ORIGINAL ARTICLE
Eligibility of Metastatic Pancreatic Cancer Patients for First-Line Palliative Intent nab-Paclitaxel Plus Gemcitabine Versus FOLFIRINOX Renata D. Peixoto, MD, Maria Ho, MD, Daniel J. Renouf, MD, Howard J. Lim, MD, Sharlene Gill, MD, Jenny Y. Ruan, BSc, and Winson Y. Cheung, MD, MPH
Objectives: The PRODIGE and MPACT trials showed superiority of FOLFIRINOX and nab-paclitaxel plus gemcitabine (NG) over gemcitabine alone, respectively. However, both had strict inclusion criteria. We sought to determine the characteristics of patients with metastatic pancreatic cancer (MPC) which inform the appropriateness of first-line chemotherapy FOLFIRINOX and NG in routine practice. Materials and Methods: Patients with MPC who initiated palliative chemotherapy with gemcitabine from 2000 to 2011 at the British Columbia Cancer Agency were identified. Clinicopathologic variables and outcomes were retrospectively collected and compared among groups. Eligibility criteria for each regimen were in accordance with the respective pivotal phase III trials. Results: A total of 473 patients were included: 25% of the patients were eligible for FOLFIRINOX versus 45% for NG. Main reasons for FOLFIRINOX ineligibility were Eastern Cooperative Oncology Group (ECOG) performance status (PS)Z2 (56.5%), age older than 75 years (19.0%), and bilirubin > 1.5 upper limit of normal (18.6%), whereas those for NG ineligibility were bilirubin > upper limit of normal (24.5%), ECOG PSZ3 (14.6%), and cardiac dysfunction (13.8%). Univariate analyses revealed that FOLFIRINOX and NG-eligible patients had longer median overall survival than their respective ineligible group (8.6 vs. 4.7 mo, P < 0.001; 6.7 vs. 4.9 mo, P = 0.008, respectively). After accounting for ECOG PS in the multivariate model, however, eligibility for either FOLFIRINOX or NG no longer predicted for better overall survival. Conclusions: The majority of patients with MPC are not candidates to either NG or FOLFIRINOX due to restrictive eligibility requirements. Specific trials addressing the unmet needs of protocol ineligible patients are warranted. Key Words: pancreatic cancer, metastasis, eligibility determination, palliative therapy, survival
(Am J Clin Oncol 2015;00:000–000)
From the Division of Medical Oncology, British Columbia Cancer Agency, University of British Columbia, Vancouver, BC, Canada. Previously presented as a poster at 2014 Gastrointestinal Cancers Symposium. Merit Award recipient. Supported by the BC Cancer Foundation, BC Cancer GI Outcomes Unit, Canadian Cancer Society Research Institute, and Celgene Canada. The authors declare no conflicts of interest. Reprints: Winson Y. Cheung, MD, MPH, Division of Medical Oncology, British Columbia Cancer Agency, University of British Columbia, Vancouver, BC, Canada V5Z 4E6. E-mail: [email protected]. Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.amjclinical oncology.com. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0277-3732/15/000-000 DOI: 10.1097/COC.0000000000000193
American Journal of Clinical Oncology
P
ancreatic cancer represents the fourth leading cause of cancer-related mortality in the developed world.1 For patients with early-stage disease, surgery followed by adjuvant therapy offers a small potential for cure and remains the treatment of choice whenever possible. However, 1.5ULN (19%). Conversely, the primary reasons for ineligibility to NG consisted of abnormal bilirubin >ULN (25%), ECOGZ3 (15%), and cardiac dysfunction (14%). Other important factors that excluded patients from both regimens as per trial eligibility were renal dysfunction (5%), no definitive histologic evidence of pancreatic cancer (7%), and prior major malignancies (14%). No patients were excluded due to brain metastases, active infection, or receipt of previous chemotherapy for MPC. Median OS for the entire cohort was 5.8 months (95% confidence interval, 5.4-6.2). Individuals who were considered eligible for both trials experienced better outcomes when compared with those deemed ineligible (median OS 8.6 vs. 4.8 mo, P < 0.001). On univariate analysis, eligible patients for FOLFIRINOX also had longer median OS than ineligible patients (8.6 vs. 4.7 mo, P < 0.001) (Fig. 2A). Likewise, eligibility to NG was associated with improved median OS (6.7 vs. 4.9 mo for eligible and ineligible groups, respectively, P = 0.008) (Fig. 2B). Additional factors that predicted for survival included ECOG, where 0 or 1 demonstrated higher median OS than Z2 (8.6 vs. 3.7 mo, P < 0.001) and primary tumor location, where head of the pancreas revealed slightly better median OS than the body or tail (6.1 vs. 5.3 mo, P = 0.025). Other variables, such as age, sex, and bilirubin levels did not appear to impact on outcomes (P = 0.63, 0.07, and 0.48, respectively). Importantly, multivariate analysis that
TABLE 1. Baseline Characteristics of Study Cohort
Characteristics
All Patients (n = 473)
Age (y) Median Sex (%) Male ECOG Performance Status Score (%) 0 1 2 Pancreatic tumor location (%) Head Body Tail Unknown Histologically proven (%) Yes Abnormal bilirubin (%) Yes Site(s) of metastasis (%) Liver only Lymph node only Peritoneum only One extrahepatic only Multiple Prior stent (%) Yes ECOG indicates Eastern Cooperative Oncology Group.
Copyright
r
66 54.5 2.7 40.8 41.9 48 24 24.9 3 93.4 24.5 35.7 3.6 7.8 4.4 48.4 31.5
FIGURE 1. A, A total of 117 patients were considered eligible for FOLFIRINOX versus 214 for nab-paclitaxel + gemcitabine. B, Main reasons for FOLFIRINOX and NG ineligibility. Top reasons for ineligibility to: (1) NG were abnormal bilirubin >ULN (25%), ECOGZ3 (15%), and cardiac dysfunction (14%); and (2) FOLFIRINOX were ECOGZ2 (57%), age older than 76 years (19%), and abnormal bilirubin >1.5 ULN (19%). ECOG indicates Eastern Cooperative Oncology Group; NG, nab-paclitaxel plus gemcitabine; ULN, upper limit of normal.
2015 Wolters Kluwer Health, Inc. All rights reserved.
www.amjclinicaloncology.com |
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of the article is prohibited.
3
American Journal of Clinical Oncology
Peixoto et al
A
-- FOLFIRINOX-eligible
-- FOLFIRINOX-ineligible
B
-- NG-eligible
-- NG-ineligible
FIGURE 2. A, Median overall survival from diagnosis was 4.7 versus 8.6 months (P < 0.001; hazard ratio [HR] 1.73) in the FOLFIRINOX-ineligible versus FOLFIRINOX-eligible patients. B, Median overall survival from diagnosis was 4.9 versus 6.7 months (P = 0.008; HR 1.28) in the NG-ineligible versus NG-eligible patients. NG indicates nab-paclitaxel plus gemcitabine.
adjusted for confounders showed that eligibility to either FOLFIRINOX or NG no longer exhibited any relationships with outcomes. Rather, only performance status and primary tumor location were predictive of survival (Supplemental Digital Content 1, Table 1, http://links.lww.com/AJCO/A88).
DISCUSSION Phase III clinical trials in oncology have typically endorsed the use of stringent inclusion and exclusion criteria to optimize their internal validity. However, this approach frequently comes at the expense of external validity, whereby study findings may have limited applicability and generalizability to patients in the “realworld” setting. In our population-based cohort of MPC patients, the majority of subjects were found to be ineligible to both FOLFIRINOX and NG, had these regimens been available during
4 | www.amjclinicaloncology.com
Volume 00, Number 00, ’’ 2015
the study time period. In comparison, almost twice as many patients were considered eligible for NG than FOLFIRINOX. Remarkably, simply being eligible for either regimen without actually receiving such treatment predicted for improved outcomes, underscoring the patient selection that occurs in the recruitment process for clinical trials. In this study, performance status was a key reason for treatment ineligibility and also the strongest driver of prognosis. The finding that more patients were eligible for NG than FOLFIRINOX is expected as the eligibility criteria in the MPACT study were more inclusive than those in the PRODIGE trial. In particular, the MPACT investigators allowed adult patients of all ages and individuals with worse performance status to participate, whereas the PRODIGE researchers did not. This is relevant as patients with MPC frequently report significant symptoms and experience rapid and progressive functional decline.21 Our data corroborate this observation as more than half of our patients presented with ECOGZ2 at the time of their metastatic diagnosis. Given our cohort was limited to those who received at least 1 cycle of palliative gemcitabine, the prevalence of poor performance status is likely even higher in an unselected and untreated population. In our analysis, almost twice as many patients were eligible for NG when compared with FOLFIRINOX (45% vs. 25%), mostly due to the inclusion of patients with KPSZ70, or ECOG 2, in the MPACT trial. Furthermore, the MPACT study recruited patients from both academic and community cancer hospitals, likely potentiating its generalizability. Although the MPACT trial demonstrated relatively better external validity than the PRODIGE trial, generalizability of both studies could be further improved if selected patients with abnormal bilirubin levels had been permitted to participate and receive treatment. It is estimated that over 50% of patients have jaundice at the time of diagnosis (especially if the tumor is located in the head of the pancreas) and exclusion of all of these patients can hinder external validity.22 In our study, 48% of the patients had their tumors located in the head of the pancreas, a slightly higher proportion when compared with the PRODIGE (38%) and the MPACT (43%) trials. This might explain why 19% and 25% of our patients were deemed ineligible for FOLFIRINOX and NG because of hyperbilirubinemia alone. While caution should certainly be exercised when systemic therapy is given in the context of abnormal blood chemistries, measures such as biliary stenting can decrease the risk of obstruction and allow for safe delivery of treatment. To this end, a significant proportion of patients (32%) underwent a stent placement in our cohort, as compared with only 14% in the PRODIGE trial and 17% in the MPACT trial. Another noteworthy finding is that eligibility in and of itself for either regimen was associated with improved survival, highlighting the selection of more prognostically favorable patients for trial participation.23–26 Without actually receiving NG and FOLFIRINOX, we observed that patients who simply fulfill the eligibility criteria of either the MPACT or PRODIGE trials experienced better median OS when compared with patients who failed to meet these eligibility criteria. The apparently longer OS observed in the FOLFIRINOX-eligible population when compared with the NG-eligible group likely reflects the exclusion of ECOG 2 patients in the former trial. Consistent with this hypothesis, eligibility was no longer a significant prognostic factor after accounting for ECOG performance status. This study should be interpreted in the context of several limitations. First, our cohort included only patients who had received at least 1 cycle of gemcitabine monotherapy. Although it would have been ideal to capture all patients
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of the article is prohibited.
American Journal of Clinical Oncology
Volume 00, Number 00, ’’ 2015
regardless of their receipt of systemic therapy, this was not possible as patients viewed as frail or unfit to receive treatment may have been seen solely by primary care physicians or other specialists in the community without a subsequent referral to the British Columbia Cancer Agency. Second, although most of our data points were collected prospectively, this was a retrospective analysis where we posed the hypothetical question of treatment decisions if certain regimens were available during the study time period. Therapeutic decisions are dependent on many patient, physician, and clinical factors. Our assumption that treatment would be given if patients were deemed eligible for the PRODIGE and MPACT trials is an oversimplification of a very complex phenomenon. Third, it was difficult to ascertain a precise correlation between KPS and ECOG scales although prior published literature suggests that KPSZ70 approximates ECOG 2, which is the definition used in this study.20 However, an unknown proportion of the ECOG 2 patients in our cohort would not have been eligible for MPACT trial due to a KPS of 60. In addition, KPS 60-70 patients represented only 7% of the MPACT study population, as compared with 42% of our cohort. In summary, the more inclusive strategy of permitting participation of worse performance status patients was a key factor in conferring better external validity in the MPACT trial when compared with the PRODIGE study. Future trials may need to consider refining their eligibility criteria such that results are generalizable to a larger number of patients. In the case of MPC, this warrants the careful incorporation of more subjects with poor function and abnormal blood chemistries as long as safety is maintained. Representation from academic and community cancer hospitals in future trials may further enhance generalizability. For the time being, carefully examining and applying the inclusion and exclusion criteria of the MPACT and PRODIGE studies may help to guide treatment decision making, especially when there is a lack of head-tohead comparison.
9. 10. 11.
12.
13.
14.
15.
16.
17.
18.
REFERENCES 1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2013;63:11–30. 2. American Cancer Society. Cancer Facts and Figures. 2013. Available at: http://www.cancer.org/research/cancerfactsfigures/ cancerfactsfigures/cancer-facts-figures-2013. Accessed July 24, 2014. 3. Burris HA III, Moore MJ, Andersen J, et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol. 1997;15:2403–2413. 4. Rothenberg ML, Moore MJ, Cripps MC, et al. A phase II trial of gemcitabine in patients with 5-FU-refractory pancreas cancer. Ann Oncol. 1996;7:347–353. 5. Carmichael J, Fink U, Russell RC, et al. Phase II study of gemcitabine in patients with advanced pancreatic cancer. Br J Cancer. 1996;73:101–105. 6. Casper ES, Green MR, Kelsen DP, et al. Phase II trial of gemcitabine (2,20 -difluorodeoxycytidine) in patients with adenocarcinoma of the pancreas. Invest New Drugs. 1994;12:29–34. 7. Lenzi R, Yalcin S, Evans DB, et al. Phase II study of docetaxel in patients with pancreatic cancer previously untreated with cytotoxic chemotherapy. Cancer Invest. 2002;20:464–472. 8. Whitehead RP, Jacobson J, Brown TD, et al. Phase II trial of paclitaxel and granulocyte colony-stimulating factor in patients
Copyright
r
19. 20. 21. 22. 23. 24. 25.
26.
2015 Wolters Kluwer Health, Inc. All rights reserved.
Eligibility of Metastatic Pancreatic Cancer
with pancreatic carcinoma: a Southwest Oncology Group study. J Clin Oncol. 1997;15:2414–2419. O’Reilly S, Donehower RC, Rowinsky EK, et al. A phase II trial of topotecan in patients with previously untreated pancreatic cancer. Anticancer Drugs. 1996;7:410–414. Wagener DJ, Verdonk HE, Dirix LY, et al. Phase II trial of CPT11 in patients with advanced pancreatic cancer, an EORTC early clinical trials group study. Ann Oncol. 1995;6:129–132. Abou-Alfa GK, Letourneau R, Harker G, et al. Randomized phase III study of exatecan and gemcitabine compared with gemcitabine alone in untreated advanced pancreatic cancer. J Clin Oncol. 2006;24:4441–4447. Oettle H, Richards D, Ramanathan RK, et al. A phase III trial of pemetrexed plus gemcitabine versus gemcitabine in patients with unresectable or metastatic pancreatic cancer. Ann Oncol. 2005;16:1639–1645. Louvet C, Labianca R, Hammel P, et al. Gemcitabine in combination with oxaliplatin compared with gemcitabine alone in locally advanced or metastatic pancreatic cancer: results of a GERCOR and GISCAD phase III trial. J Clin Oncol. 2005;23: 3509–3516. Van Cutsem E, van de Velde H, Karasek P, et al. Phase III trial of gemcitabine plus tipifarnib compared with gemcitabine plus placebo in advanced pancreatic cancer. J Clin Oncol. 2004;22: 1430–1438. Rocha Lima CM, Green MR, Rotche R, et al. Irinotecan plus gemcitabine results in no survival advantage compared with gemcitabine monotherapy in patients with locally advanced or metastatic pancreatic cancer despite increased tumor response rate. J Clin Oncol. 2004;22:3776–3783. Bramhall SR, Schulz J, Nemunaitis J, et al. A double-blind placebo-controlled, randomised study comparing gemcitabine and marimastat with gemcitabine and placebo as first line therapy in patients with advanced pancreatic cancer. Br J Cancer. 2002;87:161–167. Berlin JD, Catalano P, Thomas JP, et al. Phase III study of gemcitabine in combination with fluorouracil versus gemcitabine alone in patients with advanced pancreatic carcinoma: Eastern Cooperative Oncology Group Trial E2297. J Clin Oncol. 2002;20:3270–3275. Von Hoff DD, Ervin T, Arena FP, et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med. 2013;369:1691–1703. Conroy T, Desseigne F, Ychou M, et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med. 2011;364:1817–1825. Ma C, Bandukwala S, Burman D, et al. Interconversion of three measures of performance status: an empirical analysis. Eur J Cancer. 2010;46:3175–3183. Porta M, Fabregat X, Malats N, et al. Exocrine pancreatic cancer: symptoms at presentation and their relation to tumour site and stage. Clin Transl Oncol. 2005;7:189–197. George SL. Reducing patient eligibility criteria in cancer clinical trials. J Clin Oncol. 1996;14:1364–1370. Antman K, Amato D, Wood W, et al. Selection bias in clinical trials. J Clin Oncol. 1985;3:1142–1147. Bertelsen K. Protocol allocation and exclusion in two Danish randomised trials in ovarian cancer. Br J Cancer. 1991;64: 1172–1176. Hjorth M, Holmberg E, Ro¨djer S, et al. Impact of active and passive exclusions on the results of a clinical trial in multiple myeloma. The Myeloma Group of Western Sweden. Br J Haematol. 1992;80:55–61. Tas F, Sen F, Odabas H, et al. Performance status of patients is the major prognostic factor at all stages of pancreatic cancer. Int J Clin Oncol. 2013;18:839–846.
www.amjclinicaloncology.com |
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of the article is prohibited.
5
Eligibility of Metastatic Pancreatic Cancer Patients for First-Line Palliative Intent nab-Paclitaxel Plus Gemcitabine Versus FOLFIRINOX Renata D. Peixoto, MD, Maria Ho, MD, Daniel J. Renouf, MD, Howard J. Lim, MD, Sharlene Gill, MD, Jenny Y. Ruan, BSc, and Winson Y. Cheung, MD, MPH
Objectives: The PRODIGE and MPACT trials showed superiority of FOLFIRINOX and nab-paclitaxel plus gemcitabine (NG) over gemcitabine alone, respectively. However, both had strict inclusion criteria. We sought to determine the characteristics of patients with metastatic pancreatic cancer (MPC) which inform the appropriateness of first-line chemotherapy FOLFIRINOX and NG in routine practice. Materials and Methods: Patients with MPC who initiated palliative chemotherapy with gemcitabine from 2000 to 2011 at the British Columbia Cancer Agency were identified. Clinicopathologic variables and outcomes were retrospectively collected and compared among groups. Eligibility criteria for each regimen were in accordance with the respective pivotal phase III trials. Results: A total of 473 patients were included: 25% of the patients were eligible for FOLFIRINOX versus 45% for NG. Main reasons for FOLFIRINOX ineligibility were Eastern Cooperative Oncology Group (ECOG) performance status (PS)Z2 (56.5%), age older than 75 years (19.0%), and bilirubin > 1.5 upper limit of normal (18.6%), whereas those for NG ineligibility were bilirubin > upper limit of normal (24.5%), ECOG PSZ3 (14.6%), and cardiac dysfunction (13.8%). Univariate analyses revealed that FOLFIRINOX and NG-eligible patients had longer median overall survival than their respective ineligible group (8.6 vs. 4.7 mo, P < 0.001; 6.7 vs. 4.9 mo, P = 0.008, respectively). After accounting for ECOG PS in the multivariate model, however, eligibility for either FOLFIRINOX or NG no longer predicted for better overall survival. Conclusions: The majority of patients with MPC are not candidates to either NG or FOLFIRINOX due to restrictive eligibility requirements. Specific trials addressing the unmet needs of protocol ineligible patients are warranted. Key Words: pancreatic cancer, metastasis, eligibility determination, palliative therapy, survival
(Am J Clin Oncol 2015;00:000–000)
From the Division of Medical Oncology, British Columbia Cancer Agency, University of British Columbia, Vancouver, BC, Canada. Previously presented as a poster at 2014 Gastrointestinal Cancers Symposium. Merit Award recipient. Supported by the BC Cancer Foundation, BC Cancer GI Outcomes Unit, Canadian Cancer Society Research Institute, and Celgene Canada. The authors declare no conflicts of interest. Reprints: Winson Y. Cheung, MD, MPH, Division of Medical Oncology, British Columbia Cancer Agency, University of British Columbia, Vancouver, BC, Canada V5Z 4E6. E-mail: [email protected]. Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.amjclinical oncology.com. Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0277-3732/15/000-000 DOI: 10.1097/COC.0000000000000193
American Journal of Clinical Oncology
P
ancreatic cancer represents the fourth leading cause of cancer-related mortality in the developed world.1 For patients with early-stage disease, surgery followed by adjuvant therapy offers a small potential for cure and remains the treatment of choice whenever possible. However, 1.5ULN (19%). Conversely, the primary reasons for ineligibility to NG consisted of abnormal bilirubin >ULN (25%), ECOGZ3 (15%), and cardiac dysfunction (14%). Other important factors that excluded patients from both regimens as per trial eligibility were renal dysfunction (5%), no definitive histologic evidence of pancreatic cancer (7%), and prior major malignancies (14%). No patients were excluded due to brain metastases, active infection, or receipt of previous chemotherapy for MPC. Median OS for the entire cohort was 5.8 months (95% confidence interval, 5.4-6.2). Individuals who were considered eligible for both trials experienced better outcomes when compared with those deemed ineligible (median OS 8.6 vs. 4.8 mo, P < 0.001). On univariate analysis, eligible patients for FOLFIRINOX also had longer median OS than ineligible patients (8.6 vs. 4.7 mo, P < 0.001) (Fig. 2A). Likewise, eligibility to NG was associated with improved median OS (6.7 vs. 4.9 mo for eligible and ineligible groups, respectively, P = 0.008) (Fig. 2B). Additional factors that predicted for survival included ECOG, where 0 or 1 demonstrated higher median OS than Z2 (8.6 vs. 3.7 mo, P < 0.001) and primary tumor location, where head of the pancreas revealed slightly better median OS than the body or tail (6.1 vs. 5.3 mo, P = 0.025). Other variables, such as age, sex, and bilirubin levels did not appear to impact on outcomes (P = 0.63, 0.07, and 0.48, respectively). Importantly, multivariate analysis that
TABLE 1. Baseline Characteristics of Study Cohort
Characteristics
All Patients (n = 473)
Age (y) Median Sex (%) Male ECOG Performance Status Score (%) 0 1 2 Pancreatic tumor location (%) Head Body Tail Unknown Histologically proven (%) Yes Abnormal bilirubin (%) Yes Site(s) of metastasis (%) Liver only Lymph node only Peritoneum only One extrahepatic only Multiple Prior stent (%) Yes ECOG indicates Eastern Cooperative Oncology Group.
Copyright
r
66 54.5 2.7 40.8 41.9 48 24 24.9 3 93.4 24.5 35.7 3.6 7.8 4.4 48.4 31.5
FIGURE 1. A, A total of 117 patients were considered eligible for FOLFIRINOX versus 214 for nab-paclitaxel + gemcitabine. B, Main reasons for FOLFIRINOX and NG ineligibility. Top reasons for ineligibility to: (1) NG were abnormal bilirubin >ULN (25%), ECOGZ3 (15%), and cardiac dysfunction (14%); and (2) FOLFIRINOX were ECOGZ2 (57%), age older than 76 years (19%), and abnormal bilirubin >1.5 ULN (19%). ECOG indicates Eastern Cooperative Oncology Group; NG, nab-paclitaxel plus gemcitabine; ULN, upper limit of normal.
2015 Wolters Kluwer Health, Inc. All rights reserved.
www.amjclinicaloncology.com |
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of the article is prohibited.
3
American Journal of Clinical Oncology
Peixoto et al
A
-- FOLFIRINOX-eligible
-- FOLFIRINOX-ineligible
B
-- NG-eligible
-- NG-ineligible
FIGURE 2. A, Median overall survival from diagnosis was 4.7 versus 8.6 months (P < 0.001; hazard ratio [HR] 1.73) in the FOLFIRINOX-ineligible versus FOLFIRINOX-eligible patients. B, Median overall survival from diagnosis was 4.9 versus 6.7 months (P = 0.008; HR 1.28) in the NG-ineligible versus NG-eligible patients. NG indicates nab-paclitaxel plus gemcitabine.
adjusted for confounders showed that eligibility to either FOLFIRINOX or NG no longer exhibited any relationships with outcomes. Rather, only performance status and primary tumor location were predictive of survival (Supplemental Digital Content 1, Table 1, http://links.lww.com/AJCO/A88).
DISCUSSION Phase III clinical trials in oncology have typically endorsed the use of stringent inclusion and exclusion criteria to optimize their internal validity. However, this approach frequently comes at the expense of external validity, whereby study findings may have limited applicability and generalizability to patients in the “realworld” setting. In our population-based cohort of MPC patients, the majority of subjects were found to be ineligible to both FOLFIRINOX and NG, had these regimens been available during
4 | www.amjclinicaloncology.com
Volume 00, Number 00, ’’ 2015
the study time period. In comparison, almost twice as many patients were considered eligible for NG than FOLFIRINOX. Remarkably, simply being eligible for either regimen without actually receiving such treatment predicted for improved outcomes, underscoring the patient selection that occurs in the recruitment process for clinical trials. In this study, performance status was a key reason for treatment ineligibility and also the strongest driver of prognosis. The finding that more patients were eligible for NG than FOLFIRINOX is expected as the eligibility criteria in the MPACT study were more inclusive than those in the PRODIGE trial. In particular, the MPACT investigators allowed adult patients of all ages and individuals with worse performance status to participate, whereas the PRODIGE researchers did not. This is relevant as patients with MPC frequently report significant symptoms and experience rapid and progressive functional decline.21 Our data corroborate this observation as more than half of our patients presented with ECOGZ2 at the time of their metastatic diagnosis. Given our cohort was limited to those who received at least 1 cycle of palliative gemcitabine, the prevalence of poor performance status is likely even higher in an unselected and untreated population. In our analysis, almost twice as many patients were eligible for NG when compared with FOLFIRINOX (45% vs. 25%), mostly due to the inclusion of patients with KPSZ70, or ECOG 2, in the MPACT trial. Furthermore, the MPACT study recruited patients from both academic and community cancer hospitals, likely potentiating its generalizability. Although the MPACT trial demonstrated relatively better external validity than the PRODIGE trial, generalizability of both studies could be further improved if selected patients with abnormal bilirubin levels had been permitted to participate and receive treatment. It is estimated that over 50% of patients have jaundice at the time of diagnosis (especially if the tumor is located in the head of the pancreas) and exclusion of all of these patients can hinder external validity.22 In our study, 48% of the patients had their tumors located in the head of the pancreas, a slightly higher proportion when compared with the PRODIGE (38%) and the MPACT (43%) trials. This might explain why 19% and 25% of our patients were deemed ineligible for FOLFIRINOX and NG because of hyperbilirubinemia alone. While caution should certainly be exercised when systemic therapy is given in the context of abnormal blood chemistries, measures such as biliary stenting can decrease the risk of obstruction and allow for safe delivery of treatment. To this end, a significant proportion of patients (32%) underwent a stent placement in our cohort, as compared with only 14% in the PRODIGE trial and 17% in the MPACT trial. Another noteworthy finding is that eligibility in and of itself for either regimen was associated with improved survival, highlighting the selection of more prognostically favorable patients for trial participation.23–26 Without actually receiving NG and FOLFIRINOX, we observed that patients who simply fulfill the eligibility criteria of either the MPACT or PRODIGE trials experienced better median OS when compared with patients who failed to meet these eligibility criteria. The apparently longer OS observed in the FOLFIRINOX-eligible population when compared with the NG-eligible group likely reflects the exclusion of ECOG 2 patients in the former trial. Consistent with this hypothesis, eligibility was no longer a significant prognostic factor after accounting for ECOG performance status. This study should be interpreted in the context of several limitations. First, our cohort included only patients who had received at least 1 cycle of gemcitabine monotherapy. Although it would have been ideal to capture all patients
Copyright
r
2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of the article is prohibited.
American Journal of Clinical Oncology
Volume 00, Number 00, ’’ 2015
regardless of their receipt of systemic therapy, this was not possible as patients viewed as frail or unfit to receive treatment may have been seen solely by primary care physicians or other specialists in the community without a subsequent referral to the British Columbia Cancer Agency. Second, although most of our data points were collected prospectively, this was a retrospective analysis where we posed the hypothetical question of treatment decisions if certain regimens were available during the study time period. Therapeutic decisions are dependent on many patient, physician, and clinical factors. Our assumption that treatment would be given if patients were deemed eligible for the PRODIGE and MPACT trials is an oversimplification of a very complex phenomenon. Third, it was difficult to ascertain a precise correlation between KPS and ECOG scales although prior published literature suggests that KPSZ70 approximates ECOG 2, which is the definition used in this study.20 However, an unknown proportion of the ECOG 2 patients in our cohort would not have been eligible for MPACT trial due to a KPS of 60. In addition, KPS 60-70 patients represented only 7% of the MPACT study population, as compared with 42% of our cohort. In summary, the more inclusive strategy of permitting participation of worse performance status patients was a key factor in conferring better external validity in the MPACT trial when compared with the PRODIGE study. Future trials may need to consider refining their eligibility criteria such that results are generalizable to a larger number of patients. In the case of MPC, this warrants the careful incorporation of more subjects with poor function and abnormal blood chemistries as long as safety is maintained. Representation from academic and community cancer hospitals in future trials may further enhance generalizability. For the time being, carefully examining and applying the inclusion and exclusion criteria of the MPACT and PRODIGE studies may help to guide treatment decision making, especially when there is a lack of head-tohead comparison.
9. 10. 11.
12.
13.
14.
15.
16.
17.
18.
REFERENCES 1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA Cancer J Clin. 2013;63:11–30. 2. American Cancer Society. Cancer Facts and Figures. 2013. Available at: http://www.cancer.org/research/cancerfactsfigures/ cancerfactsfigures/cancer-facts-figures-2013. Accessed July 24, 2014. 3. Burris HA III, Moore MJ, Andersen J, et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol. 1997;15:2403–2413. 4. Rothenberg ML, Moore MJ, Cripps MC, et al. A phase II trial of gemcitabine in patients with 5-FU-refractory pancreas cancer. Ann Oncol. 1996;7:347–353. 5. Carmichael J, Fink U, Russell RC, et al. Phase II study of gemcitabine in patients with advanced pancreatic cancer. Br J Cancer. 1996;73:101–105. 6. Casper ES, Green MR, Kelsen DP, et al. Phase II trial of gemcitabine (2,20 -difluorodeoxycytidine) in patients with adenocarcinoma of the pancreas. Invest New Drugs. 1994;12:29–34. 7. Lenzi R, Yalcin S, Evans DB, et al. Phase II study of docetaxel in patients with pancreatic cancer previously untreated with cytotoxic chemotherapy. Cancer Invest. 2002;20:464–472. 8. Whitehead RP, Jacobson J, Brown TD, et al. Phase II trial of paclitaxel and granulocyte colony-stimulating factor in patients
Copyright
r
19. 20. 21. 22. 23. 24. 25.
26.
2015 Wolters Kluwer Health, Inc. All rights reserved.
Eligibility of Metastatic Pancreatic Cancer
with pancreatic carcinoma: a Southwest Oncology Group study. J Clin Oncol. 1997;15:2414–2419. O’Reilly S, Donehower RC, Rowinsky EK, et al. A phase II trial of topotecan in patients with previously untreated pancreatic cancer. Anticancer Drugs. 1996;7:410–414. Wagener DJ, Verdonk HE, Dirix LY, et al. Phase II trial of CPT11 in patients with advanced pancreatic cancer, an EORTC early clinical trials group study. Ann Oncol. 1995;6:129–132. Abou-Alfa GK, Letourneau R, Harker G, et al. Randomized phase III study of exatecan and gemcitabine compared with gemcitabine alone in untreated advanced pancreatic cancer. J Clin Oncol. 2006;24:4441–4447. Oettle H, Richards D, Ramanathan RK, et al. A phase III trial of pemetrexed plus gemcitabine versus gemcitabine in patients with unresectable or metastatic pancreatic cancer. Ann Oncol. 2005;16:1639–1645. Louvet C, Labianca R, Hammel P, et al. Gemcitabine in combination with oxaliplatin compared with gemcitabine alone in locally advanced or metastatic pancreatic cancer: results of a GERCOR and GISCAD phase III trial. J Clin Oncol. 2005;23: 3509–3516. Van Cutsem E, van de Velde H, Karasek P, et al. Phase III trial of gemcitabine plus tipifarnib compared with gemcitabine plus placebo in advanced pancreatic cancer. J Clin Oncol. 2004;22: 1430–1438. Rocha Lima CM, Green MR, Rotche R, et al. Irinotecan plus gemcitabine results in no survival advantage compared with gemcitabine monotherapy in patients with locally advanced or metastatic pancreatic cancer despite increased tumor response rate. J Clin Oncol. 2004;22:3776–3783. Bramhall SR, Schulz J, Nemunaitis J, et al. A double-blind placebo-controlled, randomised study comparing gemcitabine and marimastat with gemcitabine and placebo as first line therapy in patients with advanced pancreatic cancer. Br J Cancer. 2002;87:161–167. Berlin JD, Catalano P, Thomas JP, et al. Phase III study of gemcitabine in combination with fluorouracil versus gemcitabine alone in patients with advanced pancreatic carcinoma: Eastern Cooperative Oncology Group Trial E2297. J Clin Oncol. 2002;20:3270–3275. Von Hoff DD, Ervin T, Arena FP, et al. Increased survival in pancreatic cancer with nab-paclitaxel plus gemcitabine. N Engl J Med. 2013;369:1691–1703. Conroy T, Desseigne F, Ychou M, et al. FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer. N Engl J Med. 2011;364:1817–1825. Ma C, Bandukwala S, Burman D, et al. Interconversion of three measures of performance status: an empirical analysis. Eur J Cancer. 2010;46:3175–3183. Porta M, Fabregat X, Malats N, et al. Exocrine pancreatic cancer: symptoms at presentation and their relation to tumour site and stage. Clin Transl Oncol. 2005;7:189–197. George SL. Reducing patient eligibility criteria in cancer clinical trials. J Clin Oncol. 1996;14:1364–1370. Antman K, Amato D, Wood W, et al. Selection bias in clinical trials. J Clin Oncol. 1985;3:1142–1147. Bertelsen K. Protocol allocation and exclusion in two Danish randomised trials in ovarian cancer. Br J Cancer. 1991;64: 1172–1176. Hjorth M, Holmberg E, Ro¨djer S, et al. Impact of active and passive exclusions on the results of a clinical trial in multiple myeloma. The Myeloma Group of Western Sweden. Br J Haematol. 1992;80:55–61. Tas F, Sen F, Odabas H, et al. Performance status of patients is the major prognostic factor at all stages of pancreatic cancer. Int J Clin Oncol. 2013;18:839–846.
www.amjclinicaloncology.com |
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of the article is prohibited.
5
