination vaccine strategies Beyond mucosal vaccines, combination vaccines represent another highly attractive strategy for vaccine development, aiming to provide simultaneous protection against multiple respiratory pathogens that exhibit seasonal co-circulation—such as the aforementioned “Big 3” respiratory viruses, human metapneumovirus (HMPV), and HPIV. To this end, the concept of using RSV LAVs as vectors for heterologous antigens (e.g., SARS-CoV-2 spike) is theoretically quite attractive; however, the inherent physical instability of RSV, coupled with its typically low in vitro replication titers, poses significant technical hurdles, which may explain why such approaches remain underdeveloped. For instance, Meissa′s RSV-based COVID-19 candidate vaccine, in which the RSV gene was replaced with the SARS-CoV-2 spike gene, entered Phase 1 trials (NCT04798001) in 2021, but no further data have been released since then. Instead, viral backbones derived from viruses such as rB/HPIV3 and HMPV have been successfully engineered to express RSV F protein, demonstrating the feasibility of chimeric and multiple vaccines on more robust platforms [187], [188], [189], [190]. Notably, certain chimeric live virus vectors expressing RSV preF such as rB/HPIV3-RSV-pre-F can be administered intranasally and replicate efficiently, even in the presence of pre-existing RSV immunity. Therefore, they are also ideal candidates for a heterologous prime-boost strategy to enhance RSV immunity further. Consequently, using vectors such as rB/HPIV3, rather than RSV, to develop combination respiratory vaccines may be more feasible. Monitoring of immune escape following immunization Although the lifelong frequency of RSV reinfection mirrors that of other respiratory viruses, such as influenza virus, rhinovirus, and SARS-CoV-2, it is frequently triggered by the same viral strain with minimal antigenic drift. This unique reinfection profile sharply distinguishes RSV from these viral counterparts [8], [145], [191]. Importantly, this implies that the pronounced antigenic drift and subsequent immune escape that compromise COVID-19 and influenza vaccine efficacy do not pose similar challenges for RSV. Instead, RSV vaccine development must overcome distinct immunological barriers, as exemplified by the emergence of the preF subunit vaccine—the first successful application of structure-based vaccine design. Furthermore, the F protein exhibits remarkable evolutionary stability; its extracellular domain differs by only 25 amino acid residues between subtypes A and B, despite prolonged exposure to host immune pressure [192]. Supportive data also derive from long-term clinical experience with palivizumab, a mAb directed against antigenic site II of the F protein. Although mAb-resistant variants have been identified in approximately 5 % of breakthrough infection cases, these viruses exhibit impaired fitness compared to wild-type RSV [193]. A particularly striking example is suptavumab (REGN2222), a site V-specific mAb that failed in Phase 3 clinical trials. This failure occurred because the predominantly circulating RSV-B strains at the time harbored naturally occurring resistance mutations independent of antibody-exerted selective pressure [102], [194], [195], [196]. Predictably, rare and sporadic mutations conferring escape from nirsevimab have been observed during clinical trials and following its widespread rollout; whether these resistant variants will undergo positive selection and spread further over time warrants vigilant, long-term genomic surveillance [27], [194], [197], [198], [199], [200]. Two non-mutually exclusive hypotheses may provide key mechanistic insights into the aforementioned phenomenon: first, the F protein exhibits inherently low mutational tolerance due to structural constraints; and second, the vaccine-induced polyclonal antibody response likely targets a myriad of distinct epitopes, thereby precluding any single mutation from conferring a selective advantage upon the virus [194]. Nevertheless, given that mAb-resistant mutations do occur, the future approval and deployment of RSV LAVs will further underscore the critical necessity for continuous molecular surveillance to track the virus's evolutionary trajectories of circulating RSV strains. Global vaccination challenge The profound inequity in the global burden of RSV disease between LMICs and high-income nations underlines the imperative need for concerted international action and strategic collaborations. Globally, over 95 % of RSV-associated LRTI cases and 97 % of RSV-attributable deaths in young children occur in LMICs [2]; notably, 99 % of these deaths are concentrated in low-income countries [1], [201], [202], a stark contrast to the current accessibility and distribution of effective interventions. The rollouts of maternal RSV vaccination (Abrysvo) and the long-acting mAbs (nirsevimab and clesrovimab) have been disproportionately concentrated in high-income countries and select middle-income nations [203], [204], [205], [206]; however, these prophylactic interventions remain largely inaccessible across LMICs—precisely the regions where the burden of severe RSV infection is most acutely concentrated. This disparity in vaccine and mAb access not only perpetuates existing disparities in disease burden but also threatens to further exacerbate the global health divide. Therefore, the ultimate challenge lies not merely in scientific development of effective prophylactic tools, but—more importantly—in securing global equity in their translation and distribution. Intranasal RSV LAVs offer production cost advantages over mAbs; furthermore, their needle-free mucosal delivery not only streamlines immunization logistics but also minimizes injection-related distress, thereby enhancing clinical compliance across pediatric populations. Consequently, this mucosal immunization platform holds distinct logistical and economic promise for mitigating the global health divide. Conclusions While novel vaccine platforms such as mRNA and recombinant protein technologies have successfully mitigated RSV disease in older adults and young infants < 6 months via maternal immunization, they present a dual challenge. When administered directly to RSV-seronegative infants and young children, they carry the risk of immunopathological priming for VAERD; conversely, when utilized for maternal immunization, the resulting passive protection gradually wanes during the first year of life [121]. In contrast, RSV LAVs, built upon a well-established technological platform, offer a highly viable and safer intervention for this vulnerable pediatric cohort. Beyond eliciting robust systemic and mucosal immunity that confers durable protection, their paramount advantage lies in their well-documented inability to induce VAERD. Although balancing safe viral attenuation with robust efficacy remains a persistent challenge, continuous scientific innovation and cross-border cooperation will likely open new avenues to close this critical prophylactic gap, ultimately alleviating the staggering global burden of the disease. In doing so, these live-attenuated platforms stand at the forefront of driving the next wave of transformative public health impact, reshaping the landscape of pediatric immunization worldwide.