Neoadjuvant Treatment Crosses Cost-Effective Threshold Vs Upfront Surgery in Pancreatic Cancer

Simran Arjani, MD

Neoadjuvant chemoradiation followed by pancreaticoduodenectomy (NACRT) is more cost-effective than upfront pancreaticoduodenectomy and adjuvant chemotherapy (USR) in patients with resectable pancreatic cancer, according to a cost-effectiveness analysis published in the Journal of Clinical Oncology. Based on these findings, NACRT should be considered in all patients with resectable pancreatic cancer unless there is an absolute contraindication, study authors wrote.¹

Based on a case modeled after a male aged 65 years, NACRT yielded 1.61 quality-of-life years (QALY) at $45,483.52 USD/QALY. In comparison, USR had 1.47 QALYs at $38,642.56 USD/QALY. NACRT followed by resection was $6840.96 USD more expensive and resulted in 0.14 more QALYs. This produced an incremental cost-effectiveness ratio (ICER) of 48,130.70 for neoadjuvant therapy, which was below the predefined willingness-to-pay threshold and was the primary end point of the analysis.¹

Most one-way sensitivity analyses supported these findings. However, in the situation where 21% of patients or less were able to proceed to surgery following NACRT, USR became the most cost-effective choice. Similarly, if 85.4% or less of patents had resectable disease after NACRT, then USR was more cost-effective.

Two-way sensitivity analyses also supported the findings; however, USR became more cost-effective if many patients were unable to receive surgery after NACRT andother variables favored USR, orif patients had low NACRT resection rates and other variable favored USR. Another circumstance in which USR was more cost-effective was if the perioperative morality following USR was low and other variables favored USR.

The probabilistic sensitivity analysis supported NACRT in 94.3% of cases; and the cost-effectiveness acceptability curve showed that NACRT would be the preferred option unless the willingness to pay threshold dropped below $62,000 USD.

A secondary version of the probabilistic analysis varied costs and brand point probabilities. This version favored in NACRT in 78.4% of 100,000 simulations.

“Our study uses a cost-effectiveness analysis to compare the use of NACRT and USR in resectable [prostate cancer] and found that NACRT followed by resection is the more cost-effective approach,” wrote Simran Arjani, MD, of the Rutgers New Jersey Medical School, and co-investigators, in the study.¹ “The use of NACRT would allow physicians to better identify patients with chemoresponsive disease but would delay surgery. Given the evidence in this and similar studies, the use of NACRT in [resectable prostate cancer] should still be a discussion.”

Moreover, they noted that the increase in effectiveness was equivalent to 1.7 quality-of-life adjusted months with the added cost of $1000 to $10,000, both of which may be inconsequential to some patients and extremely important to others.

Both NACRT and USR are highlighted as possible treatments for resectable pancreatic ductal adenocarcinoma in the National Comprehensive Cancer Network guidelines.² This model divided a theoretical cohort of patients into half that received neoadjuvant FOLFIRINOX-based chemoradiation followed surgical resection and another half that received by adjuvant FOLFIRINOX-based chemotherapy.¹

This is not the first decision model or cost-effectiveness analysis to compare NACRT and USR. A decision model designed by VanHouten et al in 2012, compared USR against a gemcitabine-based NCART in patients with potentially resectable disease.³ Although the study assessed a slightly different patient population and some different treatment end points, this analysis found a slight benefit for neoadjuvant therapy, and authors simultaneously noted that the ideal treatment remained controversial. Moreover, their model showed that treatment decisions should be sensitive to the probability of tumor resectability and chemoradiation mortality, but not to the rates of other treatment complications.³

Additionally, a 2019 cost-effectiveness model ground that neoadjuvant FOLFIRINOX represented the optimal strategy compared with upfront surgery and adjuvant gemcitabine plus capecitabine/gemcitabine monotherapy.⁴ This model found the neoadjuvant treatment to be linked to a 0.30 QALY gain at a cost of $46,200 USD/QALY. Similarly, a 2019 Markov cohort decision analysis model found a slight benefit to neoadjuvant strategy over upfront pancreaticoduodenectomy. In the 2019 model, this benefit translated to a 4.6-month survival gain in most models.⁵

In the Arjani et al model, the survival and quality of life following NACRT or USR was assessed using TreeAge Pro software. The decision tree assessed the costs and survival outcomes associated with these treatments in the published literature. The time horizon was limited to 1 year.¹

In this study, the willingness-to-pay threshold was determined to be $100,000 USD/QALY gained, and anything above that priced would mean that the benefit outweighs the cost. A negative ICER meant that the option provides a benefit at a lower cost than its comparator. Because of the study was designed with a 1-year time horizon, discounts were not a factor. Moreover, to account for probability in branch points, sensitive analyses, not one-way and two-way, were conducted, to simulated 100,000 different random samplings.¹

Study authors that the model has some limitations. Data from published literature was used to population the decision tree, and, wherever, possible high levels of evidence, namely, meta-analyses, were used, according to the authors. Recent studies with greater power and more prevalent treatment regimens were prioritized. However, there were situations when the highest levels of evidence were unavailable and multiple studies were consulted to provide this data. The sensitivity analyses sought to address this issue. Moreover, the model was unable to account for the impact of a patient’s inability to complete a chemotherapy regimen, and the analysis did not comment on racial/ethnic distributions, prohibiting conclusions regarding the impact of these factors on the models.¹

“The results of our study and those in the published literature emphasize the role of resection rates after NACRT, which is a surrogate for both chemoresponsiveness of the disease and treatment toxicity,” study authors concluded. “The future of treatment decision making lies in being able to identify chemoresponsive disease. Work in genomics, proteomics, and metabolomics to identify resistance or susceptibility to chemotherapy drugs has the potential to maximize treatment benefits and minimize treatment-related toxicity.”

Reference

1. Arjani S, Prasath V, Suri N, Li S, Ahlawat S, Chokshi RJ. Neoadjuvant treatment versus upfront surgery in resectable pancreatic cancer: a cost-effectiveness analysis.JCO Oncol Pract. Published online December 22, 2022. doi:10.1200/OP.22.00536
2. National Comprehensive Cancer Network. Pancreatic adenocarcinoma. Version 2.2022). https://www.nccn.org/professionals/physician_gls/pdf/pancreatic.pdf. Accessed January 24, 2023
3. VanHouten JP, White RR, Jackson GP. A decision model of therapy for potentially resectable pancreatic cancer.J Surg Res. 2012;174(2):222-230. doi:10.1016/j.jss.2011.08.022
4. Choi JG, Nipp RD, Tramontano A, et al. Neoadjuvant FOLFIRINOX for patients with borderline resectable or locally advanced pancreatic cancer: results of a decision analysis. Oncologist. 2019;24(7):945-954. doi:10.1634/theoncologist.2018-0114
5. Rieser CJ, Narayanan S, Bahary N, et al. Optimal management of patients with operable pancreatic head cancer: a Markov decision analysis. J Surg Oncol. 2021;124(5):801-809. doi:10.1002/jso.26589