Cost of peripheral neuropathy in patients receiving treatment for multiple myeloma: a US administrative claims analysis

Abstract

Background:

Peripheral neuropathy (PN) is a common consequence of multiple myeloma (MM) among those commonly treated with older-generation proteasome inhibitors (PIs). In this study, we evaluated the economic burden attributable to PN among MM patients in real-world practice settings in the US.

Methods:

Adults diagnosed with MM and first treated (index event) between 1 July 2006 and 28 February 2017 were identified from MarketScan® Commercial and Medicare claim databases. Continuous enrollment for at least 12 months without treatment and PN diagnoses were required pre-index. Patients were followed for at least 3 months until inpatient death or end of data. The International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM) and ICD-10-CM diagnosis codes were used to identify PN. Propensity-score matching was applied to match every patient with PN to two MM patients without a PN diagnosis (controls). Healthcare utilization and expenditures per patient per month (PPPM) in the postindex period were estimated.

Results:

Of 11,851 patients meeting the study criteria, 15.5% had PN. After matching 1387 patients with PN and 2594 controls were identified. Baseline characteristics were well balanced between cohorts; mean follow up was 23–26 months. PPPM total costs were significantly higher by $1509 for patients with PN than controls, driven by higher hospitalization (PN 77.4%, controls 67.2%; p < 0.001) and emergency department rates (PN 67.8%, controls 58.4%; p < 0.001) and more outpatient hospital-based visits PPPM (PN 13.5 ± 14.7, controls 11.5 ± 18.0; p < 0.001).

Conclusions:

PN is a prevalent MM treatment complication associated with a significant economic burden adding to the complexity and cost of MM treatment. Highly effective novel treatments such as carfilzomib may reduce the overall disease burden.

Keywords

chemotherapy-induced neuropathy healthcare costs line of therapy multiple myeloma peripheral neuropathy

Introduction

Multiple myeloma (MM) is a systemic malignancy of plasma cells in bone marrow, with approximately 30,280 incident cases estimated to be diagnosed in the United States in 2017 (17,940 men and 12,790 women). MM primarily affects the elderly population since the median age at incidence is 69 years.^1,2 For several decades, the standard treatment approach has been induction therapy followed by autologous stem-cell transplantation for transplant-eligible patients and high-dose chemotherapy for other patients. While the recent introduction of many novel therapies such as immunomodulating agents, proteasome inhibitors (PIs), monoclonal antibodies, and histone deacetylase inhibitor have shown improved response, progression-free survival^3–9 and survival rates¹⁰ in clinical trials, MM remains an incurable malignancy for the majority of patients.¹¹

Peripheral neuropathy (PN) is a common complication of MM and one of the main dose-limiting iatrogenic toxicities associated with some antimyeloma treatments, including older-generation PIs such as bortezomib.^12–16 The incidence of PN associated with MM treatments during the course of the disease and treatment has been estimated to range between 21% and 70% depending on the treatment and PN severity.¹⁵ MM treatment options are often limited by the likelihood of therapies causing or exacerbating neuropathies with significant negative impacts on patient quality of life.^14,16 Treatment-induced PN, although usually reversible, can cause severe pain and affect the patient’s activities of daily living, as well as causing serious problems, such as loss of sensation, balance issues, muscle weakness, and organ failure.^15–18

Management of PN is an ongoing challenge for healthcare providers. Actions to mitigate the incidence and effects of PN may include dosage and dosing-schedule adjustments of PIs, PI treatment discontinuation, other modalities (e.g. electrical nerve stimulation, physical therapy, acupuncture, other), and medications to address the pain (e.g. topical analgesics, antidepressants, antiepileptics, opioids).^16,19 These treatment measures for managing the effects of PN also carry an economic burden for the healthcare system (e.g. increased visits to monitor or treat the PN, costs of PN treatment), as well as significant financial and lifestyle repercussions for the MM patient.

The impact of PN on healthcare utilization and costs for MM patients treated in real-world practice settings is not well understood. The purpose of this study was to use real-world data to examine the healthcare resource utilization and costs associated with PN in patients being treated for MM, comparing those diagnosed with PN to a matched control cohort without PN.

Methods

Study design and data source

This retrospective, observational cohort study was based on administrative healthcare claims data from 1 January 2006 to 28 February 2017 from the IBM® MarketScan Commercial Claims and Encounters (Commercial) and Medicare Supplemental and Coordination of Benefits (Medicare) databases. These databases include inpatient medical, outpatient medical, and outpatient pharmacy claims data, as well as insurance enrollment and demographic information collected from a wide variety of health plans across the US. The Commercial database includes information for over 20 million individuals annually who are under the age of 65 years with employer-sponsored health insurance, including the primary insured, spouses and dependents. The Medicare Supplemental database includes both the Medicare-paid and supplemental-paid components of reimbursed insurance claims information for over 2 million individuals annually with both traditional and supplemental Medicare coverage. The study databases satisfy Sections 164.514 (a)–(b)(1)(ii) of the Health Insurance Portability and Accountability Act of 1996 privacy rule (HIPAA) regarding the determination and documentation of statistically de-identified data. This study used only de-identified patient records and did not involve the collection, use, or transmittal of individually identifiable data, and therefore institutional review board approval was not required.

Patient identification

Selected adult patients, 18 years of age and older, had at least one inpatient or two outpatient claims (from 30 to 365 days of the first found outpatient claim) with a diagnosis of MM based on International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM) diagnosis codes 203.0x or ICD-10-CM (tenth revision) diagnosis codes C9000, C9001 and C9002 between 1 July 2006 and 28 February 2017, and at least one claim indicating the administration or prescription of an MM therapy (bendamustine, bortezomib, carfilzomib, cisplatin, cyclophosphamide, doxorubicin, doxorubicin liposomal, lenalidomide, melphalan, panobinostat, pomalidomide, thalidomide) on or after the date of the first MM diagnosis. Claims associated with a diagnostic workup such as laboratory tests or diagnostic X-rays were not used for patient selection. To ensure that patients were newly diagnosed and newly treated, a 12-month period with no diagnosis of MM prior to the first found (initial) MM diagnosis and a 12-month period with no MM therapy prior to the initial MM therapy was required. The index date was the date of the first claim for one of the MM therapies on or after the initial MM diagnosis. Continuous medical and prescription coverage was required for at least 12 months prior to the index date (preperiod), and for at least 3 months after the index date (postperiod). Patients were followed from the index date until the earliest evidence of inpatient death (via discharge status), end of continuous enrollment, or end of study period (28 February 2017). This process is described in Figure 1.

Figure 1.

Patient selection flowchart.

Identification of peripheral neuropathy cases and matched controls

Due to the lack of diagnosis code specificity for disease-related or treatment-induced PN, PN was identified using an algorithm from previously published studies.^20,21 PN cases were identified by a medical claim with a diagnosis for PN (codes in Table A.1) during the 9 months following their initial MM therapy and without evidence of PN during the 12-month preperiod through the 7 days following the initial MM treatment (Figure 2). Controls had no medical claims with a diagnosis of PN anytime during the 12-month preperiod and throughout the follow-up period.

Figure 2.

PN definition at the patient level.

To adjust for imbalances in demographics and clinical characteristics, patients with PN were matched to a pool of patients without PN in a ratio of 1:2 (PN:without PN) using propensity-score modeling with nearest-neighbor matching. Matching factors included patients’ demographic characteristics [age, sex, geographic region of residence, payer (Commercial or Medicare), healthplan type] and baseline clinical characteristics (Deyo–Charlson Comorbidity Index, DCI)²² and specific preindex comorbidities including cardiovascular disease, cerebrovascular disease, chronic obstructive pulmonary disease, rheumatoid arthritis, diabetes, chronic kidney disease, skeletal-related events, coagulopathies, hematologic disease, hypertension, and the index MM medication). Standardized differences in matching factors between patients with PN and patients without PN were calculated before and after the matching to examine the quality of the match.

Lines of therapy

This study used a previously published MM treatment algorithm to identify the number of lines of therapy.²¹ The first line started on the date of the first MM chemotherapy or immunotherapy treatment with bendamustine, bortezomib, carfilzomib, cisplatin, cyclophosphamide, doxorubicin, doxorubicin liposomal, lenalidomide, melphalan, panobinostat, pomalidomide, or thalidomide. A treatment regimen was defined as consisting of one or more chemotherapy with or without immunotherapy agents administered within 90 days of the start of the line of therapy. A line of therapy ended at the earliest occurrence of a 90-day gap in all MM treatments in a regimen comprising the line of therapy, initiation of a different MM treatment > 90 days after the start of current line of therapy, inpatient discharge status of death, end of enrollment, or end of data. Note that lenalidomide monotherapy initiated within 60 days of the last drug administration in the line of therapy was classified as ‘maintenance therapy’. Maintenance therapy was considered to be a continuation of the line of therapy and not a new line of therapy. Moreover, any MM therapy received within 90 days following a stem-cell transplant date was considered to be ‘consolidation therapy’ within the current line and not the start of a new line of therapy. All subsequent lines of therapy were identified using the same approach as for the first line (with the noted exception above regarding first-line maintenance). Figure 3 describes two examples of changes in treatment regimen and how lines of therapy were defined.

Figure 3.

Examples of switching in regimens.

Patients with and without PN were identified during each line of therapy. Because of the small number of patients with more than three lines of therapy with PN, the third line and subsequent lines were combined in reporting.

Covariates and study outcomes

Demographics data extracted on the index date, included age, sex, US Census Bureau geographic region, payer (Commercial insurance or Medicare), healthplan type, and index year. Baseline clinical characteristics, measured throughout the 12-month pre-index period, included the DCI (an aggregate measure of comorbidity expressed as a numeric score based on the presence of various diagnoses), specific conditions contained in the DCI, other primary cancers, and other disease-related complications.²²

Study outcomes included all-cause healthcare utilization and costs measured during the at-least-3-month follow-up period and stratified for occurrence during the first, second, or third (or higher) lines of therapy. Healthcare utilization and costs were categorized as inpatient medical, emergency department/room (ER), office visits, outpatient hospital-based visits, other outpatient services, and outpatient pharmacy. Due to the variable length of follow up for patients overall and during lines of therapy, healthcare utilization and costs were reported in per-patient-per-month units (PPPM). Costs used the total paid amounts from all payers to all providers, including plan-paid, patient-paid, and coordinated benefit payments. All dollar amounts were inflation adjusted to 2017 US dollars using the Medical Care component of the Consumer Price Index.²³

The incidence rate of PN was calculated using the unmatched sample of all treated MM patients as the number of patients with PN divided by the person-time from the index date to the first PN diagnosis for PN patients, plus the person-time from the index date to the end of follow up for patients without PN.

Statistical considerations

Pairwise descriptive statistics were used to compare demographics, comorbid conditions, healthcare utilization and costs between patient cohorts with and without PN after propensity-score matching. Descriptive statistics also evaluated these differences between the comparator cohorts during first line, second line, and third-plus-subsequent lines of therapy. Chi-squared tests were conducted for differences in dichotomous or categorical variables and t tests were conducted for comparisons of continuous variables. A p value < 0.05 was set as the threshold for statistically significant differences. Following propensity-score matching of PN patients and patients without PN, statistically significant postindex differences in results between cohorts were presumed to be associated with the effects of the key independent variable, the incidence of PN. Statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc., Cary, NC, US).

Results

A total of 9207 patients comprising the case and control pool; 15.5% (1431 patients) were identified as having PN; 7776 had no PN diagnosis anytime during the study period (Figure 1). The incidence rate for a mean (standard deviation) duration of 624 (594) days after an initial MM diagnosis until PN was identified or end of follow up was estimated as 9.1 PN cases per 100 person-years.

Following matching, the study cohorts consisted of 1387 MM patients with PN and 2594 patients without any PN diagnosis during the study period. These matched study cohorts were well balanced with no statistically significant differences in demographics or baseline clinical characteristics (Tables 1). Mean patient age was 64 years, with 60–61% males, and 43–44% covered by Medicare. The length of follow up was longer for PN patients at a mean (SD) of 788 (580) days and 693 (571) days for non-PN patients (Table 1). Among the 1387 MM patients with PN, the mean (SD) and median duration from index date to PN diagnosis was 129 (68) days and 125 days, respectively.

Table 1.

Demographic and baseline clinical characteristics.

	PN cases	Non-PN controls	p value	Standard difference
	n = 1387	n = 2594		×100^*
Age, mean (SD) years	63.9 (10.8)	64.2 (11.6)	0.314	3.39
18–44	0.4%	0.3%	0.395	2.75
45–54	2.7%	3.7%	0.084	5.89
55–64	16.4%	15.9%	0.692	1.31
65–74	37.1%	37.8%	0.670	1.42
75+	25.5%	20.7%	0.001	11.38
Sex, %
Male	60.6%	60.1%	0.725	1.17
Female	39.4%	39.9%	0.725	1.17
Payer, %
Commercial	55.6%	57.0%	0.399	2.80
Medicare	44.4%	43.0%	0.399	2.80
Insurance plan type, %
Preferred-provider organization	55.7%	56.2%	0.740	1.10
Comprehensive	19.6%	19.0%	0.623	1.63
Health-maintenance organization	11.6%	11.7%	0.945	0.23
Point of service	5.3%	5.4%	0.818	0.77
Other	7.9%	7.7%	0.867	0.55
Geographic region, %
Northeast	17.8%	17.8%	0.975	0.11
North central	28.8%	27.9%	0.552	1.97
South	35.4%	37.6%	0.166	4.62
West	17.2%	15.3%	0.114	5.22
Unknown	0.7%	1.3%	0.109	5.55
Length of follow up, mean (SD) days	788 (580)	693 (571)	<0.001	16.52
Median follow up (days)	642	522
Deyo–Charlson Comorbidity Index, mean (SD)	4.5 (2.9)	4.4 (2.8)	0.999	3.69
Comorbid conditions ^$
Hypertension	62.5%	59.9%	0.109	5.34
Skeletal-related events	48.4%	48.4%	0.980	0.08
Diabetes	30.1%	25.9%	0.004	9.43
Renal disease	23.0%	21.0%	0.155	4.71
Ischemic vascular condition^‡	22.5%	22.4%	0.922	0.32
Chronic kidney disease	19.4%	18.6%	0.532	2.07
Anemia or anemia treatment	57.0%	56.1%	0.569	1.89
Chronic pulmonary disease	16.4%	15.4%	0.435	2.59
Hypercalcemia	14.0%	13.3%	0.523	2.12
GI bleeding	5.0%	5.2%	0.871	0.54
Pneumonia	9.4%	9.6%	0.847	0.64
Congestive heart failure	7.8%	8.1%	0.732	1.14
Cerebrovascular disease	6.8%	6.6%	0.860	0.59
Thrombocytopenia	7.8%	6.9%	0.282	3.55
End-stage renal disease/renal failure	6.3%	6.5%	0.802	0.84
Venous thromboembolism	5.3%	5.7%	0.593	1.79
Prior primary cancer
Solid tumor	24.4%	24.0%	0.805	0.82
Hematologic cancer	12.4%	13.5%	0.331	3.25
Days from diagnosis to treatment, mean (SD)	158 (355)	163 (360)	0.623	1.64
Median (days)	26	27
Stem-cell transplant prior to index treatment, %	0.8%	1.0%	0.588	1.83
Index MM therapy, %
Bendamustine	0.2%	0.6%	0.081	6.22
Bortezomib	52.5%	50.0%	0.135	4.98
Carfilzomib	0.1%	0.2%	0.722	2.01
Cisplatin	0.4%	0.4%	0.902	0.41
Cyclophosphamide	9.9%	10.3%	0.733	1.14
Doxorubicin	1.7%	1.7%	0.937	0.26
Doxorubicin liposomal	0.2%	0.2%	1.000	0.32
Lenalidomide	37.1%	37.9%	0.620	1.65
Melphalan	2.1%	2.8%	0.190	4.45
Panobinostat	0.0%	0.0%	–	–
Pomalidomide	0.1%	0.1%	1.000	0.18
Thalidomide	4.7%	5.6%	0.244	3.93

GI, gastrointestinal; MM, multiple myeloma; PN, peripheral neuropathy; SD, standard deviation.

The standardized differences were multiplied by 100 to facilitate readers’ viewing results.

Comorbid condition occurring during the preperiod in <5% patients are not shown.

‡

Ischemic vascular conditions include unstable angina, stable angina, ischemic stroke, transient ischemic event, and other chronic ischemic heart disease, including coronary revascularization.

The most common baseline comorbid conditions included hypertension (60–63%), skeletal-related events (48%), diabetes (26–30%), renal disease (21–23%), and ischemic vascular conditions (22%). At index, approximately 50–53% received bortezomib; 37–38% received lenalidomide; 10% received cyclophosphamide, and 5–6% received thalidomide as their initial MM therapy, with the remaining MM therapies received as index medications in less than 3% of patients (Table 1). Of the 3981 matched patients, 1267 (32%) had a PN diagnosis and 2687 (68%) had no PN diagnosis during their first-line therapy. There were 27 patients whose PN diagnosis occurred after the end of the first line but before the start of second-line therapy. A total of 1974 patients had a second-line therapy, of which 280 patients (15%) had PN during their second line, while 1532 patients (85%) had no PN during their second line, with the remaining patients having PN during the first line, thus not eligible for the PN analysis during the second line. Of the 1103 patients with a third or subsequent line of therapy, 75 patients (7%) had a PN diagnosis and 1028 patients (93%) had no PN during their third or subsequent line. Demographic and baseline clinical characteristics measured during first-line, second-line, and third/subsequent-line therapy were similar between PN patients and patients without PN (Tables A.2 –A.4).

All-cause healthcare utilization

Healthcare utilization was significantly higher in most healthcare use categories among patients with PN compared with patients with no PN. Significantly more PN patients had a hospitalization during follow up (77.4%) compared with patients without PN (67.2%; p < 0.001), however, their length of stay was similar [PN 0.60 (0.87) days PPPM versus non-PN 0.66 (0.96) days PPPM; p = 0.052]. More patients with PN had an ER visit (67.8% versus 58.4%; p < 0.001), had significantly more outpatient hospital-based visits [mean 13.5 (14.7) visits PPPM versus 11.5 (18.0) visits PPPM; p < 0.001], fewer laboratory tests [mean 4.1 (5.1) tests PPPM versus 4.7 (5.6) tests PPPM; p < 0.001], and outpatient prescriptions [4.7 (2.5) prescription claims PPPM versus 4.2 (2.4) prescription claims PPPM; p < 0.001] compared with patients without PN (Table 2).

Table 2.

Per-patient-per-month all-cause healthcare utilization over the entire follow-up period.

	Entire follow-up period
	PN cases	Non-PNcontrols	p value
	n = 1387	n = 2594
All-cause healthcare utilization (PPPM)
Patients with an inpatient admission, %	77.4%	67.2%	<0.001
Admissions PPPM, mean (SD)	0.11 (0.14)	0.10 (0.15)	0.019
Patients with an admission, mean (SD)	0.14 (0.14)	0.15 (0.16)	0.373
Patients with an ER visit, %	67.8%	58.4%	<0.001
ER visits PPPM, mean (SD)	0.13 (0.33)	0.11 (0.22)	0.015
Patients with a visit, mean (SD)	0.19 (0.39)	0.19 (0.26)	0.629
Patients with an outpatient office visit, N%	97.9%	96.6%	0.024
Outpatient office visits PPPM, mean (SD)	1.7 (1.2)	1.6 (1.2)	0.142
Patients with a visit, mean (SD)	1.7 (1.2)	1.7 (1.1)	0.342
Patients with outpatient hospital-based visits, %	98.7%	96.3%	<0.001
Outpatient hospital-based visits PPPM, mean (SD)	13.3 (14.7)	11.1 (17.8)	<0.001
Patients with a visit, mean (SD)	13.5 (14.7)	11.5 (18.0)	<0.001
Patients with a laboratory test, %	90.9%	87.9%	0.003
Laboratory tests PPPM, mean (SD)	3.7 (5.0)	4.1 (5.5)	0.018
Patients with a test, mean (SD)	4.1 (5.0)	4.7 (5.6)	0.001
Patients filling an outpatient prescription, %	97.2%	97.1%	0.941
Outpatient prescriptions PPPM, mean (SD)	4.6 (2.6)	4.0 (2.5)	<0.001
Patients with a prescription, mean (SD)	4.7 (2.5)	4.2 (2.4)	<0.001

ER, emergency department (room); PN, peripheral neuropathy; PPPM, per patient per month; SD, standard deviation.

Use of narcotic pain medication was common and also higher among patients with PN compared with patients with no PN (during first-line therapy: 82% versus 72%; p < 0.001).

All-cause healthcare costs

Healthcare costs were also higher among patients with PN compared with control patients with no PN and increased as patients proceeded to higher lines of therapy (Table 3). Mean total costs for PN patients exceeded those of patients without PN by $1509 PPPM [PN $16,600 (SD $14,450) versus non-PN $15,090 (SD $13,399); p = 0.001] over the entire follow-up period. This difference was primarily attributable to the first 180 days postindex, where PN patients’ mean PPPM costs exceeded those of the non-PN cohort by $3317 during the first 90 days (p < 0.001), and by $5167 during the period 91–180 days postindex (p < 0.001).

Table 3.

Per-patient-per-month all-cause healthcare costs over the entire follow-up period.

	Entire follow-up period
	PN cases	Non-PN controls	p value
	n = 1387	n = 2594
All-cause healthcare costs (PPPM)
Inpatient-admission costs, mean (SD)	$4750 ($9944)	$4002 ($8993)	0.016
Outpatient medical costs, mean (SD)	$8100 ($8,341)	$7408 ($8,007)	0.010
ER costs, mean (SD)	$126 ($620)	$92 ($374)	0.029
Outpatient office-visit costs, mean (SD)	$216 ($302)	$203 ($329)	0.214
Outpatient hospital-based visit costs, mean (SD)	$4906 ($6410)	$4076 ($6280)	<0.001
Laboratory-testing costs, mean (SD)	$144 ($309)	$140 ($290)	0.687
Outpatient prescription costs, mean (SD)	$3749 ($3312)	$3681 ($3596)	0.555
Total costs, mean (SD)	$16,600 ($14,450)	$15,090 ($13,399)	0.001
Quarterly healthcare costs (PPPM)
0–90 days postindex, n	1381	2583
Total costs, mean (SD)	$22,777 ($19,074)	$19,460 ($16,064)	<0.001
91–180 days postindex, n	1299	2286
Total costs, mean (SD)	$24,402 ($24,760)	$19,235 ($19,769)	<0.001
181–270 days postindex, n	1168	1986
Total costs, mean (SD)	$15,766 ($21,032)	$14,052 ($18,286)	0.008
271–360 days postindex, n	1026	1682
Total costs, mean (SD)	$11,955 ($15,774)	$11,518 ($15,756)	0.404

ER, emergency department (room); PN, peripheral neuropathy; PPPM, per patient per month; SD, standard deviation.

Table 4 contrasts the PPPM costs for patients who had a PN diagnosis during a particular line of therapy with patients who did not have a PN diagnosis during that same line. The PN patients’ total costs were significantly higher during the first line of therapy ($23,183 (SD $22,243) versus $20,790 (SD $27,748); p = 0.007) and second line ($37,880 (SD $58,007) versus $29,694 (SD $103,457); p = 0.198) compared with patients without a PN diagnosis during those lines. This difference was primarily driven by outpatient medical costs, and particularly by outpatient hospital-based visits.

Table 4.

Per-patient-per-month all-cause healthcare costs by line of therapy stratified by patients with and without PN.

	Line of therapy
	First-line therapy
	PN in first line	No PN first line	p value
	n = 1267	n = 2687
Inpatient-admission costs, mean (SD)	$6158 ($18,095)	$5000 ($24,001)	0.127
Outpatient medical costs, mean (SD)	$12,598 ($11,306)	$10,949 ($11,319)	<0.001
ER costs, mean (SD)	$180 ($798)	$129 ($766)	0.055
Outpatient office-visit costs, mean (SD)	$318 ($659)	$281 ($522)	0.062
Outpatient hospital-based visits, mean (SD)	$7687 ($10,244)	$6049 ($9927)	<0.001
Laboratory-testing costs, mean (SD)	$212 ($552)	$197 ($454)	0.376
Outpatient prescription costs, mean (SD)	$4427 ($4103)	$4841 ($6715)	0.043
Total costs, mean (SD)	$23,183 ($22,243)	$20,790 ($27,748)	0.007
	Second-line therapy
	PN second line	No PN second line	p value
	n = 280	n = 1532
Inpatient-admission costs, mean (SD)	$15,726 ($51,561)	$11,468 ($99,972)	0.487
Outpatient medical costs, mean (SD)	$15,807 ($22,951)	$11,277 ($18,854)	<0.001
ER costs, mean (SD)	$271 ($2811)	$133 ($879)	0.119
Outpatient office-visit costs, mean (SD)	$244 ($325)	$240 ($483)	0.889
Outpatient hospital-based visits, mean (SD)	$11,250 ($21,747)	$7176 ($17,414)	0.001
Laboratory-testing costs, mean (SD)	$173 ($546)	$149 ($441)	0.428
Outpatient prescription costs, mean (SD)	$6348 ($6,192)	$6950 ($9282)	0.297
Total costs, mean (SD)	$37,880 ($58,007)	$29,694 ($103,457)	0.198
	Third and subsequent lines of therapy
	PN third+ line	No PN third+ line	p value
	n = 75	n = 1028
Inpatient-admission costs, mean (SD)	$7228 ($19,480)	$6488 ($27,591)	0.820
Outpatient medical costs, mean (SD)	$9803 ($12,438)	$10,563 ($15,657)	0.681
ER costs, mean (SD)	$148 ($347)	$126 ($784)	0.812
Outpatient office-visit costs, mean (SD)	$235 ($256)	$224 ($351)	0.787
Outpatient hospital-based visits, mean (SD)	$7414 ($11,805)	$6240 ($13,760)	0.472
Laboratory-testing costs, mean (SD)	$70 ($174)	$147 ($440)	0.132
Outpatient prescription costs, mean (SD)	$8563 ($7307)	$7777 ($11,021)	0.543
Total costs, mean (SD)	$25,594 ($26,656)	$24,827 ($34,532)	0.851

ER, emergency department (room); PN, peripheral neuropathy; PPPM, per patient per month; SD, standard deviation.

Discussion

This study found significantly higher healthcare resource utilization and costs in patients with a post-treatment diagnosis for PN during their follow up compared with a matched group of patients without PN. Patients with PN were significantly more likely to be hospitalized, had an ER visit, had an outpatient hospital-based visit, and filled more outpatient prescriptions than matched patients without a PN diagnosis. This increased use of healthcare resource was associated with $1509 higher PPPM total costs, amounting to $36,216 over PN patients’ mean 2-year follow-up time. Pike and colleagues, in a similar study using administrative US claims data (1999–2006), estimated the costs for chemotherapy-induced PN in a matched cohort comparison of cancer patients (not MM) with and without post-treatment PN, finding that PN patients had higher mean total healthcare costs by $17,344 during the 12-month study period than comparable patients without PN. They further found that more PN cases were hospitalized, had an ER visit, and had other outpatient visits.²⁴ In a 2001 pilot study of chemotherapy-induced toxicity in ovarian cancer patients, Calhoun and colleagues found indirect costs, such as the cost of caregiver time and lost wages, to be a substantial contributor to the total burden of chemotherapy-induced toxicity.²⁵ The estimated marginal healthcare expenditure of $1509 PPPM attributable to PN in our study using direct medical and prescription expenditures derived from paid healthcare claims represents only a portion of the overall societal cost. Due to the lack of information, costs associated with caregiver burden, indirect costs, or the quality of life impact were not examined.

In our study, the rate of PN in MM patients was 15.5%, which appears low compared with other studies where PN rates ranged from over 20% to as high as 70%^15,16,26,27 This may be attributable to several contributing factors. The diagnosis of PN may be under-reported in claims, as healthcare providers may only include the diagnosis when the presentation of PN is severe or substantially affecting disease management. Patients expecting drug side effects may not seek treatment for mild PN. This was confirmed by the finding of Yong et al. using chart review data. Yong and colleagues found that although more than 45% of MM patients had PN, less than 4% had grade 3 or 4 PN during the first four lines of therapy.²⁷ Clinical trials have also reported a much lower PN rate of grade 3–4 than PN rate of grade 1–2.¹⁶ Some patients indexing in the latter years of the study may have experienced PN after the end of their available data. The incidence rate reported in our study of 9.1 PN cases per 100 person-years may likewise reflect a similar underestimate. In addition, it was estimated that 3.2% of MM patients have baseline PN.²¹ In the current study, patients with baseline PN who later had another PN diagnosis post-treatment were excluded from the PN rate calculation, which may further lower the estimated PN rate. This exclusion was applied so as not to overestimate PN.

Our cohorts were demographically similar to the US MM population in terms of age and sex based on the National Cancer Institute’s Surveillance, Epidemiology and End Results (SEER) Program’s reporting most patients diagnosed between 65 and 74 years of age, median age 69 at initial diagnosis, and the SEER and American Cancer Society’s estimate of 59% male–41% female incidence. The population in this study is slightly younger overall due to larger representation from commercial carriers than from Medicare data providers.^1,2

Obtaining optimal clinical efficacy requires carefully balancing treatment effectiveness with the potential for negative consequences on the patient’s quality of life. The dosage reductions, treatment switches, or discontinuation of MM therapies to manage PN may ultimately affect response to therapy.²⁸ Thalidomide and bortezomib are associated with higher rates of PN.^12,13,15,16 Recent approval of novel therapies^29–32 hold promise for antimyeloma efficacy with reduced incidence of PN.

Limitations

There are several limitations associated with this study. Use of diagnosis coding from administrative claims data may be subject to misclassification errors, where the extent of undercoding for the selected conditions or comorbidities is unknown, and without the availability of patient charts or physician attestations. PN diagnoses may not be included on administrative claims unless the impairment significantly affects patient management, thereby under-reported or biased toward more severe cases. PN could not be identified directly in claims data because of the lack of diagnosis codes specific to disease-related and treatment-induced PN. Consequently, PN identification used an algorithm from a previously published study that has not been validated, and the PN could be due to other causes. This study used propensity-score matching to ensure cohorts had similar baseline demographic and clinical characteristics, increasing the likelihood that differences between cohorts were associated with PN. However, there is always the potential of unmeasured confounders outside of this study’s data sources. In addition, the incremental healthcare utilization and costs in the PN cohort may not be directly linked to PN; it could also be due to PN treatment, PN complications, and other conditions that could not be controlled or adjusted for in the claims data. Pharmacological treatments that were based on pharmacy prescription claims only indicated that prescriptions were received, not necessarily how the patients took the medications. This is not an issue for medications administered in the physician’s office and billed through a medical claim. Several drugs (carfilzomib, pomalidomide, panobinostat) entered the US market more recently (since 2012), resulting in limited sample sizes for these agents, so PN rates specific to individual medications were not examined. MarketScan® Commercial and Medicare databases are convenience samples of employees, retirees, and dependents with US Commercial and Medicare health-insurance coverage, therefore results from these databases may not be generalizable to populations with other healthcare coverage (e.g. Medicaid), or those lacking coverage.

Conclusion

PN was observed in 15.5% of MM patients, and was associated with a significant economic burden, adding an average of $1509 monthly per patient to the cost of MM treatment, as well as adding to the complexity of treatment with detrimental impact to patients. These results suggest that utilization of newer, more effective novel treatments might ease the economic and disease burden for MM associated with PN.

Footnotes

Appendix

Table A.4.

Demographic characteristics	Third and subsequent line of therapy
	With PN		Without PN		p value
	n = 75		n = 1028
	n/mean	%/SD	n/mean	%/SD
Age (mean, SD)	62.0	9.57	61.2	10.72	0.517
Age group (n, %)
18–34	0	0.0%	7	0.7%
35–44	2	2.7%	44	4.3%	0.764
45–54	14	18.7%	227	22.1%	0.490
55–64	31	41.3%	408	39.7%	0.779
65–74	19	25.3%	210	20.4%	0.312
75+	9	12.0%	132	12.8%	0.833
Sex (n, %)
Male	50	66.7%	617	60.0%	0.256
Female	25	33.3%	411	40.0%	0.256
Payer (n, %)
Commercial	45	60.0%	671	65.3%	0.356
Medicare	30	40.0%	357	34.7%	0.356
Insurance plan type (n, %)
Comprehensive	11	14.7%	176	17.1%	0.585
Exclusive-provider organization	0	0.0%	7	0.7%
Health-maintenance organization	9	12.0%	130	12.6%	0.871
Point of service (POS)	7	9.3%	74	7.2%	0.494
Preferred-provider organization	42	56.0%	547	53.2%	0.640
POS with capitation	0	0.0%	5	0.5%
Consumer-driven health plan	4	5.3%	66	6.4%	1.000
High-deductible health plan	2	2.7%	23	2.2%	0.685
Unknown	0	0.0%	0	0.0%
Geographic region (n, %)
Northeast	11	14.7%	188	18.3%	0.431
North central	13	17.3%	283	27.5%	0.054
South	30	40.0%	356	34.6%	0.347
West	21	28.0%	195	19.0%	0.057
Unknown	0	0.0%	6	0.6%
Population density (n, %)
Urban	64	85.3%	900	87.5%	0.577
Rural	11	14.7%	122	11.9%	0.472
Unknown	0	0.0%	6	0.6%	0.507
Duration of line of therapy (mean, SD)	244.0	275.6	196.2	239.1	0.099
DCI (mean, SD)	4.3	2.84	4.0	2.55	0.274
DCI (n, %)
0	0	0.0%	15	1.5%	0.292
1	0	0.0%	0	0.0%	NA
2	29	38.7%	399	38.8%	0.98
3+	46	61.3%	614	59.7%	0.784
DCI components (n, %)
Myocardial infarction	0	0.0%	16	1.6%
Congestive heart failure	2	2.7%	32	3.1%	1.000
Peripheral vascular disease	3	4.0%	27	2.6%	0.452
Cerebrovascular disease	6	8.0%	66	6.4%	0.625
Dementia	0	0.0%	2	0.2%
Chronic pulmonary disease	12	16.0%	150	14.6%	0.739
Rheumatologic disease	2	2.7%	13	1.3%	0.272
Peptic ulcer disease	0	0.0%	16	1.6%
Mild liver disease	3	4.0%	4	0.4%	0.009
Diabetes (mild to moderate)	13	17.3%	210	20.4%	0.519
Diabetes with chronic complications	2	2.7%	54	5.3%	0.581
Hemiplegia or paraplegia	1	1.3%	1	0.1%	0.131
Renal disease	14	18.7%	218	21.2%	0.602
Moderate or severe liver disease	0	0.0%	2	0.2%
Human immunodeficiency virus	0	0.0%	1	0.1%
Any malignancy, including lymphoma and leukemia	74	98.7%	1000	97.3%	0.716
Metastatic solid tumor	13	17.3%	140	13.6%	0.369
Prior primary cancer (n, %)
Solid tumor	19	25.3%	215	20.9%	0.366
Hematologic cancer	11	14.7%	125	12.2%	0.524
Preperiod events of interest (n, %)
Chronic kidney disease	11	14.7%	168	16.3%	0.704
End-stage renal disease/renal failure	2	2.7%	48	4.7%	0.573
Skeletal-related events	42	56.0%	486	47.3%	0.144
Hypercalcemia	10	13.3%	115	11.2%	0.571
Venous thromboembolism	1	1.3%	38	3.7%	0.512
Neutropenia	2	2.7%	28	2.7%	1.000
Pneumonia	5	6.7%	84	8.2%	0.644
Major bleeding	2	2.7%	13	1.3%	0.272
GI bleeding	3	4.0%	46	4.5%	1.000
Anemia	36	48.0%	550	53.5%	0.357
Anemia or anemia treatment	36	48.0%	554	53.9%	0.323
Thrombocytopenia	7	9.3%	62	6.0%	0.224
Amyloidosis	0	0.0%	18	1.8%	0.248

DCI, Deyo–Charlson Comorbidity Index; GI, gastrointestinal; NA, not applicaple; PN, peripheral neuropathy; SD, standard deviation.

Acknowledgements

Editorial/medical writing support was provided by Jay Margolis, an employee of IBM Watson Health.

Funding

This study was sponsored by Amgen, Inc.

Conflict of interest statement

Xue Song, Kathleen Wilson, and Jerry Kagan are employees of IBM Watson Health, which received funding from Amgen, Inc. to conduct this analysis. Sumeet Panjabi is an employee and stockholder of Amgen Inc.

ORCID iD

Xue Song

References

Howlader

Noone

Krapcho

et al . SEER cancer statistics review, 1975–2011. Bethesda, MD: National Cancer Institute, 2014, http://seer.cancer.gov/csr/1975_2011/ (accessed 14 March 2019).

American Cancer Society. What are the key statistics about multiple myeloma?, http://www.cancer.org/cancer/multiplemyeloma/detailedguide/multiple-myeloma-key-statistics (2016, accessed 1 April 2018).

Avet-Loiseau

Fonseca

Siegel

et al . Carfilzomib significantly improves the progression-free survival of high-risk patients in multiple myeloma. Blood 2016; 128: 1174–1180.

Dimopoulos

Moreau

Palumbo

et al . Carfilzomib and dexamethasone versus bortezomib and dexamethasone for patients with relapsed or refractory multiple myeloma (ENDEAVOR): a randomised, phase 3, open-label, multicentre study. Lancet Oncol 2016; 17: 27–38.

Dimopoulos

Oriol

Nahi

et al . Daratumumab, lenalidomide, and dexamethasone for multiple myeloma. N Engl J Med 2016; 375: 1319–1331.

Lonial

Dimopoulos

Palumbo

et al . Elotuzumab therapy for relapsed or refractory multiple myeloma. N Engl J Med 2015; 373: 621–631.

Palumbo

Chanan-Khan

Weisel

et al . Daratumumab, bortezomib, and dexamethasone for multiple myeloma. N Engl J Med 2016; 375: 754–766.

San-Miguel

Hungria

Yoon

et al . Panobinostat plus bortezomib and dexamethasone versus placebo plus bortezomib and dexamethasone in patients with relapsed or relapsed and refractory multiple myeloma: a multicentre, randomised, double-blind phase 3 trial. Lancet Oncol 2014; 15: 1195–1206.

San-Miguel

Hungria

Yoon

et al . Overall survival of patients with relapsed multiple myeloma treated with panobinostat or placebo plus bortezomib and dexamethasone (the PANORAMA 1 trial): a randomised, placebo-controlled, phase 3 trial. Lancet Haematol 2016; 3: e506–e515.

10.

Siegel

Oriol

Rajnics

et al . Updated results from ASPIRE and ENDEAVOR, randomized, open-label, multicenter phase 3 studies of carfilzomib in patients (pts) with relapsed/refractory multiple myeloma (RRMM). Clin Lymphoma Myeloma Leuk 2017; 17: e142.

11.

National Comprehensive Cancer Network. National comprehensive cancer network: multiple myeloma, http://www.nccn.org/professionals/physician_gls/PDF/myeloma.pdf (2013, accessed 18 August 2018).

12.

THALOMID®. Prescribing information: Thalomid (thalidomide). Summit, NJ: Celgene Corporation, https://media.celgene.com/content/uploads/thalomid-pi.pdf (2017, accessed 14 March 2019).

13.

VELCADE®. Prescribing information: Velcade (bortizumib). Cambridge, MA: Millennium Pharmaceuticals, Inc., http://www.accessdata.fda.gov/drugsatfda_docs/label/2015/021602s042lbl.pdf (2015, accessed 5 October 2018).

14.

Boland

Eiser

Ezaydi

et al . Living with advanced but stable multiple myeloma: a study of the symptom burden and cumulative effects of disease and intensive (hematopoietic stem cell transplant-based) treatment on health-related quality of life. J Pain Symptom Manage 2013; 46: 671–680.

15.

Delforge

Bladé

Dimopoulos

et al . Treatment-related peripheral neuropathy in multiple myeloma: the challenge continues. Lancet Oncol 2010; 11: 1086–1095.

16.

Mohty

El-Cheikh

Yakoub-Agha

et al . Peripheral neuropathy and new treatments for multiple myeloma: background and practical recommendations. Haematologica 2010; 95: 311–319.

17.

American Cancer Society. Peripheral neuropathy caused by chemotherapy, https://www.cancer.org/treatment/treatments-and-side-effects/physical-side-effects/peripheral-neuropathy.html (2015, accessed 1 April 2018).

18.

Grisold

Cavaletti

Windebank

AJ.

Peripheral neuropathies from chemotherapeutics and targeted agents: diagnosis, treatment, and prevention. Neuro Oncol 2012; 14(Suppl. 4): iv45–iv54.

19.

Trivedi

Hershman

Crew

KD.

Management of chemotherapy-induced peripheral neuropathy. Am J Hematol Oncol 2015; 11: 4–10.

20.

Cavaletti

Cornblath

Merkies

et al . The chemotherapy-induced peripheral neuropathy outcome measures standardization study: from consensus to the first validity and reliability findings. Ann Oncol 2013; 24: 454–462.

21.

Song

Cong

Wilson

Real-world treatment patterns, comorbidities, and disease-related complications in patients with multiple myeloma in the United States. Curr Med Res Opin 2016; 32: 95–103.

22.

Deyo

Cherkin

Ciol

MA.

Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992; 45: 613–619.

23.

United States Bureau of Labor Statistics. Consumer price index, http://www.bls.gov/cpi/ (2018, accessed 1 June 2018).

24.

Pike

Birnbaum

Muehlenbein

et al . Healthcare costs and workloss burden of patients with chemotherapy-associated peripheral neuropathy in breast, ovarian, head and neck, and nonsmall cell lung cancer. Chemother Res Pract 2012; 2012: 913848.

25.

Calhoun

Chang

C-H

Welshman

et al . Evaluating the total costs of chemotherapy-induced toxicity: results from a pilot study with ovarian cancer patients. Oncologist 2001; 6: 441–445.

26.

Seretny

Currie

Sena

et al . Incidence, prevalence, and predictors of chemotherapy-induced peripheral neuropathy: a systematic review and meta-analysis. Pain 2014; 155: 2461–2470.

27.

Yong

Delforge

Driessen

et al . Multiple myeloma: patient outcomes in real-world practice. Br J Haematol 2016; 175: 252–264.

28.

Martin III

Panjabi

Kerr

et al . Association of treatment induced peripheral neuropathy (TIPN) with treatment patterns and outcomes in patients (pts) with newly diagnosed multiple myeloma (NDMM). Blood 2013; 122: 1750.

29.

POMALYST®. Prescribing information: Pomalyst (pomalidomide). Summit, NJ: Celgene Corporation, http://www.celgene.com/content/uploads/pomalyst-pi.pdf (2018, accessed 18 April 2018).

30.

KYPROLIS®. Prescribing information: Kyprolis (carfilzomib). Thousand Oaks, CA: Amgen Inc., http://pi.amgen.com/~/media/amgen/repositorysites/pi-amgen-com/kyprolis/kyprolis_pi.ashx (2018, accessed 18 April 2018).

31.

DARZALEX®. Prescribing information: Darzalex (daratumumab), https://www.accessdata.fda.gov/drugsatfda_docs/label/2016/761036s004lbl.pdf (2016, accessed 18 April 2018).

32.

EMPLICITI®. Prescribing information: Empliciti (elotuzumab). Princeton, NJ: Bristol-Myers Squibb Company, http://packageinserts.bms.com/pi/pi_empliciti.pdf?&utm_source=bing&utm_medium=cpc&utm_campaign=managementbnd&utm_term=prescribinginformation&utm_content=managementbnd_sitelink_pi_empliciti.pdf_text (2017, accessed 18 April 2018).