A total of 3,052 publications were identified after data screening. As demonstrated in Figure 2, the annual global output from 1995 to 2024 reflected a fluctuating upward trend, indicating the increasing attention of this field over the past 3 decades. The number of annual global output was below 100 during 1995–2009. The number of publications began to exceed 100 in 2010 (n = 104), and subsequently peaked in 2021 with 263 publications.

A total of 86 countries have published studies related to cannabinoids and endocannabinoids in cancer. Figure 3 depicted the overview of countries on this research theme. In Figure 3A, the size of the node reflected the scientific outputs of countries, while the color represented their total citations. The top three productive countries were the USA (n = 892), Italy (n = 447), and China (n = 319), whereas the top three most-cited countries were the USA (n = 52,754), Italy (n = 27,837), and the UK (n = 16,979). The visualization of country collaboration networks and academic activities were illustrated in Figure 3B. The node size indicated the link strength between countries, whereas the color represented their academic activity. We found the thickest link was between the USA and Italy, reflecting their close collaborations. Concerning academic activity, Sweden exhibited an earlier contribution in this field, whereas South Africa displayed a more recent activity.

A total of 3,362 institutions participated in this research domain, with Italian National Research Council (n = 93), Complutense University of Madrid (n = 87), and University of Naples Federico II (n = 61) ranked as the leading three most productive institutions (Table 1). As summarized in Table 2, Complutense University of Madrid (n = 9,983) was the institution with the highest citations, followed by Italian National Research Council (n = 9,418) and University of Naples Federico II (n = 5,329).

In this study, we found 14,087 authors contributed to the research of cannabinoids and endocannabinoid system in cancer. Figure 4 depicted the co-authorship visualization atlas. The outputs of authors were indicated by the size of the nodes, while the connecting lines between nodes denoted collaboration strength among authors. The color of nodes indicated distinct collaboration clusters (Figure 4A) or the academic activities of authors (Figure 4B). Vincenzo Di Marzo was one of the most fruitful scholars in this field, ranking first in publication (n = 61) and total citation (n = 4,364). Moreover, Vincenzo Di Marzo, Maurizio Bifulco, Mauro Maccarrone, Manuel Guzman, and Marco Macchia were the leading scholars in this domain and formed close collaborations.

In total, we detected 1,023 journals published literature in this theme. Table 3 listed the leading journals in publications. The most fruitful journal in this research area was International Journal of Molecular Sciences (n = 119), followed by British Journal of Pharmacology (n = 60) and Molecules (n = 59). There were eight of the top 10 most productive journals located in Q1 region according to the Journal Citation Reports 2023, and British Journal of Pharmacology had the highest impact factor (IF) of 6.8. As shown in Table 4, British Journal of Pharmacology (n = 6,690), International Journal of Radiation Oncology Journal of Biological Chemistry (n = 3,550), and European Journal of Pharmacology (n = 2,471) were the top three journals in citations. Moreover, seven of the top 10 most-cited journals were distributed in Q1 region, and Cancer Research had the highest IF of 12.5. Together, these results of the journal analysis indicated their high relevance and quality in the field. Table 5 presented the top 10 most-cited documents in this field. These documents were published between 2003 and 2022, and they were all cited more than 500 times. The review article titled “The endocannabinoid system as an emerging target of pharmacotherapy” published in Pharmacological Reviews by Pál Pacher et al. had the highest citation of 1,570.

A total of 11,137 relevant keywords were collected, and the top 30 keywords with the strongest occurrence were demonstrated in Figure 5. The minimum threshold of keywords was adjusted to 120 to ensure readability of the graph. The keyword “cannabinoids” had the highest occurrence and the strongest link strength. The keyword with the earliest appearance was “anandamide”, and the newly emerged keyword was “cannabidiol”. The identified keywords were clustered into three distinct groups labeled by different colors. The red cluster encompassed keywords such as “endocannabinoid system”, “anandamide”, “cannabinoid receptor”, “endocannabinoids”, “inflammation”, “oxidative stress”, “cb2 receptor”, and “acid amide hydrolase”. The blue cluster included keywords of “cannabinoids”, “cannabidiol”, “cancer”, “cannabis”, “marijuana”, “neuropathic pain”, and “delta-9-tetrahydrocannabinol”. In the green cluster, “activation”, “expression”, “apoptosis”, “inhibition”, “receptor”, “cb1”, and “proliferation” were the main nodes included. The average publication year of the keywords demonstrated that the research frontiers were shifted from basic functional characterization of cannabinoid and endocannabinoid (e.g., “anandamide” and “acid amide hydrolase”) as well as cannabinoid receptors (e.g., “cb1” and “cb2 receptor”), to mechanisms of diseases (e.g., “neuropathic pain”, “proliferation”, “apoptosis”, and “oxidative stress”), and ultimately the translational applications within multidisciplinary framework of molecular biology, pathophysiology, pharmacology, immunology, and oncology.

In this study, we conducted a comprehensive bibliometric analysis in research concerning cannabinoids, endocannabinoid system, and cancer. The increasing tendency in global output reflected the growing recognition of this field, which was likely driven by the rapid developments in disciplines of physiology, pharmacology, biochemistry, and oncology. However, in recent years, the advances in anticancer therapies, including immune checkpoint inhibitors (ICIs), antibody-drug conjugates, and chimeric antigen receptor T cells, have attracted a substantial amount of attention and resources. In contrast, the limited progress in new target and unelucidated clinical efficacy of cannabinoids might have led to the recent slowdown in the development of this field (Pramesh et al., 2022; Skórzewska and Gęca, 2024; Chen et al., 2022).

As for productive countries, although the USA led in national contributions, Italy was more prominently in institutional output, possessing half of the top 10 institutions. A more in-depth analysis revealed that the majority of the top 100 institutions (n = 33) were still located in the USA, exhibiting a scattered and evenly distributed institutional output. This pattern might be attributed to differences among states in funding sources, research scale, and cannabis-related legal restrictions (Dickson et al., 2019; Cooper et al., 2021). Meanwhile, due to cultural acceptance and regulatory differences, cannabis research was more prevalent in European countries, such as Italy and Spain (Krcevski-Skvarc et al., 2018; Ransing et al., 2022). As the core of scientific research system in Italy, the Italian National Research Council plays a fundamental role in coordinating research activities among institutions and has contributed significantly to the research output (Tuzi, 2005). On the other hand, although Spain ranked sixth in the number of publications (n = 212), the Complutense University of Madrid held second in institutional output (n = 87), contributing 41.0% of the national production. The fruitful researchers, including Manuel Guzman, Guillermo Velasco, Mar Lorente, Cristina Sánchez, and Maria Salazar Roa, have formed close collaborations and contributed significantly to the academic status of this institution (Velasco et al., 2012; Maccarrone et al., 2014; Guzmán, 2003; Velasco et al., 2016a).

From the perspective of author contribution, the research groups led by Vincenzo Di Marzo, Maurizio Bifulco, Mauro Maccarrone, Manuel Guzman, and Marco Macchia, excelled in this domain and greatly advanced its development. Vincenzo Di Marzo pioneered in this topic and provided profound insights into the metabolic pathways, physiological functions, and anticancer actions, of endocannabinoids (Di Marzo, 2008; Marzo et al., 2004). Maurizio Bifulco had close connection with Vincenzo Di Marzo and primarily investigated the endocannabinoid system, cannabinoid receptor, cannabinoid, and their roles in cancer (Ligresti et al., 2003; Ligresti et al., 2006a; Pisanti et al., 2017). Mauro Maccarrone was a key scholar in cannabinoid research, and focused on endocannabinoid signaling, lipoxygenase pathway, and neuroinflammation (Maccarrone et al., 2000; Maccarrone et al., 2015). Manuel Guzman made fruitful achievements in the mechanisms of cannabinoid and cannabinoid receptors in oncology and neurology, and collaborated closely with Mauro Maccarrone (Velasco et al., 2012; Maccarrone et al., 2014; Guzmán, 2003). Marco Macchia was active more recently in this field, and concentrated on cannabinoid receptor-targeting pharmacological strategies in cancer treatment (Manera et al., 2015; Bertini et al., 2016).

Literature on this topic tended to be published in specialized journals of molecular sciences, pharmacology, biochemistry, and oncology. Moreover, the majority of journals were Q1 journals, and ranged 2.9 to 12.5 in IF. These findings indicated the high quality of the literature, and would be helpful for researchers seeking the core journals in the field. Besides, the most-cited publication was the review by Pál Pacher et al. titled “The endocannabinoid system as an emerging target of pharmacotherapy”. This review comprehensively summarized the knowledge status of the endocannabinoid system as a target of pharmacotherapy (Pacher et al., 2006).

Publications on anticancer effects of cannabinoids and endocannabinoid system have accumulated over the last few decades. In the early 1970s, by using Lewis lung adenocarcinoma mice model, Munson et al. first discovered that orally administration of tetrahydrocannabinol inhibited tumor growth, suggesting the antiproliferative effects of cannabinoids (Munson et al., 1975). In the 1980s, Howlett et al. systematically introduced the cannabinoid receptors in their identification, distribution, and pharmacological properties (Howlett et al., 2002). The research of cannabinoid regulation of cell growth signaling was advanced after the characterization of cannabinoid receptors. From the 1990s, numerous findings from preclinical animal studies and human clinical work revealed the antitumor effects of endogenous cannabinoids and synthetic cannabinoids among various cancers, including lymphoma, pancreatic cancer, glioma, breast cancer, and prostate cancer (Carracedo et al., 2006a; Blázquez et al., 2006; Ligresti et al., 2006b). Follow-up studies from the 2000s focused on the mechanisms of cannabinoid-induced apoptosis and growth inhibition (Galve-Roperh et al., 2000a). Nowadays, the interest in cannabinoids and endocannabinoid system have increased due to their effects of antibacterial, anti-inflammatory, anti-anxiety, and neuroprotective (Luz-Veiga et al., 2023; Aljobaily et al., 2022). However, considering the psychoactivity and side effects of cannabinoids (Irrera et al., 2021; Abrams, 2016), their pharmacological potential is yet to be fully realized.

The endocannabinoid system is a vital mechanism controlling multiple cellular growth and developmental processes in organisms (Moreno et al., 2019). Dysregulation of this system may result in abnormal proliferation, vascularization, and tumor invasion (Cherkasova et al., 2022). The endocannabinoid system consists of G protein-coupled cannabinoid receptors, endocannabinoids, and enzymes that produce and degrade endocannabinoids (Fonseca et al., 2018). On the other hand, cannabinoids are compounds derived from the marijuana plant, and showed inhibitory effects of tumor proliferation, invasion, metastasis, and angiogenesis (Ramer et al., 2021; Velasco et al., 2012). Delta-9-tetrahydrocannabinol is the major component of phytocannabinoid, responsible for the psychoactive effects of marijuana (Maia et al., 2023). Previous reviews have summarized the evidence from animal experiments, clinical studies, and epidemiological research, indicating that non-psychoactive cannabinoids, such as cannabidiol, has potential therapeutic effects of anti-inflammatory, analgesic, and anxiolytic (Pertwee, 2006; Blessing et al., 2015). Besides phytocannabinoids, endocannabinoids also have the ability to activate cannabinoid receptors (Fonseca et al., 2013). The two well-known endocannabinoids are anandamide and 2-arachidonoylglycerol (Lim et al., 2023), and they are closely related to the regulation of pain, emotion, and immunity (Popescu-Spineni et al., 2022).

Effects of cannabinoids are thought to be mediated via 2 G protein-coupled receptors, namely, cannabinoid receptors type 1 (CB1) and cannabinoid receptors type 2 (CB2) (Howlett, 1995). The former is primarily expressed in central nervous system, whereas the latter is mostly found in the peripheral nervous system and immune cells (Devane et al., 1992). Moreover, an earlier in vitro study has reported the antitumor effect of cannabinoids via transient receptor potential vanilloid type 1 (TRPV1), reflecting the attractive research prospect of selective targeting TPRV1 (Fonseca et al., 2018). Aside from the direct activation of cannabinoid receptors, another approach is to increase endocannabinoid levels by inhibiting endocannabinoid-degrading enzymes (Schwarz et al., 2018). Therefore, we envision that selective inhibitors of endocannabinoid degrading enzyme would have promising research potential.

Numerous findings from rodent and in vitro models have suggested that cannabinoid receptors exert anticancer effects through diverse mechanisms, such as cell death stimulation, cell proliferation inhibition, and angiogenesis inhibition (Garmpis et al., 2022). Several pathology studies using surgical specimens from cancer patients have indicated that upregulated levels of CB1 or CB2 were associated with reduced survival and increased tumor metastasis and recurrence (Hashemi et al., 2020; Klein Nulent et al., 2013; Larrinaga et al., 2013; Wu et al., 2012). Other findings from clinical immunohistochemistry studies and rodent model experiments reported that higher CB1 and CB2 levels were linked to poor clinical outcomes in patients of renal cell carcinoma (Wang et al., 2018), head and neck squamous carcinoma (Klein et al., 2013), breast cancer (Blasco-Benito et al., 2019), and colorectal cancer (Martínez-Martínez et al., 2015). However, others reported the association between CB1/CB2 upregulation and longer survival time in hepatocellular carcinoma (Xu et al., 2006) and non-small cell lung cancer (Preet et al., 2011; McAllister et al., 2021). Thus, these controversial findings on the effects of cannabinoid receptors in cancer might suggest the necessity to distinguish exact cell types, such as tumor tissue, molecular tumor subtype, or immune cells, in order to explore more accurate prognostic property from the parameters. Further endeavors are required to thoroughly evaluate the potential role of cannabinoid receptors as biomarkers in cancer prevention and treatment.

Cannabinoids have been applied clinically to treat various cancer-associated symptoms, such as nausea, vomiting, and cancer pain, which brought additional benefits for cancer patients and their quality of life (Tramèr et al., 2001; Khasabova et al., 2012). As for antitumor properties of cannabinoids, several mechanisms have been proposed, including cytostatic effects, apoptosis induction, and inhibition of neo-angiogenesis (Bifulco et al., 2006). Extensive studies using animal tumor models have revealed an association between cannabinoid-induced autophagy and apoptosis (Lorente et al., 2011). Salazar et al. first demonstrated that tetrahydrocannabinol induced autophagy in glioma cells via CB1 phosphorylation and endoplasmic reticulum stress (Salazar et al., 2009). Other preclinical work revealed that classic anticancer agents may benefit from such endoplasmic reticulum stress and autophagy inhibitors (Sui et al., 2013; Xu et al., 2014). In addition to cell death promoting effects, cannabinoids also showed antiangiogenic effects by inhibition of vascular endothelial growth factor receptors (VEGFR), VEGFR one and VEGFR2 (Casanova et al., 2003). Likewise, synthetic cannabinoids exhibited sustained inhibition of angiogenic process rather than inducing angiogenesis (Maia et al., 2023).

Besides, cannabinoids have been reported to greatly affect immune cells by altering multiple genes involved in immune response, immune cell apoptosis, and cell proliferation (Hu et al., 2020). Findings from immunodeficient mice model have shown a strong antitumor effect of the synthetic cannabinoid (Blázquez et al., 2006). The anti-inflammatory and immunosuppressive effects of cannabinoids have prompted the combination therapy with ICIs, though the clinical outcomes remained unclear. A few clinical studies strongly suggested that medicinal cannabis can improve the tolerability of cancer patients to ICIs treatment (Vigano et al., 2025), while no significant benefits on overall survival or progression free survival were noted (Bar-Sela et al., 2020; Biedny et al., 2020). On the contrary, some studies have indicated that the concurrent use of cannabinoids with ICIs can induce adverse effects on clinical outcomes (Sarsembayeva and Schicho, 2023).

The endocannabinoid system has emerged as a promising new target for pharmacological intervention in growth suppression and apoptosis induction in many cancer cells (Prather et al., 2013; Galve-Roperh et al., 2000b; Carracedo et al., 2006b; Massi et al., 2004; Tomko et al., 2019). Of note, although cannabinoids were found to display apoptotic effects in breast cancer cells (Caffarel et al., 2012), others studies reported protumorigenic effects of cannabinoids in breast cancer (McKallip et al., 2005; Zhu et al., 2000). Due to the complexity of endocannabinoid system in regulating cancer biology pathways, it may exert different effects on antiproliferative, antiangiogenic, anti-metastatic, and anti-inflammatory, which might explain its uncertainty in anticancer treatments (Mandrika et al., 2010). Interestingly, regarding lung cancer, although cannabis contains similar toxins and carcinogens to tobacco, there is currently no conclusive evidence that cannabis is associated with an increased risk of lung cancer (Jett et al., 2018; Yarlagadda et al., 2019). On the other hand, while some clinical evidence supported the benefits of cannabis in the treatment of lung cancer, it remained debatable to be included in official cancer treatment guidelines (Smith et al., 2015).

Regarding combined therapies, the combined use of cannabinoids was found to enhance the antiproliferative effects of various classical chemotherapeutic agents, including temozolomide, gemcitabine, paclitaxel, and of 5-fluorouracil (Shrivastava et al., 2011; Donadelli et al., 2011; Miyato et al., 2009; Gustafsson et al., 2009). However, the pharmacokinetic properties of cannabinoids need to be carefully considered as well. Previous findings have indicated that cannabinoids inhibited cytochrome P450 activity, which may consequently increase the toxicity and reduce the efficacy, of chemotherapeutic agents (Jiang et al., 2013; Yamaori et al., 2011). Therefore, the interaction between chemotherapeutic agents and cannabinoids would be a noteworthy issue that requires further clarification in future clinical studies. Additionally, the interaction between cannabinoids and ICIs has also gained attention. Clinical trial evidence has revealed the immunosuppressive effects of cannabinoids during concurrent nivolumab treatment (Taha et al., 2019; Bar-Sela et al., 2020). Thus, careful consideration should be given on patients treated with ICIs and cannabinoids. In targeted therapies, Velasco et al. found that cannabinoids is able to enhance the antineoplastic activity of targeted therapies by regulating protein kinases, such as anaplastic lymphoma receptor tyrosine kinase, epidermal growth factor receptor kinase, and extracellular signal-regulated kinase, demonstrating promising anticancer potential (Velasco et al., 2016b).

Interestingly, our results of keywords analysis revealed that several hotpots were more frequently occurred in recent years. First, with the legislation changes in the USA that medicinal use of cannabis has become more normalized, its potential in managing cancer-related symptoms and improving quality of life for cancer patients are becoming particularly attractive these years (McClure et al., 2023; Amin et al., 2024). Moreover, there has been growing evidence supporting the antitumor role of cannabis in an expanding list of cancers (Alsalamat et al., 2024). Although cannabidiol and tetrahydrocannabinol are the primary chemical compounds derived from cannabis, a scientific favor of cannabidiol over tetrahydrocannabinol was observed due to the multi-targeted property and less psychotropic effects of cannabidiol (Maayah et al., 2020; Mashabela and Kappo, 2024). Furthermore, evidence suggested that cannabidiol promotes apoptosis in cancer cells by targeting oxidative stress through oxidative cellular damage (Mokoena et al., 2022). Hence, cannabidiol would be a promising non-intoxicating target which is crucial for optimizing the therapeutic potential in cancer treatment.

Overall, cannabinoids and endocannabinoid system have demonstrated potential therapeutic benefits for cancer. However, challenges have simultaneously emerged due to their extensive and complex influence on cancer progression (Skórzewska and Gęca, 2024). One of the main challenges lies in the step of clinical translation. Current evidence from clinical trials suggested that studies of cannabinoids and endocannabinoid system primarily focused on the management of cancer and/or treatment-related symptoms, with variable quality of evidence (Skórzewska and Gęca, 2024; Häuser et al., 2023). Therefore, it is necessary to conduct well-designed and high-quality clinical trials in the future. In addition, it would be of great value to combine preclinical mechanisms of action into clinical practice, such as optimal ratio of cannabinoid agents with chemotherapy and ICIs therapy (Vigano et al., 2025). Importantly, we should emphasize that the legislative issues with cannabis have constrained the collection of human trial data, thereby limiting our understanding of its side effects and drug interactions (Hayry, 2004).Thus, clinicians should carefully weigh the legal and ethical issues, such as informed patient consent and potential risks.

This study should be interpreted within its limitations. First, although the use of single database facilitated our result management, the comprehensiveness of data might be potentially limited. Currently, data merging of multiple databases is often manual and requires significant effort. Future research utilizing newly developed open source tools would be highly beneficial for integrating the essence of multiple databases (Nikolić et al., 2024). In addition, due to the intrinsic limitation of bibliometrics that recent high-quality documents generally have low citation counts, these studies might be underappreciated. Future studies combining alternative metric method, such as early citation trends, would better capture the impact of emerging research in a timely manner (Wang et al., 2013).