The prognostic value of visible hematuria is only significant in T1a renal cell carcinoma: a single-center retrospective study

Abstract

Objectives

To investigate the prognostic value of visible hematuria in T1a renal cell carcinoma (RCC).

Materials and methods

In the RCC database of the Chinese People’s Liberation Army General Hospital Department of Urology, we assembled the records of patients with unilateral RCC over 18 years of age diagnosed between 2008 and 2019. The clinical stage was cT1, and the tumors ranged in size from 0 to 7 cm. The primary treatments were partial nephrectomy (PN) or radical nephrectomy (RN). Logistic regression analysis, Cox regression, interaction analysis, and Kaplan-Meier survival analysis were used to study the correlation between visible hematuria and progression-free survival (PFS), and cancer-specific survival (CSS).

Results

A total of 7,610 patients with cT1 RCC comprised the study population, including 505 RCC patients with visible hematuria. The average follow-up time was 64.6 months (range: 12–144 months). Visible hematuria was significantly associated with the prognosis (PFS, hazard ratio [HR] = 2.7, P < 0.001; CSS, HR = 4.2, P < 0.001) of T1a RCC, but was more significant for CSS in cases of a tumor size ≤ 2 cm (HR = 26.8, P = 0.026). This effect was not significant in T1b RCC (PFS, HR = 0.7, P = 0.153; CSS, HR = 1.1, P = 0.862). The interaction between visible hematuria and tumor size was significant (P = 0.001).

Conclusions

This study showed that visible hematuria was an independent risk factor for PFS and CSS in T1a RCC. The predictive value of visible hematuria for CSS was more significant in RCCs ≤ 2 cm, but did not reach statistical significance in T1b RCC. T1a RCC patients with visible hematuria should be intensively monitored during follow-up.

Introduction

Renal cell carcinoma (RCC) accounts for 2–3% of adult malignancies and is one of the most common cancers of the urinary system [1]. More than 50% of RCCs have no clinical symptoms but are occasionally detected by imaging [2]. The classic triad of flank pain, visible hematuria, and palpable abdominal mass are rare (6–10%) [3]. With improvements in imaging screening techniques, and more patients undergoing physical examinations, the incidence of incidental RCC has increased, the number of clinically manifested RCCs has decreased, and patients with T1 RCC are generally less likely to be symptomatic than patients with later-stage tumors [4]. However, the small number of patients with T1 RCC and clinical symptoms should not be overlooked, and the clinical manifestations are closely related to tumor progression [5, 6]. Current studies have concluded that patients with localized symptoms, including flank pain, visible hematuria, and a palpable abdominal mass have poorer survival than asymptomatic patients. Patients with cancer symptoms associated with changes in health status, including anorexia, fatigue, weight loss, or symptoms due to metastasis have significantly shorter survival times than patients with localized symptoms [7]. Visible hematuria is the highest individual risk factor for developing kidney cancer [8]. Visible hematuria is a relatively common local symptom, regardless of whether it represents an advanced-stage tumor. No accurate correlation has been reported between visible hematuria and the T1 RCC prognosis. Our study captured the prognostic role of visible hematuria in patients with stage T1 RCC.

Thus, we retrospectively collected the records of patients with T1 RCC from the Chinese People’s Liberation Army (PLA) General Hospital urological RCC database to better understand the general clinical features, history, pathological subtypes, Fuhrman grade, distribution of pathological features, and prognosis of patients with RCC with or without visible hematuria. We also investigated the prognostic significance of visible hematuria in patients with T1 RCC.

Materials and methods

Patients and tumors

We retrospectively reviewed the clinical data of patients undergoing renal surgery at the Chinese PLA General Hospital from 2008 to 2019. We collected the records of patients with unilateral renal tumors aged ≥ 18 years. The clinical stage was T1 (cT1). The tumor size ranged from 0 to 7 cm. The surgical approaches included open surgery, laparoscopy, and robot-assisted laparoscopy surgery. A partial nephrectomy (PN) or radical nephrectomy (RN) was performed via the transperitoneal or retroperitoneal approach. We excluded patients with flank pain, a palpable abdominal mass, multiple renal tumors, bilateral renal tumors, RCC associated with a tumor thrombus in the inferior vena cava, benign renal tumor, renal pelvic cancer, RCC associated with preoperative distant metastasis, and patients who lacked follow-up data. Patients with postoperative pathological diagnosis indicated invasion of the renal sinus fat, collecting system, or perinephric fat were excluded. Patients with hematuria for other reasons were also excluded: RCC associated with urolithiasis, urinary tract infection, recent strenuous exercise, menstrual period, recent use of anticoagulant drugs, other urinary system tumors, prostatitis, and benign prostatic hyperplasia. Patients with intermittent or continuous painless visible hematuria who presented previously or currently were defined as cases of visible hematuria. The aforementioned conditions were confirmed by integrating the results of patients' MRI, CT, ultrasound, and other imaging examinations with medical record information(For instance, the presence of a tumor thrombus in the inferior vena cava was confirmed through renal MRI, and preoperative distant metastases were confirmed via PET-CT, among other assessments). The urine examination after admission further confirmed hematuria when three or more red blood cells were in each high-powered field of the microscope [9]. Using these criteria, 7,610 patients with RCC were identified, including 505 patients with RCC and visible hematuria. The average follow-up time was 64.6 months (range: 12–144 months).

Tumor size was defined in centimeters (cm) of diameter. The histological subtypes were stratified according to the World Health Organization [10] (WHO classification 2016). All patients underwent preoperative staging, including enhanced abdominal computed tomography (CT) scans or enhanced abdominal magnetic resonance imaging (MRI) and chest X-rays. The imaging data were reviewed by two radiologists in the urology department. All patients were monitored for more than 1 year after the surgery. During the initial 2-year follow-up, all patients were re-examined every 3 months for history, chest X-ray, abdominal ultrasound or CT scan, blood examination, and a biochemical examination. The follow-up was performed every 6 months for the first 3 years and then every year for 5 years. The perirenal fat overlying the tumor was part of the submitted specimen for PN. Renal sinus fat was sampled primarily for lesions near the collecting system if the collecting system was encountered during extirpation. The survival outcome was determined at the end of the follow-up, and divided into the categories of progression-free survival (PFS) and cancer-specific survival (CSS). Tumor recurrence was defined by specimens of recurring tumors compared with the original tumor pathology to confirm histological consistency as verified by experienced urologists, or recurrence diagnosed via an imaging examination. Multiple tumors in distant organs were considered metastatic. Tumor-specific death was that caused by metastasis and multiple organ failure or recurrence in RCC, not by disease of other organs.

Statistical analysis

Descriptive statistics were used for all patients in our study cohort, and the categorical variables are expressed as frequencies and proportions, whereas the continuous variables are expressed as means and standard deviations. Statistical analyses of the qualitative and quantitative variables were carried out with chi-square and Kruskal-Wallis tests.

We performed a logistic univariate analysis using the visible hematuria classification for all T1 patients to assess the relationship between visible hematuria and the study variables. Cox multivariate regression analysis was used to assess the correlation between visible hematuria and the prognosis (PFS and CSS) in the subgroups (≤ 2 cm, T1a, T1b, and T1) after adjusting for confounding factors such as age, sex, body mass index (BMI), smoking, alcohol consumption, hypertension, diabetes, tumor size, pathology, necrosis, renal sinus fat invasion, perirenal fat invation, renal pelvic invasion and Fuhrman-grade to identify independent risk factors of visible hematuria in the T1 subgroups. The interaction between the tumor-size subgroups was examined.

The time from surgery to recurrence was estimated and plotted using the Kaplan-Meier method. The survival rates of RCC patients with and without visible hematuria were compared between the tumor subgroups. All statistical tests were two-sided, and a P-value < 0.05 was considered significant. Data were analyzed with the R statistical package (The R Foundation for Statistical Computing, Vienna, Austria; http://www.r-project.org; version R3.4.3) and Empower (R) (www.empowerstats.com).

Results

Baseline patients and disease characteristics

The study comprised 5,490 males (72.1%) and 2,120 females (27.9%). The mean age was 53.1 ± 11.7 years, the mean body BMI was 25.6 ± 4.9 (kg/m²), and the average tumor size was 3.5 ± 1.5 cm. There were 1,060 patients with diabetes mellitus (13.9%), 2,455 patients with hypertension (32.3%), 1,066 patients with a smoking history (14.0%), and 984 patients with a drinking history (12.9%). There were 505 RCC patients with visible hematuria (6.6%), among which 4,198 (55.2%) patients had PN; 3,412 (44.8%) patients had RN, including 6,864 (90.2%) clear-cell RCC (ccRCC), 307 (4.0%) chromophobe carcinoma, and 274 (3.6%) papillary carcinoma; other types of cancer were found in 165 (2.2%). There were 5,506 (72.4%) with stage T1a RCC and 2,104 (27.6%) with stage T1b RCC, of which 1,503 (19.8%) were ≤ 2 cm tumors. The pathological features were 575 (7.6%) with necrosis, 77 (1.60%) with calcification, 341 (4.5%) with renal sinus fat invasion, 147 (1.9%) with perinephric fat invasion, and 325 (4.3%) with renal pelvic invasion. The Fuhrman-grade distribution was 897 (11.8%) grade 1, 5,132 (67.4%) grade 2, 701 (9.2%) grade 3, 44 (0.6%) grade 4, and 836 uncertain (11.0%). The patients develop a recurrence or metastasis 313 (4.1%), and to have cancer-specific death 158 (2.1%) with an average follow-up time of 64.6 months (range: 12–144 months) (Table 1).

Table 1 Baseline patients and disease characteristics

Characteristics of patients with or without visible hematuria

According to the visible hematuria classification, the patients were classified as RCC with or without visible hematuria. The characteristics of the two groups are compared in Table 2. Significant differences were detected between the two groups except for the gender ratio, hypertension, and the pathological calcification.

Table 2 Characteristics of patients with or without visible hematuria

Patients with visible hematuria were older (53.0 vs. 54.4 years, P = 0.009), had a lower BMI (25.6 vs. 25.1 kg/m², P = 0.037), a larger tumor size (3.4 vs. 4.3 cm, P < 0.001), less alcohol consumption (13.3% vs. 8.1%, P < 0.001), and were more likely to receive RN (42.7% vs. 75.2%, P < 0.001) than RCC patients without visible hematuria. Patients with visible hematuria were less likely to have the ccRCC histological subtype (90.6% vs. 84.4%, P < 0.001), but more likely to have papillary RCC (3.4% vs. 6.1%, P < 0.001), renal sinus fat invasion (3.8% vs. 13.5%, P < 0.001), perinephric fat invasion (1.8% vs. 3.8%, P < 0.001), renal pelvic invasion (2.8% vs. 24.2%, P < 0.001), tumor necrosis (TN) (7.0% vs. 15.2%, P < 0.001), or a higher Fuhrman-grade (HG 9.2% vs. 18.8%, P < 0.001). The patients with visible hematuria were more likely to be develop a recurrence or metastasis (3.8% vs. 8.1%, P < 0.001), and were more likely to have cancer-specific death (1.8% vs. 5.7%, P < 0.001). In total, visible hematuria was significantly correlated with tumor size, alcohol consumption, histological subtype, TN, renal sinus fat invasion, perinephric fat invasion, renal pelvic invasion, Fuhrman grade, and prognosis (PFS and CSS) (P < 0.001) (Table 2).

Multivariate analysis of hematuria and prognosis

We adjusted for age, sex, BMI, hypertension, diabetes, smoking, alcohol consumption and tumor size in the subgroup analysis. Other extra adjusted factors included pathology, necrosis, renal sinus fat invasion, perirenal fat invation, renal pelvic invasion and Fuhrman-grade (Table 3).

Table 3 Multivariate analysis of hematuria and prognosis

The multivariate analysis showed that visible hematuria was an independent risk factor for the prognosis (PFS, HR = 2.7, 95% confidence interval [CI] = 1.5–4.9, P < 0.001; CSS, HR = 4.2, 95% CI = 1.8–9.8, P < 0.001) only in T1a RCC, but not in T1b RCC (PFS, HR = 0.7, 95% CI = 0.4–1.1, P = 0.153; CSS, HR = 1.1, 95% CI = 0.6–1.9, P = 0.862; Fig. 1). The statistical result for CSS was even more significant in tumors ≤ 2 cm (HR = 26.8, 95% CI = 1.4–486.2, P = 0.026) (Table 3). In addition, visible hematuria had a different relationship with different tumor-size subgroups. Based on the P-interaction of the RCC subgroups shown in Table 3, visible hematuria does not have a significant influence on the prognosis in patients with tumors ≤ 2 cm or at the T1a stage (PFS, P-interaction = 0.391; CSS, P-interaction = 0.085), but a significant difference was observed between T1a and T1b for prognosis (PFS, P-interaction = 0.001, CSS, P-interaction = 0.011).

Fig. 1

Multivariate analysis of hematuria and prognosis in the subgroups. CI: confidence interval, HR: hazard ratio. PFS = Progression-free survival, CSS = Cancer-specific survival

As shown in Fig. 2a–d, patients with visible hematuria had shorter survival times than T1a RCC patients without visible hematuria (PFS, P < 0.001; CSS, P < 0.001). However, no significant differences were observed in the survival times of patients with and without visible hematuria and T1b RCC (PFS, P = 0.367; CSS, P = 0.521).

Fig. 2

a Progression-free survival (T1a RCC patients). b Progression-free survival (T1b RCC patients). c Cancer-specific survival (T1a RCC patients). d Cancer-specific survival (T1b RCC patients)

We also conducted an analysis on patients with microscopic hematuria (Tables 4 and 5), and by categorizing the results of microscopic hematuria into five grades (0-3, 3-5, 5-10, 10-20, > 20), we performed non-parametric statistical analysis and found a clear correlation between the condition of microscopic hematuria and the RENAL score of the patients (Kendall’s tau coefficient 0.216, P < 0.001). Moreover, aside from the R score, which represents tumor size and showed no correlation, there was a significant correlation with the exophytic score E, the collecting system separation score N, and the level score L (R, Kendall's tau coefficient 0.032, P = 0.639; E, Kendall's tau coefficient 0.214, P = 0.001; N, Kendall's tau coefficient 0.191, P = 0.003; L, Kendall's tau coefficient 0.173, P = 0.007). Regarding patient prognosis, patients with strongly positive microscopic hematuria (> 20/HPF) had a significantly shorter cancer-specific survival compared to negative patients (0–3/HPF), rather than progression-free survival (PFS, HR = 2.1, 95% CI = 0.4–10.5, P = 0.378; CSS, HR = 20.3, 95% CI = 2.6–156.4, P = 0.004; Fig. 3).

Table 4 Correlation analysis of microscopic hematuria and RENAL score

Table 5 Correlation analysis of microscopic hematuria and prognosis

Fig. 3

a Progression-free survival (microscopic hematuria). b Cancer-specific survival (microscopic hematuria)

Due to the possibility that gross hematuria may indicate an invasion of the collecting system, we analyzed the relationship between gross hematuria and pathological grading. It was found that among patients without gross hematuria, 44.87% were pT1, and 55.13% were pT3. In contrast, among patients with gross hematuria, 35.26% were pT1, and 64.74% were pT3, the difference was statistically significant (P < 0.001) (Table 6). In the domain of survival analysis, the presence of gross hematuria in both pT1 and pT3 patient cohorts was significantly associated with inferior outcomes, with particularly pronounced effects observed in the pT3 cohort (Fig. 4). Specifically, within the pT1 cohort, gross hematuria was significantly correlated with diminished progression-free survival (PFS), whereas no significant associations were observed with overall survival (OS) or cancer-specific survival (CSS) (PFS, HR = 2.7, 95% CI = 1.4–5.2, P = 0.003; OS, HR = 1.6, 95% CI = 0.9–2.9, P = 0.116; CSS, HR = 2.0, 95% CI = 0.9–4.2, P = 0.069). Conversely, within the pT3 cohort, gross hematuria was significantly associated with adverse outcomes across all considered endpoints, including PFS, OS, and CSS (PFS, HR = 4.5, 95% CI = 2.9–6.9, P < 0.001; OS, HR = 4.0, 95% CI = 2.6–6.2, P < 0.001; CSS, HR = 4.9, 95% CI = 3.0–8.1, P < 0.001).

Fig. 4

a Overall survival (pT1/pT3 with or without hematuria). b Progression free survival (pT1/pT3 with or without hematuria). c Cancer specific survival (pT1/pT3 with or without hematuria)

Discussion

The detection rate of early RCC has increased, the proportion of RCCs detected by physical examination has increased, the number of RCC patients with clinical symptoms has decreased, and the overall RCC survival rate has improved [4, 7]. Previous studies have demonstrated that clinical symptoms are related to the RCC prognosis. The UK general practice database records suggest rates of hematuria in RCC patients as low as 18% [8]. Vasudev reported that 23.5% of RCC cases in the UK had hematuria [5], Szendroi reported that 35.8% of RCC cases in Hungary had hematuria [11] and Seo reported that 9.4% of RCC cases in Korea had haematuria [12]. However, in our study, the rate of RCC with hematuria was 6.6%. This lower rate of hematuria may be due to our patient population. Most of the patients of our center, one of the largest urology departments in Beijing, China, are urban residents who get regular medical check-ups, and only a few patients had not undergone physical examinations and ignored localized symptoms, such as visible hematuria. T1 RCC is an early-stage RCC with fewer clinical symptoms than RCC of stages T2-4.

The most common nephrogenic diseases associated with hematuria are urolithiasis and urothelial carcinoma [13]. Visible hematuria is the most common and powerful single predictor of kidney cancer [8]. Our study considered the mechanisms of visible hematuria from RCC for the following reasons. First, in early RCC, local tumor cells destroy the surrounding nephrons after the early renal tubular epithelial lesions cause necrosis and release inflammatory factors, and the broken red blood cells in the renal collecting system are not completely reabsorbed, resulting in hematuria. Second, the visible hematuria may originate from the tumor growth position near the renal pelvis, and invade the collecting system, resulting in rupture of the renal pelvis mucosa. Lastly, the bigger the tumor, the more malignant, because it is more likely that the tumor cells are in an anoxic microenvironment, and there is more likely to be necrosis with hematuria. Because this study mainly investigated the independent risk factors of visible hematuria in patients with T1a RCC, the tumors were smaller and detected earlier. These reasons result in a lower rate of visible hematuria.

Predictive models for the clinical manifestations of RCC are not uncommon. The symptomatic presentation is correlated with aggressive histology and advanced disease, and incidental tumors may be frequently detected in female and elderly patients [4, 5]. Patard et al. reported that the symptom classification remains an independent prognostic marker together with the TNM stage and Furhman grade [7]. Szendroi et al. demonstrated that weight loss and flank pain are frequently neglected and possible symptoms of renal cancer, and the presence of symptoms is independently correlated with intraoperative complications and survival in renal cancer patients [11]. These studies focused primarily on the overall clinical predictive model of each stage of RCC, but no study has analyzed the clinical significance of visible hematuria in cases of T1 RCC. Thus, visible hematuria should be paid more attention. We conclude that visible hematuria is an independent risk factor for the prognosis (PFS and CSS) in patients with T1a RCC. The risk is more significant for CSS in ≤ 2 cm tumors with visible hematuria than in T1b RCC. In renal cell carcinoma, T1a represents an earlier stage within the T classification, whereas T1b denotes a relatively more advanced stage, with tumors of larger volume and higher malignancy. This may have been one of the reasons why gross hematuria did not confer a significant prognostic difference among T1b patients. The National Comprehensive Cancer Network (NCCN) guidelines define active surveillance as the initial monitoring of tumors using abdominal imaging techniques with delayed intervention when indicated. Elderly patients and those with small renal masses (≤ 2 cm) and other comorbidities often have low RCC-specific mortality [14, 15]. Our research suggested that patients presenting with small renal tumors (≤ 2 cm) and visible hematuria might be candidates for a more proactive treatment strategy.

Interestingly, the incidence of the prognosis (PFS and CSS) in T1a RCC patients with visible hematuria was similar to that in T1b RCC patients without visible hematuria (not shown in the Table). Visible hematuria may increase the risk of T1a RCC patients to T1b, but the risk remains much lower than that of T3a patients. This observation indicates that visible hematuria is not a sign of T3a, but is more likely to be a more severe manifestation of local tumor destruction, which may also be worth considering. Many studies have confirmed that tumor stage, grade, size, and pathological characteristics have a strong influence on the prognosis [16,17,18,19]. In our study, T1a RCC patients with visible hematuria had worse tumor characteristics, including necrosis, renal sinus fat invasion, perinephric fat invasion, renal pelvic invasion, and a higher Fuhrman-grade than patients without visible hematuria. However, patients with these signs are not a large proportion of all T1a RCC cases with visible hematuria. Thus, visible hematuria should be considered a risk of T1a updated to a risk of T1b in the NCCN guidelines to better guide clinical treatment and follow-up.

Table 6 Distribution of gross hematuria across pathological stages

Renal sinus fat invasion was intensively monitored after surgery during the follow-up and was significantly associated with a poor CSS compared to patients with perinephric fat invasion [20, 21]. Tumors with visible hematuria were more correlated to renal sinus fat invasion and renal pelvic invasion than T1 RCC without hematuria, indicating that the tumor directly invaded the collecting system, or tumor growth near the invasion of the renal pelvis resulted in hematuria and upstaging of the pathology. Visible hematuria was closely associated with postoperative pathological upstaging, so whether visible hematuria is necessary for selecting surgery is unclear and needs further study.

Some studies have reported that PN does not compromise the chance for a cancer cure in patients with cT1 tumors that are upstaged pathologically to pT2 or pT3 or high-grade renal masses compared with RN [22]. Both PN and RN provide excellent oncological results in cases of T1 [23]. Another study demonstrated that a subgroup of patients with clinical stage T1 RCC was pathologically upstaged to T3a. Among these patients, those undergoing PN appear to have inferior recurrence-free survival relative to those undergoing RN [24]. In our study, T1 patients with RCC and visible hematuria were more likely to suffer renal sinus fat invasion, perinephric fat invasion, and renal pelvic invasion. The surgeon should consider the following points during the PN. First, tumors surrounding perinephric fat should be completely resected to prevent a positive tumor margin. Second, the suspected renal sinus fat invasion should be prepared to be transformed to RN, because the visible hematuria is more likely to invade the renal pelvis. Finally, the surgeon should be prepared to repair the renal pelvis to prevent postoperative urinary leakage after PN.

The present study was limited by its retrospective nature, which can cause selection bias. Moreover, the T1 RCC with visible hematuria sample was relatively small in our center, caused by the limited number of T1 RCC cases. We should continue to study the mechanisms of visible hematuria, not just the clinical symptoms. Furthermore, a large proportion of T1 RCC cases with visible hematuria underwent RN and a small proportion underwent PN, resulting in an underestimate of the true incidence of upstaged tumor characteristics in our study. In addition, although we adjusted our multivariate models for the patient and tumor characteristics, we cannot rule out the possibility of residual confounding. Visible hematuria was an independent risk factor for T1a RCC, but the clinical treatment of visible hematuria was not fully addressed by this study.

Conclusion

The present study showed that visible hematuria was an independent risk factor for T1a RCC. The predictive value of visible hematuria for CSS was more significant in cases of RCC tumors ≤ 2 cm, but did not reach significance in cases of T1b RCC. Visible hematuria may increase the risk of T1a patients to be similar to those with T1b, but the risk was still much lower than that of T3a patients. T1a RCC patients with visible hematuria should be intensively monitored after surgery and during follow-up.