Identification of prognostic factors for postoperative recurrence-free survival in T2-stage renal cell carcinoma: a retrospective cohort study

Abstract

Background

T2-stage renal cell carcinoma (RCC) is associated with high postoperative recurrence risk and heterogeneous outcomes, but stage-specific prognostic factors remain insufficiently explored. This study aimed to systematically analyze the clinicopathological factors influencing postoperative recurrence-free survival (RFS) (defined as the time from surgical resection to the first occurrence of local recurrence, distant metastasis, or death from any cause) in patients with T2-stage RCC and explore reliable independent prognostic indicators for this specific subgroup.

Methods

We retrospectively collected clinicopathological data and follow-up records of T2-stage RCC patients who underwent surgery at the Second Hospital of Lanzhou University from January 2018 to December 2023. Postoperative RFS was the primary endpoint. The Kaplan-Meier method was used for survival analysis and survival curve plotting; univariate and multivariate Cox proportional hazards regression models were applied to identify independent prognostic factors. Internal stability of the identified prognostic factors was evaluated using bootstrap resampling (1,000 iterations) and Harrell’s concordance index (C-index).

Results

A total of 182 patients were included (mean age: 55±13 years; median follow-up: 40 months). After rigorous variable screening (including collinearity elimination and biological independence assessment), five independent prognostic factors for postoperative RFS were identified: the aspartate transaminase to alanine transaminase ratio (AST/ALT ratio) [hazard ratio (HR) =3.295, 95% confidence interval (CI): 1.899–5.719, P<0.001], Fuhrman grade (HR =2.380, 95% CI: 1.412–4.012, P=0.001), systemic immune-inflammation index (SII) (HR =3.009, 95% CI: 1.567–5.776, P<0.001), sarcomatoid differentiation (HR =3.463, 95% CI: 1.981–6.054, P<0.001), and carbonic anhydrase IX (CAIX) (HR =5.425, 95% CI: 2.555–11.519, P<0.001). Internal validation showed a mean C-index of 0.835 (95% CI: 0.780–0.885), indicating good and stable discriminatory ability of the identified factors.

Conclusions

The AST/ALT ratio, Fuhrman grade, SII, sarcomatoid differentiation, and CAIX are independent prognostic indicators for postoperative RFS in T2-stage RCC patients. These factors can facilitate accurate risk stratification and provide a scientific basis for formulating individualized follow-up and treatment strategies. Given the limitations of single-center retrospective design and small sample size, future large-scale prospective multicenter studies are needed to validate these findings.

Highlight box.

Key findings

• This study identified five independent prognostic factors for postoperative recurrence-free survival in T2-stage renal cell carcinoma (RCC): aspartate transaminase to alanine transaminase (AST/ALT) ratio, Fuhrman grade, systemic immune-inflammation index (SII), sarcomatoid differentiation, and carbonic anhydrase IX (CAIX).

What is known and what is new?

• Existing prognostic models for RCC lack specificity for T2-stage disease, and stage-specific prognostic factors remain insufficiently explored.

• This manuscript systematically confirms the prognostic value of AST/ALT ratio, SII, and CAIX in T2-stage RCC, providing a non-invasive and practical risk stratification tool.

What is the implication, and what should change now?

• Combined application of the five factors improves the accuracy of postoperative recurrence risk assessment for T2-stage RCC.

• These findings support individualized follow-up and treatment strategies, especially for high-risk patients with elevated AST/ALT ratio, high Fuhrman grade, or positive CAIX expression.

Introduction

Renal cell carcinoma (RCC), which accounts for roughly 90% of all malignant kidney tumors, is one of the most common urological cancers worldwide (1). Recent epidemiological data show an annual global incidence of about 435,000 new RCC cases, with 11% initially diagnosed as T2-stage disease (2,3). According to the current tumor-node-metastasis (TNM) classification, T2-stage RCC is defined as tumors larger than 7 cm but still confined within the renal capsule (4), and is further subdivided into T2a (7–10 cm) and T2b (>10 cm) subgroups based on maximum tumor diameter. However, controversy persists regarding the prognostic relevance of this substaging: some studies have demonstrated that T2b-stage tumors are associated with higher recurrence rates and poorer survival outcomes due to increased tumor heterogeneity and invasiveness (5), while others have failed to confirm a significant prognostic difference between T2a and T2b subgroups (6). This inconsistency underscores the need for a more precise understanding of T2-stage RCC’s biological behavior and prognostic characteristics.

Compared to T1-stage disease, T2-stage RCC has a significantly larger tumor burden, increased biological aggressiveness, and accordingly higher risks of postoperative recurrence and metastasis. In China, the increasing incidence of RCC combined with limited public awareness of early screening has led to a large number of patients presenting with T2-stage disease at diagnosis, posing significant challenges for clinical management (7). Although surgery (including radical nephrectomy and nephron-sparing procedures) remains the main treatment for T2-stage RCC, there is still considerable variation in postoperative outcomes despite improvements in surgical techniques and perioperative care. Clinical data show notable prognostic differences, with some patients achieving long-term survival while others experience rapid disease progression.

Existing prognostic models for RCC, such as the Stage, Size, Grade, and Necrosis (SSIGN) score and University of California, Los Angeles Integrated Staging System (UISS), were developed based on heterogeneous cohorts including all disease stages (8,9). These models prioritize generalizability across RCC stages but lack specificity for T2-stage disease. For instance, the SSIGN score emphasizes tumor size and necrosis but underestimates the prognostic impact of inflammatory markers and molecular biomarkers that are particularly relevant to T2-stage RCC (9,10). The UISS model integrates clinical stage, histological grade, and performance status, but fails to account for T2-specific pathological features such as sarcomatoid differentiation and intravascular tumor thrombus, which are key determinants of outcomes in this subgroup (9,11). This limitation results in inadequate risk stratification for T2-stage patients, hindering the development of personalized treatment strategies.

This significant variability in prognosis, coupled with the limitations of existing models, highlights the urgent need for detailed investigation of T2-stage-specific prognostic factors. However, current research is often limited by the common practice of analyzing T2-stage RCC together with other disease stages, which hides stage-specific prognostic factors and hampers the development of personalized treatment strategies. This retrospective cohort study analyzed comprehensive clinical data from T2-stage RCC patients treated at the Second Hospital of Lanzhou University. We systematically evaluated prognostic determinants, incorporating clinical parameters, pathological characteristics, molecular biomarkers, therapeutic regimens, and baseline patient demographics. Through multivariate analysis of these factors, we aimed to identify independent prognostic factors for T2-stage RCC and provide evidence-based insights for clinical decision-making and individualized follow-up strategies. We present this article in accordance with the TRIPOD reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-aw-837/rc).

Methods

Ethical approval and informed consent

This study was performed in accordance with the principles of the Declaration of Helsinki and its subsequent amendments. The study protocol was reviewed and approved by the Institutional Review Board (IRB) of the Second Hospital of Lanzhou University (approval No. 2025A-983). Due to the retrospective nature of the study, the requirement for informed consent was waived by the IRB.

Study population

A total of 213 consecutive patients with histologically confirmed T2-stage RCC who underwent surgical intervention (radical nephrectomy or partial nephrectomy) at the Department of Urology, Second Hospital of Lanzhou University between January 2018 and December 2023 were initially screened. After applying the inclusion and exclusion criteria, 182 patients were finally enrolled, with 31 patients excluded (detailed exclusion reasons are reported in “Inclusion and exclusion criteria” section). All enrolled patients met strictly defined inclusion and exclusion criteria through a comprehensive medical record review.

Inclusion and exclusion criteria

Inclusion criteria:

1. Pathologically confirmed primary RCC meeting the American Joint Committee on Cancer (AJCC) eighth edition [2017] criteria for T2-stage disease (tumors >7 cm in maximal diameter with confinement within the renal capsule) (12);

2. Underwent radical nephrectomy or partial nephrectomy;

3. Availability of complete baseline clinicopathological data and ≥12 months’ postoperative follow-up;

4. Received no preoperative anti-tumor treatment.

Exclusion criteria:

1. Non-T2-stage RCC or non-RCC renal malignancies (n=5);

2. Concurrent malignancies of other organ systems (n=3);

3. Complicated with hematological or immune system diseases (n=2);

4. Lost to follow-up or with incomplete clinical documentation (n=31, including 12 with missing postoperative immunohistochemical results, 10 with follow-up duration <12 months, and nine with missing preoperative laboratory test data);

5. Presented with recurrence or had a history of repeated admissions (n=4);

6. Previously received targeted therapy or immunotherapy for renal cancer (n=2);

7. Evidence of metastatic disease at presentation (n=4) (note: the total number of excluded patients exceeds the sum of individual exclusion categories due to overlapping exclusion criteria in some patients).

Study design and methodology

Data collection

Relevant patient data were collected through review of electronic medical records, outpatient follow-up visits, and telephone follow-ups, including:

1. Demographic and clinical baseline information: gender, age, height, weight, ethnicity, presenting symptoms, history of diabetes and hypertension, perioperative blood transfusion, tumor location, and surgical approach;

2. Preoperative laboratory test results: hemoglobin, globulin, albumin, monocyte count (MC), white blood cell count (WBC), neutrophil count (NC), lymphocyte count (LC), platelet count (PLT), preoperative levels of aspartate transaminase (AST) and alanine transaminase (ALT), and the AST/ALT ratio;

3. Pathological data: tumor diameter, renal score, pathological subtype, presence of capsular invasion, presence of necrosis, presence of vascular invasion, and presence of sarcomatoid differentiation;

4. Postoperative immunohistochemical findings: expression of Ki-67, carbonic anhydrase IX (CAIX), paired box gene 8 (PAX-8), epithelial membrane antigen (EMA), cluster of differentiation 10 (CD10), cytokeratin 7 (CK7), Vimentin, cluster of differentiation 117 (CD117), and paired box gene 2 (PAX-2);

5. Follow-up data: survival status, presence of recurrence or metastasis, administration of postoperative adjuvant therapy, and adherence to regular re-examinations.

Definition of recurrence and metastasis

Recurrence and metastasis were rigorously defined based on combined radiological and pathological evidence to ensure accuracy:

1. Local recurrence: new lesions detected at the surgical resection site via contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), or ultrasound. Pathological confirmation was performed if tissue sampling was feasible (e.g., via biopsy or reoperation);

2. Distant metastasis: lesions identified in extrapulmonary organs (lung, bone, adrenal gland, peritoneum, etc.) or distant lymph nodes through imaging modalities including CT, MRI, bone scan, or positron emission tomography-CT (PET-CT). For resectable lesions, pathological verification was required; for unresectable lesions, consistent radiological progression on serial follow-up imaging was considered confirmatory.

Analysis methods

Baseline demographic and clinical characteristics were systematically recorded for all enrolled patients. Peripheral blood samples collected during the initial post-admission evaluation were analyzed to calculate the following inflammatory indices: neutrophil-to-lymphocyte ratio (NLR) = neutrophil count ÷ lymphocyte count; platelet-to-lymphocyte ratio (PLR) = platelet count ÷ lymphocyte count; monocyte-to-lymphocyte ratio (MLR) = monocyte count ÷ lymphocyte count; systemic immune-inflammation index (SII) = (platelet count × neutrophil count) ÷ lymphocyte count; systemic inflammation response index (SIRI) = (neutrophil count × monocyte count) ÷ lymphocyte count; pan-immune-inflammation value (PIV) = (platelet count × monocyte count) ÷ lymphocyte count. All hematological parameters were derived from complete blood count (CBC) analysis performed using standardized automated hematology analyzers.

Follow-up methods

All enrolled patients were followed up regularly after surgical resection. The primary endpoint of this study was recurrence-free survival (RFS). RFS was explicitly defined as the time interval from the date of surgical resection of T2-stage RCC to the first occurrence of any of the following events: radiologically confirmed local tumor recurrence, distant metastasis (including but not limited to lung, bone, liver, and brain metastasis), or death from any cause. For patients who did not experience any of the above events during the follow-up period, RFS data were censored at the date of the last effective follow-up. The overall survival (OS) endpoint was defined as the time from the date of surgery to death from any cause. The follow-up period concluded on December 31, 2024.

Statistical analysis

Sample size and power analysis

This study was an exploratory retrospective cohort study focusing on identifying prognostic factors for T2-stage RCC. The sample size was determined based on the number of eligible T2-stage RCC patients treated at our institution during the study period (January 2018 to December 2023). A post-hoc power analysis was performed using G*Power 3.1 software to verify the statistical power. Taking the key prognostic factor (Fuhrman grade) identified in the univariate analysis as an example, with an estimated hazard ratio (HR) of 2.380, α=0.05 (two-sided), and β=0.2, the calculated minimum sample size required was 124 patients. The actual enrolled sample size (182 patients) exceeded the minimum requirement, indicating sufficient statistical power to detect the hypothesized effect size for exploratory analysis.

Missing data handling

A total of 31 patients were excluded based on the predefined exclusion criteria (accounting for 14.5% of the initially screened 213 patients). Missing data were evaluated using the Little’s Missing Completely at Random (MCAR) test, which yielded a P-value of 0.33 (>0.05), indicating that the missing data were randomly distributed and unlikely to introduce significant selection bias. Given the low missing rate (<15%) and random distribution, a complete-case analysis was adopted for data processing.

Exploratory prognostic factor analysis

Statistical analyses were performed using SPSS 29.0 and R (version 4.5.1) software. Receiver operating characteristic (ROC) curves were utilized to determine the optimal cut-off values for NLR, PLR, MLR, PIV, SII, SIRI, and AST/ALT ratio, based on which patients were stratified into low- and high-risk groups. The Kaplan-Meier method was applied to calculate survival rates, perform survival analyses, and generate survival curves, with the Log-Rank test used to compare intergroup survival differences (two-sided, P<0.05 considered statistically significant).

To identify independent prognostic factors for T2-stage RCC and address methodological limitations, a simplified and rigorous multivariate analysis strategy was adopted:

1. Candidate variable screening: univariate Cox regression analysis was conducted for all included variables (two-sided test). Variables with P<0.20 were retained as candidate factors to avoid missing potential confounders, while highly correlated inflammatory indices (PIV, SII, SIRI) were prioritized for collinearity assessment due to shared calculation components (NC, LC, and PLT);

2. Multicollinearity resolution: Pearson correlation analysis and variance inflation factor (VIF) testing were performed. Variables with a correlation coefficient >0.70 or VIF >5 were considered to have severe multicollinearity. Among collinear variables, SII was retained as the representative inflammatory index based on its wide application in RCC prognostic research and strongest univariate association with RFS;

3. Simplified multivariate Cox regression: a maximum of 6 biologically distinct candidate variables (after collinearity exclusion) were incorporated into the Cox proportional hazards model, ensuring compliance with the “10 events per variable (EPV)” rule (64 outcome events, EPV >10). A backward elimination method was used to further optimize the model, with two-sided P<0.05 as the final inclusion criterion for independent prognostic factors;

4. Internal stability assessment: given the exploratory nature, internal validation was limited to bootstrap resampling (1,000 iterations) to assess the stability of identified prognostic factors. Discriminative ability was quantified using Harrell’s concordance index (C-index).

All statistical tests were two-sided, and P<0.05 was considered statistically significant.

Results

Clinical and postoperative pathological characteristics

A total of 213 consecutive patients with histologically confirmed T2-stage RCC who underwent surgical treatment at our department between January 2018 and December 2023 were initially screened. After applying the strict inclusion and exclusion criteria, 182 patients were finally enrolled in the study, with 31 patients excluded. Among the enrolled patients, 115 were male (63.19%) and 67 were female (36.81%), with a male-to-female ratio of approximately 2:1. The mean age was 55±13 years (range, 1–81 years). Tumors were detected incidentally in 93 patients (51.10%) during physical examinations or other tests, while 89 patients (48.90%) presented with clinical symptoms. Specifically, 48 patients had hematuria, 57 experienced waist or abdominal pain/discomfort, and 5 had abdominal masses. There were 127 patients with T2a-stage disease (69.78%) and 55 with T2b-stage disease (30.22%). Regarding treatment, most patients (92.31%) underwent radical nephrectomy, while a small proportion (7.69%) underwent nephron-sparing surgery. Among these procedures, 145 cases (79.67%) were laparoscopic, 21 (11.54%) were robot-assisted, and 16 (8.79%) were open surgeries. The demographic, clinical, and pathological characteristics of the patients are summarized in Table 1.

Table 1. Baseline clinical and pathological characteristics of patients with T2-stage RCC.

Variable	Recurrence, metastasis or death (n=64)	No recurrence, metastasis or death (n=118)
Gender
Male	42 (65.63)	73 (61.86)
Female	22 (34.37)	45 (38.14)
Tumor location
Left	36 (56.25)	56 (47.46)
Right	28 (43.75)	62 (52.54)
BMI (kg/m2)
>23.9	27 (42.19)	60 (50.85)
18.5–23.9	31 (48.43)	48 (40.68)
<18.5	6 (9.38)	10 (8.47)
Surgical method
Radical resection	62 (96.88)	106 (89.83)
Nephron-sparing	2 (3.12)	12 (10.17)
Diabetes
Yes	5 (7.81)	5 (4.24)
No	59 (92.19)	113 (95.76)
Perioperative blood transfusion
Yes	12 (18.75)	16 (13.56)
No	52 (81.25)	102 (86.44)
Fuhrman grade
Grade 1–2	33 (51.56)	104 (88.14)
Grade 3–4	31 (48.44)	14 (11.86)
Globulin (g/L)
≥34	25 (39.06)	21 (17.80)
<34	39 (60.94)	97 (82.20)
MLR
≥0.22	51 (79.69)	61 (51.69)
<0.22	13 (20.31)	57 (48.31)
PLR
≥138	30 (46.88)	50 (42.37)
<138	34 (53.12)	68 (57.63)
SII
≥430	52 (81.25)	43 (36.44)
<430	12 (18.75)	75 (63.56)
Tumor necrosis
Yes	22 (34.38)	13 (11.02)
No	42 (65.62)	105 (88.98)
Sarcomatoid differentiation
Yes	25 (39.06)	7 (5.93)
No	39 (60.94)	111 (94.07)
Pathological type
Clear cell carcinoma	51 (79.68)	81 (68.65)
Papillary cell carcinoma	2 (3.13)	4 (3.39)
Chromophobe cell carcinoma	2 (3.13)	22 (18.64)
Other types of carcinoma	9 (14.06)	11 (9.32)
CAIX
Positive	54 (84.37)	44 (39.29)
Negative	10 (15.63)	74 (62.71)
PAX-2
Positive	16 (25.00)	34 (28.81)
Negative	48 (75.00)	84 (71.19)
CD10
Positive	58 (90.63)	98 (83.05)
Negative	6 (9.37)	20 (16.95)
Vimentin
Positive	49 (76.56)	78 (66.10)
Negative	15 (23.44)	40 (33.90)
MelanA
Positive	7 (10.94)	4 (3.39)
Negative	57 (89.06)	114 (96.61)
Age (years)
≥60	32 (50.00)	39 (33.05)
<60	32 (50.00)	79 (66.95)
Tumor diameter (cm)
7–10	47 (73.44)	80 (67.80)
>10	17 (26.56)	38 (32.20)
Ethnicity
Han	57 (89.06)	110 (93.22)
Hui	3 (4.68)	5 (4.24)
Tibetan	2 (3.13)	2 (1.69)
Other	2 (3.13)	1 (0.85)
Hypertension
Yes	20 (31.25)	25 (21.19)
No	44 (68.75)	93 (78.81)
Preoperative anemia
Yes	14 (21.88)	12 (10.17)
No	50 (78.12)	106 (89.83)
Renal score
4–6 points	0 (0.00)	4 (3.39)
7–9 points	19 (29.69)	35 (29.66)
10–12 points	45 (70.31)	79 (66.95)
Albumin (g/L)
≥35	13 (20.31)	104 (88.14)
<35	51 (79.69)	14 (11.86)
AST/ALT
≥1.46	26 (40.63)	26 (22.03)
<1.46	38 (59.37)	92 (77.97)
NLR
≥2.7	39 (60.94)	35 (29.66)
<2.7	25 (39.06)	83 (70.34)
PIV
≥176.65	36 (56.25)	23 (19.49)
<176.65	28 (43.75)	95 (80.51)
SIRI
≥1.27	42 (65.63)	25 (33.90)
<1.27	22 (34.37)	93 (66.10)
Capsular invasion
Yes	20 (31.25)	26 (22.03)
No	44 (68.75)	92 (77.97)
Intravascular tumor thrombus
Yes	28 (43.75)	11 (9.32)
No	36 (56.25)	107 (90.68)
Ki-67
≤10%	38 (59.38)	91 (77.12)
>10%	26 (40.62)	27 (22.88)
PAX-8
Positive	51 (79.69)	76 (64.41)
Negative	13 (20.31)	42 (35.59)
EMA
Positive	54 (84.38)	98 (83.05)
Negative	10 (15.62)	20 (16.95)
CK7
Positive	10 (15.62)	39 (33.05)
Negative	54 (84.38)	79 (66.95)
CD117
Positive	5 (7.81)	23 (19.49)
Negative	59 (92.19)	95 (80.51)

Data are presented as n (%). ALT, alanine transaminase; AST, aspartate transaminase; BMI, body mass index; CAIX, carbonic anhydrase IX; CD10, cluster of differentiation 10; CD117, cluster of differentiation 117; CK7, cytokeratin 7; EMA, epithelial membrane antigen; MLR, monocyte-to-lymphocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PAX-2, paired box gene 2; PAX-8, paired box gene 8; PIV, pan-immune-inflammation value; PLR, platelet-to-lymphocyte ratio; RCC, renal cell carcinoma; SII, systemic immune-inflammation index; SIRI, systemic inflammation response index.

The results of missing data assessment showed that Little’s MCAR test was performed on the missing data of the 31 excluded patients, yielding a P value of 0.33 (>0.05). This indicates that the missing data in this study were completely randomly distributed, and the use of complete-case analysis would not introduce significant selection bias. Post-hoc power analysis results showed that taking Fuhrman grade as the key prognostic factor (HR =2.380 in univariate analysis), with a significance level α=0.05 (two-sided) and power 1−β=0.8, the minimum sample size required for this study was 124 cases. The actually enrolled 182 cases met the statistical power requirements, which could effectively detect the target effect size and ensure the statistical reliability of the study results.

Postoperative survival outcomes

The postoperative survival status of the patients is shown in Table 2. A total of 21 postoperative deaths occurred, with 16 attributed to tumor recurrence or metastasis and five due to other causes. There were 58 cases of in situ tumor recurrence or metastasis, including 29 lung metastases, 15 bone metastases, 12 in situ recurrences, and five adrenal metastases. The median follow-up duration for all patients was 40 months (range, 1–77 months). As of the last follow-up, the mean OS was 37 months. Kaplan-Meier survival analysis (Figure 1A) indicated that the 1-, 3-, and 5-year RFS rates for the entire cohort were 82.4%, 65.2%, and 53.2%, respectively, with a median RFS of 25 months. Analysis of OS (Figure 1B) demonstrated that the 1-, 3-, and 5-year postoperative OS rates were 97.2%, 89.6%, and 78.3%, respectively, with a median OS of 35 months.

Table 2. Postoperative survival status of patients with T2-stage RCC.

Item	Number of cases (%)
Survival status
Deceased	21 (11.54)
Survived	161 (88.46)
Recurrence and/or metastasis	58 (31.87)
Local recurrence	12 (6.59)
Bone metastasis	15 (8.24)
Lung metastasis	29 (15.93)
Adrenal metastasis	5 (2.75)
Peritoneal metastasis	7 (3.85)
Metastasis to other sites	4 (2.20)

RCC, renal cell carcinoma.

Figure 1.

Kaplan-Meier survival curves after surgery for T2-stage RCC. (A) Kaplan-Meier curve of RFS. (B) Kaplan-Meier curve of OS. OS, overall survival; RCC, renal cell carcinoma; RFS, recurrence-free survival.

Results of univariate Cox regression analysis

Univariate analysis was performed to identify prognostic factors in patients with T2-stage RCC, with results (Table 3) showing that age (P=0.03), preoperative anemia (P=0.03), Fuhrman grade (P<0.001), globulin (P<0.001), AST/ALT ratio (P=0.003), MLR (P<0.001), NLR (P<0.001), PIV (P<0.001), SII (P<0.001), SIRI (P<0.001), tumor necrosis (P<0.001), sarcomatoid differentiation (P<0.001), intravascular tumor thrombus (P<0.001), Ki-67 (P=0.01), CAIX (P<0.001), PAX-8 (P=0.01), CK7 (P=0.009), CD117 (P=0.04), and MelanA (P=0.03) were significantly associated with postoperative RFS in T2-stage RCC (P<0.05). Conversely, gender, ethnicity, tumor location, tumor size, body mass index (BMI), history of diabetes and hypertension, surgical approach, perioperative blood transfusion history, renal score, albumin, PLR, capsular invasion, pathological subtype, PAX-2, EMA, CD10, and Vimentin showed no significant relationship with RFS (P>0.05).

Table 3. Results of univariate Cox regression analysis for RFS in patients with T2-stage RCC.

Variable	1-year survival rate (%)	3-year survival rate (%)	5-year survival rate (%)	HR (95% CI)	P value
Gender				1.009 (0.601–1.693)	0.97
Male	82.6	64.2	52.4
Female	82.0	68.1	54.6
Age (years)				1.756 (1.075–2.869)	0.03
≥60	78.8	53.9	48.8
<60	84.6	72.6	60.0
Tumor location				1.167 (0.712–1.914)	0.54
Left	85.9	63.6	45.0
Right	81.1	69.7	62.5
Tumor diameter (cm)				0.755 (0.433–1.316)	0.32
7–10	81.8	61.5	49.5
>10	83.6	72.5	60.7
BMI (kg/m2)				0.868 (0.601–1.255)	0.45
>23.9	86.1	67.5	55.2
18.5–23.9	77.2	64.4	49.4
<18.5	87.5	57.4	45.2
Ethnicity				1.304 (0.885–1.920)	0.18
Han	82.6	67.0	54.0
Hui	75.0	60.0	55.1
Tibetan	75.2	65.4	50.0
Other	66.7	55.4	50.2
Surgical method				2.713 (0.663–11.094)	0.17
Radical resection	82.1	63.8	51.4
Nephron-sparing	95.7	72.5	52.5
Hypertension				0.631 (0.371–1.073)	0.09
Yes	84.4	51.0	35.3
No	81.7	69.5	57.9
Diabetes				0.511 (0.205–1.276)	0.15
Yes	60.0	48.0	40.1
No	83.7	66.2	53.8
Preoperative anemia				1.924 (1.062–3.485)	0.03
Yes	73.1	48.4	32.3
No	83.9	68.4	56.6
Perioperative blood transfusion				1.265 (0.675–2.370)	0.46
Yes	78.6	58.4	46.8
No	83.1	66.6	54.4
Renal score				1.313 (0.789–2.185)	0.30
4–6 points	–	–	–
7–9 points	85.2	63.2	56.5
10–12 points	82.3	64.2	49.9
Fuhrman grade				3.865 (2.364–6.321)	<0.001
Grade 1–2	89.7	77.0	63.8
Grade 3–4	60.0	32.0	24.7
Albumin (g/L)				0.629 (0.340–1.161)	0.14
≥35	83.9	66.9	56.7
<35	65.7	57.0	24.9
Globulin (g/L)				2.451 (1.481–4.056)	<0.001
≥34	69.9	47.4	33.2
<34	88.2	72.2	59.7
AST/ALT				2.142 (1.300–3.531)	0.003
≥1.46	69.2	51.9	40.4
<1.46	87.6	70.6	58.7
MLR				2.809 (1.527–5.169)	<0.001
≥0.22	76.7	55.8	40.1
<0.22	91.4	81.1	73.8
NLR				2.900 (1.748–4.812)	<0.001
≥2.7	70.2	46.9	30.6
<2.7	90.7	78.3	66.7
PLR				1.299 (0.715–2.013)	0.41
≥138	81.2	62.8	40.9
<138	83.3	67.0	58.9
PIV				3.846 (2.329–6.351)	<0.001
≥176.65	64.4	39.2	17.9
<176.65	91.0	78.1	68.0
SII				4.472 (2.387–8.381)	<0.001
≥430	70.5	47.6	36.2
<430	95.4	88.2	76.8
SIRI				4.398 (2.621–7.381)	<0.001
≥1.27	64.0	37.4	29.3
<1.27	93.0	82.3	68.4
Tumor necrosis				2.814 (1.671–4.738)	<0.001
Yes	68.6	41.5	13.0
No	85.7	69.7	61.9
Capsular invasion				1.505 (0.884–2.561)	0.13
Yes	82.6	55.6	43.3
No	82.3	68.7	57.2
Sarcomatoid differentiation				4.579 (2.761-7.593)	<0.001
Yes	53.1	26.8	16.7
No	88.6	74.1	61.9
Intravascular tumor thrombus				3.404 (2.076–5.581)	<0.001
Yes	66.7	38.6	15.6
No	86.7	73.5	68.2
Pathological type				0.886 (0.688–1.141)	0.35
Clear cell carcinoma	82.2	61.3	45.0
Papillary cell carcinoma	83.3	66.7	50.1
Chromophobe cell carcinoma	90.9	85.6	67.4
Other types of carcinoma	73.7	66.3	56.8
Ki-67				1.868 (1.134–3.079)	0.01
≤10%	85.2	72.6	56.7
>10%	68.9	52.1	43.4
CAIX				9.308 (4.517–19.178)	<0.001
Positive	74.5	44.2	8.3
Negative	94.0	88.2	86.3
PAX-8				2.189 (1.180–4.062)	0.01
Positive	80.2	59.7	41.5
Negative	87.3	77.7	71.1
PAX-2				0.901 (0.510–1.594)	0.72
Positive	88.0	63.3	60.2
Negative	80.3	66.0	52.3
EMA				1.071 (0.545–2.105)	0.84
Positive	81.5	64.7	52.7
Negative	86.7	68.0	54.4
CD10				1.814 (0.781–4.212)	0.17
Positive	82.6	63.3	48.6
Negative	80.8	75.7	59.1
CK7				0.401 (0.203–0.792)	0.009
Positive	85.7	79.7	75.5
Negative	81.1	59.8	41.9
Vimentin				1.635 (0.913–2.927)	0.10
Positive	83.4	60.9	43.8
Negative	80.0	73.5	68.9
CD117				1.432 (1.173–3.079)	0.04
Positive	82.1	79.8	66.7
Negative	82.4	61.0	47.6
MelanA				2.423 (1.099–5.345)	0.03
Positive	63.6	33.9	10.2
Negative	83.6	67.1	56.3

ALT, alanine transaminase; AST, aspartate transaminase; BMI, body mass index; CAIX, carbonic anhydrase IX; CD10, cluster of differentiation 10; CD117, cluster of differentiation 117; CI, confidence interval; CK7, cytokeratin 7; EMA, epithelial membrane antigen; HR, hazard ratio; MLR, monocyte-to-lymphocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PAX-2, paired box gene 2; PAX-8, paired box gene 8; PIV, pan-immune-inflammation value; PLR, platelet-to-lymphocyte ratio; RCC, renal cell carcinoma; RFS, recurrence-free survival; SII, systemic immune-inflammation index; SIRI, systemic inflammation response index.

Stratified survival analysis based on univariate Cox regression results

To further verify the prognostic differences of the 19 statistically significant variables identified by univariate analysis, stratified survival analysis was performed using the log-rank test (two-sided). The results (Table 4) showed that there were significant differences in RFS between subgroups for all 19 variables (all P<0.05). For instance, the 5-year survival rate was significantly lower in patients aged ≥60 years (48.8%) than in those aged <60 years (60.0%) (P=0.02). Similarly, patients with high Fuhrman grade (grade 3–4) had a significantly lower 5-year survival rate (24.7%) compared to those with low grade (grade 1–2) (63.8%, P<0.001). These results further confirmed the prognostic association of the 19 variables, providing a reliable basis for subsequent multivariate regression analysis.

Table 4. Prognostic comparison between groups for significant variables from univariate analysis in T2-stage RCC.

Variables	Recurrence, metastasis or death (n=64)	No recurrence, metastasis or death (n=118)	Log-rank χ2	P value
Age (years)			5.256	0.02
≥60	32 (50.00)	39 (33.05)
<60	32 (50.00)	79 (66.95)
Preoperative anemia			4.880	0.03
Yes	14 (21.88)	12 (10.17)
No	50 (78.12)	106 (89.83)
Fuhrman grade			34.075	<0.001
Grade 1–2	33 (51.56)	104 (88.14)
Grade 3–4	31 (48.44)	14 (11.86)
Globulin (g/L)			13.160	<0.01
≥34	25 (39.06)	21 (17.80)
<34	39 (60.94)	97 (82.20)
AST/ALT			9.483	0.002
≥1.46	26 (40.63)	26 (22.03)
<1.46	38 (59.37)	92 (77.97)
MLR			12.169	<0.001
≥0.22	51 (79.69)	61 (51.69)
<0.22	13 (20.31)	57 (48.31)
NLR			18.810	<0.01
≥2.7	39 (60.94)	35 (29.66)
<2.7	25 (39.06)	83 (70.34)
PIV			32.332	<0.001
≥176.65	36 (56.25)	23 (19.49)
<176.65	28 (43.75)	95 (80.51)
SII			26.551	<0.01
≥430	52 (81.25)	43 (36.44)
<430	12 (18.75)	75 (63.56)
SIRI			37.955	<0.01
≥1.27	42 (65.63)	25 (33.90)
<1.27	22 (34.37)	93 (66.10)
Tumor necrosis			16.713	<0.01
Yes	22 (34.38)	13 (11.02)
No	42 (65.62)	105 (88.98)
Sarcomatoid differentiation			42.384	<0.01
Yes	25 (39.06)	7 (5.93)
No	39 (60.94)	111 (94.07)
Intravascular tumor thrombus			26.989	<0.01
Yes	28 (43.75)	11 (9.32)
No	36 (56.25)	107 (90.68)
Ki-67			6.289	0.01
≤10%	38 (59.38)	91 (77.12)
>10%	26 (40.62)	27 (22.88)
CAIX			50.121	<0.001
Positive	54 (84.37)	44 (39.29)
Negative	10 (15.63)	74 (62.71)
PAX-8			6.541	0.01
Positive	51 (79.69)	76 (64.41)
Negative	13 (20.31)	42 (35.59)
CK7			7.474	0.006
Positive	10 (15.62)	39 (33.05)
Negative	54 (84.38)	79 (66.95)
CD117			3.463	0.04
Positive	5 (7.81)	23 (19.49)
Negative	59 (92.19)	95 (80.51)
MelanA			5.199	0.02
Positive	7 (10.94)	4 (3.39)
Negative	57 (89.06)	114 (96.61)

ALT, alanine transaminase; AST, aspartate transaminase; CAIX, carbonic anhydrase IX; CD117, cluster of differentiation 117; CK7, cytokeratin 7; MLR, monocyte-to-lymphocyte ratio; NLR, neutrophil-to-lymphocyte ratio; PAX-8, paired box gene 8; PIV, pan-immune-inflammation value; RCC, renal cell carcinoma; SII, systemic immune-inflammation index; SIRI, systemic inflammation response index.

Results of multivariate Cox regression analysis

Based on a four-step variable selection strategy, candidate variables were gradually refined to address multicollinearity and variable redundancy, with the specific process as follows:

1. Initial screening of candidate variables: bilateral univariate Cox regression analysis was performed for all included variables, and 27 variables with P<0.20 were included in the initial candidate pool. This pool included 19 variables with original P<0.05 and eight potential confounding variables (0.05≤P<0.2), to avoid missing potential prognosis-related factors.

2. Multicollinearity elimination: multicollinearity among all 27 initial candidate variables was assessed using Pearson correlation analysis (high correlation threshold: |r|≥0.7) and VIF testing (severe multicollinearity threshold: VIF ≥10). The results indicated no severe multicollinearity. Among the inflammatory indices, all pairwise correlations between PIV, SII, and SIRI were below 0.7 (range, 0.169–0.388), with VIF values all below five (range, 1.059–1.423). Similarly, the correlations and VIF values for MLR and NLR with SII were below the thresholds. For other clinicopathological variables (e.g., Fuhrman grade, AST/ALT ratio, intravascular tumor thrombus), all VIFs were below five. Despite the absence of severe statistical collinearity, given the inherent mathematical coupling among the inflammatory indices (MLR, NLR, PIV, SII, SIRI) derived from shared blood components (neutrophils, lymphocytes, platelets), and to ensure the parsimony and statistical robustness of the final multivariate model given the limited number of events, we decided to retain only one representative indicator from this group for subsequent analysis. The selection was based on the univariate Cox regression results, prioritizing the statistical significance (P value) of the association with RFS, while also considering the magnitude of the HR. Accordingly, SII was retained, while MLR, NLR, PIV, and SIRI were excluded. Following this screening step, a total of 23 variables proceeded to the next stage for assessment of biological independence and the final multivariable analysis.

3. Biological independence screening: based on the principles of “no overlap in biological significance, sufficient clinical evidence, and specific correlation with the prognosis of T2-stage RCC”, further screening was performed on the 23 non-collinear variables. Three categories of variables were excluded as follows: first, variables with overlapping biological functions: Ki-67 (both Ki-67 and Fuhrman grade reflect tumor proliferation ability, with Fuhrman grade being more widely used in clinical practice), CK10 (both CK10 and CAIX are immunohistochemical indicators, with CAIX being a classic specific marker for RCC); second, variables with low clinical priority: 10 variables including age, preoperative anemia, ethnicity, and others (all were clinical confounding factors, with HR close to one in univariate Cox regression analysis, indicating weak independent effects on prognosis); third, variables with insufficient prognostic evidence: five variables including CK7, MelanA, and others (no clear clinical research evidence supports their correlation with the prognosis of T2-stage RCC). Finally, six biologically independent core candidate variables were identified: AST/ALT ratio, Fuhrman grade, SII, intravascular tumor thrombus, sarcomatoid differentiation, and CAIX.

4. Simplified multivariate regression analysis: the above six core candidate variables were incorporated into the Cox proportional hazards regression model, and stepwise optimization was performed using the two-sided backward elimination method (inclusion criterion: P<0.05; exclusion criterion: P≥0.05). Model fitting results showed that intravascular tumor thrombus was excluded due to an adjusted P value of 0.097, which did not reach statistical significance for independent prognostic value. Ultimately, the five remaining variables all had P values <0.05 and were identified as independent prognostic factors for postoperative RFS in T2-stage RCC patients (Table 5): AST/ALT ratio [HR =3.295, 95% confidence interval (CI): 1.899–5.719, P<0.001], Fuhrman grade (HR =2.380, 95% CI: 1.412–4.012, P=0.001), SII (HR =3.009, 95% CI: 1.567–5.776, P<0.001), sarcomatoid differentiation (HR =3.463, 95% CI: 1.981–6.054, P<0.001), and CAIX (HR =5.425, 95% CI: 2.555–11.519, P<0.001).

Table 5. Results of multivariate Cox regression analysis for RFS in patients with T2-stage RCC.

Variable	B value	Wald	P value	HR	95% CI
AST/ALT
≥1.46	1.192	17.977	<0.001	3.295	1.899–5.719
<1.46	1
Fuhrman grade
Grade 1–2	0.867	10.596	0.001	2.380	1.412–4.012
Grade 3–4	1
SII
≥430	1.102	10.962	<0.001	3.009	1.567–5.776
<430	1
Intravascular tumor thrombus
Yes	0.458	2.758	0.10	1.581	0.921–2.715
No	1
Sarcomatoid differentiation
Yes	1.242	18.995	<0.001	3.463	1.981–6.054
No	1
CAIX
Positive	1.961	19.372	<0.001	5.425	2.555–11.519
Negative	1

1 indicates the reference level. ALT, alanine transaminase; AST, aspartate transaminase; CAIX, carbonic anhydrase IX; CI, confidence interval; HR, hazard ratio; RCC, renal cell carcinoma; RFS, recurrence-free survival; SII, systemic immune-inflammation index.

Internal stability assessment of prognostic factors

Given the exploratory nature of this study, 1,000 bootstrap resamplings were performed to evaluate the internal stability of the identified independent prognostic factors. The results showed that the mean Harrell’s concordance index (C-index) validated by bootstrap was 0.835, with a 95% CI of 0.780–0.885. The Bootstrap C-index distribution histogram further confirmed this (Figure 2), demonstrating that the C-index values obtained from 1,000 resamplings presented a concentrated normal distribution without obvious skewness. These results indicate that the five independent prognostic factors identified in this study have good and stable discriminatory ability for postoperative RFS in patients with T2-stage RCC, and can effectively distinguish patient groups with different recurrence risks.

Figure 2.

Bootstrap sampling distribution of Harrell’s C-index. The red dotted line represents the mean value of Harrell’s C-index (0.835), and the blue dotted line represents the 95% confidence interval boundaries (0.780–0.885), indicating the stability of the prognostic factors.

Discussion

As one of the relatively common stages in RCC, T2-stage RCC is characterized by a large tumor volume, often accompanied by local invasion or regional lymph node metastasis. While some patients achieve long-term tumor-free survival postoperatively, others experience recurrence or metastasis within a short period. The 5-year survival rate ranges from 60% to 80% (13), and accurate clinical prognostic evaluation remains challenging. Although surgical resection currently remains the primary treatment for T2-stage RCC, significant variability in postoperative survival rates persists. This variability not only reflects the complexity of the tumor’s biological behavior but also underscores the limitations of the existing prognostic evaluation system. In recent years, with the advancement of precision medicine, an increasing number of studies have begun exploring prognostic indicators for RCC across multiple dimensions, encompassing clinical biochemical markers, imaging features, treatment modalities, and pathological molecular markers, among other fields (14,15).

The AST/ALT ratio emerged as an independent prognostic factor in T2-stage RCC (HR =3.295, 95% CI: 1.899–5.719, P<0.001), which is consistent with the findings of recent studies on solid tumors (16,17). More importantly, the AST/ALT ratio exerts its prognostic effect through multiple biological mechanisms in RCC progression. Firstly, AST and ALT are not only traditional indicators of hepatocellular function but also closely related to systemic metabolic disorders and immune microenvironment imbalance (18). In T2-stage RCC, tumor cells can induce persistent systemic inflammation, which in turn leads to liver function damage and changes in the AST/ALT ratio. Specifically, inflammatory factors [such as tumor necrosis factor (TNF)-α, interleukin (IL)-6] secreted by tumor-associated macrophages can promote hepatocyte mitochondrial damage, resulting in increased release of AST (mainly located in mitochondria) into the blood, thereby increasing the AST/ALT ratio (19,20). Secondly, the elevated AST/ALT ratio can further promote tumor progression by regulating the redox balance of tumor cells. Mitochondrial damage caused by high AST levels leads to the accumulation of reactive oxygen species (ROS) in tumor cells. Moderate ROS accumulation can activate signal transduction pathways such as PI3K/Akt and MAPK, promoting tumor cell proliferation and invasion; on the other hand, it can inhibit the function of lymphocytes and promote the release of pro-inflammatory cytokines, weakening the body’s anti-tumor immune response (21-23). Thirdly, RCC cells are prone to glycolysis even under aerobic conditions (Warburg effect), and the elevated AST/ALT ratio can enhance glycolytic enzyme activity, providing energy and material basis for tumor cell growth and metastasis (24). Compared to other inflammatory indicators, the AST/ALT ratio is more easily accessible in clinical practice (as part of routine liver function tests) and has higher clinical applicability. Our findings suggest that the preoperative AST/ALT ratio can serve as a convenient and reliable prognostic marker for T2-stage RCC, providing a non-invasive tool for risk stratification. However, studies have demonstrated that dynamic changes in the AST/ALT ratio from pre- to postoperative periods have prognostic significance for postoperative survival in RCC patients, particularly those with a low preoperative ratio that becomes elevated postoperatively (25). The current study highlights the prognostic value of the preoperative AST/ALT ratio in T2-stage RCC, however, the impact of its dynamic changes on prognosis requires further verification through prospective studies.

Fuhrman grade was identified as a crucial independent prognostic factor for postoperative RFS in T2-stage RCC (HR =2.380, 95% CI: 1.412–4.012, P=0.001), consistent with findings from multiple previous studies (13,26). As a key indicator for assessing histological differentiation in RCC, a higher Fuhrman grade (grades III–IV) typically indicates more pronounced tumor cell atypia, stronger proliferative activity, and greater clinical invasiveness (27). Patients with high Fuhrman grades (grades III–IV) had a significantly lower 5-year survival rate (24.7%) compared to those with low grade (grades I–II) (63.8%), which may be attributed to the enhanced ability of high-grade tumor cells to remodel the tumor microenvironment. Specifically, high-grade tumor cells can secrete a variety of pro-angiogenic factors to promote neovascularization, providing sufficient nutrients and oxygen for tumor growth and metastasis (28). Meanwhile, they can inhibit the function of cytotoxic T lymphocytes and natural killer cells, escaping anti-tumor immune surveillance. However, the World Health Organization (WHO) recommended the updated WHO/ISUP nuclear grading system for tumor nuclear grading in 2016 (29). While the clinical application of Fuhrman grading has declined in recent years, its prognostic value in clear cell renal cell carcinoma (ccRCC) remains widely acknowledged, especially in T2-stage tumors with large volume and complex biological behavior. Emerging evidence has confirmed that combined evaluation strategies integrating computed tomography (CT), proliferative markers such as Ki-67, or molecular typing can further optimize the prognostic utility of Fuhrman grading (30,31). Therefore, in the clinical management of T2-stage RCC, comprehensive assessment—incorporating Fuhrman grading alongside other prognostic factors identified in this study (such as vascular tumor thrombus, SII, etc.)—will more effectively facilitate accurate prognostic prediction and the formulation of individualized treatment strategies.

Inflammatory processes participate in tumor initiation, progression, and metastasis through multiple mechanisms (32), and their association with the prognosis of RCC has gradually attracted attention. In this study, SII was identified as independent prognostic factors for the prognosis of T2-stage RCC (HR =3.009, 95% CI: 1.567–5.776, P<0.001), which further confirms the core role of inflammatory and immune regulation in the tumor microenvironment. As a composite inflammatory marker integrating platelet, neutrophil, and lymphocyte counts, SII reflects the dynamic balance between the tumor-induced immunosuppressive microenvironment and the host anti-tumor immune response. Studies have demonstrated that a high preoperative SII is significantly associated with poorer OS and cancer-specific survival in patients with non-metastatic RCC; that is, the higher the preoperative SII value, the higher the risk of death (33). Elevated SII levels typically indicate enhanced immunosuppression mediated by platelets and neutrophils—platelets can promote tumor cell adhesion and angiogenesis, while neutrophils may facilitate tumor invasion by secreting matrix metalloproteinases and ROS (34). Moreover, patients with higher SII levels had an elevated mortality rate, which may be attributed to the fact that chronic inflammation can induce the generation of reactive nitrogen and oxygen species, leading to genomic instability and cellular senescence, thereby increasing the risk of death (35). For T2-stage RCC, which is characterized by relatively localized lesions but potential high invasiveness, SII serves as a non-invasive, readily measurable indicator to evaluate the systemic immune status of patients. This provides a reliable basis for clinical risk stratification and the formulation of personalized immunomodulatory therapy strategies. However, it should be noted that the prognostic value of SII in RCC still needs to be verified in large-scale multicenter prospective studies, and the specific molecular mechanisms underlying its association with T2-stage RCC recurrence require further exploration.

The present study identified sarcomatoid differentiation as independent prognostic factor for T2-stage RCC through multivariate analysis (P<0.05). Extensive literature has established that sarcomatoid features correlate with aggressive tumor biology, manifested by shortened OS and enhanced malignant potential (36,37). Our findings corroborate these observations, demonstrating markedly worse prognosis in patients with sarcomatoid differentiation compared to those without this feature, with a 5-year survival rate of 16.7% consistent with the overall 5-year survival rate of 18.1% reported in the literature (38). Sarcomatoid differentiation is a histopathological variant characterized by the transformation of epithelial tumor cells into mesenchymal cell morphology, which is widely recognized as a marker of high aggressiveness in RCC. This morphological transformation is accompanied by the activation of epithelial-mesenchymal transition (EMT) pathways, which endow tumor cells with enhanced migratory and invasive capabilities, as well as resistance to apoptosis and chemotherapy (39). In T2-stage RCC, although the tumor is confined to the kidney and has not yet invaded adjacent organs or metastasized distantly, the emergence of sarcomatoid differentiation implies a higher malignant potential. The presence of sarcomatoid components fundamentally alters tumor behavior by augmenting invasive capacity and metastatic propensity, thereby drastically diminishing survival prospects. Additionally, given the strong association between sarcomatoid differentiation and immunosuppression, targeted immunotherapy may be a promising treatment option for such patients (40). However, due to the relatively rare incidence of sarcomatoid differentiation in T2-stage RCC, further studies with larger sample sizes are needed to clarify its optimal treatment strategies and prognostic evaluation criteria.

In recent years, molecular markers have garnered increasing attention in tumor prognostic analysis, emerging as a prominent research focus. Our multivariate analysis identified CAIX as an independent prognostic factor for T2-stage RCC, which aligns with findings from previous studies (41). CAIX is a transmembrane protein encoded by the CA9 gene, and its expression is significantly upregulated through the hypoxia-inducible factor 1-alpha (HIF-1α) pathway under hypoxic conditions (42). In ccRCC, frequent Von Hippel Lindau (VHL) gene mutations or deletions lead to the accumulation of HIF-1α, which further upregulates the expression of CAIX (43). Notably, we found that T2-stage RCC patients with elevated CAIX expression exhibited significantly worse prognosis. Mechanistically, CAIX overexpression may drive tumor progression through extracellular matrix degradation, matrix metalloproteinase activation, and extracellular acidification, collectively fostering a metastatic niche (44,45). Beyond CAIX, current research on RCC prognosis involves various other molecular markers. For instance, vascular endothelial growth factor (VEGF) and its receptor (VEGFR), as well as TP53, have been demonstrated to be closely associated with tumor angiogenesis and progression, and their expression levels can reflect the invasive potential of tumors and patients’ survival status (46). Additionally, programmed death ligand 1 (PD-L1) and microRNAs have gradually become prominent research hotspots (47). These markers delineate the prognostic characteristics of RCC from diverse biological perspectives, and together with CAIX identified in this study, provide multifaceted references for constructing a more comprehensive prognostic evaluation system for T2-stage RCC.

This study has several main limitations that need to be acknowledged. Firstly, the sample size is relatively small, and it is a single-center retrospective cohort study. The selection bias caused by the retrospective design and the differences in clinical practice between different institutions may limit the generalizability of the study results. Secondly, this study lacks an external validation cohort, so the applicability of the identified prognostic factors in diverse patient populations cannot be confirmed, which affects the reliability and promotion of the research results. Thirdly, due to the limitation of retrospective data, we did not obtain the dynamic monitoring data of biomarkers (such as AST/ALT ratio, CAIX) during follow-up, so we cannot analyze the dynamic changes of these biomarkers and their potential prognostic value. In future research, we will try to make up for these limitations and further improve the research design.

Conclusions

In conclusion, this study identifies five independent prognostic factors (AST/ALT ratio, Fuhrman grade, SII, sarcomatoid differentiation, CAIX) for postoperative RFS in T2-stage RCC, highlighting the AST/ALT ratio as a readily accessible marker for this high-recurrence subgroup and filling the research gap in prognostic factors for T2-stage RCC. Combined application of these factors improves risk stratification accuracy, providing a scientific basis for individualized management and optimized prognosis. Limitations include small sample size, single-center retrospective design and lack of external validation. Future large-scale prospective multicenter studies are needed to validate these factors and explore their molecular mechanisms, facilitating precision management of T2-stage RCC.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was performed in accordance with the principles of the Declaration of Helsinki and its subsequent amendments. The study protocol was reviewed and approved by the Institutional Review Board of the Second Hospital of Lanzhou University (approval No. 2025A-983). Due to the retrospective nature of the study, the requirement for informed consent was waived by the IRB.

Data Sharing Statement

Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-aw-837/dss

tau-15-02-57-dss.pdf ^{(138.9KB, pdf)}

DOI: 10.21037/tau-2025-aw-837