Relationship of Cisplatin-Related Adverse Health Outcomes With Disability and Unemployment Among Testicular Cancer Survivors

Abstract Background Few data exist on the relationship of cisplatin-related adverse health outcomes (AHOs) with disability, unemployment, and self-reported health (SRH) among testicular cancer survivors (TCS). Methods A total of 1815 TCS at least 1 year postchemotherapy underwent clinical examination and completed questionnaires. Treatment data were abstracted from medical records. A cumulative burden of morbidity score (CBMPt) encompassed the number and severity of platinum-related AHOs (peripheral sensory neuropathy [PSN], hearing loss, tinnitus, renal disease). Multivariable regression assessed the association of AHOs and CBMPt with employment status and SRH, adjusting for sociodemographic and clinical characteristics. Unemployment was compared with a male normative population of similar age, race, and ethnicity. Results Almost 1 in 10 TCS was out of work (2.4%, disability leave; 6.8%, unemployed) at a median age of 37 years (median follow-up = 4 years). PSN (odds ratio [OR] = 2.89, 95% confidence interval [CI] = 1.01 to 8.26, grade 3 vs 0, P = .048), renal dysfunction defined by estimated glomerular filtration rate (OR = 12.1, 95% CI = 2.06 to 70.8, grade 2 vs 0, P = .01), pain (OR = 10.6, 95% CI = 4.40 to 25.40, grade 2 or 3 vs 0, P < .001), and CBMPt (OR = 1.46, 95% CI = 1.03 to 2.08, P = .03) were associated with disability leave; pain strongly correlated with PSN (r2 = 0.40, P < .001). Statistically significantly higher percentages of TCS were unemployed vs population norms (age-adjusted OR = 2.67, 95% CI = 2.49 to 3.02, P < .001). PSN (OR = 2.44, 95% CI = 1.28 to 4.62, grade 3 vs 0, P = .006), patient-reported hearing loss (OR = 1.82, 95% CI = 1.04 to 3.17, grade 2 or 3 vs 0, P = .04), and pain (OR = 3.75, 95% CI = 2.06 to 6.81, grade 2 or 3 vs 0, P < .001) were associated with unemployment. Increasing severity of most cisplatin-related AHOs and pain were associated with statistically significantly worse SRH. Conclusions Our findings have important implications regarding treatment-associated productivity losses and socioeconomic costs in this young population. Survivorship care strategies should include inquiries about disability and unemployment status, with efforts made to assist affected TCS in returning to the workforce.

Behavioral Risk Factor Surveillance System (BRFSS) [22], accessed October 2018) and were used to compare unemployment rates by age groups between testicular cancer survivors (TCS) and the general U.S. population, matching on race/ethnicity. BRFSS is a cross-sectional, random-digit-dialed telephone survey of >400,000 non-institutionalized U.S. adults aged ≥18 years in the 50 states, the District of Columbia (DC), Puerto Rico, the U.S. Virgin Islands, and Guam that uses trained interviewers and standard core and optional questionnaire modules to collect uniform data (details at http://www.cdc.gov/brfss/index.htm). Surveys are conducted with U.S. households each month using disproportionate, stratified, random sampling in all states and Washington D.C. and simple random sampling in Guam, Puerto Rico, and the U.S. Virgin Islands. Employment status in BRFSS is assessed as follows: "Are you currently?: Employed for wages; Self-employed; Out of work for 1 year or more; Out of work for less than 1 year; A homemaker; A student; Retired; Unable to work; Refused; Not asked or Missing". Responses of "Out of work for 1 year or more" were used for comparison to unemployed TCS. Analyses excluded BRFSS participants from Guam, Puerto Rico, and the U.S. Virgin Islands.
To assess health behaviors, we used validated questionnaires to query current or former smoking status, alcohol consumption, and physical activity over the past year. The validated physical activity questionnaire [57,58] asked participants to report their average time per week (over the past year) spent at each of nine recreational activities: walking or hiking (including walking to work); jogging (slower than 10-minute miles); running (10-minute miles or faster); bicycling (including stationary bike); aerobic exercise or dance/exercise machines; lower intensity exercise/yoga/stretching/toning; tennis, squash or racquetball; lap swimming; and weight lifting/strength training. Each physical activity was assigned a metabolic equivalent (MET) value. MET values are a commonly used metric for describing the relative energy expenditure of a specific type of physical activity (1 MET=1 kcal/kg/hour, or the energy cost of sitting quietly). The MET values for each activity were then used to calculate MET-hours/week for each participant, and these were grouped into categories of vigorous or non-vigorous physical activity applying standard definitions grading took into account current prescription medication use (i.e. peripheral sensory neuropathy, pain, kidney disease, hypercholesterolemia, hypertriglyceridemia, hypertension, diabetes, peripheral artery disease, thromboembolic event, thyroid disease, anxiety/depression, erectile dysfunction, and hypogonadism), we only took into account current prescription medication use (with usage for more than one month), using data provided by the patient at the time of clinical evaluation. Pearson's correlation coefficient (r 2 ) was used to assess the correlation between patient-reported peripheral sensory neuropathy and pain.

Objective Assessment of AHOs
Body mass index (BMI) was derived from height and weight assessed at the time of clinical evaluation using the following equation: weight (kg)/height (m) 2 . Audiometric data assessed at 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 10 and 12 kHz were categorized according to the severity of hearing loss as defined previously [13] following American Speech-Language-Hearing Association criteria [17].
Kidney disease was based on the estimated glomerular filtration rate (eGFR) that is used in clinical The following were considered as potential explanatory (independent) variables in multivariable logistic regression models of employment status (dependent variable) reported in Table 5: time since end of chemotherapy, attained age at clinical assessment, number of chemotherapy cycles, race/ethnicity, educational status, marital status, health insurance status, current use of antipsychotropic medication use, and engagement in vigorous physical activity. All were considered in the full model and backward model selection chose variables with p<0.05 for retention as adjustment factors in the final model.  . For each item, participants were asked to report the age (in years) at first occurrence. If onset of symptoms was prior to age of TC diagnosis, those responses were not considered when assigning severity grade. ¶ Item is from the SF36 questionnaire [35]. # Prescription medications taken for at least the past 4 weeks were only used to assign grade if the participant reported that the indication was for the given adverse health outcome and the medication was started during or after chemotherapy. ** eGFR was calculated using the following formula for males: eGFR = 141 x min(SCr/0.9, 1) -0.411 x max(SCr /0.9, 1) -1.209 x 0.993 Age x 1.159 [if Black], where SCr is standardized serum creatinine (in mg/dL), min indicates the minimum of SCr/0.9 or 1, max indicates the maximum of SCr/0.9 or 1, and age is in units of years [18].

Supplementary
† b Includes 5 cycles (n=29), 6 cycles (n=38), 7 cycles (n=5), and 8 or more cycles (n=6) of cisplatin-based chemotherapy. All chemotherapy was completed ≥1 year before study entry (see Methods). ‡ c Patient-reported outcomes are used to define all variables except obesity, which is based on body mass index measured at time of clinical assessment. § d CBMPt score was calculated using patient-reported adverse health outcomes previously related to cisplatin exposure (i.e. peripheral sensory neuropathy, hearing loss, tinnitus, renal disease) using a modification of Kerns et al.

11
|| e Assessed with the European Organisation for Research and Treatment of Cancer Chemotherapy-Induced Peripheral Neuropathy 20item (EORTC/CIPN-20) questionnaire, 14 the Scale for Chemotherapy-Induced Long-Term Neurotoxicity (SCIN) questionnaire, 15 and patient-reported current prescription medication use. Prescription medications were only considered if the respondent stated that the indication was for neuropathy.
¶ f Assessed with the SCIN questionnaire 15 based on symptoms experienced over the past 4 weeks. If participants responded that the symptom started before chemotherapy, those responses were not considered when assigning severity grade. # g Assessed using the Hearing Handicap Inventory 48 and assessed symptoms at the time of clinical evaluation. For each item, participants were asked to report the age (in years) at first occurrence. If onset of symptoms was prior to age of germ cell tumor diagnosis, those responses were not considered when assigning severity grade. ** h A total of 1,321 patients underwent audiometric assessment, among whom 1,202 received one of the chemotherapy regimens shown in this table. Hearing loss was defined following methods in Frisina et al. 13 using American Speech-Language-Hearing Association criteria.
17 † † i Creatinine measurements were available for the first 976 enrolled TCS, of whom 369 received BEPX3, 299 received EPX4, 181 received BEPX4, 27 received VIPX4/VIPX5, and 14 received cisplatin-based chemotherapy ≥5 cycles. The estimated glomerular filtration rate was calculated following methods in Levy et al. 18 See Table A1 for details.