Factors influencing postoperative of hypertension remission in adrenal cavernous hemangioma: A multicenter retrospective cohort study

Table 1 Baseline characteristics and clinical features of patients in hypertension cure group and non-cure group after adrenal cavernous hemangioma surgery, mean ± SD

Indicator	Positive group (n = 62)	Negative group (n = 40)	P value
Age (years)	48.1 ± 9.8	50.5 ± 8.9	0.213
Gender (male/female)	24/38	16/24	0.576
Tumor diameter (cm)	3.3 ± 1.2	4.7 ± 1.4	< 0.001
Hypertension duration (years)	3.6 ± 1.3	7.5 ± 1.2	< 0.001
RAAS hormones (pg/mL)	128.5 ± 34.2	182.3 ± 43.8	< 0.001
Adrenal medullary hormones (pg/mL)	85.6 ± 21.3	132.4 ± 30.5	< 0.001
Surgical approach (laparoscopy/open)	52/10	24/16	0.008
Total adrenalectomy (laparoscopy/open)	48/8	16/12	0.012
Partial adrenalectomy (laparoscopy/open)	4/2	8/4	0.659

Measurement data were expressed as mean ± SD, and comparisons between groups were performed using the independent samples t-test. Enumeration data were presented as case numbers, and inter-group comparisons were conducted using the χ² test. Basis for selecting surgical methods: In the laparoscopic group, 93.5% (58/62) of tumors were ≤ 6 cm, while in the open surgery group, 77.5% (31/40) of tumors were > 6 cm. The proportions of adhesion grade ≥ III in the laparoscopic group and the open surgery group were 58% vs 82.5%, respectively (P < 0.001). Positive group: Patients whose hypertension was cured 24 months after surgery (24-hour ambulatory blood pressure monitoring daytime < 135/85 mmHg and nighttime < 120/70 mmHg without antihypertensive medication). Negative group: Patients with persistent hypertension not meeting the above criteria. RAAS: Renin-angiotensin-aldosterone system.

Table 2 Comparison of central vein preservation rate between surgical approach

Surgical approach	Total cases	Central vein-preserved	Preservation	P value
Laparoscopic	82	75	91.2	< 0.001
Open	28	19	68.3

Central venous injury was defined as dislodgement of vascular clips or accidental injury by ultrasonic scalpel. The laparoscopic group adopted the fine dissection technique with vascular clips/ultrasonic scalpel, while the open surgery group used the ligation + transection method. Fisher’s exact test was applied, and the difference in retention rate between the two groups was of highly statistical significance (P = 0.003), suggesting that the laparoscopic approach provides better vascular protection.

Table 3 Univariate logistic regression analysis of hypertension cure after adrenal cavernous hemangioma surgery

Influencing factor comparison group	Comparison group	OR	95%CI	P value
Tumor diameter (cm)	Per 1-cm increase	0.68	0.51-0.90	< 0.01
Hypertension duration (year)	Per 1-year increase	0.59	0.43-0.81	< 0.01
RAAS hormones (pg/mL)	Per 1-SD increase	3.63	1.78-7.40	< 0.01
Adrenal medullary hormones (pg/mL)	Per 1-SD increase	2.98	1.31-6.76	< 0.001
Surgical approach (laparoscopy/open)	Laparoscopy vs open	2.8	1.1-7.2	< 0.01
Age (years)	> 50 years/< 50 years	1.05	0.48-2.30	0.90
Gender (male/female)	Male/female	1.12	0.49-2.55	0.79

Group comparisons for surgical approach used both χ² test (Table 2) and logistic regression. P values varied across models, but both methods indicated significant associations. RAAS: Renin-angiotensin-aldosterone system; OR: Odds ratio; CI: Confidence interval.

Table 4 Subgroup analysis of patients with tumor diameter ≥ 6 cm and adhesion grade III, n (%)/mean ± SD

Variable	Laparoscopic group (n = 17)	Open group (n = 17)	OR (95%CI)	P value
Hypertension cure	12 (70.6)	7 (41.2)	2.68 (1.15-6.23)	0.024
Tumor diameter (cm)	7.2 ± 1.3	7.5 ± 1.1	0.89 (0.56-1.41)	0.62
RAAS hormones (pg/mL)	195.3 ± 42.7	198.6 ± 40.5	0.97 (0.71-1.33)	0.86
Adrenal medullary hormones (pg/mL)	142.5 ± 31.8	145.8 ± 29.7	0.96 (0.68-1.35)	0.81
Hypertension duration (years)	8.2 ± 2.5	8.5 ± 2.3	0.94 (0.61-1.45)	0.77

Subgroup inclusion criteria: Tumor diameter ≥ 6 cm and modified Adler adhesion grade III, with a total of 34 cases (accounting for 33.3% of the total sample). The baseline characteristics (tumor diameter, hormone levels, disease duration) of the two groups showed no significant differences by t-test or χ² test (all P > 0.05), indicating comparability. The cure rate in the laparoscopic group was 29.4% higher than that in the open surgery group. Logistic regression analysis showed that the surgical approach was an independent influencing factor (odds ratio = 2.68), suggesting that laparoscopic surgery remains advantageous in complex cases. RAAS: Renin-angiotensin-aldosterone system; OR: Odds ratio; CI: Confidence interval.

Table 5 Statistical power analysis for subgroup comparisons

Subgroup	Sample size	Cure rate difference (%)	Statistical power (1-β)	P value
Open surgery	28	33.3 (vs laparoscopic)	0.72	< 0.008
Laparoscopic partial	18	33.4 (vs total)	0.58	< 0.659
Tumor > 6 cm + adhesio III	34	29.4	0.75	< 0.024

Cure rate difference refers to the difference in cure rates between the subgroup and the control group (e.g., the open surgery group vs the total laparoscopic resection group). Statistical power calculation was based on Cohen’s h formula (α = 0.05). In the subgroup with tumor ≥ 6 cm and grade III adhesions, the statistical power (1-β) is 0.75 (with α = 0.05), which necessitates cautious interpretation of the conclusion regarding the advantages of laparoscopy.

Table 6 Baseline characteristics of patients with and without preoperative drug withdrawal

Indicator with	Withdrawal group (n = 90)	Non-withdrawal group (n = 12)	P value
Age (years)	48.5 ± 10.2	50.1 ± 9.8	0.612
Gender (male/female)	36/54	4/8	0.735
Tumor diameter (cm)	3.5 ± 1.3	3.8 ± 1.2	0.478
Hypertension duration (years)	4.2 ± 1.8	4.5 ± 1.9	0.631
RAAS hormones (pg/mL)	146.3 ± 38.5	152.7 ± 40.2	0.524
Adrenal medullary hormones (pg/mL)	96.8 ± 24.1	101.5 ± 25.3	0.497
Surgical approach (laparoscopy/open)	72/18	8/4	0.689

Patients unable to discontinue medications were defined as those who failed to stop taking diuretics or calcium channel blockers within 4 weeks before surgery (n = 12, accounting for 11.8%), among whom 8 cases were diuretic-dependent and 4 cases were calcium channel blocker-dependent. Baseline characteristics of the two groups showed no significant differences by independent samples t-test or χ² test (all P > 0.05), indicating comparability. RAAS: Renin-angiotensin-aldosterone system.

Table 7 Effect of non-withdrawal status as covariate in multivariate model

Variable	β	OR	95%CI	Ward value	P value
Tumor diameter (cm)	1.258	3.5	1.82-6.91	10.48	0.001
RAAS hormones (pg/mL)	1.285	3.63	1.78-7.40	10.62	0.001
Adrenal medullary hormones (pg/mL)	1.089	2.98	1.31-6.76	6.74	0.009
Hypertension duration (years)	3.012	20.23	4.15-98.27	16.53	< 0.001
Surgical approach (laparoscopy/open)	2.769	16.01	2.78-91.82	9.36	0.002
Tumor × surgery	0.587	1.80	1.12-2.90	4.52	0.031
Diuretic/CCB non-withdrawal	0.173	1.19	0.58-2.45	2.56	0.672

In the multivariate model, “inability to discontinue medications” was included as a covariate (1 = inability to discontinue, 0 = discontinued). The results showed that it had no significant effect on the postoperative cure of hypertension (P = 0.672). The equivalent dose of diuretics (standardized to furosemide mg/day) and calcium channel blocker dose (standardized to amlodipine mg/day) were included in the model as continuous variables, which did not significantly alter the odds ratio values of other variables (fluctuation range < 3%). RAAS: Renin-angiotensin-aldosterone system; CCB: Calcium channel blocker; OR: Odds ratio; CI: Confidence interval.

Table 8 Baseline characteristics comparison between laparoscopic and open surgery groups after propensity score matching, n (%)/mean ± SD

Variable	Group	n	Mean/proportion	SD	Statistical test (post-matching)
Tumor diameter (cm)	Laparoscopic	30	3.5 ± 1.2	0.06	t = 0.58, P = 0.56
	Open	30	3.8 ± 1.3
RAAS hormones (pg/mL)	Laparoscopic	30	145.2 ± 36.1	0.07	t = 0.69, P = 0.49
	Open	30	149.8 ± 37.5
Adhesion grade (≥ III)	Laparoscopic	30	3 (10.0)	0.05	χ2 = 0.42, P = 0.52
	Open	30	4 (13.3)
Hypertension duration (years)	Laparoscopic	30	4.3 ± 1.4	0.08	t = 0.74, P = 0.46
	Open	30	4.5 ± 1.5
Adrenal medullary hormones (pg/mL)	Laparoscopic	30	94.7 ± 23.1	0.09	t = 0.87, P = 0.39
	Open	30	98.5 ± 24.3

After matching, the standardized mean differences of all variables were < 0.1, indicating that the baseline characteristics were balanced. The cure rate in the laparoscopic group (80.0%, 24/30) was significantly higher than that in the open surgery group (53.3%, 16/30). Logistic regression analysis showed that the surgical method was an independent risk factor (odds ratio = 2.65, 95% confidence interval: 1.17-6.03, P = 0.021). RAAS: Renin-angiotensin-aldosterone system.

Table 9 Diagnostic efficacy comparison between combined predictor and independent factors

Variable	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)
Combined predictor	82.7	86.1	85.7	83.3
Hypertension duration (years)	78.8	79.4	76.5	81.2
Tumor diameter (cm)	76.9	77.6	74.3	79.8
Surgical approach (laparoscopy/open)	73.1	75.3	71.1	77.5
RAAS hormones (pg/mL)	65.4	68.2	63.5	70.1
Adrenal medullary hormones (pg/mL)	61.5	64.7	60.2	66.3

Sensitivity, specificity, positive predictive value and negative predictive value are expressed as percentages. The combined predictor corresponds to the multivariable model integrating tumor diameter, hypertension duration, renin-angiotensin-aldosterone system hormone, adrenal medullary hormone and surgical approach. RAAS: Renin-angiotensin-aldosterone system; PPV: Positive predictive value; NPV: Negative predictive value.

Table 10 Receiver operating characteristic curve analysis of combined predictor and independent factors

Variable	AUC	95%CI (lower limit)	95%CI (upper limit)	P value	Sensitivity	Specificity	Youden’s index	Cutoff value
Combined predictor	0.872	0.791	0.934	< 0.001	0.827	0.861	0.616	12.57
Tumor diameter (cm)	0.813	0.712	0.895	0.003	0.788	0.794
Surgical approach (laparoscopy/open)	0.796	0.688	0.882	0.005	0.769	0.776
RAAS hormones (pg/mL)	0.725	0.598	0.831	0.021	0.731	0.753
Adrenal medullary hormones (pg/mL)	0.65	0.513	0.768	0.045	0.654	0.682
Tumor diameter (cm)	0.632	0.487	0.763	0.057	0.615	0.647
Combined predictor (excluding AMH)	0.854	0.768	0.915	0.036	0.805	0.842	0.647	11.89

Surgical approach was coded as laparoscopy = 0 and open = 1 in the combined predictor. Delong test showed the combined predictor had a significantly larger area under the curve than any single factor (P < 0.05). RAAS: Renin-angiotensin-aldosterone system; AMH: Anti-Mullerian hormone; CI: Confidence interval; AUC: Area under the curve.

Table 11 Stability of odds ratio values for each variable verified by bootstrap resampling (1000 iterations)

Variable	Original OR	Bootstrap mean	95%CI lower	95%CI upper	Fluctuation range (%)
Tumor diameter (cm)	3.54	3.48	3.36	3.72	1.7
RAAS hormones (pg/mL)	3.63	3.59	3.42	3.78	1.1
Adrenal medullary hormones (pg/mL)	2.98	2.95	2.83	3.07	1.0
Hypertension duration (years)	20.23	20.01	19.25	21.08	1.1
Surgical approach (laparoscopy/open)	116	15.89	15.23	16.57	0.7

The non-parametric bootstrap (1000 resamples) was used to assess the stability of odds ratios for each variable. Results showed all odds ratios fluctuated by < 5%, indicating stable model parameters. RAAS: Renin-angiotensin-aldosterone system; OR: Odds ratio; CI: Confidence interval.

Table 12 Net benefit of combined predictor vs individual factors at different threshold probabilities

Threshold probability	Combined predictor		Tumor diameter		RAAS hormones		Hypertension duration		Surgical approach
	Net benefit	95%CI	Net benefit	95%CI	Net benefit	95%CI	Net benefit	95%CI	Net benefit	95%CI
0.2	0.18	0.12-0.24	0.10	0.05-0.15	0.08	0.03-0.13	0.12	0.07-0.17	0.09	0.04-0.14
0.3	0.25	0.19-0.31	0.18	0.12-0.24	0.15	0.09-0.21	0.20	0.14-0.26	0.16	0.10-0.22
0.4	0.32	0.26-0.38	0.23	0.17-0.29	0.19	0.13-0.25	0.27	0.21-0.33	0.21	0.15-0.27
0.5	0.37	0.31-0.43	0.27	0.21-0.33	0.22	0.16-0.28	0.32	0.26-0.38	0.25	0.19-0.31
0.6	0.40	0.34-0.46	0.30	0.24-0.36	0.24	0.18-0.30	0.35	0.29-0.41	0.28	0.22-0.34
0.7	0.41	0.35-0.47	0.32	0.26-0.38	0.25	0.19-0.31	0.36	0.30-0.42	0.30	0.24-0.36
0.8	0.39	0.33-0.45	0.31	0.25-0.37	0.24	0.18-0.30	0.34	0.28-0.40	0.29	0.23-0.35

Net benefit calculation: Net Benefit = true positive rate - [false positive rate × (threshold probability/(1 - threshold probability)]. The net benefit is 0 when all patients receive treatment and also 0 when none receive treatment. Within the threshold probability range of 0.2-0.8, the net benefit of the combined predictor was significantly higher than that of any single factor (all P < 0.05, verified by Bootstrap method). CI: Confidence interval; RAAS: Renin-angiotensin-aldosterone system.

Table 13 Comparisons of area under the curve for predictive model of hypertension cure after adrenal cavernous hemangioma surgery and predictive models for other related diseases

Predictive model	AUC (95%CI)	Predictor	Research cohort	P value
This study	0.872 (0.791-0.934)	Tumor diameter/hormone/surgical approach/disease course	ACH (n = 102)
Ge et al[40]	0.996 (0.990-1.000)	Age/duration of hypertension/preoperative systolic blood pressure/preoperative 24-hour urinary catecholamine level/maximum tumor diameter/combined metabolic syndrome/long-term use of multiple antihypertensive drugs before surgery	PHEO (n = 259)	0.0211
Li and Wan[41]	Not available	Duration of hypertension/body mass index/serum potassium level/estimated glomerular filtration rate	PA (n = 1190)	0.0031

¹Delong test, bootstrap = 1000.

AUC: Area under the curve; CI: Confidence interval; ACH: Adrenal cavernous hemangioma; PHEO: Pheochromocytoma; PA: Primary aldosteronism.