Improving cancer care while reducing its carbon footprint: GHG-mitigating strategies highlighted at ESMO 2025 Authors: Matthieu Delaye, Paul Matte, Anthony Turpin, Coline Ducrot, Sarah Dumont, Pierre-Florent Petit, Max Piffoux Published: 2026 ABSTRACT AbstractBackgroundThe environmental impact of cancer care is an emerging concern, yet therapeutic strategies that may incidentally reduce greenhouse gas (GHG) emissions remain poorly described. The 2025 ESMO Congress, where nearly 3000 abstracts were presented, offered an opportunity to identify clinically validated approaches that could lower the carbon footprint of oncology.MethodsAll abstracts from the 2025 ESMO Congress were reviewed. Studies suggesting meaningful downstream reductions in drug use, visit frequency, treatment duration, or overall care intensity were selected. For each, GHG emissions were estimated using the ECOVAMED drug carbon-footprint database and standard emission factors.ResultsTen abstracts were identified. They spanned primary prevention, therapeutic de-escalation, reduced dose or duration of systemic therapy, biomarker-guided treatment selection, and non-pharmacological interventions. Several strategies demonstrated non-inferior or improved clinical outcomes while substantially lowering treatment-related emissions. Notably, none of the selected studies had explicitly assessed environmental impact.ConclusionsMultiple strategies presented at ESMO 2025 appear capable of improving patient outcomes while reducing the environmental burden of cancer care. Systematically integrating environmental endpoints into clinical research would help identify such win-win approaches and advance a more sustainable oncology practice. Introduction The healthcare sector contributes to about 5% of global greenhouse gas (GHG) emissions [1]. Among specialties, oncology is expected to be particularly emissive due to its intensity of care delivery and the expected increase in cancer incidence in Europe and worldwide. Climate change might impact hospital structures, limit access to medical equipment, or limit patient access to hospital facilities, and also has an impact on the overall health of populations [2]. Despite ESMO’s proactive policy (creation of a task force, work on reducing the environmental impact of the ESMO congress, sessions dedicated to climate change), the environmental impact of oncology, including climate change, is still rarely discussed at congresses. This could be explained by a lack of awareness among decision-makers and healthcare professionals regarding the impacts of climate change and the identification of strategies for GHG reduction. Documenting the environmental impact of healthcare enables us to take a critical look at how we provide care and how we could provide it differently. For 50 years, the ESMO Congress has been a must-attend event in global oncology, with 2927 abstracts presented and 37,000 participants this year in Berlin. The latest groundbreaking results in oncology are presented and discussed, aiming to shape the future of medical oncology practice. Among the therapeutic strategies presented at 2025 ESMO Congress, none was evaluated for its potential to mitigate the environmental impact of cancer care. Here, we present a selection of abstracts from the congress that evaluate treatment strategies with improved or non-inferior efficacy and potentially lower environmental impact. Focusing on GHG emissions, we estimated the reductions associated with these strategies. Methods All abstracts presented at the 2025 ESMO Congress were reviewed by the authors, who then selected the most relevant ones (each author presented the abstracts they had identified to the rest of the group that secondary selected the abstracts that seemed to present strategies that could lead to a significant reduction in GHG emissions). GHG emissions of therapeutic strategies were then calculated for selected abstracts. Abstracts that were initially selected but whose GHG emissions calculations did not indicate a reduction were not retained and presented in the article (N = 8). The details of the GHG emissions calculation for each analyzed abstract are presented in the supplementary material. Briefly, drug and other items consumed in the care pathway were counted, and multiplied by their associated emission factor (in kgCO2e/item) using the ECOVAMED drug carbon-footprint database [3, 4]. Uncertainty corresponds to the 95% confidence interval. Since our work is based on study results presented at the ESMO Congress and published, no specific ethics approval was required. Results The selected abstracts focused on strategies for therapeutic de-escalation, dose reduction or treatment duration reduction, improved patient selection, use of non-pharmacological treatments, and prevention (Table 1). Table 1Abstracts selected for presenting GHG emission mitigation strategies.Therapeutic strategyAbstracts concernedPrinciplePrimary endpoint presented at the congressTherapeutic de-escalation2464 PIntermittent ARPIs in mPCTreatment-free survival (TFS)Reduced dose or treatment durationLBA15Low-dose pembrolizumab in TNBCPathological Complete Response.LBA61Early discontinuation of immunotherapy in case of clinical response in melanomaRate of ongoing response at 24 months from start of treatmentLBA1093 versus 6 cycles of gemcitabine + platinum in BCPRO and OS727 MOMaintenance treatment with 5FU alone after FOLFOX + panitumumab in CRCDuration of disease control (DDC)2800 OReducing the use of BMA after 2 years in BrC and PCQuality of lifeBetter patient selectionLBA8ctDNA-guided adjuvant atezolizumab in BCDisease-Free SurvivalNon-pharmacological treatments2915ePMindfulness-based stress reduction programMedian Distress Score DropLBA28Physical activity program in operated CRCAdherence to a structured exercise programaPrevention2274 OReduce the incidence of cancerCancer incidencemPC metastatic prostate cancer, ARPIs Androgen Receptor Pathway Inhibitors, TNBC triple-negative breast cancer, BC bladder cancer, PRO patient reported outcomes; OS: overall survival, CRC colorectal cancer, BMA bone-modifying agents, BrC breast cancer, GCSF granulocyte colony-stimulating factor. Treatment-free survival was defined as time from treatment cessation to progression; Duration of disease control: time from inclusion to radiological progression under EGFR therapy with chemotherapy or death.aAdditional analysis of the Challenge study, presented at a previous conference, showed an increase in DFS with a structured physical activity program compared to standard care. Abstracts selected for presenting GHG emission mitigation strategies. mPC metastatic prostate cancer, ARPIs Androgen Receptor Pathway Inhibitors, TNBC triple-negative breast cancer, BC bladder cancer, PRO patient reported outcomes; OS: overall survival, CRC colorectal cancer, BMA bone-modifying agents, BrC breast cancer, GCSF granulocyte colony-stimulating factor. Treatment-free survival was defined as time from treatment cessation to progression; Duration of disease control: time from inclusion to radiological progression under EGFR therapy with chemotherapy or death. aAdditional analysis of the Challenge study, presented at a previous conference, showed an increase in DFS with a structured physical activity program compared to standard care. Therapeutic de-escalation In patients with metastatic prostate cancer, intermittent hormone therapy could be an option to mitigate adverse events and decrease the carbon footprint as a co-benefit, as presented in abstract 2464 P (retrospective study) [5]. Patients included in the study had a median time without treatment of 25 months (95%CI 20.8–28.8), corresponding to about 3485 ± 1323 kgCO2e/patient for enzalutamide, 2334 ± 1634 kgCO2e/patient for abiraterone, 3625 ± 1378 kgCO2e/patient for apalutamide and 4459 ± 2497 kgCO2e/patient for darolutamide. Although 95% of patients responded to the rechallenge of Androgen Receptor Pathway Inhibitor (ARPI), prospective data are needed to confirm the non-inferiority of this strategy, and trials are ongoing to address that question. Reduction in doses or duration of treatment In the PLANeT phase II single-center study, the authors compared neoadjuvant chemotherapy (4 cycles of doxorubicin + cyclophosphamide followed by 4 cycles of paclitaxel) with or without low-dose pembrolizumab (50 mg for 3 cycles, versus the standard 200 mg for 8 infusions) in patients with triple-negative breast cancer (N = 157) [6]. The lower dose was justified by the limited access to pembrolizumab in low-income countries and financial toxicity. The addition of low-dose pembrolizumab was associated with better pathological complete response (53.8% versus 40.5%, p = 0.047) which is aligned with the results of the Keynote-522 study (absolute difference of 14.5%). No deterioration of quality of life or additional toxicity was observed. In terms of carbon footprint, decreasing the dose from 1600 mg to 150 mg per patient (assuming similar number of outpatient and inpatient visits) mitigates emissions by an estimated 2129 ± 1171 kgCO2e/patient. In melanoma, the Safe Stop Trial studied whether early discontinuation of first-line anti-PD-1 antibody was safe in patients with metastatic melanoma who achieve a confirmed objective tumor response [7]. Instead of treating until disease progression, limiting toxicity or 24 months, patients that achieved objective response (29% of complete responses and 71% of partial responses) discontinued the PD-1 inhibitor earlier (median: 24 weeks, interquartile range [IQR]: 20–26). The 2-years ongoing response rate was 81.8% (90% CI 76.7–85.8%) and 68.8% (90% CI 62.7–74.1%) at 4 years. The 4-year melanoma overall survival rate was 90.9% (90% CI 86.5–93.9%). These results suggest that early discontinuation of PD-1 blockade could be an option in patients who achieve an objective response. In terms of carbon footprint, the shorter treatment duration (by ~80 weeks for ~90% of patients) corresponds to avoiding ~26.7 three-weekly 200 mg pembrolizumab doses ( ≈ 7830 ± 4,307 kgCO2e per patient) or ~26.7 three-weekly 360 mg nivolumab doses ( ≈ 6115 ± 3363 kgCO2e per patient), to which must be added an additional 1.068 kgCO₂e per patient, corresponding to the environmental costs of hospital days avoided. Concerning chemotherapy, the DISCUS phase II study compared administering 3 cycles versus 6 cycles of platinum–gemcitabine chemotherapy followed by avelumab maintenance in patients with locally advanced or metastatic urothelial cancer who were naïve to prior systemic treatment [8]. Patients who received 6 cycles experienced a greater deterioration of their quality of life (primary endpoint) compared to those who received only 3 cycles, with no difference in terms of overall survival (median 18.9 months versus 18.9, p = 0.56) or PFS (median: 9.0 months versus 8.0 months, p = 0.788). In terms of carbon footprint, avoiding 3 injections of Cisplatin/gemcitabine (30.5 ± 11.9 kgCO2e/injection) reduces GHG emissions by 91.4 ± 35.8 kgCO2e per patient, to which must be added an additional 120 kgCO2e per patient, corresponding to the environmental costs of hospital days avoided. In colorectal cancer, a maintenance strategy after FOLFOX chemotherapy + panitumumab was studied in the OptiPRime study (maintenance with 5-FU alone after 6 cycles, and reintroduction of oxaliplatin and panitumumab upon progression) [9]. The study met its primary endpoint (duration of disease control) and the median OS was 36.1 months (95%CI: 27.5–39.9). One month of 5-FU alone produces ~12.3 ± 5.2 kgCO2e/patient, which is about a 10-fold reduction compared to one month of 5-FU + panitumumab (125.4 ± 62.4 kgCO2e), and ~19-fold lower than one month of FOLFOX + panitumumab (238.2 ± 112.8 kgCO2e), to which must be added the environmental costs of hospital days avoided ( ≈ 40 kgCO2e/patient). In abstract 2800O, authors evaluated the non-inferiority of reducing the use of bone-modifying agents (BMA) after 2 years in patients with bone metastases from breast cancer or castration-resistant prostate cancer [10]. They showed that using BMA (zoledronate, pamidronate or denosumab) every 24 weeks rather than every 4 to 12 weeks was non-inferior in terms of physical functioning (primary endpoint) or symptomatic skeletal events. Such a strategy would allow for more than halve the carbon footprint of prolonged BMA treatment (87.05 ± 47.84 kgCO2e/year for a Q4W treatment, versus 29.02 ± 15.95 kgCO2e/year for a Q12W treatment). Better patient selection Post-surgery ctDNA testing could more effectively select patients for adjuvant therapy while avoiding overtreatment in others. The results of a phase III trial of ctDNA-guided adjuvant atezolizumab in muscle-invasive bladder cancer were presented [11]. Of 761 patients enrolled, 250 were ctDNA-positive and received atezolizumab (N = 167) or placebo (N = 83). The median disease-free survival (mDFS) was 9.9 months with atezolizumab, as compared with 4.8 months with placebo (HR: 0.64; 95% CI 0.47-0.87; P = 0.005). The median overall survival (mOS) was 32.8 months with atezolizumab, as compared with 21.1 months with placebo (HR: 0.59; 95% CI 0.39-0.90; P = 0.01). Compared to administering adjuvant immunotherapy to all patients, the ctDNA-guided strategy results in an average saving of approximately 2,147 kgCO2e per patient. Several abstracts presented prognostic or predictive scores based that could guide treatment escalation or de-escalation to better tailor treatment intensity to disease risk. However, these retrospective results require prospective validation. For example, the HEROES trial is evaluating the feasibility of therapeutic de-escalation in HER2-positive mBC with disease controlled after ≥2 years of anti-HER2 targeted therapy; its study design was presented as poster 620TiP. Non-pharmacological treatments The use of non-pharmacological treatments can help manage symptoms without resorting to medication, limiting potential side effects and environmental impact. For example, mindfulness-based stress reduction program can reduce patients’ stress levels and could easily be incorporated into routine practice (session duration 20–25 min) [12]. The most convincing results relate to physical activity as part of a sustained support program. In the CHALLENGE trial, this intervention improved DFS (HR 0.72, p = 0.017) and OS (HR 0.63, p = 0.022) compared to standard care in patients who had undergone surgery for colorectal cancer [13]. Prevention Prevention may be one of the key strategies to reduce the environmental impact of healthcare, as it may help to avoid long, costly, and polluting cancer care pathways. In abstract 2274 O, the authors presented the accumulated data on primary cancer prevention using pharmacological compounds published since 1980 [14]. Of the various strategies, the HPV vaccine for preventing cervical cancer was the most well-documented (9 studies including 8 positives). Calculating the environmental impact of this type of prevention strategy is difficult due to the variability in the care pathways for patients with the cancers concerned. Concerning HPV vaccination, if applied to a country with an incidence of 94 cases per 100,000 persons (e.g. Sweden [15]), the GHG emissions from a vaccination program would be at least offset by the emissions avoided due to a significant reduction in CIN 2–3 lesions and invasive cervical cancers, as well as other HPV related cancers requiring treatment (details in Supplementary). Discussion Although not exhaustive, this work highlights several abstracts presented at the 2025 ESMO congress that are expected to be GHG mitigating, in addition to being clinically non-inferior or more effective. Although our calculations remain rough estimates in the absence of source data, they allow us to better understand the order of magnitude of GHG emissions reduction. These studies mainly involved therapeutic de-escalation, reduction of treatment doses or duration, improved patient selection, and prevention, where most of the abstracts presented at the congress correspond to therapeutic escalation strategies and/or the use of families of molecules, such as antibody-drug conjugates, tyrosine kinase inhibitors, and immunotherapy, which have higher greenhouse gas (GHG) emissions than conventional chemotherapy. Interestingly, none of these abstracts calculated or presented their GHG impact results at the congress. Moreover, they were all presented with purely clinical objectives (no explicit environmental endpoint). Ideally, the environmental footprint must be included as an endpoint in clinical trials, just as patient quality-of-life and cost-effectiveness metrics have been integrated over time. This underlines the fact that there is still room for education in this area, which ESMO can help with. The ESMO congress usually mostly reports results regarding therapeutic interventions but other aspects like diagnostic pathways, telemedicine, or supply chain deserve to be addressed more extensively, as this would broaden oncologists’ perspective on the overall care of patients and may also offer additional avenues for emission reduction in oncology. The medical community is increasingly aware of the environmental impact of oncology care and the urgency of reducing it, and a large majority of oncologists would be willing to implement strategies with a lower environmental impact in their daily practice [16]. Oncologists are ready to implement this approach even in clinical trials. In fact, the environmental impact of trials themselves is now being studied [17]. Criteria based on environmental impact calculations should be integrated to better assess the short- and long-term benefits of new therapies [18]. Studies on therapeutic de-escalation, reduction of treatment dose/duration/frequency, and improved patient selection must be supported by cancer societies. This is particularly true since these studies are academic and often difficult to fund. This could be achieved by several means; by more effectively promoting such studies, by creating a dedicated label or award at the congress, by providing specific signposts to sustainability-related talks, by increasing the role and visibility of the ESMO Climate Change Task Force, or by establishing funding mechanisms specifically for this type of study. Evaluating the environmental impact of cancer care must no longer be considered a separate, esoteric concern, but rather a legitimate scientific approach aimed at improving the quality of cancer care for our patients now and in the future, and help to avoid dramatic climate change and associated dire health consequences [19]. Supplementary information Supplementary Author contributions MD, PM, and P-FP designed the study; all authors selected the abstracts; MP and CD calculated the GHG emissions; MD wrote the first draft of the manuscript; all authors proofread and corrected the manuscript. Data availability The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Competing interests The author declared no competing interests Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary information The online version contains supplementary material available at 10.1038/s44276-026-00231-z. References Associated Data Supplementary Materials Supplementary Data Availability Statement The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.