The most important factor affecting the prognosis of penile cancer is ILN metastasis [3, 11–13]. Several studies have indicated that the rate of ILN metastasis in patients with penile cancer is 30–60% . The latest meta-analysis  selected 42 eligible studies that included a total of 4,802 patients, of whom 1,706 (36%) were diagnosed with ILN metastasis. This finding was corroborated by our results, in which 37.3% of patients had ILN metastasis (41/110).
In our multiple regression analysis, the maximum diameter of the enlarged ILNs, pathological stage, pathological differentiation, and LVI were the only predictors of ILN metastasis. We conducted a stratified analysis and interaction tests on the four factors and did not find any obvious interaction, further proving the stability of our results.
A maximum ILN diameter of > 1.0 cm is usually considered abnormal, while a diameter > 1.5 cm, with a relatively hard texture, strongly indicates tumor metastasis . The present study suggested that the largest diameter of the enlarged ILNs was an independent risk factor, corroborating studies by Tang et al.  and Zhou et al. . However, in another study, 50% of enlarged ILNs were inflamed or reactive, rather than metastatic , indicating that ILN metastasis cannot be reliably detected using imaging or clinical evaluation. It is essential to predict ILNM in combination with the pathological characteristics of the primary tumor.
The pathological stage of the primary tumor is generally considered the most important parameter for predicting ILN metastasis in patients with penile cancer [14, 18]. Depending on whether there is infiltration of the urethra or corpus cavernosum, the tumor stage can be classified as T1, T2, or higher . In the present study, the rate of ILN metastasis in patients with T1 stage was 22.7% (10/44), while it was 47.0% (31/66) in those with T2 stage and above. The ILN metastasis rate in patients with T2 stage and above was significantly higher than that of patients with T1 stage disease. Several researchers have opined that, if the corpus cavernosum is infiltrated, ILN dissection should be performed, even if there are no obvious enlarged ILNs [19, 20]. However, deciding to perform ILN dissection on all patients with T2 and above only based on tumor stage, which will lead to overtreatment. In the present study, 53.0% of patients with T2 stage and above showed no obvious signs of ILN metastasis. Therefore, other factors should be taken into account when screening high-risk patients.
The degree of tumor differentiation under pathology is negatively correlated with tumor pathological grade or malignancy, with lower tumor differentiation indicating higher grade, higher malignancy, and greater risk of metastasis . Solsona et al.  found that the grade of tumor differentiation correlates well with ILN metastasis, corroborating our multiple regression analysis (HR = 3.0, 95% CI: 1.3–7.1). According to Horenblas et al. , the lymph node metastasis rates of G1, G2, and G3 were 29%, 46%, and 82%, respectively. Our results were similar to these, with 20% (8/40), 45% (30/66), and 75% (3/4), respectively. Theodoresu et al.  asserted that tumor grade is the only predictor of ILN metastasis, and they recommended ILN dissection in patients with G2 and G3, which coincides with our point of view.
The existing literature has different opinions on whether LVI is a predictor of ILN metastasis. Some studies have ranked LVI as one of the most important factors of metastasis [24–27], while others have not . Our findings suggest that LVI is significantly associated with ILN metastasis (HR = 3.6, 95% CI: 1.1–11.6) among patients with penile cancer.
The present study had several strengths. Firstly, although some potential confounding factors were unavoidable, we used strict statistical adjustment to minimize residual confounding. Secondly, the effect modifier factor analysis took full advantage of the data; no interaction was found, indicating that the results are more stable.
Conversely, there were some limitations of our study. The patients with penile cancer were collected from a large single medical center. Therefore, external validation of the results is required. The study was cross-sectional, and no information was available about the degree of risk factors prior to ILN metastasis, because the earlier pathology reports contained no such data. Moreover, no data were available regarding lymph node extranodal transfer and the growth pattern of tumors in some patients (papillary, ulcerated, invasive, etc.) As such, we could not consider these variables in the final results, although there was evidence that these factors had prognostic significance. Despite these limitations, the findings of this study have implications for clinicians when formulating further treatment plans in patients with penile cancer who have undergone surgery. However, prospective studies with more samples are needed.