Introduction

Bladder cancer (BC) ranks as the ninth most common cancer worldwide for both sexes combined, and the age-adjusted incidence varies between 1.2 per 100,000 in Middle Africa and 15.2 per 100,000 in Southern Europe.1, 2 Whereas neoadjuvant chemotherapy and radical cystectomy are recommended for muscle-invasive BC, high-grade nonmuscle-invasive BC (NMIBC) should be treated with intravesical bacillus Calmette-Guerin (BCG) immunotherapy.3 After a weekly induction course, BCG is administered for up to 3 years as maintenance therapy. A meta-analysis demonstrated that BCG reduced the odds of progression to muscle-invasive BC by 27%.4 Interestingly, clinical trials for NMIBC are currently combining BCG with other immunotherapies, underlining the relevance of BCG for BC therapy.5 However, even after BCG instillations, patients bear a 5-year risk of recurrence of up to 67.6% in the highest risk group according to the Spanish Urological Club for Oncological Treatment (CUETO) risk calculator.6, 7 Consequently, patients with NMIBC need to undergo lifelong follow-up, including imaging, cystoscopy, and, importantly, urinary cytology.

Cytology is noninvasive, bears no radiation exposure, and can detect carcinoma in situ, which neither is visible on computed tomography scans nor can be reliably detected during white-light cystoscopy. Since 2015, The Paris System for Reporting Urinary Cytology (TPS) has defined the criteria for malignant urothelial cells.8, 9 For this reason, the TPS uses key cytomorphologic criteria, including the nuclear:cytoplasmic ratio, hyperchromasia, nuclear borders, and chromatin structure. For high-grade urothelial carcinoma (HGUC) in unselected cohorts, sensitivity and specificity reportedly range between 70.5% to 84% and 78.5% to 94%, respectively.10-12 Importantly, despite significant scientific efforts, there is still no guideline-supported, commercially available diagnostic biomarker for BC that outperforms urinary cytology.13

Even 40 years after the approval of BCG, the underlying biologic mechanism remains elusive.14 Urothelial cells in this setting undergo morphologic alterations secondary to a complex interaction between BCG and the local microenvironment. Notably, changes, including prominent nucleoli, reactive atypia, cytoplasmic vacuolization and inclusions, and an increase in granulocytes, can be observed even months after the last BCG instillation.15, 16 Previous studies investigating the diagnostic accuracy of post-BCG cytology demonstrated wide variation in sensitivity, ranging from 33% to 67%.17, 18 However, none of those studies consistently used the TPS.

For the current study, we primarily investigated the diagnostic performance of post-BCG urinary cytology for different TPS cutoff definitions and time intervals. In addition, the impact of common incidental findings, including granulocytes and lymphocytes, was assessed, and oncologic follow-up of false-positive cases was analyzed.

Materials and Methods Study Design and Patient Cohort

We retrospectively included patients who underwent transurethral resection of bladder tumors (TUR-BT) between October 2011 and May 2018 in our institution. Only patients who had a history of high-grade NMIBC and previous induction or maintenance BCG intravesical therapy were included. Indications for a TUR-BT after BCG therapy were guideline-conformed3 and required a control biopsy after induction BCG therapy, high-grade UC (HGUC) in a urinary cytology sample, and an intravesical tumor-suspect mass in imaging modalities or suspicious cystoscopy during follow-up. If no visible tumor or suspicious urothelium was detected during cystoscopy, systematic cold-cup biopsies were performed. Histopathologic staging and grading of bladder tissue was conducted according to the 2004 and 2016 World Health Organization/International Society of Urological Pathology tumor (T), nodes (N), and metastases (M) systems.19 The study was carried out in accordance with the ethical principles for medical research of the Declaration of Helsinki and its later amendments. All patient data were anonymized before analysis, and the institutional ethics committee approved the study.

Urinary Cytology Classification

A washing cytology was routinely conducted directly after cystoscopy and before resection or biopsies during the TUR-BT procedure. Cytology samples were centrifuged, stained according to Papanicolaou and Marshall,20 fixed, and archived. All samples were processed in the institutional cytology laboratory. Two board-certified urologists extensively trained in urinary cytology analysis (G.B.S. and S.T.) independently reviewed all anonymized cytology samples according to the TPS classification using transmitted light microscopy and ×400 magnification: the samples were classified as non-HGUC (NHGUC), atypical urothelial cells (AUC), suspicious for HGUC (SHGUC) and, HGUC. In case of discordant results, the higher classification was used for statistical analysis. The urinary cytology review was blinded to histopathology and clinical parameters. The presence of significant hematuria, lymphocytes, granulocytes, or bacteria in the urinary cytology specimen was assessed on a nominal scale level. Furthermore, we identified typical cytopathologic findings after BCG therapy, including reactive atypia and signs of cellular degeneration like vacuolization or cell lysis. Cases with an insufficient amount of urothelial cells for urinary cytology classification according to the TPS (nondiagnostic) were excluded.

Statistical Analysis

Interobserver reliability was analyzed using the Cohen κ method. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), including 95% confidence intervals (95% CIs), were assessed for the whole cohort using 3 different cutoff definitions for positive urinary cytology: HGUC (including only HGUC), SHGUC (including HGUC and SHGUC) and AUC (including HGUC, SHGUC, and AUC). For further analysis, both SHGUC and HGUC were counted as positive urinary cytology. Sensitivity, specificity, PPV, and NPV were compared between cases in which the time interval between the last BCG instillation and TUR-BT/cytology was <100 days versus ≥100 days. A binary regression analysis was conducted for all false, false-positive, and false-negative urinary cytology classifications; and common incidental findings like hematuria, lymphocytes, bacteria, granulocytes, reactive atypia, and cellular degeneration were included as independent variables. Follow-up data were based on subsequent diagnostic procedures, including histologic assessment conducted in our institution. IBM SPSS Statistics 25.0 (IBM Corporation) was used for statistical analyses.

Results Diagnostic Performance of Urinary Cytology After BCG Correlates With Cutoff Definitions

In total, 197 patients who underwent 389 TUR-BT procedures and had matched urinary cytology after BCG therapy were included in the current study. Six cases were excluded because of insufficient numbers of urothelial cells in the cytologic specimens (nondiagnostic). In total, 335 of 383 TUR-BT procedures (87.5%) were performed using photodynamic diagnosis (PDD) guidance, and HGUC in the tissue specimen was diagnosed in 60 of 383 pathologic reports (15.7%). Urinary cytology detected non-HGUC in 191 of 383 cases (49.9%), AUC in 80 of 383 cases (20.9%), SHGUC in 56 of 383 cases (14.6%), and HGUC in 56 of 383 cases (14.6%) (Fig. 1, Table 1). The Cohen κ assessment of interrater reliability indicated substantial concordance (κ = 0.660) between the 2 investigators for HGUC, applying the interpretation of Landis and Koch.21

Urinary cytology graded according to The Paris System for Reporting Urinary Cytology is illustrated. Urinary cytology findings were based on the Papanicolaou stain. (A) These typical cells were classified as high-grade urothelial cells according to the Paris system. Several malignancy criteria can be seen, including a nuclear:cytoplasmic ratio >0.7, hyperchromasia, clumped chromatin, irregular nuclear borders, and nuclei with irregular shape. (B) These typical benign urothelial cells were classified as negative for high-grade urothelial carcinoma according to the Paris system. Scale bars = 25 µm.

TABLE 1. Clinicopathologic Characteristics of the Patient Cohort Parameter No. of Matched Cases/Total Cases (%) Age: Median [IQR], y 71 [61-76] Sex Male 323/383 (84.3) Female 60/383 (15.7) Histopathologic staging of the TUR-BT specimen HGUC 60/383 (15.7) pTa 24/383 (6.3) pT1 20/383 (5.2) pT≥2 6/383 (1.6) Concomitant/primary carcinoma in situ/pTis 38/383 (9.9) Grade 1 10/383 (2.6) Grade 2 8/383 (2.1) Grade 3 59/383 (15.4) PDD-assisted TUR-BT 335/383 (87.5) Urinary cytology result: Paris classification NHGUC 191/383 (49.9) AUC 80/383 (20.9) SHGUC 56/383 (14.6) HGUC 56/383 (14.6) Urinary cytology findings Granulocytes 88/383 (23.0) Lymphocytes 80/383 (20.9) Bacteria 6/383 (1.6) Hematuria 16/383 (4.2) Typical cytopathologic finding after BCG therapy Reactive atypia 52/383 (13.6) Cellular degeneration 26/383 (6.8) Abbreviations: AUC, atypical urothelial cells; BCG, bacillus Calmette-Guerin; HGUC, high-grade urothelial carcinoma; IQR, interquartile range; NHGUC, negative for high-grade urothelial carcinoma; PDD, photodynamic diagnosis; pT, pathologic tumor classification; pTis, pathologic carcinoma in situ; SHGUC, suspicious for high-grade urothelial carcinoma; TUR-BT, transurethral resection of bladder tumors.

First, we investigated the diagnostic quality criteria for urinary cytology after BCG therapy using the TPS classification system. Because the clinical relevance of SHGUC is still unclear, we tested for different cutoff definitions. Sensitivity for HGUC after BCG therapy was low (0.45; 95% CI, 0.32-0.58) when HGUC was used as the cutoff. However, sensitivity increased by 66.7% to 0.75 (95% CI, 0.64-0.86) when the threshold was lowered and SHGUC cases were counted in addition to HGUC cases as positive urinary cytology findings. This was accompanied by a moderate relative decrease of 13.1% regarding specificity from 0.91 (95% CI, 0.88-0.94) to 0.79 (95% CI, 0.75-0.84). There was only a negligible increase of 4% in sensitivity when we also added AUC cases, with specificity dropping from 0.79 (95% CI, 0.75-0.84) to 0.55 (95% CI, 0.50-0.61) in this scenario (Table 2). Overall, the diagnostic quality of urinary cytology after BCG instillations can be significantly improved by adding the SHGUC classification to the cutoff definition of a positive cytology result.

TABLE 2. Impact of Different Cutoff Categories on the Diagnostic Accuracy of Urinary Cytology After Bacillus Calmette-Guerin Therapy Cutoff Urinary Cytology Classificationa Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) HGUC 0.45 (0.32-0.58) 0.91 (0.88-0.94) 0.48 (0.35-0.61) 0.90 (0.87-0.93) SHGUC 0.75 (0.64-0.86) 0.79 (0.75-0.84) 0.40 (0.31-0.49) 0.94 (0.92-0.97) AUC 0.78 (0.68-0.89) 0.55 (0.50-0.61) 0.24 (0.18-0.31) 0.93 (0.90-0.97) Abbreviations: AUC, atypical urothelial cells; HGUC, high-grade urothelial carcinoma; CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; SHGUC, suspicious for high-grade urothelial carcinoma. BCG Therapy Decreases the Diagnostic Quality of Urinary Cytology in the Short Term

Previous studies demonstrated that morphologic changes in urothelial cells persisted for several months after BCG therapy. However, to date, no study has been conducted to investigate the time-dependent effect of BCG intravesical instillations on the diagnostic accuracy of urinary cytology. Therefore, we compared sensitivity, specificity, NPV, and PPV between patients who underwent BCG therapy within the last 100 days before TUR-BT/urinary cytology versus patients for whom this time interval was longer. Because of our results (shown in Table 2), we included SHGUC in the definition of positive cases. Interestingly, the sensitivity of urinary cytology for HGUC was 24.4% lower (0.62; 95% CI, 0.41-0.83) within the first 100 days after BCG instillation therapy compared with the sensitivity >100 days after intravesical immunotherapy (0.82; 95% CI, 0.70-0.94) (Table 3). Taken together, BCG seems to negatively affect the diagnostic quality of urinary cytology.

TABLE 3. Time-Dependent Impact of Bacillus Calmette-Guerin Instillation Therapy on Urinary Cytology Efficiency Time Between Last BCG Instillation and Combined Urinary Cytology/TUR-BT Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) <100 d 0.62 (0.41-0.83) 0.82 (0.73-0.90) 0.48 (0.29-0.67) 0.80 (0.81-0.96) ≥100 d 0.82 (0.70-0.94) 0.78 (0.73-0.84) 0.38 (0.27-0.48) 0.97 (0.94-0.99) Abbreviations: BCG, bacillus Calmette-Guerin; CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value; TUR-BT, transurethral resection of bladder tumors. Reactive Atypia and Cellular Degeneration of Urothelial Cells, but Not Common Incidental Findings in Urinary Cytology Samples, Impede its Diagnostic Accuracy

Hematuria, bacteriuria, lymphocyturia, and granulocyturia are frequently observed when analyzing urinary cytology. Because of the mechanism of intravesical immunotherapy, these incidental findings are more common after BCG. Therefore, we investigated the impact of the presence of these cells on the diagnostic quality of urinary cytology for HGUC. Granulocytes were identified in 88 of these cases (23%), lymphocytes were identified in 80 cases (20.9%), hematuria was detected in 16 cases (4.2%), and bacteria were observed in 6 cases (1.6%). Reactive atypia and signs of cellular degeneration, which are typical cytopathologic findings after BCG therapy, were reported in 52 (13.6%) and 26 (6.8%) cases (Table 1, Fig. 2). Intriguingly, reactive atypia (odds ratio, 4.155; 95% CI, 2.136-8.085; P < .001) and cellular degeneration (odds ratio, 5.050; 95% CI, 2.094-12.175; P < .001) were associated with false-positive results. There was no impact of the other more nonspecific cytologic findings on any false, false-positive, or false-negative urinary cytology results (Table 4).

Typical urinary cytology findings after bacillus Calmette-Guerin therapy are shown. (A,B) Toxic vacuolization (asterisks) is observed in the cytoplasm of urothelial cells. (C,D) Cellular degeneration/cell lysis of urothelial cells (arrows) and reactive atypia of urothelial cells (arrowheads) are shown. (E,F) Note the massive presence of granulocytes. Scale bars = 25 µm.

TABLE 4. Impact of Common Incidental Cytology Findings on False Urinary Cytology Results Variable Odds Ratio 95% CI P False urinary cytology Erythrocytes 1.287 0.331-5.006 .716 Lymphocytes 1.013 0.450-2.282 .974 Bacteria 1.385 0.147-13.030 .776 Granulocytes 0.585 0.254-1.348 .208 Reactive atypia 2.945 1.544-5.618 .001a Cellular degeneration 5.818 2.498-13.551 <.001a False-negative urinary cytologyb Erythrocytes 1.179 0.134-10.364 .882 Lymphocytes 1.670 0.405-6.891 .478 Granulocytes 1.194 0.283-5.044 .809 Cellular degeneration 3.448 0.883-13.469 .075 False-positive urinary cytology Erythrocytes 1.244 0.253-6.109 .788 Lymphocytes 0.846 0.335-2.138 .724 Bacteria 2.184 0.244-21.293 .501 Granulocytes 0.479 0.183-1.251 .133 Reactive atypia 4.155 2.136-8.085 <.001a Cellular degeneration 5.050 2.094-12.175 <.001a Abbreviation: CI, confidence interval. Patients With Positive Urinary Cytology Findings but Negative Histology Have a High Risk of Recurrence

Currently, there is no consensus on surveillance strategies for patients who have a positive urinary cytology finding but a negative biopsy specimen. Therefore, we assessed follow-up data from patients who had false-positive (negative histology/positive cytology) urinary cytology findings. There was surveillance information on 38 of the 50 patients (76%) who had false-positive urinary cytology. The median follow-up was 227 days. Notably, 17 of these 38 patients (44.7%) experienced an HGUC recurrence of the upper or lower urothelial tract. Tumor stages of relapses were distributed as follows: muscle-invasive UC (2 of 17 patients; 11.8%), metastatic UC (2 of 17 patients; 11.8%), carcinoma in situ (8 of 17 patients; 47.1%), pathologic Ta (pTa) HGUC (2 of 17 patients; 11.8%), pT1 HGUC (2 of 17 patients; 11.8%), and upper tract urothelial carcinoma (1 of 17 patients; 5.9%). In sum, patients who have unremarkable histology but positive urinary cytology are at high risk for high-grade recurrence.

Discussion

Patients who have BC with HGUC have a high risk of relapse, even when treated with intravesical BCG instillation. A timely diagnosis is mandatory because there are curative salvage treatments, mainly radical cystectomy, but also BCG rechallenges, chemotherapy instillations, or, recently, PD-1–based immune checkpoint-inhibitors.22 Because BCG induces inflammatory processes of the bladder mucosa, the efficacy of white-light cystoscopy and PDD-guided biopsies is diminished. Urinary cytology applying the TPS classification is recommended for follow-up of HGUC, but data on the efficacy after BCG therapy are sparse.

The first objective of our study was to access the impact of different TPS cutoff definitions for positive cytology. Our study demonstrates high specificity but only mediocre sensitivity of post-BCG urinary cytology, especially when using HGUC as a cutoff for positive cytology. These results are in line with a previous study by Gupta et al that included 56 post-BCG patient samples.17 Those authors reported 53% sensitivity and 95% specificity, comparable to 45% and 91%, respectively, reported in our current study. Of note and in contrast to our study, that investigation did not consistently apply the TPS classification for all samples, and cystoscopy did not include PDD guidance. The study by Hara et al included 127 patients who underwent transurethral biopsy in addition to urinary cytology 4 weeks after BCG instillation.23 In that investigation, the PPV was 40.7%, but specificity and sensitivity were not reported. The study by Nurminen et al focused on post-BCG urinary cytology and included 267 biopsies. In that publication, slightly lower sensitivity and specificity values, 39% and 77%, respectively, were reported.18 Guy et al included 130 patients, and both sensitivity (56%) and specificity (92%) of post-BCG cytology were reported.24 Notably, neither investigation applied the TPS classification system. It is clinically relevant that sensitivity markedly increased in our current study when we lowered the threshold of positive cytology by including SHGUC cases, which is in line with a previous study.17 Also, in cohorts that did not receive BCG therapy, urinary cytology seemed to be highly dependent on the cutoff used. A cytologic study that included 110 cases with histopathologic correlation reported a sensitivity of 90.9% when atypical cytology cases were included compared with 70% when SHGUC and HGUC cases were included.25 Sensitivity for HGUC in cohorts that did not receive BCG therapy ranges between 70.5% and 84%.10, 12 Therefore, it seems to be of utmost importance to consider differences in cutoff definitions, classification systems applied, and cohort characteristics when comparing the efficacy of urinary cytology between different studies.

The interrater reliability of our study was substantial for HGUC when we applied the validated, 5-category scale described by Landis and Koch.21 This seems to be acceptable because it is within the range of previously published studies investigating user-dependency of urinary cytology.26-28 In those investigations, interobserver agreement was moderately good26: a κ coefficient of 0.67 was reported for HGUC,27 or a predictive accuracy for BC between 63% and 89% was demonstrated, depending on the institution.28 However, 2 other studies that applied the TPS system reported disappointing interrater reliability, with κ values of 0.3229 and 0.36.30 Taken together, based on the vast increase in sensitivity with only a minimal decrease in specificity, we suggest including SHGUC in the definition of positive cytology.

Next, we addressed the question of whether BCG might have a direct impact on the efficacy of urinary cytology. Interstudy comparisons bear significant limitations, including variations regarding experience with cytology, disparate grading systems, and cohorts. Therefore, we analyzed different time intervals between the last BCG therapy and urinary cytology within the same study. Interestingly, the sensitivity of urinary cytology was >30% higher when urinary cytology assessment was conducted >100 days after the last BCG instillation therapy compared with shorter intervals. Because there were no other possible confounders that could explain this observation, we conclude that BCG seems to have an impact on the efficacy of urinary cytology, especially in the short term.

There are only few reports on BCG-induced morphologic alterations in urothelial cells. Takashi et al compared cytology findings from 22 patients undergoing BCG therapy versus findings from a control group without any intravesical therapy.15 Several reactive morphologic findings, including prominent nucleoli, translucent nuclei, cytoplasmic inclusions, vacuolization of cytoplasm, and an increased nuclear:cytoplasmic ratio, were found with different frequencies. Interestingly, most of these observations occurred within the first 3 months after BCG therapy. Mack et al assessed 20 patients after BCG therapy.31 Those authors reported severe morphologic changes that lasted >2 years, including enlarged nuclei and nucleoli, an increased nuclear:cytoplasmic ratio, hyperchromatic nucleoli, and multiple nucleoli, which impaired the correct interpretation of cytology. The study by Betz et al also identified several morphologic changes after BCG therapy.16 Notably, those alterations were mainly seen within the first 3 months after the last BCG instillation and decreased in the following months. Interestingly, in a cytology study, 24 of 29 patients who had SHGUC/HGUC diagnoses and negative follow-up biopsies had a history of BCG or chemotherapy instillation, which points to the implications of BCG therapy in urinary cytology.32 Taken together, the results indicate that BCG seems to induce morphologic alterations in urothelial cells that impair the diagnostic accuracy of urinary cytology, prevail for several months, but are reversible in the long term. Therefore, our current results underline the need for cystoscopy eventually followed by histologic confirmation in all patients who recently underwent BCG therapy.

BCG induces a sterile inflammation of the urothelium,14 which finds its cytologic correlate in the frequently found presence of inflammatory cells like granulocytes, lymphocytes, and erythrocytes. In patients undergoing BCG therapy, these cells commonly can outnumber urothelial cells. Furthermore, reactive atypia of urothelial cells and signs of cellular degeneration like vacuolization and cell lysis can be found in patients undergoing BCG instillation therapy. These findings may also hamper the sensitivity and specificity of urinary cytology. Notably, we observed a significant association of reactive atypia and cellular degeneration with false-positive results in our study. Reactive atypia is a phenomenon in which inflammatory stress induces morphologic alterations that can mimic dysplastic characteristics. Importantly, these changes cannot be cytologically distinguished. The different grades of cellular degeneration, especially at the beginning of karyolysis, also can imitate dysplastic urothelial cells well, which might explain our observation. We did not identify any association of other findings with discordances between histology and cytology, including false-positive and false-negative cytology results. These results are of clinical relevance because the sheer presence of granulocytes, erythrocytes, lymphocytes, or bacteria does not seem to lessen the validity of the cytology classification and thus does not automatically justify repeat cytology.

Next, we analyzed the oncologic outcome of patients who had negative histology but positive cytology results because this population presents the physician with a dilemma regarding the optimal management.33 Notably, nearly one-half of patients who had a false-positive cytology result were diagnosed with a high-grade recurrence during surveillance—mostly carcinoma in situ. This is remarkable because nearly 90% of the biopsies were performed with PDD guidance. However, based on the inflammatory processes and hypervascularization after BCG instillation therapy, the efficacy of PDD might be severely impaired. On the basis of these results, close follow-up, eventually with repeat biopsies and diagnostic clarification of the upper urinary tract, seems to be necessary in all patients after BCG therapy and positive urinary cytology.

Taken together, our study strengthens the role of urinary cytology after BCG therapy. For high sensitivity, the inclusion of SHGUC cases into the classification of positive cytology should be considered, especially because this suggests only a minimal loss in specificity. Of note, BCG seems to have an adverse effect on cytology within the first months after the last instillation, with diagnostic accuracy resolving thereafter. Finally, patients who have positive cytology but negative biopsies should undergo close follow-up because the frequency of recurrences in this subgroup is exceedingly high.

Our study has several key strengths. First, we present the largest cohort to date of patients who underwent urinary cytology assessment after BCG therapy. Second, the guideline-recommended TPS classification system was consistently applied for all cases. Third, an intrastudy comparison of diagnostic accuracy between different time intervals after BCG-therapy has been addressed for the first time in this study. There are also limitations within this study. Despite the retrospective design, we do not have follow-up information for all patients who had false-positive cytology results.

Currently, several studies are focusing on urinary biomarkers based on the analysis of RNA, DNA, or extracellular vesicles.34 Although the preliminary data are promising, currently, there is no method capable of outperforming urinary cytology. Therefore, performing urinary cytology after BCG therapy is crucial to ensure an optimal oncologic outcome for patients with NMIBC.

Funding Support

No specific funding was disclosed.

Conflict of Interest Disclosures

The authors made no disclosures.

Author Contributions

Julian Hermans: Conceptualization, data curation, formal analysis, validation, and investigation. Friedrich Jokisch: Investigation, validation, and data curation. Yannik Volz: Investigation, validation, and data curation. Lennert Eismann: Investigation, validation, and data curation. Paulo Pfitzinger: Investigation, validation, and data curation. Benedikt Ebner: Investigation, validation, and data curation. Philipp Weinhold: Investigation, validation, and data curation. Boris Schlenker: Investigation, validation, and data curation. Christian G. Stief: Investigation, validation, and project administration. Stefan Tritschler: Conceptualization, data curation, formal analysis, project administration, supervision, and investigation. Gerald B. Schulz: Conceptualization, data curation, formal analysis, writing–original draft, project administration, supervision, visualization, and Investigation.

Acknowledgment

Open access funding enabled and organized by ProjektDEAL.