Impact of antiangiogenic treatment on the erectile function in patients with advanced renal cell carcinoma

We longitudinally assessed erectile function as well as the willingness to use pro‐erectile treatment in a cohort on AAT for advanced RCC. Thirty‐seven patients with advanced RCC completed the five‐item version of the International Index of Erectile Function (IIEF‐5) and other interview items before (T0) and 12 weeks into therapy (T12) with AAT. Patients were further asked if they were willing to use and pay out‐of‐pocket for on‐demand treatment with phosphodiesterase‐5‐inhibitors (PDE‐5i). Statistical analysis was performed using nonparametric hypothesis testing. The IIEF‐5 score at T12 was significantly decreased compared with T0 (p < .001). Subjective patient satisfaction regarding their sexual lives was associated with higher IIEF‐5 scores at both time points (p = .006 and p = .03, respectively). At T12, subjective sexual contentment showed a nonsignificant trend towards decline (p = .074). Patients who opted for medical treatment of ED showed significantly better IIEF‐5 scores at both time points compared with the rest of the cohort (p < .001 and p = .005, respectively). In summary, AAT seems to have a negative effect on erectile function in RCC patients, however, the role of psychosocial issues warrants further elucidation. Affected patients may benefit from a proactive approach promoting medical treatment of erectile dysfunction during AAT.

domains of sexuality; however, no association with AAT-related fatigue or biological parameters was found (Denouel et al., 2018). An earlier study also reported a sexual decline, including erectile function, in 38 male patients on AAT for advanced RCC (Bessede et al., 2011).
Sexual function in men can be assessed with the previously validated International Index of Erectile Function (IIEF) (Rosen et al., 2002(Rosen et al., , 2006). An abridged five-item version of the questionnaire (IIEF-5) is a widely used tool to evaluate erectile function (Rosen et al., 1999). The score has previously been used to evaluate erectile dysfunction (ED) in patients undergoing cancer therapy (Huang et al., 2016;Schoentgen et al., 2019).
The gold standard for the medical treatment of ED is phosphodiesterase-5-inhibitors (PDE-5i). All available PDE-5i drugs yielded a significant increase of erectile function in large randomised studies (Brock et al., 2002;Goldstein et al., 1998Goldstein et al., , 2012Hellstrom et al., 2002). However, high medication costs can lead to drug discontinuation in nearly one third of patients (Kim et al., 2014).
In this study, we prospectively addressed sexual activity and the dynamics of erectile function in male patients treated with first-line AAT for advanced RCC at our institution. Moreover, the willingness to use and pay for PDE-5i and the association with erectile function during AAT was assessed.

| Patient selection
This study was approved by the Institutional Review Board under the premise to analyse anonymised data. Inclusion criteria comprised the diagnosis of locally advanced or metastatic renal cell carcinoma and treatment initiation with antiangiogenic drugs. In the urologic oncology outpatient clinic of our tertiary care academic centre, 85 patients with metastatic RCC were approached to participate in the study between September 2012 and December 2018. They were asked to complete anonymised questionnaires before initiation of first-line AAT. Assessments were carried out at the beginning (T0) and 12 weeks after the start of systemic treatment (T12). A flow chart displaying the patient selection process is shown in Figure 1.

| Patient interview items
In order to maintain anonymity, patients received questionnaires to fill out privately after appropriate instruction. Three interview items were included to assess the patients' subjective perspective on their sexual lives and whether they noticed an improvement or a decline of erectile function at T12. The questionnaire further included social variables, including relationship status and whether regular sexual intercourse (>1×/month) was practised. Additionally, patients were asked whether they desired medical treatment of ED and if they were willing to pay for it out-of-pocket.
To evaluate erectile function, we used the standardised five-item version of the 'International Index of Erectile Function' scale (IIEF-5). The complete questionnaire in English, including the IIEF-5 scale and the aforementioned interview items, is provided in Appendix S1. Erectile dysfunction was based on the calculated score and was categorised as follows: no ED (score: 22-25), mild ED (score: 17-21), mild-to-moderate ED (score: 12-16), moderate ED (score: 8-11) and severe ED (score: <8) (Rosen et al., 1999). Following completion of the survey patients were offered personal counselling, in-depth assessment and upon request, medical treatment of erectile dysfunction, in the form of an on-demand treatment with sildenafil, with dose titrations performed depending on patients' responses. This offer was F I G U R E 1 Flow chart displaying the process of patient selection and percentage of men who requested PDE-5i possible, as the treating urologic oncologists in our outpatient clinic are urologists with expertise in andrology and sexual medicine.

| Statistical analysis
All statistical analyses were performed using the R platform v3.5.3.
Nonparametric hypothesis testing was performed as data were nonnormally distributed. For unpaired data, Fisher's exact test was used for binary and the Mann-Whitney U test for numeric outcomes. With regard to paired data, the Wilcoxon signed-rank test was performed for numeric and McNemar's test for binary outcomes. RCC subtypes were divided into clear cell (ccRCC) and nonclear cell RCC (nccRCC).
Patients with concurrent ccRCC and nccRCC lesions were counted as nccRCC. To assess a potential correlation between IIEF-5 score variation between T0/T12 and patient age, IIEF-5 values at T0 were subtracted from scores at T12 and the Spearman rank correlation test was carried out. Tests were performed two-sided; statistical significance was defined by a p-value of <.05. However, 2/37 (5.4%) patients displayed both ccRCC and pRCC lesions in their pathological report and were counted as nccRCC variants. Thirty-three (89.2%) patients underwent TKI therapy, while the rest of the group received an mTOR-inhibitor (n = 2, 5.4%) or a TKI/ mTOR-inh combination regimen (n = 1, 2.7%) as first-line treatment. Figure 2 illustrates the distribution of IIEF-5 scores for the two time points of interest. The median IIEF-5 score at baseline was 13 (IQR: 6-20). Twelve weeks into therapy, the median IIEF-5 score significantly decreased to 9 (IQR: 2-16, p < .001). The number of patients in each of the IIEF-5 categories at both time points is listed in Table 2.

| Erectile function at T0 and T12
As the patient age ranged from 40 to 84.2 years, we performed a Spearman rank correlation test to explore a potential correlation between age and variations of erectile function scores between T0 and T12. However, no such correlation was found (Spearman ρ = .11, p = .51). There also was no significant association between IIEF-5 scores at T0/T12 and RCC subtype (p = .97 and p = .72, respectively).

| Subjective assessment of sexual function
At T0, 21 (56.8%) patients reported to be satisfied with their sexual lives.
For both time points, a satisfactory sexual life was associated with higher IIEF-5 scores (p = .006 and p = .03, respectively). At T12, 4 (10.8%) patients described an improvement of their erectile function, while 28 (75.7%) men reported a deterioration; the remaining 5 (13.5%) patients described neither of both. Patients reporting an improvement of erectile function showed a statistical trend regarding an increase of IIEF-5 scores compared with the rest of the cohort (median score increase: +0.5 vs. −1, p = .078). In patients who reported a subjective decline of erectile function, a nonsignificant decrease of IIEF-5 scores was found with comparisontotheothercases(medianscoredecrease:−1.5vs.−1,p = .27); no significant association between subjective deterioration and sexual satisfaction was found (p = .14). Strong consideration of a PDE-5i therapy (presuming the willingness to pay for medication) was only observed in 1/4 patients with subjective improvement and no significant association was shown (p = 1). Subjective improvement was also not associated with patients' subjective satisfaction with their sexual lives at T12 (p = .3).
Regular sexual intercourse was significantly associated with higher IIEF-5 scores at both time points (p = .001 and p = .002, respectively) and with subjective sexual satisfaction at T0 (p = .02) ( Table 3). There was no significant association between IIEF-5 scores and relationship status and no significant association between histological variants (ccRCC vs. nccRCC) and sexual satisfaction at T0 or T12 (p = .49 and p = .74, respectively) .

| Willingness to use and pay for PDE-5i
A total of 21/37 (56.8%) men in our cohort were interested in willing or able to pay for medication at T0 and 13/21 (61.9%) at T12. Both T0 and T12 IIEF-5 scores were significantly higher in patients who were willing to pay for PDE-5i (p < .001 and p = .005, respectively, Figure 3). Sexual satisfaction in these patients was significantly higher at both time points (p = .048 and p = .036, respectively).

| D ISCUSS I ON
Previous studies have shown that sexuality plays an essential role for all age groups, including the elderly population, and can be con-

T0 T12
Median IIEF-5 scores in patients with/ without authors reported that age was not associated with a variation of the IIEF score between baseline and 1-month follow-up time point. This is relatively consistent with our study, in which no correlation between erectile function and patient age was found at either time point.
A major aspect, that is still insufficiently elucidated, is the direct effect of antiangiogenic drugs employed in RCC on sexual function.
In this regard, mTOR-inh may exert a negative effect on gonadal hormones, as previously suggested in patients following renal transplantation (Huyghe et al., 2007). However, in an earlier study looking into the association between mTOR-inh and erectile function, including 66 male renal transplant recipients, no significant difference in IIEF scores could be found (Lee et al., 2005). In the present study, only three F I G U R E 3 Boxplots visualising median IIEF-5 scores of patients willing to use and pay for PDE-5i therapy at both T0 (a) and T12 (b). Interquartile ranges are marked by the hinges, values within 1.5× of the IQR are displayed by the whiskers patients received mTOR-inh, precluding a statistically conclusive attribution of the effects of this substance group. TKIs target the VEGF pathway and earlier studies hypothesised that TKIs could lead to sexual dysfunction by causing an endocrine imbalance (Bessede et al., 2011;Wong et al., 2007). However, studies of imatinib, a TKI inhibiting the platelet-derived growth factor receptor (PDGFR) and used in patients with chronic myeloid leukaemia, have shown a favourable impact on erectile and vascular function. In vitro and animal tests with imatinib led to smooth muscle relaxation via the nitric oxide/guanosine monophosphate pathway in the human tissue (Gur et al., 2010(Gur et al., , 2013. These contradicting results with respect to the in vivo impact of AAT in humans, impede to draw causal links between AAT use and ED. More so, as erections are caused and influenced by the complex interplay of neurological,vascular,andpsychologicalfactors (Marcon&Stief,2020).
In this context, cancer-related symptoms, such as pain, fatigue, or altered body image may also negatively influence sexuality. However, in the present study, only 70% of the patients interested in ED treatment were willing to pay for it out-of-pocket. A main concern for patients is the treatment price which is usually not covered by health insurance policies (Kim et al., 2014).
Recent patent expirations of PDE-5i drugs have led to the launch of affordable generics. This enormous cost alleviation has ignited discussions in different countries, whether reimbursement for PDE-5i should be permitted (Hansen et al., 2020). Both price reductions due to upcoming patent expirations and potential reimbursement may increase ED therapy acceptance in affected patients.
Limitations of our study include the large proportion of patients not completing the questionnaire, the missing long-term follow-up as well as the absence of a control arm. This study was focused on male sexuality and in particular erectile function, while female sexual dysfunction on AAT was not contemplated. Anonymisation of patient data prevented assessment of detailed clinicopathologic parameters, influence of biological variables, detailed comorbidities (e.g. by using the Charlson Comorbidity Index, Charlson et al., 1987) and observation of adverse events related to either targeted agents or PDE-5i. A detailed documentation of dosages used after the common starting dose of 50 mg sildenafil was also not carried out, as the focus of this work was put on patients' motivation to use PDE-5i.
Another aspect regards the missing long-term follow-up data which, in combination with information on therapy response (e.g. by performing a RECIST-based imaging analysis), could shed light on the association between response and sexual function.
Lastly, the study started before the advent of immune check-

| CON CLUS ION
After 12 weeks of AAT, patients with advanced RCC showed a decline of erectile function scores. With regard to patients' contentment with their sexual lives, the results after 12 weeks trended towards a lower satisfaction. Furthermore, patients with both higher objective and subjective erectile function were more inclined towards medical treatment of erectile dysfunction. The impact of the psychosocial burden in affected patients warrants further elucidation.

ACK N OWLED G EM ENT
Open access funding enabled and organized by Projekt DEAL.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available on request from the authors.