A total of 154 patients diagnosed with BPH were included in the current analysis. Table 1 presents the characteristics of the entire study population. The subjects had a mean age of 66.8 ± 7.4 years, a height of 170.6 ± 4.7 cm, a weight of 70.9 ± 9.1 kg, a BMI of 24.3 ± 3.0, a PSA of 7.5 ± 10.9 ng/ml, an IPSS of 15 ± 8, a Qmax of 11.4 ± 7.2 ml/s, a PV of 40.1 ± 27.7 ml, an IPP of 11.1 ± 6.5 mm, a PUL of 5.2 ± 0.9 cm, a PCAR of 0.83 ± 0.04, and a PSVR of 0.74 ± 0.04. At 12 month follow-up, 74(48.1%) patients underwent BPH-related prostatic surgery, while among the 58 (37.6%) patients who did not undergo surgery, 10 (17%) experienced an increase in IPSS greater than or equal to 4 point. 3 patients were diagnosed with prostate cancer, and 19 patients were lost to follow-up.

Pearson’s correlation analysis revealed that PV (r = 0.189, P = 0.019), IPP (r = 0.433, P < 0.001), PUL (r = 0.245,  = 0.003), PCAR (r = 0.362, P < 0.001) and PSVR (r = 0.308, P < 0.001 P) increased with increasing IPSS (Figure S1). In univariate and multivariate linear regression analysis, we evaluated the relationships between IPSS and potential variables, including Age, Height, Weight, BMI, PSA, PV, IPP, PUL, PCAR, and PSVR. Multivariate analysis showed that IPP [B:0.428, 95% confidence interval: (0.221, 0.635), p < 0.001] and PSVR [B:38.109, 95% confidence interval: (3.926, 72.292), p = 0.029] were the two most significant parameters associated with IPSS (Table 2).

Pearson’s correlation analysis showed that PV (r = −0.203, P = 0.012), IPP (r = −0.462, P < 0.001), PUL (r = −0.312, P < 0.001), PCAR (r = −0.293, P < 0.001), and PSVR (r = −0.391, P < 0.001) increased with declining Qmax (Figure S2). In univariate linear regression analyses, PV, IPP, PUL, PCAR, and PSVR were negatively associated with Qmax, but age, height, weight, BMI, and PSA were not. PV [B:0.067, 95% confidence interval: (0.019, 0.115), p = 0.007], IPP [B:−0.432, 95% confidence interval: (−0.603, −0.262), p < 0.001] and PSVR [B:−53.510, 95% confidence interval: (−80.898, −26.122), p < 0.001] were significant parameters associated with Qmax in multivariate linear regression analysis (Table 3).

In univariate logistic regression analyses, PV, IPP, PUL, PCAR, and PSVR were statistically correlated with BPH-related prostatic surgery, but only PSVR was statistically significant with BPH-related prostatic surgery(OR = 5.153 P = 0.02) in multivariate logistic regression. (Table 4). In univariable and multivariate logistic regression analysis, only PSVR remained statistically significant with an increase in the IPSS of at least 4 points, the p-values were 0.041 and 0.016. (Table 5).

To assess the predictive power of prostate anatomical parameters for surgery. We compared the area under the curve(AUCs) between PV, IPP, PUL, PCAR,PSVR and PSVR combined IPP( PSVR&IPP)with BPH-related prostatic surgery in the present analysis(N = 132).AUCs ranged from 0.617 to 0.744 (PCAR, 0.617; PV, 0.713; IPP, 0.723; PSVR, 0.735; PUL, 0.744; PSVR&IPP, 0.776) Fig. 3.

Receiver operating characteristics of PV, IPP, PUL, PCAR, PSVR and PSVR&IPP with BPH-related prostatic surgery. PV Prostate volume, IPP Intravesical prostatic protrusion, PUL Prostatic urethral length, PCAR Presumed circle area ratio, PSVR Prostate spherical volume ratio, PSVR&IPP PSVR combined IPP

To assess the predictive power of prostate anatomical parameters for increased IPSS. We also compared AUCs between PV, IPP, PUL, PCAR, PSVR and PSVR combined IPP( PSVR&IPP) with an increase in the IPSS of at least 4 points in the present analysis(N = 58).AUCs ranged from 0.423 to 0.724 (PV, 0.423; IPP, 0.480; PUL, 0.475; PCAR, 0.0.483; PSVR, 0.724; PSVR&IPP, 0.760) Fig. 4.

Receiver operating characteristics of PV, IPP, PUL, PCAR, PSVR and PSVR&IPP with an increase in the IPSS of at least 4 points. PV Prostate volume, IPP Intravesical prostatic protrusion, PUL Prostatic urethral length, PCAR Presumed circle area ratio, PSVR Prostate spherical volume ratio, PSVR&IPP PSVR combined IPP