Economic cost-utility analysis of stage-directed oesophageal cancer treatment

Le résumé fourni par la source

Oesophageal cancer (OC) treatment is guided by radiological diagnostic stage1, with prognosis worsening as stage advances2. Potentially curative treatment is possible in 30–40% of patients2,3. Treatments include definitive chemoradiotherapy or surgery, with or without neoadjuvant therapy4, and patients not considered suitable for curative treatment receive palliative treatments or Best Supportive Care (BSC)2,5. Clinical effectiveness of treatments can be estimated in terms of Quality-Adjusted Life Years (QALYs) to price a treatment’s cost-effectiveness6. The aim of this study was to estimate the cost-utility of curative treatment related to OC stage compared with BSC. The primary hypothesis was that OCs of earlier stage would prove cheaper to treat in fiscal terms than OCs of more advanced later stage. Consecutive patients undergoing surgical treatment for OC diagnosed according to established protocols with curative intent within a regional cancer network from 2010 to 2020 were included in the analysis. The cost of 1-year’s treatment from referral was calculated based on current management standards. Primary outcome was overall survival (OS). Detailed methods can be found in Supplementary methods. 365 patients (median age 65 years (range 38–80), 308 male, 57 female, 263 neoadjuvant therapy) who underwent surgical treatment with curative intent for OC were included. Based on pathological and intraoperative assessment, 111 (30.4%) were stage I, 65 (17.8%) stage II, 118 (32.3%) stage III and 71 (19.5%) were analysed as stage IV. Of these, 331 had adenocarcinoma, 32 squamous cell carcinoma and two high-grade dysplasia. Median follow-up was 36 (interquartile range (i.q.r.) 34.9–39.0) months and median OS was 42.9 (95% c.i. 35.6 to 53.2) months with an average cost of the first year’s treatment of €30 916. This resulted in a QALY-adjusted survival of 34.3 months, with cost per QALY of €10 817. In patients who underwent curatively intended surgery, median survival in the patients receiving neoadjuvant chemotherapy followed by surgery (CS) was 35.8 (95% c.i. 26.5 to 45.1) months compared with 45.6 (95% c.i. 37.7 to 46.7) months in the patients receiving neoadjuvant chemoradiotherapy followed by surgery (CRS) and 50.8 (95% c.i. 39.4 to 47.2) months in patients receiving surgery (S) alone. The QALY-adjusted survival was 28.6 months in the CS cohort, compared with 36.5 in the CRS cohort and 40.6 in the S cohort. The cost per QALY for CS was €14 448, CRS €13 040 and S €5276. The CS cohort had a significantly lower proportion of patients with pTNM stage I and II disease (28.6%) compared with the CRS and S cohorts (66.7 and 69.6% respectively, P < 0.001). Data relating to QALY-adjusted survival and the cost per QALY, stratified by tumour stage, can be found in Table 1. The cost analysis of treating OC related to TNM stage and treatment modality can be found in Fig. 1. Median OS for patients receiving BSC reported in the literature is around 4 months7, with a Health State Utility Value (HSUV) of 0.56, equating to a QALY-adjusted survival of 2.24 months and a cost per QALY of €70 463. Cost-effectiveness of oesophageal cancer treatment stratified by stage and treatment modality BSC, best supportive care; stage 1, pathology tumour node and metastasis (pTNM) stage I; stage 2, pTNM stage II; stage 3, pTNM stage III; stage 4, pTNM stage IV; S1, CS1 and CRS1, pTNM stage I oesophageal cancer (OC) treated with surgery alone, perioperative chemotherapy and perioperative chemoradiotherapy respectively; S2, CS2 and CRS2, pTNM stage II OC treated with surgery alone, perioperative chemotherapy and perioperative chemoradiotherapy respectively; S3, CS3 and CRS3, pTNM stage III OC treated with surgery alone, perioperative chemotherapy and perioperative chemoradiotherapy respectively; S4, CS4 and CRS4, pTNM stage IV OC treated with surgery alone, perioperative chemotherapy and perioperative chemoradiotherapy respectively (intention to treat). CRS, chemoradiotherapy followed by surgery; CS, chemotherapy followed by surgery; S, surgery alone; QALY, quality-adjusted life years. Cost-utility analysis of treatment of oesophageal cancer related to disease stage Values in parentheses are 95% confidence intervals. QALY, quality-adjusted life year. Cost-utility analysis of treatment of oesophageal cancer related to disease stage Values in parentheses are 95% confidence intervals. QALY, quality-adjusted life year. In patients undergoing CS, CRS and S, the QALY OS gains were 26.4, 34.3 and 38.4 months, with an associated increased cost of €20 110, €26 510 and €4694 respectively. This equates to an Incremental Cost-Effectiveness Ratio (ICER) of €9672/QALY for CS, €9275/QALY for CRS and €1467/QALY for S. When stratified by tumour stage, the ICER for TNM stage I was €3960/QALY, for stage II it was €11 365/QALY, for stage III it was €10 226/QALY and for stage IV it was €20 783/QALY. When further stratified by tumour stage and treatment modality, the ICER for TNM stage I S, CS and CRS was €1231/QALY, €5581/QALY and €6952/QALY respectively. The ICER for stage II S, CS and CRS was €2553/QALY, €7542/QALY and €8749/QALY respectively. The ICER for stage III S, CS and CRS was €2097/QALY, €12 604/QALY and €14 292/QALY respectively. The ICER for stage IV was €3765/QALY, €26 435/QALY and €22 789/QALY respectively. Further results can be found in Supplementary results. This is the first cost-utility analysis of stage-directed potentially curative OC therapy. The salient findings were that surgery improved OS six-fold compared with BSC and was cost-effective at nationally accepted thresholds of readiness to pay per QALY, supporting the primary hypothesis. Costs per QALY increased incrementally and proportionately with the stage of OC so that stage I treatment cost per QALY was a fifth of that associated with BSC, and stage III treatment cost per QALY less than half that of BSC. Similarly, regarding ICER-defined cost-effectiveness comparisons, treatment of patients diagnosed with stage I cancer was between three- and four-fold cheaper than treatment of patients diagnosed with stage IV disease. Economic cost-utility analyses regarding potentially curative treatment for cancers from other anatomical sites, namely breast8, colorectal8 and prostate8 cancers, have reported similar associations between greater costs and more advanced stage at presentation8,9. Powell et al. in a related study from the same regional cancer network reported similar findings in a cost-utility analysis related to gastric cancer, with costs per QALY gained of €8335, €8952, €11 317 and €25 669 related to stages I through IV respectively9. The poorer survival seen in advanced disease is a major contributor to cost per QALY10. Current evidence suggests that only 15–20% of patients undergoing neoadjuvant therapy and surgery for OC show significant pathological tumour regression11. Improving response to chemotherapy might, therefore, offer the best cost benefit11. This study showed at least a 10-month QALY-adjusted OS benefit associated with perioperative chemotherapy in patients with stage pTNM II. However, this was not the case for stage III cancer, where perioperative chemotherapy was associated with a poorer QALY-adjusted OS. This likely reflects the differential extent of downstaging versus no response in these patients. These findings support a precision-medicine approach to OC. A cost-utility analysis of the Keynote 559 trial, Pembrolizumab plus 5-Fluorouracil and Cisplatin-based chemotherapy in advanced OC, revealed differential cost-effectiveness based on the expression of PD-L1 (programmed death-ligand 1), the target antigen of Pembrolizumab12. Given the risk profiles and costs associated with using biological therapies to treat OC, it would seem beneficial to reserve these for patients most likely to benefit. This study has inherent limitations in that key working assumptions were made. Operational efficiency meant excluding the

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.