UTUC is a rare disease with poor prognosis. More than 40% of patients have advanced-stage cancer at diagnosis, and their prognosis is poor [
1]. To improve survival, perioperative chemotherapy is performed. The efficacy of neoadjuvant chemotherapy (NAC) for urinary bladder cancer had been confirmed by randomized study. Immediate AC for patients with advanced urinary bladder cancer did not improved overall survival over that of patients who underwent deferred chemotherapy, but it might benefit a subgroup of urinary bladder cancer patients, especially pN-positive patients [
9]. After RNU, many patients lose nearly 50% of their renal function and can be ineligible to receive chemotherapy [
10]. From these points of view, NAC might be preferred for the patients with advanced UTUC. However, the precise preoperative diagnosis of tumor stage or LVI status is difficult, although one study showed the usefulness of magnetic resonance imaging for the prediction of tumor stage [
11]. Unlike urinary bladder cancer, for which pathological stage can be accurately diagnosed by transurethral resection of the bladder tumor before radical cystectomy, the accurate staging of UTUC is difficult even with a ureteroscopic biopsy [
12].
Because UTUC is a rare malignancy comprising 5% of all urothelial cancer, it is difficult to enroll enough UTUC patients to adequately perform a prospective, randomized study to prove the efficacy of perioperative chemotherapy. For lymph node-positive UTUC patients, the efficacies of adjuvant chemotherapy were reported. Retrospective analysis of 74 lymph node-positive UTUC patients showed the AC improved CSS compared with RNU alone (HR 0.52, 95%CI 0.24–0.82,
P = 0.014) [
2]. Retrospective analysis of 263 lymph node-positive UTUC patients showed that AC did not improve CSS in overall patients (HR 0.89,
P = 0.49), but improved CSS in the subgroup of patients with pT3–4 N+ (HR 0.67,
P = 0.022) [
13]. Retrospective analysis of 109 locally advanced UTUC patients (pT3–4pN0/xM0) showed that cisplatin-based AC improved recurrence-free survival (HR = 0.41,
P = 0.017) and CSS (HR 0.33,
P = 0.037) [
14]. Propensity-matched analysis of 1544 UTUC patients with pT2-4 N0 or lymph node-positive showed that AC did not improve overall survival compared with RNU alone (HR 1.14, 95%CI 0.91–1.43,
P = 0.268). The largest study recently reported used data from the National Cancer Database [
5]. This retrospective analysis of the 3253 high-risk UTUC patients showed that AC was statistically associated with an overall survival benefit. A meta-analysis based on this retrospective analysis showed that AC could improve overall survival, CSS, and disease-free survival, but neoadjuvant chemotherapy was more favorable for UTUC than AC in disease-specific survival [
3]. The systematic review and meta-analysis of 24 retrospective analysis studied the efficacy of NAC and AC in UTUC [
15]. Across 2 retrospective studies about NAC, NAC improved CSS, with a pooled HR of 0.41 (95%CI 0.22–0.76,
P = 0.005). Across three cisplatin-based studies about AC, the pooled HR for overall survival was 0.43 (95% CI, 0.21–0.89,
P = 0.023) compared with those who received RNU alone. For disease-free survival, the pooled HR across two studies of AC was 0.49 (95% CI, 0.24–0.99;
p = 0.048). Benefit was not seen for non- cisplatin–based regimens in AC. Meta-analysis of 31 retrospective studies with 8100 UTUC patients who underwent perioperative treatments also showed that AC improved overall survival (HR 0.71, 95%CI 0.51–0.89), CSS (HR 0.71, 95%CI 0.54–0.89), and recurrence-free survival (HR 0.49, 95%CI 0.23–0.85) [
16]. We adopted propensity score-matching analysis, which can reduce the differences between patient characteristics in each group, and the results were consistent with those of this previous study. Furthermore, we identified the patients who benefitted from AC. We previously reported that patients with serum low sodium or hemoglobin levels have a poor prognosis. The supposed mechanism of these markers may be that cells in UTUC with a poor prognosis may secrete inflammatory cytokines such as interleukin-6 that cause anemia and low serum sodium levels. This preoperative prognostic marker may also be useful in the selection of patients to receive AC. AC did not improve the prognosis of patients with normal sodium and hemoglobin levels because these patients already had a better prognosis with or without AC.
There are several limitations in this study. Although we matched the cohorts by propensity scores, this is the retrospective study. A multi-institutional, prospective, randomized study should be performed to prove the efficacy of AC. In this study, a median of 2 cycles of AC were administered, but the optimal number of cycles was not determined. We entered only serum sodium and hemoglobin levels into the Cox proportional analysis, but other prognostic markers might exist to predict the benefit of AC.