The impact of single nucleotide polymorphisms on return-to-work after taxane-based chemotherapy in breast cancer

Purpose Breast cancer treatment is associated with adverse effects, which may delay return-to-work. Single nucleotide polymorphisms (SNPs) may influence the risk and severity of treatment toxicities, which in turn could delay return-to-work. We examined the association of 26 SNPs with return-to-work in premenopausal women with breast cancer. Methods Using Danish registries, we identified premenopausal women diagnosed with non-distant metastatic breast cancer during 2007‒2011, assigned adjuvant combination chemotherapy including cyclophosphamide and docetaxel. We genotyped 26 SNPs in 20 genes (ABCB1, ABCC2, ABCG2, CYP1A1, CYP1B1, CYP3A, CYP3A4, CYP3A5, GSTP1, SLCO1B1, SLCO1B3, ARHGEF10, EPHA4, EPHA5, EPHA6, EPHA8, ERCC1, ERCC2, FGD4 and TRPV1) using TaqMan assays. We computed the cumulative incidence of return-to-work (defined as 4 consecutive weeks of work) up to 10 years after surgery, treating death and retirement as competing events and fitted cause-specific Cox regression models to estimate crude hazard ratios (HRs) and 95% confidence intervals (CIs) of return-to-work. We also examined stable labor market attachment (defined as 12 consecutive weeks of work). Results We included 1,964 women. No associations were found for 25 SNPs. The cumulative incidence of return-to-work varied by CYP3A5 rs776746 genotype. From 6 months to 10 years after surgery, return-to-work increased from 25 to 94% in wildtypes (n = 1600), from 17 to 94% in heterozygotes (n = 249), and from 7 to 82% in homozygotes (n = 15). The HR showed delayed return-to-work in CYP3A5 rs776746 homozygotes throughout follow-up (0.48, 95% CI 0.26, 0.86), compared with wildtypes. Estimates were similar for stable labor market attachment. Conclusion Overall, the SNPs examined in the study did not influence return-to-work or stable labor market attachment after breast cancer in premenopausal women. Our findings did suggest that the outcomes were delayed in homozygote carriers of CYP3A5 rs776746, though the number of homozygotes was low. Supplementary Information The online version contains supplementary material available at 10.1007/s00280-022-04499-z.


Introduction
Advances in breast cancer diagnosis and treatment have enlarged the pool of breast cancer survivors [1], emphasizing the need to better understand breast cancer survivorship. Return-to-work may be a marker of recovery and return to daily activities after concluding breast cancer treatment [2].
Up to 80% of women with breast cancer return-to-work during or after adjuvant treatment, but some may have delayed or may never return-to-work [3][4][5]. Return-to-work after breast cancer may be hindered by the type of work involved and by the working environment, but may also be affected by survivor well-being, health and functional impairment, societal factors, socioeconomic position, and family support [6,7]. Research suggests that chemotherapy may impede return-to-work in breast cancer survivors, probably due to adverse effects during or after chemotherapy [5], this may be affected by socioeconomic position.
Premenopausal women receive taxane-based adjuvant chemotherapy as guideline treatment [8]. In the absence of their cancer, these women are likely to contribute substantial person-years to the workforce [9]. While improving survival [10], taxanes induce a number of potentially severe adverse effects [11][12][13]. Single nucleotide polymorphisms (SNPs) in genes related to taxane transport, drug metabolism, neural function/repair or DNA-repair mechanisms have been linked to increased risks of severe adverse effects, including chemotherapy-induced peripheral neuropathy [14][15][16][17]. Accordingly, such SNPs may be associated with slower recovery and delayed return to normal daily activities, including work. Tamoxifen treatment, given to premenopausal women with estrogen receptor (ER) positive tumors, may modify these associations [18]. Tamoxifen shares metabolizers and transporters with taxanes, and is associated with adverse effects [18,19], which may also influence return-to-work.
No studies have explored whether SNPs connected to taxane effectiveness or adverse effects influence returnto-work after cancer. Therefore, we examined this in premenopausal breast cancer patients treated with taxane-based chemotherapy. Furthermore, we examined the mediating role of estrogen receptor (ER) status and indicators of socioeconomic position.

Data sources
Denmark has a tax-supported population-wide health care system [20]. We linked individual-level electronic data from Danish administrative and medical registries with biological data using a unique ten-digit personal identifier assigned to all Danish residents at birth or immigration [21]. The Danish Breast Cancer Group (DBCG) registers all incident breast cancers, along with clinical information and follow-up data on recurrences and other malignancies [22]. The Danish National Pathology Registry [23] has routinely recorded information on all histopathological analyses and the whereabouts of associated formalin-fixed paraffin-embedded (FFPE) tissue blocks. The Cause of Death Registry records date of death along with underlying and contributory causes registered by the inspecting doctor [24]. In addition, we summarized comorbidities via the Charlson Comorbidity Index using diagnoses identified in the Danish National Patient Registry [25]. Information on childbirths after breast cancer diagnosis was collected from the Danish Medical Birth Registry [26]. Highest attained education level at date of breast cancer diagnosis was collected from the Danish Population's Education Registry [27], household income from the Danish Income Statistics Registry [28], and cohabitation status from Danish Civil Registration System [20].
The Danish labor market model, also known as the flexicurity model, favors employers with flexible hiring and firing rules, while it safeguards employees with a generous social system and security net [29]. The Danish state provides substantial subsistence payments, unemployment benefits, and a range of social and health-related benefits. Since 1991, these payments have been registered on a weekly basis in the Danish Register for Evaluation of Marginalization (DREAM) [30]. During the study period (see below), the length of employer-paid sick leave ranged between 14 and 31 days; employer-paid sick leave is not registered in DREAM. Assuming breast cancer patients undergoing surgery and chemotherapy have longer periods of sick leave, DREAM can capture the length of absence from the labor market, and hence return-to-work.

Study cohort
Our study cohort was nested in the ProBe CaRe (Predictors of Breast Cancer Recurrence) cohort [31]. This cohort includes premenopausal women diagnosed with incident non-distant metastatic breast cancer in Denmark during 2002-2011 (n = 5959), registered in DBCG. We restricted to women who were diagnosed during 2007-2011, during which period most premenopausal breast cancer patients were recommended three cycles of epirubicin and cyclophosphamide every third week, followed by three cycles of docetaxel, while some received sequential docetaxel and cyclophosphamide [32]. We included the women who at diagnosis were aged ≤ 55 years, had chemotherapy as intended adjuvant treatment, and were employed any time during the 2 months before breast cancer primary surgery. We excluded women who were on maternity leave during the week of surgery (Supplemental Fig. S1). We expected all women to be not working at least 1 day during the week of surgery regardless of any payouts.
Seven SNPs were genotyped in a previous project [18] and nineteen SNPs were genotyped for this project using commercially available TaqMan assays on a StepOne Plus real-time instrument (Applied Biosystems, Thermo Fisher Scientific, Foster City, California, USA). Genotyping was performed using 2 µL genomic DNA (10 ng/µL) extracted from FFPE tissue, 5 µL TaqMan Genotyping Master Mix, and 0.5 µL TaqMan allelic discrimination assay (VIC-and FAM-labeled probes) in a final volume of 10 µL. Thermal cycling conditions were: 95 °C for 10 min followed by 50 cycles of 95 °C for 15-s, and 60 °C for 60-s. Genotype calling was performed using the QuantStudio Software V1.3 with automatic calling. After automatic calling genotype results were manually inspected, acceptance was overridden manually if irregular amplification curves were observed. We compared the observed genotype frequencies with those expected under Hardy-Weinberg equilibrium (HWE), and allele frequencies with those reported in European non-Finnish female populations reported in the Genome Aggregation Database (gnomAD) [33].

Outcomes
We assumed the women were working if they did not receive any social benefits, as done in other studies [30,[34][35][36][37]. We also included women receiving substituted unemployment benefits for part-time work or educational grants (see coding list, Supplemental Table S1). We defined return-to-work as 4 consecutive weeks of work. We examined stable labor market attachment defined as 12 consecutive weeks of work.

Covariates
Patient, tumor, and treatment characteristics included age group, comorbidities, education level, cohabitation/marital status, household income, ER status combined with endocrine therapy, double/triple negative tumors, TNM stage [38], grade (in ductal and lobular tumors), surgery type, and intended radiotherapy. Detailed categorizations of the covariates are listed in Supplemental Tables S1 and S2.

Statistical analyses
We examined the cumulative incidence of return-to-work and stable labor market attachment using the Nelson-Aalen estimator, treating death and retirements as competing risks [39]. To examine time to return-to-work and stable labor market attachment, we fitted cause-specific Cox regression models to calculate crude hazard ratios (HRs) and associated 95% confidence intervals (CIs) of return-to-work and stable labor market attachment by genotypes within time periods all counting from day of surgery. Follow-up continued until return-to-work or stable labor market attachment or until maternity leave, childbirth, recurrence, death, emigration, other malignancies, early or normal retirement, or 25th September 2017. All statistical analyses were conducted using SAS software (Cary, NC).

Additional analyses
We examined effect-measure modification stratifying the univariate models by ER/endocrine therapy status, income, education level and cohabitation/marital status. We performed several sensitivity analyses by alternative pre-surgery employment criteria: narrowing the assessment window to 4 weeks pre-surgery and applying a stricter criterium of at least 4 weeks of employment up to 2 months before surgery. Breast cancer survivors with physical or psychological sequelae may qualify for a flexible job schedule [40]. We, therefore, reran analyses including flexible job schedules in the return-to-work outcome. As suggested by others [41], we stratified our assessment of CYP3A4 by CYP3A5 genotype, considering any variant carriers versus wildtype.
The cumulative incidence of return-to-work was lower in CYP3A5 rs776746 homozygotes (n = 15) than seen in wildtypes (n = 1600) and heterozygotes (n = 249). Illustrated in Fig. 2A, the cumulative incidence of returnto-work in wildtypes increased from 25% at 6 months to 94% at 10 years, in heterozygotes from 17 to 94%, and in homozygotes from 7% at 6 months and 82% at 10 years. We observed a similar delay in the cumulative incidence of stable labor market attachment among homozygotes (Fig. 2B).
We observed associations for other SNPs (see Supplemental Table S4), but these had limited numbers of homozygotes, and inconsistencies between cumulative incidence curves and HRs suggesting these were probably chance findings.
Although based on small strata, we observed no effectmeasure modification by ER status or socioeconomic position. In the analyses factoring in flexible job schedules, the cumulative incidence of return-to-work reached 97%, but relative estimates were similar in all sensitivity analyses.

Discussion
In this study, premenopausal women had a high cumulative incidence of return-to-work and stable labor market attachment, reaching 94% and 93% at 10 years after treatment for early breast cancer. These levels were reached already within 3-4 years. Homozygote carriers of CYP3A5 rs776746 had delayed return-to-work and stable labor market attachment, compared with wildtypes.
Our overall assessment of return-to-work extends previous research. Arndt et al. [4] studied the cumulative incidence of return-to-work among 1,070 women with breast cancer living in Germany. Compared with our findings, they reported a slightly lower 10-year cumulative incidence of 85%, presumably due to their older study cohort including postmenopausal breast cancer survivors. Their study was prone to selection bias as they only included 5-year survivors, and therefore could have overestimated return-towork as women dying within 5 years after surgery were then excluded. However, the higher incidence of return-to-work in our study may reflect long work life expectancy in premenopausal women and successful rehabilitation.
We investigated genes that theoretically could influence treatment effectiveness/adverse effects or that previously had been associated with such. The lack of association of the SNPs with return-to-work is encouraging for patients and contrasts with some other studies. Kus et al. [17] investigated 219 Turkish breast cancer patients and found that the CYP3A4 rs2740574 variant allele was associated with lower risk of neuropathy, especially for chemotherapy-induced peripheral neuropathies that interfered with activities of daily living. We note that our HRs for CYP3A4 rs2740574 were somewhat consistent with increased return-to-work, but increased return-to-work was not evident in our cumulative incidence curves. Furthermore, the Turkish study had a substantially higher frequency of CYP3A4 rs2740574 homozygotes of 42% compared with less than 1% in our Danish cohort. A study by Eckhoff et al. [16] of 150 Danish early stage breast cancer patients, and thus similar allele frequencies to our population, found that GSTP1 rs1138272 variant carriers had increased risk of chemotherapy-induced peripheral neuropathy during docetaxel treatment, also when examining neuropathies graded ≥ 2 [5]. Still, we found no evidence of an influence of GSTP1 on return-to-work.
Our findings of delayed return-to-work in CYP3A5 rs776746 homozygote women may indicate poorer recovery compared with their wildtype counterparts. CYP3A5 is a phase 1 enzyme involved in the metabolism of docetaxel in the liver. CYP3A5 rs776746 is highly polymorphic and  [41]. As such, expressors could be expected to have higher drug clearance. This has been found in one study including Caucasian cancer patients (27% of whom had breast cancer) treated with docetaxel [41]. The CYP3A5 non-expressing variant has been associated with reduced risk of neurotoxicity during treatment in 118 Spanish cancer patients (1/3 breast cancer) treated with paclitaxel [42], corresponding to an increased risk in expressors. Although the Spanish study did not examine long-term adverse effects, these findings could potentially explain our observed delayed return-towork in CYP3A5 expressors in our study. Studies on chemotherapy-induced peripheral neuropathy in taxanes suggest that the symptoms during therapy resolve or diminish after the end of treatment [43,44]. Still, some symptoms may persist [43]. Our findings may indicate longer term adverse effects in CYP3A5 rs776746 homozygotes. Nonetheless, only 15 women in our cohort were homozygotes for CYP3A5 rs776746; therefore, our findings may also be attributable to chance. The major strength of our population-based study was the nationwide genotyping of premenopausal breast cancer survivors, and the linkage to validated clinical and individual data with high completeness [20,27,30,[45][46][47]. We incorporated several quality-control measures to ensure highquality SNP data. We did not only rely on Hardy-Weinberg disequilibrium, as this can be influenced by sample size [48,49]. Instead, we inspected (and present) congruence between observed and expected frequencies. We excluded SNPs with call rates < 95% and with overlapping genotype clusters. DNA was derived from FFPE tumor-infiltrated tissue, which previously has been proven suitable for studies of breast cancer prognosis [50]. Moreover, studies report high genotype concordance between FFPE breast tumors and both FFPE normal lymph nodes [51,52] and plasma [51]. Fig. 1 Cumulative incidence of return-to-work and stable labor market attachment Fig. 2 Cumulative incidence of return-to-work (A) and stable labor market attachment (B) by CYP3A5 genotype. Curves were smoothed using loess function. The shaded bands represent associated 95% confidence intervals Some limitations must be considered. We had no information on adverse effects and individual information on administered chemotherapy, including docetaxel plasma concentrations or cumulative dose. Adverse effects are poorly registered in Danish health registries, and chemotherapy dose capping could have been associated with treatment toxicity. Information on treatment toxicity and later adverse effects could have substantiated our interpretation but are unlikely to confound our estimates. None of the women in our study were intended to receive docetaxel monotherapy. Our findings could, therefore, be influenced by interaction as CYP3A5 is also involved in cyclophosphamide metabolism [53].
Despite high validity of social benefit payments registered in DREAM [30,47], this database has its limitations. Our outcomes relied on the assumption that no evidence of a payment record was equivalent to employment. Validated against self-reporting, self-supportiveness (defined by no DREAM entry, student grant, leave-ofabsence schemes including maternity leave) has a positive predictive value of 98% [30]. However, this may not always indicate employment as withdrawal from the work force could be supported by savings or spouse earnings. A Danish study examining income changes after breast cancer found that those married had a lower income up to 9 years after diagnosis, while this was 6 years for those who were single [54]. This suggests that married breast cancer survivors are supported economically by their spouse.
In a study examining return-to-work after maternity leave, we validated employment (defined as no payment, or vacation from employment payouts) against records of salary payments and found an agreement of 94% (unpublished). It is likely that some women choose not to returnto-work and also avoid the bureaucracy associated with registering for social benefit payments from the public sector. In that case, we may have overestimated the cumulative incidence of return-to-work.
Our study provides novel insights that argue for more research on the impact of CYP3A5 rs776746 on recovery in women treated with taxanes and cyclophosphamide. Such research could help identify women at risk of poor recovery after taxane-based chemotherapy. As we only examined the associations of single SNPs, future studies should include Bayesian pathway analysis considering the entire complex metabolic pathway of docetaxel [18,55].

Conclusion
In this population-based cohort of premenopausal breast cancer survivors with non-distant metastatic breast cancer, homozygous carriers of CYP3A5 rs776746 had delayed return-to-work and stable labor market attachment after breast cancer. These associations-and their underlying mechanisms-need to be investigated further. Still, if validated elsewhere, these findings may indicate the utility of CYP3A5 rs776746 to identify women at risk of a poor clinical course, who may benefit from enhanced supportive care during treatment and follow-up. Author contributions CFH: conceptualization, methodology, formal analysis, investigation, writing-original draft, visualization, funding acquisition, project administration. PD: conceptualization, resources, investigation, data curation, supervision, methodology, writingreview and editing. TBS: conceptualization, supervision, methodology, writing-review and editing. SF: resources, data curation, software, writing-review and editing. SHD: data curation, resources, investigation, writing-review and editing. BE: resources, supervision, writing-review and editing. HB: supervision, writing-review and editing. TLL: conceptualization, supervision, investigation, funding acquisition, writing-review and editing. HTS: investigation, resources, writing-review and editing. DCF: conceptualization, supervision, data curation, methodology, investigation, writing-original draft, writing-review and editing, funding acquisition, project administration.
Funding This work was supported by grants from the Danish Cancer Society (R167-A11045-17-S2 to DCF); Aarhus University (CFH); the Danish Cancer Research Foundation (FID1839672 to CFH); the Lundbeck Foundation (R167-2013-15861 to DCF) and the Novo Nordisk Foundation (NNF19OC0058710 to DCF). The ProBe CaRe cohort infrastructure was supported by the US National Cancer Institute (R01CA166825 to TLL). The funders had no role in the design, conduct or publication of the present study.

Data availability statement
The dataset generated and analyzed during the current study are not publicly available due to Danish legislations but are available from the corresponding author and Statistics Denmark on reasonable request and permissions from the below mentioned third parties. All the data used for the current study are stored at Statistics Denmark. Researchers can apply to Statistics Denmark for data access, conditional on obtained permissions from the DBCG, the Danish Health Authorities, the Danish Data Protection Agency, and the Central Jutland Region Committee on Health Research Ethics.

Declarations
Conflict of interest All authors declare no support in relation to the present study. TBS receives consultancy fees from Pfizer and teaching fees from Orifarm, Eisai, Novartis, and Astellas Pharma. TLL receives consulting fees and travel support for his participation in the Amgen Methods Advisory Council. BE receives institutional grants from AstraZeneca, Eli Lilly, Merck Sharpe & Dohme, Novartis, Pfizer, Roche, and Samsung Bioepis. The Department of Clinical Epidemiology, Aarhus University Hospital, receives funding for other studies from European Medicines Agency and from companies in the form of research grants to (and administered by) Aarhus University.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.