Upper urinary tract urothelial carcinoma (UTUC) is a rare cancer and accounts for only 5%–10% of genitourinary urothelial cancers (UC) in Western countries (1). UTUC with nontransitional-cell variant histopathology (vUTUC) is an even rarer situation, accounting for 8%~24% among historical UTUC series where the patients were managed with radical nephroureterectomy (RNU) (1–9). vUTUC is commonly associated with more adverse pathological features and advanced disease status at presentation when compared with those of UTUC with pure urothelial pathology (pUTUC) (1–4, 6–9). In addition, patients with vUTUC are usually associated with worse outcomes in overall (OS), cancer-specific (CSS), and disease-free survival (DFS) (1). Therefore, a variant histopathologic feature in UTUC is an important prognostic factor that should be recognized for subsequent treatment planning and disease surveillance.

Bladder UC shares several similar histopathological and prognostic features with UTUC. Variant histology in bladder UC is also associated with adverse pathological features and poor outcomes (10–12). The poor outcomes of variant histology bladder UC raised the speculation of whether perioperative chemotherapy will improve outcomes in this distinct subset of bladder UC. Neoadjuvant chemotherapy has been reported to improve survival outcomes with variant histologic features in bladder UC (10, 13). A recent meta-analysis revealed that good outcomes are associated with chemotherapy for small-cell and micropapillary variants, while chemotherapy has a potential role in squamous cell and adenocarcinoma variants (13).

The success of adjuvant chemotherapy in variant bladder UC certainly led to the speculation whether vUTUC would benefit from adjuvant chemotherapy, but the study of the more aggressive vUTUC is usually limited by its even rare presentation. Although the safety and efficacy of adjuvant chemotherapy for UTUC had been confirmed in the Perioperative Chemotherapy Versus Surveillance in Upper Tract urothelial cancer) trial, the outcomes of adjuvant chemotherapy for patients with vUTUC remain scarce (14). There is no randomized controlled trials (RCTs) or comparative trial focusing on the even rarer and more aggressive vUTUC. Therefore, we conducted a propensity score-matched study to examine the impact of adjuvant chemotherapy on vUTUC.

UTUC is an uncommon malignancy in Western countries, and UTUC associated with variant histology is even rarer; with an overall incidence of 13.4% among the worldwide UTUC cohort (1). The variant histology UTUC is an important adverse prognostic factor affecting most survival domains of UTUC (1, 17, 18). The squamous or/and glandular variant histology was associated with an even worse CSS compared with the other UTUC variants (1). Therefore, vUTUC is a warning and prognostic indicator that could require upfront systemic therapy which usually means chemotherapy before or after RNU for UTUC. Based on recent evidence, perioperative chemotherapy (neoadjuvant or adjuvant) is beneficial in prolonging survivals of UTUC (19–21). Although perioperative chemotherapy had been proved beneficial in bladder UC variants, currently available literature failed to reveal its benefit in vUTUC (3, 4, 8, 22).

Current evidence regarding the efficacy of systemic chemotherapy on vUTUC remained scarce in the literature. As UTUC and bladder UC shared several histological and prognostic features, the regimen of chemotherapy for UTUC commonly follows the classic regimen for bladder urothelial carcinoma. Hence, most urologists/oncologists adapt classic chemotherapy regimens of bladder UC for vUTUC. Hajiran et al. conducted a cohort study to compare the result of cystectomy followed by neoadjuvant chemotherapy. The patients with variant histology (micropapillary, plasmacytoid, nested variant, sarcomatoid, or neuroendocrine variant histology), had a comparable response rate as the pure urothelial carcinoma cohort (23). A National Cancer Database study showed neoadjuvant chemotherapy was associated with significant pathological downstaging for all histological variants (sarcomatoid, micropapillary, squamous, neuroendocrine, and adenocarcinoma) and overall survival improvement for patients with variant bladder UC (24). In addition, MIBC with micropapillary, plasmacytoid variant, or squamous/glandular differentiation should be treated with neoadjuvant chemotherapy according to the recommendation of the 2020 EAU-ESMO Consensus Statements (25). Therefore, there is a clear role of chemotherapy in bladder UC with histological variants. Hence, the current cisplatin-based chemotherapy regimen is potentially effective in urothelial carcinoma with variant histology, as UTUC deserved a prospective randomized trial to confirm our speculation.

There was a national cancer database study that analyzed survivals of the UTUC with variant histology in the renal pelvis after surgery (26). Being restricted to the limited clinicopathological information in the cancer database, it only revealed the survival benefit of adjuvant chemotherapy in patients with pure UTUC but not for the patients with UTUC variants. Indeed, several limitations would hamper the power of cancer database study, such as significant variations in regimen and cycles of chemotherapy, comorbidity of patients, and intergroup bias of adverse histologic factors. Based on our findings, several important clinicopathological factors were independently associated with the survivals of the variant UTUC cohort ( Tables 3 – 5 ). The important clinicopathological data that include comorbidities, surgical margin status, tumor size, tumor location, and lymphovascular invasion were generally not available in the cancer database study, therefore, limited the power of a national cancer database study. Hence, a comprehensive clinicopathological database could help in clarifying the real-world strategies of cancer treatment.

Murakami and their colleagues reported a multi-institutional study where those patients with UTUC with variant histology is an independent risk factor for recurrence-free survival but not for CSS (6). In addition, those vUTUC with pT3 or higher T stage and/or positive lymph node status were indicated for adjuvant chemotherapy. However, due to its retrospective study design and very-limited case number (37 UTUC cases with variant histology), significant selection bias and low power to reveal the real impact of adjuvant chemotherapy. To the best of our knowledge, our study is currently the limited PSM cohort study for the impact of chemotherapy on vUTUC. The current study did not identify the beneficial effect of adjuvant chemotherapy in CSS and DFS for vUTUC ( Tables 4 and 5 ). In addition, the presence of LVI, positive lymph node status, or a positive surgical margin are independent risk factors for poor survival outcomes, and adjuvant chemotherapy should be considered in this vUTUC after RNU.

The impact of the subtypes of variant histology has been reported in the literature without a definite conclusion. The micropapillary, squamous, and/or glandular subtypes were potentially associated with a worse CSS among patients with UTUC in a prior meta-analysis (1). However, the variant histology of adenocarcinoma had been reported to be associated with a better OS compared with pure UTUC in a cancer database study (26). This inconclusive result could be derived from excluding metastatic disease during case enrollment, therefore, excluding advanced variant adenocarcinoma subtype, which is commonly an important factor of lethal disease (27). According to our analysis, different variant subtypes had comparable survival outcomes ( Supplementary Figure S1 ). However, limited by the small case number in each subtype, the real impact of vUTUC subtypes should be clarified with a large-scale prospective study.

Lymphadenectomy has been recommended in RNU for UTUC for its potential survival benefit. In our matched cohort, 23 (46%) of 50 patients without adjuvant chemotherapy and 24 (48%) of 50 patients with chemotherapy underwent lymphadenectomy during RNU. As the role of extended lymphadenectomy remained controversial in UTUC treatment, extended or regional lymphadenectomy was only performed in selected cases who were clinically suspected of having advanced or nodal diseases; otherwise, lymphadenectomy was not performed in low-risk diseases. According to a large population cohort of 16,619 UTUCs, only 15.4% of cases underwent LND; therefore, the proportion of lymphadenectomy in our cohort is clearly not low when compared with the historical series (28).

Our cohort was extracted from the Taiwan UTUC registry which is a multicenter UTUC cohort that enrolled more than 4,000 UTUC cases in Taiwan. In the currently enrolled cohort, we identified 245 (7.2%) vUTUC among 3,043 Taiwan UTUC patients from 16 centers. The incidence of variant UTUC varied among centers in Taiwan; it generally ranges from 2% to 13% with only two exceptions (only two centers had a high incidence of 20% which accounted for only 15% of the vUTUC cohort). According to a recent meta-analysis, vUTUC generally accounted for about 13% of all UTUC worldwide, and the incidence varied significantly among different centers (1). Therefore, significant interobserver variation in making the diagnosis of vUTUC is common and clearly inevitable in a multicenter study. Histology review has been recommended for the multicenter study of vUTUC; however, relevant studies remained extremely scarce in the literature. Last year, we randomly enrolled 154 UTUC cases from the Taiwan UTUC registry for histology review (29). Based on our review, 7.8% and 30.5% variant UTUC were identified by the local or the reviewing pathologist, respectively. Only a slight interobserver agreement was achieved with a kappa value of 0.168. However, whether the vUTUC has an impact on disease outcome, according to the univariate analysis, the presentation of variant histology was the only risk factor of DFS in the local pathology reviewed cohort, but not in the review pathology cohort. This could relate to an overdiagnosis of clinically nonsignificant vUTUC in the review pathologist’s cohort, therefore, certainly underestimating the impact of vUTUC on disease outcome. In summary, whether histology review of vUTUC is beneficial in disease outcome prediction for multicenter study remained controversial and need a prospective large cohort study in the future.

This study has some limitations. First, the current study is still limited by the small sample size, the lack of the desired power level, and the effect size that does not correspond to the required sample size. Second, the pathology was not centrally reviewed; therefore, interobserver reporting bias was also considered one of the limitations. To minimize the impact of in-concordance of pathology between observers, we used a standardized histological report format which was approved by the Pathology Society of Taiwan based on the AJCC TNM staging system, and the principles of pathology management for urothelial cancer in the NCCN guidelines to ensure a standardized management protocol. In addition, genitourinary pathologists in Taiwan followed the same training program, specimen manipulation protocol, diagnostic criteria, standardized report template, and peer review system in each local institution to minimize the interobserver bias. Third, the retrospective nature of the study design is subject to selection bias. We attempted to account for this by the multi-institution enrollment, PSM, and multivariate Cox regression analyses with adjustment of confounding factors. In addition, cases with incomplete or missing information were excluded, except for some minor variables. Fourth, a lymphadenectomy was mainly performed in patients with clinical suspicious nodal diseases or advanced clinical stages, therefore, not routinely performed during RNU in Taiwan. Hence, the rate of pNx was as high as 45%~76%, which significantly undermined our ability to have further interpretation regarding the influence of the nodal status. Fifth, overall mortality and cancer-specific mortality could be partially overlapped in the primary comparison of matched cohorts. To minimize this effect, multivariate analysis was performed to adjust the confounders that could impact the survival. Finally, the lack of standard templates in reporting specific variant histology inevitably introduced further bias in the reported results (24).

In summary, adjuvant chemotherapy following RNU significantly improved CSS and DFS in patients with UTUC with a variant histology in the current propensity score-matched study. Hence, the effect of adjuvant systemic chemotherapy deserves a further prospective, multi-institutional study to elucidate the optimal care for these rare and challenging patients.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

Conceptualization: H-YW and C-WL. Data curation: W-ML, H-LK, Y-HC, H-CW, I-HC, J-TL, C-YH, C-HC, J-ST, W-RL, Y-HJ, Y-KL, C-YT, S-DC, TH, AC, Y-CJ, I-SC, Y-TC, J-SC, B-JC, C-CY, WL, C-CW, C-SC, and H-YW. Formal analysis: Y-CT. Methodology: W-ML and H-LK. Supervision: J-TL, AC, and Y-CJ. Writing—original draft: Y-CT. Writing—review and editing: Y-CT. All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.