Figures
Abstract
Background
Immunotherapy using high dose interleukin-2 (HD IL2) in patients with renal cell carcinoma (RCC) and melanoma is associated with severe toxicities. The association between annual hospital volume of HD IL2 and inpatient mortality is not well studied. In this study we aim to quantify the impact of annual hospital volume of HD IL2 on inpatient mortality using National Inpatient Sample (NIS) data.
Methods
We did a cross-sectional study using NIS, one of the largest inpatient datasets in United States, from 2003 to 2011. Patients with melanoma and RCC receiving HD IL2 were identified by ICD9 procedure code 00.15. The primary outcome was inpatient mortality. Using Joinpoint regression, which detects change in trend of inpatient mortality with change in annual volume, the hospitals were classified in three volume categories (low: 1–40, medium: 41–120, high: >120). Multivariate logistic regression was used to identify predictors of inpatient mortality controlling for confounders.
Results
From 2003 to 2011, 29,532 patients with RCC or melanoma who received HD IL2 were identified, and 124 died during the hospitalization (0.4%). The hospitals with low, medium and high annual volume had significant difference in inpatient mortality (0.83%, 0.29% and 0.13% respectively, p = 0.0003). On multivariate analysis, low volume hospitals were associated with significantly higher odds of inpatient mortality (OR 6.1, 95% CI 1.6–23.2, p = 0.003) as compared to high volume hospitals. Additionally, the hospitals with annual volume of 1–20 had even higher rates (1.31% vs. 0.13%, p<0.0001) and multivariate odds (OR 8.9, 95% CI 2.4–33.2, p = 0.0006) of inpatient mortality as compared to high volume hospitals.
Conclusions
Lower annual hospital volume of HD IL2 is associated with worse outcomes. Annual hospital volume of 1–40 and 1–20 treatments per year is associated with 6 and 9 times higher odds of inpatient mortality respectively as compared to high volume hospitals. Our findings provide preliminary evidence for a volume-outcome relationship for RCC and melanoma patients undergoing HD IL2 treatment. They support future volume-outcome analyses in relation to other anti-cancer therapies that require special training and expertise.
Citation: Mehta K, Appleman L, Wang H, Tarhini AA, Parikh RA (2016) Annual Hospital Volume of High Dose Interleukin-2 and Inpatient Mortality in Melanoma and Renal Cell Carcinoma Patients. PLoS ONE 11(1): e0147153. https://doi.org/10.1371/journal.pone.0147153
Editor: Gregory Lesinski, The Ohio State University, UNITED STATES
Received: October 20, 2015; Accepted: December 29, 2015; Published: January 22, 2016
Copyright: © 2016 Mehta et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: NIS data are available from the Healthcare Cost and Utilization Project (https://www.hcup-us.ahrq.gov/tech_assist/centdist.jsp). Data are available to all researchers following a standard application process and signing of a data use agreement.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: HD IL2, High dose interleukin-2; RCC, Renal cell carcinoma; NIS, National inpatient sample; U.S., United States; HCUP, Healthcare cost and utilization project; ICD9-CM, International classification of diseases 9th edition–clinical modification; CCI, Charlson co-morbidity index; FDA, Food and Drug Administration
Introduction
HD IL2 is used in treatment of selected patients with metastatic renal cell carcinoma (RCC) and metastatic melanoma due to its ability of inducing complete and durable responses in these patients.[1] Initial studies had shown that HD IL2 leads to partial response in 10% of patients and complete response in 6% of patients with metastatic melanoma and an overall response rate of 14% in renal cell carcinoma.[2,3] In a large prospective, multicenter study with one hundred and twenty seven evaluable subjects (Select), HD IL2 led to partial responses in 23% and complete response in 3% of patients with metastatic RCC.[4] These responses were seen in both good and poor risk subjects and positively associated with tumor PD-L1 expression.[4] HD IL2 use is associated with hypotension, cardiac arrhythmias, myocardial infarction, capillary-leak syndrome, reduced organ perfusion and rarely death. Death related to sepsis during HD IL2 was seen in approximately 2% of melanoma patients in a retrospective analysis.[5] These significant HD IL2 associated toxicities have limited its use to specialized centers. Despite several new options in the treatment of metastatic melanoma and RCC, HD IL2 continues to be widely used as salvage therapy. There are several ongoing clinical trials testing HD IL2 in combination with newer agents in patients with metastatic melanoma and RCC. The 2014 expert consensus on HD IL2 states that "Treating a minimum number of patients per year is important, as quality depends upon familiarity and repetition".[6] A clear association between annual volume of HD IL2 and clinical outcomes has not been studied. We hypothesized that higher annual HD IL2 volume is associated with lower inpatient death rates during HD IL2 administration.
Material and Methods
We used the National Inpatient Sample (NIS), one of the largest publicly available inpatient dataset in United States (U.S.) for our analysis. NIS is published by Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality.[7] NIS represents a 20% stratified random sample of discharges from all hospitals, excluding rehabilitation and long-term acute care hospitals. NIS is drawn from all States participating in HCUP, representing more than 95 percent of the U.S. population and uses National Health Survey Strata to weigh each participating hospital. Discharge weights are provided for each entry and are used to project to a nationally representative population. The details of NIS methodology have been previously published by the HCUP.[8]
We queried NIS database between 2003 and 2011 for patients with melanoma and RCC by using international classification of diseases 9th edition–clinical modification (ICD9-CM) diagnostic codes published by HCUP.[9] From this sample, patients receiving IL2 were identified by ICD9-CM procedure code 00.15. Annual hospital volume was calculated by the unique hospital number available in the dataset. Using join-point regression software, which detects change in trend of inpatient mortality with change in annual hospital volume, the hospitals were classified in three volume categories (low, medium and high). Join-point regression uses the grid search method to detect points at which significant changes in the direction and magnitude of trend of dependent variable (mortality) with reference to the independent variable (hospital volume) occur, under the assumption of constant variance and uncorrelated errors.[10] The severity of co-morbid conditions was quantified by calculating the Deyo’s modification of Charlson co-morbidity index (CCI).[11,12] The index ranges from 0–33, with higher scores indicating severe co-morbid conditions. Since all patients in our sample had cancer, we excluded cancer from the calculation of index.
Chi-square test was used to test significance of difference in categorical variables. One-way ANOVA was used to test significance of difference in continuous variables. Multivariate logistic regression was used to identify predictors of inpatient mortality controlling for possible confounders (Age, sex, co-morbid conditions, type of malignancy (RCC vs. melanoma), calendar year, urban location of the hospital and teaching status of the hospital). Due to large amount of missing data, race was not included in the regression model. Using SAS 9.2 (SAS Institute Inc., Cary, NC, USA), SURVEY procedures were used to adjust for stratified design of NIS.
Results
From 2003 to 2011, the NIS contained 29,532 patients (weighted number) with RCC (weighted N = 8,969) or melanoma (weighted N = 20,563) who received HD IL2. The annual number of HD IL2 treatments captured by NIS is described in Fig 1. Utilization of HD IL2 for RCC and melanoma appears to be declining after 2005 and 2009 respectively (Fig 1). The annual hospital volume for HD IL2 ranged from 1 treatment per year to 286 per year with the median of 71 per year. The Joinpoint regression classified the hospitals in 3 volume categories (low: 1–40, medium: 41–120, high: >120) based on change in trend of mortality. There was no difference in age, sex, and race of the patients in different hospital volume categories (Table 1). The high volume hospitals had lesser proportion of RCC patients (16.9% vs. 36.2% vs. 37.9%, p = 0.01) as compared to medium and low volume hospitals. The patients admitted to low volume hospitals had higher burden of co-morbid conditions (CCI of 0.3 vs. 0.1 vs. 0.1, p = 0.0004) as compared to medium and high volume hospitals. High volume hospitals were more likely to be in urban location as compared to medium and low volume hospitals (100% vs. 85.9% vs. 96.1%, p<0.001). There was no difference in teaching status of the hospitals among three volume categories.
Among the 29,532 patients receiving HD IL2, 124 died during the hospitalization (0.4%). The higher hospital volume cut-offs showed consistent reduction in inpatient mortality (Fig 2). The hospitals with low, medium and high annual volume had significant difference in inpatient mortality (0.83%, 0.29% and 0.13% respectively, p = 0.0003). On multivariate analysis, the low volume hospitals were associated with higher odds of inpatient mortality (OR 6.1, 95% CI 1.6–23.2, p = 0.003) as compared to high volume hospitals. Additionally, the hospitals with annual volume of 1–20 had even higher rates (1.31% vs. 0.13%, p<0.0001) and multivariate odds (OR 8.9, 95% CI 2.4–33.2, p = 0.0006) of inpatient mortality as compared to high volume hospitals. Age, sex, co-morbid conditions, type of malignancy (RCC vs. melanoma), calendar year, urban location of the hospital and teaching status of the hospital were not associated with increased inpatient mortality (Table 2). Additionally, the patients admitted to low volume hospitals had higher length of stay (5.5 days vs. 4.7 days vs. 4.4 days, p<0.0001) and total charges ($93,376 vs. $73,332 vs. $61,813, p = 0.0002) as compared to medium and high volume hospitals.
Annual hospital volume of high dose interleukin– 2 admissions is plotted as cumulative frequency on X–Axis.
Discussion
Since the Food and Drug Administration’s (FDA) approval of HD IL2 for RCC in 1992[13] and for melanoma in 1998,[14] it has been widely used in treatment of both malignancies. Since approval of Sorafenib (2005)[15] and Sunitinib (2006)[16] by FDA, the utilization of HD IL2 has reduced for RCC patients (Fig 1). Ipilimumab, which inhibits cytotoxic T-lymphocyte-associated antigen 4 was shown to provide survival advantage, with or without a gp100 peptide vaccine, as compared with gp100 alone, in patients with previously treated metastatic melanoma in 2010;[17]. This lead to FDA approval for ipilimumab in 2011.[18] FDA also approved Vemurafenib in 2011 and dabrafenib for melanoma patients with the BRAF V600E mutation.[19–21] Currently, the standard of care for patients with metastatic melanoma with the BRAF V600E mutation is a combination of BRAF and MEK inhibitors.[22,23] After introduction of these new therapies the utilization of HD IL2 has reduced for melanoma patients. While most of initial experience with HD IL2 was at National Cancer Institute, gradually more and more hospitals have started treating patients with HD IL2 therapy. Currently, there are more than 60 centers in the U.S. offering IL2 therapy.[6] The management of HD IL2 therapy and treatment related adverse events vary considerably among centers. It has been argued that the centers with more experience in HD IL2 may perform better patient selection and manage toxicities with more diligence leading to lower treatment related mortality.[6] Data presented in this study clearly shows that higher hospital volume is associated with lower mortality. Moreover, the patients selected for HD IL2 treatment at lower volume centers may have higher co-morbid conditions, which may have contributed to higher mortality. The higher length of stay and higher total charges at low volume hospital could be attributed to higher complication rates. Thus, the results of this study lend evidence to support the hypothesis of volume-outcome relationship of HD IL2 treatment.
The volume-outcome relationship is well studied in surgical procedures like pancreatectomy and esophagectomy.[24] Prior studies have shown similar volume outcome relationship in allogeneic bone marrow transplant.[25,26] However, the data on volume-outcome relationship of other medical treatments for cancers is very limited. As a result, the Institute of Medicine’s workshop on volume-outcome relationship identified cancer treatment and non-surgical interventions as a new area of research.[27] To our knowledge, this study is the first study to report volume-outcome relationship of HD IL2 therapy. Our study has important implications for RCC and melanoma patients who are considering HD IL2 treatment and physicians who are referring their patients for HD IL2 therapy at specialty centers. With decrease in overall utilization of HD IL2 and increase in number of centers offering HD IL2 therapy, the annual hospital volume of HD IL2 treatments may decline in near future. This may pose a unique challenge to our healthcare system, to provide safe and high quality care to our patients who choose to undergo HD IL2 treatments.
Our study has several limitations. The NIS does not contain the number of the HD IL2 doses received or number of doses missed by the patient. While the NIS contains a variable for the attending physician, we could not study the impact of annual physician volume of HD IL2 treatment on outcomes due to large amount of missing data. Recently, there is a shift to offer a number of therapies including HD IL2 as outpatient. Due to the nature of the NIS database, we are unable to capture this information and compare it to inpatient HD IL2 use. Due to lack of post-discharge follow up data, we could not study the overall survival. Finally, the NIS does not contain data on vital signs, laboratory values, radiological data, EKG or weight of the patient. Despite these limitations, our study represents the largest sample of patients, who have received HD IL2 for melanoma or RCC.
Conclusions
Lower annual hospital volume of HD IL2 is associated with worse outcomes. Annual hospital volume of 1–40 and 1–20 treatments per year is associated with 6 and 9 times higher odds of inpatient mortality respectively as compared to high volume hospitals. Our findings provide preliminary evidence for a volume-outcome relationship for RCC and melanoma patients undergoing HD IL2 treatment. They support future volume-outcome analyses in relation to other anti-cancer therapies that require special training and expertise.
Author Contributions
Conceived and designed the experiments: KM LA AT RP. Analyzed the data: KM HW. Contributed reagents/materials/analysis tools: KM HW RP. Wrote the paper: KM LA HW AT RP.
References
- 1. Rosenberg SA, Yang JC, Topalian SL, Schwartzentruber DJ, Weber JS, et al. (1994) Treatment of 283 consecutive patients with metastatic melanoma or renal cell cancer using high-dose bolus interleukin 2. JAMA 271: 907–913. pmid:8120958
- 2. Atkins MB, Kunkel L, Sznol M, Rosenberg SA (2000) High-dose recombinant interleukin-2 therapy in patients with metastatic melanoma: long-term survival update. Cancer J Sci Am 6 Suppl 1: S11–14. pmid:10685652
- 3. Fyfe G, Fisher RI, Rosenberg SA, Sznol M, Parkinson DR, et al. (1995) Results of treatment of 255 patients with metastatic renal cell carcinoma who received high-dose recombinant interleukin-2 therapy. J Clin Oncol 13: 688–696. pmid:7884429
- 4. McDermott DF, Cheng SC, Signoretti S, Margolin KA, Clark JI, et al. (2015) The high-dose aldesleukin "select" trial: a trial to prospectively validate predictive models of response to treatment in patients with metastatic renal cell carcinoma. Clin Cancer Res 21: 561–568. pmid:25424850
- 5. Atkins MB, Lotze MT, Dutcher JP, Fisher RI, Weiss G, et al. (1999) High-dose recombinant interleukin 2 therapy for patients with metastatic melanoma: analysis of 270 patients treated between 1985 and 1993. J Clin Oncol 17: 2105–2116. pmid:10561265
- 6. Dutcher JP, Schwartzentruber DJ, Kaufman HL, Agarwala SS, Tarhini AA, Lowder JN, et al. (2014) High dose interleukin-2 (Aldesleukin)–expert consensus on best management practices-2014. Journal for ImmunoTherapy of Cancer 2: 1–23.
- 7.
(2012) HCUP National Inpatient Sample (NIS). Rockville, MD: Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality.
- 8. Steiner C, Elixhauser A, Schnaier J (2002) The healthcare cost and utilization project: an overview. Eff Clin Pract 5: 143–151. pmid:12088294
- 9.
HCUP (2015) Appendix A—Clinical Classification Software-DIAGNOSES (January 1980 through September 2015). Healthcare Cost and Utilization Project.
- 10. Kim HJ, Luo J, Kim J, Chen HS, Feuer EJ (2014) Clustering of trend data using joinpoint regression models. Stat Med 33: 4087–4103. pmid:24895073
- 11. Deyo RA, Cherkin DC, Ciol MA (1992) Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 45: 613–619. pmid:1607900
- 12. Charlson M, Szatrowski TP, Peterson J, Gold J (1994) Validation of a combined comorbidity index. J Clin Epidemiol 47: 1245–1251. pmid:7722560
- 13. Dutcher J (2002) Current status of interleukin-2 therapy for metastatic renal cell carcinoma and metastatic melanoma. Oncology (Williston Park) 16: 4–10.
- 14. Bhatia S, Tykodi SS, Thompson JA (2009) Treatment of metastatic melanoma: an overview. Oncology (Williston Park) 23: 488–496.
- 15.
FDA (2013) Approved Drugs > Sorafenib (NEXAVAR). Food and Drug Administration.
- 16.
FDA (2006) 2006 > FDA Approves New Treatment for Gastrointestinal and Kidney Cancer. Food and Drug Administration. 2006
- 17. Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA, et al. (2010) Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 363: 711–723. pmid:20525992
- 18.
FDA (2011) Press Announcements > FDA approves new treatment for a type of late-stage skin cancer. U.S. Food and Drug Administration.
- 19.
FDA (2011) Press Announcements > FDA approves Zelboraf and companion diagnostic test for late-stage skin cancer Food and Drug Administration.
- 20. Hauschild A, Grob JJ, Demidov LV, Jouary T, Gutzmer R, et al. (2012) Dabrafenib in BRAF-mutated metastatic melanoma: a multicentre, open-label, phase 3 randomised controlled trial. Lancet 380: 358–365. pmid:22735384
- 21. Chapman PB, Hauschild A, Robert C, Haanen JB, Ascierto P, et al. (2011) Improved survival with vemurafenib in melanoma with BRAF V600E mutation. N Engl J Med 364: 2507–2516. pmid:21639808
- 22. Long GV, Stroyakovskiy D, Gogas H, Levchenko E, de Braud F, et al. (2014) Combined BRAF and MEK inhibition versus BRAF inhibition alone in melanoma. N Engl J Med 371: 1877–1888. pmid:25265492
- 23. Larkin J, Ascierto PA, Dreno B, Atkinson V, Liszkay G, et al. (2014) Combined vemurafenib and cobimetinib in BRAF-mutated melanoma. N Engl J Med 371: 1867–1876. pmid:25265494
- 24. Begg CB, Cramer LD, Hoskins WJ, Brennan MF (1998) Impact of hospital volume on operative mortality for major cancer surgery. JAMA 280: 1747–1751. pmid:9842949
- 25. Horowitz MM, Przepiorka D, Champlin RE, Gale RP, Gratwohl A, et al. (1992) Should HLA-identical sibling bone marrow transplants for leukemia be restricted to large centers? Blood 79: 2771–2774. pmid:1586723
- 26. Hows J, Bradley BA, Gore S, Downie T, Howard M, et al. (1993) Prospective evaluation of unrelated donor bone marrow transplantation. The International Marrow Unrelated Search and Transplant (IMUST) Study. Bone Marrow Transplant 12: 371–380. pmid:8275037
- 27.
Board IoMUaNRCUNCP (2001) Interpreting the Volume–Outcome Relationship in the Context of Cancer Care. In: Hewitt M PD, editor. Washington (DC): National Academies Press (US).