Evaluation of Drug–Drug Interactions Using Drug-PIN® in Patients Receiving Avelumab Maintenance for Advanced Urothelial Carcinoma: Evidence from the MALVA Study (Meet-URO 25) Authors: Giandomenico Roviello, Mattia Alberto Di Civita, Daniele Santini, Elisabetta Gambale, Lidia Strigari, Simone Scagnoli, Laura Pappalardo, Andrea Torchia, Andrea Botticelli, Serena Pillozzi, Lorenzo Antonuzzo Published: 2026 ABSTRACT AbstractIntroductionPolypharmacy is common in patients with advanced urothelial carcinoma (UC) receiving avelumab maintenance therapy, raising concerns about potential drug–drug interactions (DDIs).MethodsWe retrospectively analyzed patients enrolled in the multicenter MALVA study treated with avelumab maintenance. Concomitant medications were assessed using the Drug Interaction and Pharmacogenomics Information Network (Drug-PIN®). Drug-PIN® scores were calculated with and without avelumab and correlated with overall survival (OS) and progression-free survival (PFS).ResultsAmong 115 patients, Drug-PIN® scores showed wide inter-patient variability but were not significantly influenced by inclusion of avelumab. No significant differences in OS or PFS were observed across Drug-PIN® risk categories.ConclusionIn this real-world cohort, polypharmacy and Drug-PIN® score did not impact survival outcomes, suggesting that, within the limitations of this exploratory analysis, Drug-PIN®-assessed DDI burden was not associated with survival outcomes in patients receiving avelumab maintenance therapy.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s40487-026-00448-5.Keywords: Avelumab, Urothelial carcinoma, Maintenance immunotherapy, Drug–drug interactions, Polypharmacy, Drug-PIN® Key Summary Points Why carry out this study?Patients with urothelial carcinoma (UC) receiving avelumab maintenance are frequently exposed to multiple concomitant medications due to age-related comorbidities in real-world clinical settings.The impact of drug–drug interaction (DDI) burden in this population remains uncertain.The study explored whether DDI burden, as assessed using the Drug Interaction and Pharmacogenomics Information Network (Drug-PIN®), influences clinical outcomes and whether inclusion of avelumab contributes to interaction risk.What was learned from the study?Among 115 patients, the burden of DDIs showed substantial inter-patient variability, and was not significantly influenced by the inclusion of avelumab.No significant association was observed between Drug-PIN® score and clinical outcomes, as overall survival and progression-free survival were comparable across different interaction risk categories.DDI burden assessed by Drug-PIN® was not associated with survival outcomes; however, while the study cohort reflects the known male predominance of UC, further studies may help explore potential sex- and gender-related differences. Introduction Urothelial carcinoma (UC) represents one of the most common malignancies worldwide, with a markedly higher incidence in men than in women, accounting for approximately three to four times more cases in male individuals [1]. For years, platinum-based chemotherapy has been the standard of care for locally advanced or metastatic UC [2]. Recently, the treatment of advanced UC has undergone significant changes due to the immune checkpoint inhibitors (ICIs) and antibody–drug conjugates. Indeed, UC has high genomic instability and programmed death-ligand 1 (PD-L1) protein expression; these features make it sensitive to immunotherapy [3]. Avelumab, a human IgG1 monoclonal antibody directed against the PD-L1, is used as maintenance therapy after 4–6 cycles of platinum-based first-line chemotherapy in patients without disease progression, regardless of PD-L1 expression [4]. Despite these results, outcomes with avelumab maintenance vary considerably among patients. In fact, clinical or pharmacological factors may influence clinical outcomes [5]. On the basis of the pharmacological characteristics of monoclonal antibodies, we hypothesized that avelumab would have a minimal impact on drug–drug interaction (DDI) burden. In parallel, we explored whether baseline polypharmacy and interaction burden could be associated with clinical outcomes. Considering the average age of patients with UC, approximately 70 years, and increasing life expectancy, the management of this disease represents a significant challenge [6]. The scenario is further complicated by the fact that patients often take medications for various comorbidities (diabetes, hypertension, etc.) that may interact with each other [7]. In fact, DDIs, either pharmacokinetic (absorption, metabolism, and elimination) or pharmacodynamic (synergistic or antagonistic), affect the efficacy of a treatment or exacerbate adverse events [8, 9]. This problem affects more than one-third of patients with cancer undergoing treatment [10]. Drug-PIN® [11] can help clinicians identify drug interactions, provide clinical recommendations, integrate genetic data, and support treatment personalization. In this study, data on the baseline concomitant medications taken by patients treated with avelumab in the MALVA trial were gathered and then analyzed using Drug-PIN® with the aim to assess potential interactions between avelumab and regular drugs. The aim of the present study was to evaluate the impact of polypharmacy and the overall burden of DDIs, as assessed by the Drug-PIN®, on clinical outcomes in patients with advanced urothelial carcinoma receiving avelumab maintenance therapy in a real-world setting. In addition, we sought to explore the specific contribution of avelumab to the overall DDI profile by comparing Drug-PIN® scores calculated with and without the inclusion of immunotherapy. The study then examined how the Drug-PIN® score could influence avelumab treatment in terms of overall survival (OS) and progression-free survival (PFS). Methods Study Design and Patient Population This study is a retrospective analysis conducted within the observational, multicenter MALVA study, which enrolled patients with advanced UC treated with avelumab as maintenance therapy after achieving stable disease or an objective response following first-line platinum-based chemotherapy. The design of the MALVA study and its primary clinical outcomes have been previously reported [12, 13]. Eligible patients were adults with locally advanced or metastatic UC who initiated avelumab maintenance according to standard clinical practice. For the present analysis, demographic, anthropometric, and clinical variables were extracted from the study database, including age, sex, height, weight, and metastatic sites at diagnosis. In addition, detailed information on baseline concomitant medications, defined as drugs being taken at the time of avelumab initiation, was collected whenever available and constituted the basis for the DDI analysis. This study was approved by the local Ethics Committee (Tuscany Area Vasta Centro, protocol number 20817_oss), which served as the coordinating (“master”) ethics committee for the study. Given the retrospective nature of the analysis, clinical data were collected and analyzed in anonymized form, and no directly identifiable patient information was accessed by the investigators. All patients had previously provided written informed consent for the use of their clinical data for research purposes. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki (1964) and its subsequent amendments. Assessment of Concomitant Medications Baseline concomitant medications were retrieved from the available clinical records at the time of avelumab initiation. Because medication data were not uniformly available for all patients included in the MALVA cohort, the Drug-PIN® analysis was restricted to patients with sufficiently complete baseline pharmacological information. Concomitant medications included all non-oncological drugs reported in the medical records and were categorized according to therapeutic class. No restrictions were applied with regard to drug class, dosage, or duration of treatment. For each patient, the total number of concomitant medications was recorded and summarized descriptively. Only patients with complete information on baseline concomitant therapies were included in the Drug-PIN® analysis. Patients reported as having no concomitant medications were included in the descriptive analysis; however, only patients with complete and reliable pharmacological data were included in the Drug-PIN® analysis. Because medication data were retrospectively extracted from routine clinical records and no formal standardized medication reconciliation process was prospectively implemented, completeness of concomitant medication capture could not be guaranteed. Drug-PIN® Analysis Potential DDIs were evaluated using Drug-PIN® (Drug Interaction and Pharmacogenomics Information Network), a clinical decision support system designed to assess interactions among multiple drugs and to integrate them with patient-specific demographic, clinical, and biochemical characteristics, including age, body weight, and liver and renal function parameters [8]. The output of the Drug-PIN® software consists of a numerical score (Drug-PIN® score) reflecting the overall interaction burden, along with a corresponding color-coded risk tier (Drug-PIN® tier) indicating the potential clinical relevance of DDIs and the need for therapeutic intervention. Drug-PIN® scores are classified as follows: tier 1 (green), 0–20, indicating no clinically relevant DDIs; tier 2 (yellow), 21–30, indicating low-risk DDIs; tier 3 (dark yellow), 31–70, indicating intermediate- to high-risk DDIs; and tier 4 (red), > 71, indicating high-risk DDIs. As reported in previous studies validating its use in oncology, the software evaluates the pharmacological effectiveness and interactions between drugs (both pharmacokinetic, such as CYP450 induction/inhibition, and pharmacodynamic) through its internal algorithms. Because Drug-PIN® is a proprietary clinical decision support system, the exact mathematical equations and the specific weighting assigned to individual interactions are intellectual property and are not accessible to the users. While this internal calculation process cannot be manually reproduced, the reliability of the system’s output and its clinical utility have been established in prior published literature. According to available evidence, the use of Drug-PIN® improves patient safety by facilitating the evaluation of DDIs in clinical practice and reducing the risk of drug-related toxicity and treatment non-compliance [8, 14–17]. Drug-PIN® is a proprietary clinical decision support system used under authorized access within routine clinical practice. No proprietary algorithms or software components are disclosed in this manuscript. To specifically assess the contribution of avelumab to the overall interaction profile, two Drug-PIN® scores were calculated for each patient. The first score included all baseline concomitant medications together with avelumab, while the second score was calculated using the same medication list excluding avelumab. This approach allowed a direct intra-patient comparison of Drug-PIN® values with and without the inclusion of immunotherapy. Survival Analysis The primary aim of this analysis was to explore whether baseline polypharmacy and drug interaction burden, as assessed by Drug-PIN®, were associated with survival outcomes in patients receiving avelumab maintenance therapy, from the perspective of supporting pretreatment clinical evaluation. OS was defined as the time from initiation of avelumab maintenance therapy to death from any cause or last follow-up. PFS was defined as the time from avelumab initiation to documented disease progression or death, whichever occurred first. Survival outcomes were estimated using the Kaplan–Meier method. Patients were stratified according to Drug-PIN® score categories using different cutoffs, including tier 1 versus tier > 1. Univariate analysis (UVA) was conducted using the Kaplan–Meier curves of PFS and OS. The differences between the Kaplan–Meier curves were assessed using the log-rank test. Patients were stratified according to pre-specified Drug-PIN® risk categories based on the established Drug-PIN® tier classification (tier 1 vs tier > 1). Given the exploratory nature of the analysis and the limited sample size, no outcome-driven optimization of cutoff values was performed. The multivariate Cox proportional hazard regression model (MVA) was not implemented because of the exploratory design of this study and the small sample size. Data were collected in an anonymous database and analyzed with R-package and SPSS v.26. Statistical Analysis Drug-PIN® scores calculated with and without avelumab were compared using the Wilcoxon signed-rank test. Continuous variables were summarized using median and range or mean values, as appropriate, while categorical variables were reported as absolute numbers and percentages. All statistical analyses were performed using standard statistical software, and a two-sided p value < 0.05 was considered statistically significant. Results A total of 115 patients from the MALVA database were evaluated, and their baseline characteristics are summarized in Table 1. Of the 115 patients included in the overall cohort, 78 had sufficiently complete baseline medication data and were therefore included in the Drug-PIN® analysis and for these patients we calculated OS and PFS. Table 1baseline characteristics of the patientsAge, years, median (range)71.5 (54–87)Sex, n (%)Total: 115 Female21 (18.3%) Male94 (81.7%)Weight, kg, mean (range)72.3 (50–115)Height, cm, mean (range)170.2 (150–192)Metastatic at diagnosis47Bone metastases30Liver metastases21Lung metastases40Lymph node metastases84Brain metastases0Other metastases24Histology Adenocarcinoma1 Squamous7 Small cell2 Pure urothelial101Objective response (PR, CR) to platinum59SD to platinum56 baseline characteristics of the patients Overall, Drug-PIN® scores showed wide inter-patient variability, reflecting the heterogeneous burden of comorbidities and polypharmacy in this population. When avelumab was included in the calculation, the median Drug-PIN® score was 9.3 (range 0.5–147.9), whereas the median score calculated without avelumab was slightly higher at 11.5 (range 0.8–148.8). Corresponding mean values were 18.3 and 19.5, respectively. Although a trend toward lower scores was observed when avelumab was incorporated into the algorithm, suggesting that avelumab itself does not substantially contribute to DDI burden, this difference did not reach statistical significance. These findings are consistent with the known pharmacological profile of ICIs, whose clearance mechanisms are associated with a low likelihood of clinically relevant pharmacokinetic interactions. To explore the potential clinical impact of DDIs in patients receiving avelumab maintenance therapy, we assessed the association between Drug-PIN® scores and survival outcomes. When patients were stratified according to pre-specified Drug-PIN® risk categories (tier 1 vs tier > 1), no significant differences in OS or PFS were observed (Fig. 1). Median OS was not reached in both groups (DPS tier = 1 and DPS tier > 1) and hazard ratio was not significant (HR 0.44; 95% CI 0.15–1.26, p = 0.125). In term of PFS, median PFS was 6.3 months for DPS tier 1 and 7.8 months for DPS tier > 1 without significative differences (HR 0.64; 95% CI 0.31–1.32, p = 0.229). Kaplan–Meier estimates of a PFS and b OS stratified by Drug-PIN® risk category. The analysis includes the 78 patients with complete baseline pharmacological data. Patients were stratified into two groups: “DPS = 1” indicates patients with a low interaction risk (tier 1, Drug-PIN® score 0–20, n = 61), whereas “DPS > 1” indicates patients with an intermediate-to-high interaction risk (tier > 1, Drug-PIN® score > 20, n = 17). a For PFS, hazard ratio (HR) 0.64 (95% CI 0.31–1.32), p = 0.229; total events 40 (tier 1) vs 9 (tier > 1). b For OS, HR 0.44 (95% CI 0.15–1.26), p = 0.125; total events 25 (tier 1) vs 4 (tier > 1) Among the 115 patients included in the analysis, information on concomitant non-oncological medications highlighted a heterogeneous pattern of drug use across different therapeutic classes (Table 2). Cardiovascular drugs were the most frequently reported, with beta blockers and statins or other lipid-lowering agents each used by 13 patients. Gastroprotective agents, mainly proton pump inhibitors, were used by 12 patients, while antiplatelet agents were reported in 11 cases. Oral antidiabetic drugs were prescribed to nine patients. Opioids and antigout agents were each used by eight patients. Other antihypertensive treatments included ACE inhibitors and combination antihypertensive therapies, each reported in six patients, as well as calcium channel blockers in five patients and diuretics in two patients. Analgesics or neuromodulators were used by five patients, antipsychotics by four patients, and direct oral anticoagulants and low molecular weight heparins by three patients each. Additional drug classes, including immunosuppressive or antineoplastic agents, urological drugs for benign prostatic hyperplasia, and vitamins or supplements, were each reported in three patients, while vitamin K antagonists were used in one patient. Table 2Baseline concomitant medicationsTherapeutic classNumber of patientsBeta-blockers13Statins/lipid-lowering agents13Gastroprotective agents (PPIs)12Antiplatelets11Oral antidiabetics9Opioids8Antigout agents8ACE inhibitors6Antihypertensives (combos)6Calcium channel blockers5Analgesics/neuromodulators5Antipsychotics4DOACs (direct oral anticoagulants)3Heparins (LMWH)3Immunosuppressants/antineoplastics3Urologicals (BPH)3Vitamins/supplements3Diuretics (incl. combos)2Anticoagulants (VKA)1 Baseline concomitant medications The distribution of the number of concomitant drugs per patient is summarized in Table 3. Thirty-one patients were not receiving any concomitant medications at baseline. Twenty patients were receiving one concomitant drug, four patients were receiving two drugs, and three patients were receiving three drugs. Twenty patients were receiving four or more concomitant medications. Table 3Number of concomitant medications per patientNumber of concomitant drugsNumber of patients03112024334 or more20 Number of concomitant medications per patient Discussion Patients with advanced UC represent a clinical population in which polypharmacy is highly prevalent, driven both by the advanced age at diagnosis and by the burden of chronic comorbidities such as cardiovascular disease, diabetes, and metabolic disorders. In this context, maintenance immunotherapy with avelumab is frequently administered alongside multiple concomitant medications, raising clinically relevant questions regarding the potential impact of DDIs on treatment efficacy and safety [18]. The present analysis provides a real-world evaluation of DDI burden in patients receiving avelumab maintenance therapy, using the Drug-PIN® system within the multicenter MALVA study. The clinical effectiveness of avelumab maintenance in advanced UC has been firmly established by the JAVELIN Bladder 100 trial and further supported by real-world evidence. Recently, the Italian multicenter MALVA study confirmed the effectiveness of avelumab maintenance in routine clinical practice, demonstrating survival outcomes comparable to those observed in the pivotal randomized trial, despite a broader and more heterogeneous patient population. These real-world data underline the robustness of avelumab efficacy across different clinical settings and patient profiles, including older and more comorbid individuals. The clinical relevance of our findings lies in the fact that avelumab does not appear to increase the overall drug interaction burden, supporting its integration into treatment strategies in patients with complex pharmacological profiles. In this setting, Drug-PIN® may still represent a potentially useful tool for structured assessment of baseline polypharmacy, although its prognostic utility was not supported in the present study. Against this background, our analysis focused on whether the burden of concomitant medications, as captured by the Drug-PIN® score, could influence clinical outcomes during avelumab maintenance therapy. Drug-PIN® is designed to integrate information on polypharmacy into a composite score reflecting the overall risk of DDIs. In our cohort, Drug-PIN® scores showed wide inter-patient variability, reflecting the heterogeneity of concomitant medication use observed in everyday practice. However, stratification of patients according to Drug-PIN® score did not reveal any significant differences in OS or PFS. These findings are consistent with the pharmacological characteristics of ICIs. Unlike cytotoxic agents and most targeted therapies, anti-PD-L1 monoclonal antibodies such as avelumab are not metabolized by cytochrome P450 enzymes and are not eliminated through hepatic, biliary, or renal pathways. Their clearance relies predominantly on target-mediated and non-target-mediated proteolytic catabolism within the reticuloendothelial system, resulting in classical pharmacokinetics. DDIs are unlikely to occur, even in the presence of extensive polypharmacy. This pharmacological profile likely explains why inclusion of avelumab in the Drug-PIN® computation did not meaningfully increase interaction scores in our analysis. Nevertheless, growing evidence suggests that interactions involving ICIs may occur at a pharmacodynamic or immunological level rather than through traditional pharmacokinetic mechanisms. Recent studies have highlighted that several commonly prescribed drug classes, including corticosteroids, antibiotics, proton pump inhibitors, opioids, and other immunomodulatory agents, may influence antitumor immune responses by altering immune cell function, cytokine signaling, or the gut microbiota. These so-called immunological DDIs have been associated with altered efficacy and toxicity of ICIs across different tumor types [19]. In this context, the absence of an association between Drug-PIN® score and survival outcomes in our study deserves careful interpretation. Drug-PIN® is primarily designed to capture pharmacokinetic and clinically actionable DDIs, and although it integrates multiple patient-related variables, it does not specifically quantify immunological interactions that may subtly modulate ICI activity. Therefore, it is possible that certain immunologically relevant interactions are not fully reflected in the Drug-PIN® score, particularly when they do not translate into overt clinical toxicity or measurable differences in survival within the available follow-up period. Importantly, our findings do not suggest that concomitant medications are clinically irrelevant in patients receiving avelumab maintenance therapy. Rather, they indicate that, within the limits of our analysis, the overall burden of DDIs as measured by Drug-PIN® does not appear to compromise the survival benefit of avelumab. This observation is reassuring for clinicians managing older and comorbid patients, in whom discontinuation or modification of chronic therapies may not always be feasible. Another relevant aspect emerging from our study is that the variability in Drug-PIN® scores was largely driven by background comorbidities and non-oncological treatments rather than by avelumab itself. This finding reinforces the concept that ICIs can be safely integrated into complex therapeutic regimens without substantially increasing interaction burden. From a practical standpoint, this supports the continued use of avelumab maintenance therapy in real-world populations characterized by polypharmacy, as also reflected in the MALVA effectiveness study. Several limitations of this analysis should be acknowledged. First, the retrospective nature of the study limits causal inference and exposes the analysis to potential selection bias and missing data. Second, the relatively limited sample size, particularly within the higher Drug-PIN® risk categories and among patients with greater levels of polypharmacy, may have reduced the statistical power to detect modest but clinically meaningful differences in survival outcomes. Therefore, the absence of statistically significant associations should be interpreted with caution. In addition, given the exploratory nature of the analysis and the limited sample size, survival analyses were intentionally restricted to descriptive comparisons across pre-defined Drug-PIN® categories and should not be interpreted as inferential evidence of prognostic effects. An important methodological limitation relates to the ascertainment of concomitant medications. Baseline pharmacological data were retrospectively collected from routine clinical records, and no standardized medication reconciliation process was prospectively implemented. Consequently, incomplete capture of concomitant therapies cannot be excluded, potentially leading to underascertainment of polypharmacy and systematic underestimation of Drug-PIN® scores. This limitation is particularly relevant considering the age and comorbidity profile of the study population and may partly explain the relatively low prevalence of polypharmacy observed in our cohort. The unexpectedly low prevalence of concomitant medications observed in this older metastatic population further supports the possibility of underascertainment in routine clinical records. Therefore, the reliability of the primary exposure variable may have been affected, and the present findings should be regarded as exploratory and hypothesis-generating. Therefore, the reliability of the primary exposure variable may have been affected, and the present findings should be regarded as exploratory and hypothesis-generating. Another limitation is that the present analysis was specifically designed to explore the potential impact of polypharmacy and DDIs on survival outcomes and was not intended to systematically assess clinically meaningful DDI-related outcomes, including treatment-related toxicities, treatment modifications, discontinuation, or immune-related adverse events. Moreover, as a result of the limited sample size and exploratory design, multivariable analyses adjusting for key confounding factors such as ECOG performance status and comorbidity burden were not performed, and residual confounding related to frailty and baseline clinical characteristics cannot be excluded. Consequently, although no association between Drug-PIN® score and survival outcomes was observed, we cannot exclude that higher DDI burden may still influence toxicity risk or treatment adherence, even in the absence of measurable survival differences. Finally, an additional limitation concerns the proprietary nature of the Drug-PIN® algorithm. Although previous studies have supported its clinical utility in oncology and pharmacogenomics settings, the exact weighting system and mathematical calculations underlying the software are not publicly accessible, limiting independent reproducibility and external verification of the generated scores. Furthermore, the present study was not designed to externally validate Drug-PIN® performance or compare it with alternative DDI assessment approaches. For these reasons, our findings warrant confirmation in future prospective studies incorporating standardized medication reconciliation, systematic toxicity assessment, and independently verifiable methodologies. Finally, another important limitation of this study is the absence of sex- and gender-stratified analyses. This was primarily due to the limited sample size and the marked imbalance between male and female patients, which would have reduced the statistical robustness and interpretability of subgroup analyses. As a result, potential differences related to sex or gender in terms of DDI burden, immune response, and clinical outcomes could not be explored. This limitation may affect the interpretation and generalizability of our findings, particularly for female patients and underrepresented gender groups. Future studies with larger and more balanced populations are needed to better elucidate the role of sex and gender in patients receiving avelumab maintenance therapy. Despite these limitations, this study represents, to our knowledge, one of the first real-world analyses specifically evaluating DDI burden in patients receiving avelumab maintenance therapy for advanced UC. By integrating Drug-PIN® assessment within a well-characterized multicenter cohort, our findings provide exploratory real-world data regarding the feasibility of evaluating DDI burden in patients receiving avelumab maintenance therapy. Future prospective studies are warranted to further elucidate the clinical relevance of DDIs in the context of immune checkpoint inhibition. In particular, studies incorporating systematic toxicity assessment, immune-related adverse event profiling, microbiome analyses, and dedicated evaluation of immunological DDIs may help clarify whether specific concomitant medications influence the efficacy or safety of avelumab beyond what is captured by current interaction scoring systems. Conclusions In this retrospective exploratory analysis of patients with advanced UC receiving avelumab maintenance therapy, Drug-PIN®-assessed DDI burden was not significantly associated with OS or PFS. However, because Drug-PIN® does not directly evaluate clinically meaningful DDI outcomes such as toxicity, treatment modification, or treatment discontinuation, and considering the limited sample size and retrospective design, these findings should be interpreted cautiously. Prospective studies incorporating systematic toxicity assessment and standardized medication reconciliation are warranted to better define the clinical relevance of DDIs in this setting. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (PDF 132 KB) Acknowledgments We thank the participants of the study. We also thank the participating investigators and clinical staff from all contributing centers for their valuable collaboration. A complete list of collaborators and their affiliations is provided in the Supplementary Material. Medical Writing/Editorial Assistance The authors did not receive any external medical writing or editorial assistance. AI tools were used solely for language editing and formatting support, and all content was reviewed and approved by the authors. Author Contribution Giandomenico Roviello: Conceptualization, methodology, formal analysis, investigation, data curation, writing – original draft. Mattia Alberto Di Civita: Data curation. Daniele Santini: Investigation, writing—review & editing, supervision. Elisabetta Gambale: Writing—original draft. Lidia Strigari: Methodology. Simone Scagnoli: Data curation. Laura Pappalardo: Investigation. Andrea Torchia: Investigation. Andrea Botticelli: Data curation. Serena Pillozzi: Writing—original draft, writing—review & editing. Lorenzo Antonuzzo: Investigation, writing—review & editing, supervision, project administration. Giandomenico Roviello, Mattia Alberto Di Civita, Daniele Santini, Elisabetta Gambale, Lidia Strigari, Simone Scagnoli, Laura Pappalardo, Andrea Torchia, Andrea Botticelli, Serena Pillozzi, and Lorenzo Antonuzzo made substantial contributions to the work, critically reviewed the manuscript, and approved the final version for publication in accordance with ICMJE guidelines. Giandomenico Roviello, Mattia Alberto Di Civita, Daniele Santini, Elisabetta Gambale, Lidia Strigari, Simone Scagnoli, Laura Pappalardo, Andrea Torchia, Andrea Botticelli, Serena Pillozzi, and Lorenzo Antonuzzo meet the authorship criteria according to the ICMJE guidelines. Funding No funding or sponsorship was received for this study or publication of this article. The Rapid Service Fee was funded by the authors. Data Availability The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Conflict of Interest Giandomenico Roviello reports scientific advisory boards, consulting, and travel expenses from MSD, Merck Serono, Bayer, BMS, AAA, AstraZeneca, Pfizer, and Astellas. Lorenzo Antonuzzo reports payment or honoraria for lectures and presentations from AstraZeneca, Roche, Astellas, Novartis, MSD, BMS, Ipsen, Merck Serono, Amgen, and Bayer; support for attending meetings and/or travel grants from AstraZeneca, Novartis, Ipsen, and Merck Serono; and research funding to his institution from Novartis and AstraZeneca. Mattia Alberto Di Civita has nothing to disclose. Daniele Santini has nothing to disclose. Elisabetta Gambale has nothing to disclose. Lidia Strigari has nothing to disclose. Simone Scagnoli has nothing to disclose. Laura Pappalardo has nothing to disclose. Andrea Torchia has nothing to disclose. Andrea Botticelli has nothing to disclose. Serena Pillozzi has nothing to disclose. Ethical Approval This study was approved by the local Ethics Committee (Tuscany Area Vasta Centro, protocol number 20817_oss), which served as the coordinating (“master”) ethics committee for the study. Given the retrospective nature of the analysis, clinical data were collected and analyzed in anonymized form, and no directly identifiable patient information was accessed by the investigators. All patients had previously provided written informed consent for the use of their clinical data for research purposes. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki (1964) and its subsequent amendments. References Associated Data Supplementary Materials Supplementary file1 (PDF 132 KB) Data Availability Statement The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.