- Email: [email protected]
Dose Dense Chemotherapy for Front-line Ovarian Cancer Treatment: The Price is Right?
Gynecologic Oncology 145 (2017) 1–2
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Editorial
Dose Dense Chemotherapy for Front-line Ovarian Cancer Treatment: The Price is Right?
Determining the cost of cancer care is a valuable component of therapy assessment, and provides further context for determining true patient benefit that includes evaluation of efficacy and toxicity endpoints as well as incorporating patient reported outcomes. Determining treatment value, especially in lieu of the rapidly escalating costs of oncology care with novel treatments exceeding $10,000/month, has become a major initiative with numerous organizations offering input including: ASCO Value Framework, ESMO Magnitude of Clinical Benefit Scale, NCCN Evidence Blocks, Memorial Sloan Kettering’s DrugAbacus, and Institute for Clinical and Economic Review Value Assessment Framework. Cost determination is a challenging undertaking whereby multiple methodologies have been utilized, but arguably none fully capture the totality of economic impact for individuals or society exposed to cancer treatment. Ovarian cancer is a complex disease, requiring individualized therapy to maximize patient outcomes. Over many decades, the overall survival of patients with ovarian cancer has increased substantially, and completed and ongoing clinical trials aim to continue to improve upon these outcomes. In certain circumstances, improved outcomes do come at a cost, which can be measured as both patient side effects and financial cost. New approaches to ovarian cancer care have led to increased toxicity with the use of intraperitoneal (IP) chemotherapy [1], and to increased financial toxicity with the use of novel therapies such as bevacizumab [2]. GOG 262 attempted to define the best schedule for the administration of paclitaxel in patients with newly cytoreduced ovarian cancer, randomizing 692 patients to either day 1 (“standard”) versus days 1, 8, and 15 (“dose dense”) every 21 days administration of paclitaxel [3]. Results showed that in the overall population studied, there was no difference in the primary outcome of progression-free survival between arms; however, patients who chose not to be discretionally treated with bevacizumab (defined this way since it was not mandated in the protocol) and were randomized to the dose dense arm had a 3.9 month improved PFS compared with those randomized to the standard arm. This subset analysis of 112 patients (16% of the overall population) serves as the basis that has led some clinicians to argue that dose dense paclitaxel should be the standard of care for women with advanced ovarian cancer, especially when bevacizumab is not administered concurrently. In this edition of Gynecologic Oncology, Seagle and Shahabi model this exact subset of patients treated on GOG 262, aiming to establish the cost effectiveness of dose dense versus standard paclitaxel delivered without concurrent bevacizumab. They find that “while dose dense chemotherapy is not cost-saving”, it “is robustly cost-effective” [4]. This finding is certainly interesting and exciting, bolstering support for the
http://dx.doi.org/10.1016/j.ygyno.2017.03.001 0090-8258/© 2017 Elsevier Inc. All rights reserved.
dose dense treatment arm of GOG 262 that, in addition to being more effective than standard therapy (when either arm is administrated without concurrent bevacizumab), is also cost effective. This finding is, however, not entirely unexpected, since the more heavily-weighted increase in PFS in the dose-dense paclitaxel arm is counterbalanced by a very modest increase in the cost of drug for weekly administration. In their model, the cost of toxicity was calculated only for the use of GCSF for neutropenia, transfusion for anemia, and hospitalization. In the overall population treated on GOG 262, there were significant differences in the rates of G-CSF use (30% vs. 22%) and red-cell transfusion (55% vs. 23%), all higher in the dose dense arm. The rate of hospitalization in GOG 262 was not reported, so it is unclear how this parameter was used in the model. Overall, this publication from Seagle and Shahabi is well executed, with an appropriately described model where costs are justified, and uncertainties handled with appropriate one-way and probabilistic sensitivity analyses. However, like any modeling study, a cost analysis such as this is limited by its structure and assumptions. And, as noted by the authors, their model has several limitations. Most striking is the lack of incorporation of quality of life (QoL) data (from GOG 262 or modeled), without which a more robust (and patient-centered) “cost utility analysis” cannot be performed. For those readers not familiar with cost-utility, a little background is necessary to help in interpreting their manuscript. In the simplest sense, one can think of a cost-utility analysis as a balance; on one side is the "cost," made up of the money it costs for a treatment, the money it costs to treat a complication, and the "cost" to a patient who experiences a decrement to their QOL (this is known as a negative "utility score"). On the other side of the balance is the "effectiveness," which is usually made up of the cancer survival, and occasionally an improvement in QOL (a positive "utility score"). The balance between the cost and the effectiveness is used to calculate a statistic comparing the cost effectiveness of one strategy to another. The authors note that in GOG 262, overall QOL was not different between the arms in patients who did not receive bevacizumab; however, the overall study did demonstrate worse overall QOL and worse sensory neuropathy in women on the dose dense arm. What is not considered in the model is the societal impact of weekly therapy. It has been demonstrated that chemotherapy can lead to a significant impact on “financial toxicity” including out of pocket expenses, childcare, transportation, employment absenteeism, facility fees [5]. These parameters potentially impact both the patient and her support team. While unproven, one only can assume that these effects are more pronounced with weekly versus every three weeks treatment schedules. This cost effectiveness analysis is from a
2
Editorial
healthcare system perspective, and thus is not designed to assess these important societal parameters. It would be interesting to remodel the impact of weekly therapy to include the societal perspective. Another limitation of the model presented by Seigel and Shahabi is the fact that they use data from the overall study arms as a surrogate for informing their model for the minority of patients who did not receive bevacizumab. Since the actual data for this small group of patients is not reported in GOG 262, we will not know whether the authors’ assumption is correct (but can imagine that if they are not correct, any errors would be amplified in a model that expands this small subset of the larger population). The generalizability of this unplanned post hoc analysis in a small subset of patients, upon which the economic analysis herein is based, should be interpreted with caution. Another important concern with this economic analysis of GOG 262 is that the endpoint assessed was not overall survival (OS), but rather progression free survival (PFS). While efforts have been made to consider surrogate endpoints such as PFS in the context of clinical trial design and regulatory approval, use of OS has been the most common variable for economic advantage assessment using quality adjusted life year (QALY) methodology, which usually references a year of survival, not PFS. [6] Obviously, when OS is not statistically different, as seen in GOG 218 and likely with GOG 262, then some insight may be sought from cost modeling the PFS gain, but the limitation of using PFS should be acknowledged. The economic assessment of JGOG weekly versus every three week paclitaxel schedule was also based on PFS, likely due to the fact that the paper was published prior to the long-term survival outcome availability. [7] The correlation between PFS and OS has been demonstrated, but this correlation has been less robust in the era of targeted therapies. [8] So what does the Seigel et. al paper mean to our readers? Does this analysis establish the dose dense paclitaxel regimen as “the” standard of care? These are confusing times for clinicians and patients in deciding the optimal care of ovarian cancer patients in the front-line setting. There are multiple acceptable standards of care that alter the traditional upfront cytoreductive surgery followed by every 3 weeks platinum and taxane model, including neoadjuvant, IP, bevacizumab, and dose dense taxane therapies. This paper, along with the existing literature on outcomes and costs of weekly taxanes, certainly add to the evidence that the dose dense strategy is a standard of care that should be strongly considered for patients who are willing and able to incur the inconvenience and toxicity of weekly therapy when bevacizumab is not administered concurrently. Further data such as long-term follow-up for OS and late toxicity effects are needed from GOG 262, and future cost modeling of dose dense versus bevacizumab after cost reductions from off-patent or biosimilar availability is warranted. Nonetheless, refined cost effectiveness analysis is critical in helping patients, providers, insurers, and governments make informed decisions. Continued investigation of what contributes to the cost effectiveness of cancer therapy is critical in improving the outcomes for our patients while mitigating the escalating costs of an increasingly expensive healthcare system.
Conflict of Interest Dr. Herzog reports personal fees from Roche during the conduct of the study for advisory boards; personal fees from J & J, Clovis, AstraZeneca, and Tesaro, outside the submitted work for advisory boards. Dr. Cohn has nothing to disclose. Transparency document The Transparency document associated with this article can be found, in online version. References [1] K. Deborah, M.D. Armstrong, Brian Bundy, Lari Wenzel, Helen Q. Huang, Rebecca Baergen, Shashikant Lele, Larry J. Copeland, Joan L. Walker, Robert A. Burger, for the Gynecologic Oncology Group*, Intraperitoneal Cisplatin and Paclitaxel in Ovarian Cancer, N Engl J Med 354 (2006) 34–43 January 5, 2006 10.1056/NEJMoa052985. [2] D.E. Cohn, K.H. Kim, K.E. Resnick, D.M. O’Malley, J.M. Straughn Jr., At what cost does a potential survival advantge of bevacizumab make sense for the primary treatment of ovarian cancer? A cost-effectiveness analysis, J Clin Oncol 29 (2011) 1247–1251. [3] J.K. Chan, M.F. Brady, R.T. Penson, et al., Weekly vs. every-3-week paclitaxel and carboplatin for ovarian cancer, N Engl J Med 374 (2016) 738–748. [4] B.L. Seagle, S. Shahabi, Cost effectiveness analysis of dose-dense versus standard intravenous chemotherapy for ovarian cancer: An economic analysis of results from the Gynecologic Oncology Group protocol 262 randomized controlled trial, Gynecol Oncol 145 (2017) 9–14. [5] S.Y. Zafar, J.M. Peppercorn, D. Schrag, D.H. Taylor, A.M. Goetzinger, X. Zhong, A.P. Abernethy, The financial toxicity of cancer treatment: a pilot study assessing outof-pocket expenses and the insured cancer patient's experience, Oncologist. 18 (4) (2013) 381–390, http://dx.doi.org/10.1634/theoncologist.2012-0279 Epub 2013 Feb 26. [6] T.J. Herzog, R.D. Alvarez, A. Secord, B.A. Goff, R.S. Mannel, B.J. Monk, R.L. Coleman, SGO guidance document for clinical trial designs in ovarian cancer: a changing paradigm, Gynecol Oncol. 135 (1) (2014 Oct) 3–7, http://dx.doi.org/10.1016/j.ygyno.2014.08. 004. [7] H.J. Dalton, X. Yu, L. Hu, et al., An economic analysis of dose dense weekly paclitaxel plus carboplatin versus every-3-week paclitaxel plus carboplatin in the treatment of advanced ovarian cancer, Gynecol Oncol 124 (2012) 199–204. [8] M.F. Brady, First-line therapy in ovarian cancer—surrogate endpoints for accelerated approval. FDA/ASCO/AACR Ovarian Cancer End Points Workshop[Retrieved Feb 28, 2017, from http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/ DevelopmentResources/CancerDrugs/ucm094640.pdf 2006.
Thomas J. Herzog Deputy Director- University of Cincinnati Cancer Institute, University of Cincinnati Medical Center, Medical Sciences Bldg, Suite 2005H, ML0662, 231 Albert Sabin Way, Cincinnati, OH 45267-0662. Corresponding author. Tel. +1 513 584 6373, 558 2177 (Admin); fax: +1 513 558 2666. E-mail address: [email protected] David E. Cohn Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Ohio State University College of Medicine, Columbus, OH, United States
Contents lists available at ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Editorial
Dose Dense Chemotherapy for Front-line Ovarian Cancer Treatment: The Price is Right?
Determining the cost of cancer care is a valuable component of therapy assessment, and provides further context for determining true patient benefit that includes evaluation of efficacy and toxicity endpoints as well as incorporating patient reported outcomes. Determining treatment value, especially in lieu of the rapidly escalating costs of oncology care with novel treatments exceeding $10,000/month, has become a major initiative with numerous organizations offering input including: ASCO Value Framework, ESMO Magnitude of Clinical Benefit Scale, NCCN Evidence Blocks, Memorial Sloan Kettering’s DrugAbacus, and Institute for Clinical and Economic Review Value Assessment Framework. Cost determination is a challenging undertaking whereby multiple methodologies have been utilized, but arguably none fully capture the totality of economic impact for individuals or society exposed to cancer treatment. Ovarian cancer is a complex disease, requiring individualized therapy to maximize patient outcomes. Over many decades, the overall survival of patients with ovarian cancer has increased substantially, and completed and ongoing clinical trials aim to continue to improve upon these outcomes. In certain circumstances, improved outcomes do come at a cost, which can be measured as both patient side effects and financial cost. New approaches to ovarian cancer care have led to increased toxicity with the use of intraperitoneal (IP) chemotherapy [1], and to increased financial toxicity with the use of novel therapies such as bevacizumab [2]. GOG 262 attempted to define the best schedule for the administration of paclitaxel in patients with newly cytoreduced ovarian cancer, randomizing 692 patients to either day 1 (“standard”) versus days 1, 8, and 15 (“dose dense”) every 21 days administration of paclitaxel [3]. Results showed that in the overall population studied, there was no difference in the primary outcome of progression-free survival between arms; however, patients who chose not to be discretionally treated with bevacizumab (defined this way since it was not mandated in the protocol) and were randomized to the dose dense arm had a 3.9 month improved PFS compared with those randomized to the standard arm. This subset analysis of 112 patients (16% of the overall population) serves as the basis that has led some clinicians to argue that dose dense paclitaxel should be the standard of care for women with advanced ovarian cancer, especially when bevacizumab is not administered concurrently. In this edition of Gynecologic Oncology, Seagle and Shahabi model this exact subset of patients treated on GOG 262, aiming to establish the cost effectiveness of dose dense versus standard paclitaxel delivered without concurrent bevacizumab. They find that “while dose dense chemotherapy is not cost-saving”, it “is robustly cost-effective” [4]. This finding is certainly interesting and exciting, bolstering support for the
http://dx.doi.org/10.1016/j.ygyno.2017.03.001 0090-8258/© 2017 Elsevier Inc. All rights reserved.
dose dense treatment arm of GOG 262 that, in addition to being more effective than standard therapy (when either arm is administrated without concurrent bevacizumab), is also cost effective. This finding is, however, not entirely unexpected, since the more heavily-weighted increase in PFS in the dose-dense paclitaxel arm is counterbalanced by a very modest increase in the cost of drug for weekly administration. In their model, the cost of toxicity was calculated only for the use of GCSF for neutropenia, transfusion for anemia, and hospitalization. In the overall population treated on GOG 262, there were significant differences in the rates of G-CSF use (30% vs. 22%) and red-cell transfusion (55% vs. 23%), all higher in the dose dense arm. The rate of hospitalization in GOG 262 was not reported, so it is unclear how this parameter was used in the model. Overall, this publication from Seagle and Shahabi is well executed, with an appropriately described model where costs are justified, and uncertainties handled with appropriate one-way and probabilistic sensitivity analyses. However, like any modeling study, a cost analysis such as this is limited by its structure and assumptions. And, as noted by the authors, their model has several limitations. Most striking is the lack of incorporation of quality of life (QoL) data (from GOG 262 or modeled), without which a more robust (and patient-centered) “cost utility analysis” cannot be performed. For those readers not familiar with cost-utility, a little background is necessary to help in interpreting their manuscript. In the simplest sense, one can think of a cost-utility analysis as a balance; on one side is the "cost," made up of the money it costs for a treatment, the money it costs to treat a complication, and the "cost" to a patient who experiences a decrement to their QOL (this is known as a negative "utility score"). On the other side of the balance is the "effectiveness," which is usually made up of the cancer survival, and occasionally an improvement in QOL (a positive "utility score"). The balance between the cost and the effectiveness is used to calculate a statistic comparing the cost effectiveness of one strategy to another. The authors note that in GOG 262, overall QOL was not different between the arms in patients who did not receive bevacizumab; however, the overall study did demonstrate worse overall QOL and worse sensory neuropathy in women on the dose dense arm. What is not considered in the model is the societal impact of weekly therapy. It has been demonstrated that chemotherapy can lead to a significant impact on “financial toxicity” including out of pocket expenses, childcare, transportation, employment absenteeism, facility fees [5]. These parameters potentially impact both the patient and her support team. While unproven, one only can assume that these effects are more pronounced with weekly versus every three weeks treatment schedules. This cost effectiveness analysis is from a
2
Editorial
healthcare system perspective, and thus is not designed to assess these important societal parameters. It would be interesting to remodel the impact of weekly therapy to include the societal perspective. Another limitation of the model presented by Seigel and Shahabi is the fact that they use data from the overall study arms as a surrogate for informing their model for the minority of patients who did not receive bevacizumab. Since the actual data for this small group of patients is not reported in GOG 262, we will not know whether the authors’ assumption is correct (but can imagine that if they are not correct, any errors would be amplified in a model that expands this small subset of the larger population). The generalizability of this unplanned post hoc analysis in a small subset of patients, upon which the economic analysis herein is based, should be interpreted with caution. Another important concern with this economic analysis of GOG 262 is that the endpoint assessed was not overall survival (OS), but rather progression free survival (PFS). While efforts have been made to consider surrogate endpoints such as PFS in the context of clinical trial design and regulatory approval, use of OS has been the most common variable for economic advantage assessment using quality adjusted life year (QALY) methodology, which usually references a year of survival, not PFS. [6] Obviously, when OS is not statistically different, as seen in GOG 218 and likely with GOG 262, then some insight may be sought from cost modeling the PFS gain, but the limitation of using PFS should be acknowledged. The economic assessment of JGOG weekly versus every three week paclitaxel schedule was also based on PFS, likely due to the fact that the paper was published prior to the long-term survival outcome availability. [7] The correlation between PFS and OS has been demonstrated, but this correlation has been less robust in the era of targeted therapies. [8] So what does the Seigel et. al paper mean to our readers? Does this analysis establish the dose dense paclitaxel regimen as “the” standard of care? These are confusing times for clinicians and patients in deciding the optimal care of ovarian cancer patients in the front-line setting. There are multiple acceptable standards of care that alter the traditional upfront cytoreductive surgery followed by every 3 weeks platinum and taxane model, including neoadjuvant, IP, bevacizumab, and dose dense taxane therapies. This paper, along with the existing literature on outcomes and costs of weekly taxanes, certainly add to the evidence that the dose dense strategy is a standard of care that should be strongly considered for patients who are willing and able to incur the inconvenience and toxicity of weekly therapy when bevacizumab is not administered concurrently. Further data such as long-term follow-up for OS and late toxicity effects are needed from GOG 262, and future cost modeling of dose dense versus bevacizumab after cost reductions from off-patent or biosimilar availability is warranted. Nonetheless, refined cost effectiveness analysis is critical in helping patients, providers, insurers, and governments make informed decisions. Continued investigation of what contributes to the cost effectiveness of cancer therapy is critical in improving the outcomes for our patients while mitigating the escalating costs of an increasingly expensive healthcare system.
Conflict of Interest Dr. Herzog reports personal fees from Roche during the conduct of the study for advisory boards; personal fees from J & J, Clovis, AstraZeneca, and Tesaro, outside the submitted work for advisory boards. Dr. Cohn has nothing to disclose. Transparency document The Transparency document associated with this article can be found, in online version. References [1] K. Deborah, M.D. Armstrong, Brian Bundy, Lari Wenzel, Helen Q. Huang, Rebecca Baergen, Shashikant Lele, Larry J. Copeland, Joan L. Walker, Robert A. Burger, for the Gynecologic Oncology Group*, Intraperitoneal Cisplatin and Paclitaxel in Ovarian Cancer, N Engl J Med 354 (2006) 34–43 January 5, 2006 10.1056/NEJMoa052985. [2] D.E. Cohn, K.H. Kim, K.E. Resnick, D.M. O’Malley, J.M. Straughn Jr., At what cost does a potential survival advantge of bevacizumab make sense for the primary treatment of ovarian cancer? A cost-effectiveness analysis, J Clin Oncol 29 (2011) 1247–1251. [3] J.K. Chan, M.F. Brady, R.T. Penson, et al., Weekly vs. every-3-week paclitaxel and carboplatin for ovarian cancer, N Engl J Med 374 (2016) 738–748. [4] B.L. Seagle, S. Shahabi, Cost effectiveness analysis of dose-dense versus standard intravenous chemotherapy for ovarian cancer: An economic analysis of results from the Gynecologic Oncology Group protocol 262 randomized controlled trial, Gynecol Oncol 145 (2017) 9–14. [5] S.Y. Zafar, J.M. Peppercorn, D. Schrag, D.H. Taylor, A.M. Goetzinger, X. Zhong, A.P. Abernethy, The financial toxicity of cancer treatment: a pilot study assessing outof-pocket expenses and the insured cancer patient's experience, Oncologist. 18 (4) (2013) 381–390, http://dx.doi.org/10.1634/theoncologist.2012-0279 Epub 2013 Feb 26. [6] T.J. Herzog, R.D. Alvarez, A. Secord, B.A. Goff, R.S. Mannel, B.J. Monk, R.L. Coleman, SGO guidance document for clinical trial designs in ovarian cancer: a changing paradigm, Gynecol Oncol. 135 (1) (2014 Oct) 3–7, http://dx.doi.org/10.1016/j.ygyno.2014.08. 004. [7] H.J. Dalton, X. Yu, L. Hu, et al., An economic analysis of dose dense weekly paclitaxel plus carboplatin versus every-3-week paclitaxel plus carboplatin in the treatment of advanced ovarian cancer, Gynecol Oncol 124 (2012) 199–204. [8] M.F. Brady, First-line therapy in ovarian cancer—surrogate endpoints for accelerated approval. FDA/ASCO/AACR Ovarian Cancer End Points Workshop[Retrieved Feb 28, 2017, from http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/ DevelopmentResources/CancerDrugs/ucm094640.pdf 2006.
Thomas J. Herzog Deputy Director- University of Cincinnati Cancer Institute, University of Cincinnati Medical Center, Medical Sciences Bldg, Suite 2005H, ML0662, 231 Albert Sabin Way, Cincinnati, OH 45267-0662. Corresponding author. Tel. +1 513 584 6373, 558 2177 (Admin); fax: +1 513 558 2666. E-mail address: [email protected] David E. Cohn Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Ohio State University College of Medicine, Columbus, OH, United States