This retrospective study was approved by the institutional review board. The clinical data of patients with entire-limb DVT receiving CDT treatment in our department from January 2018 to December 2022 were collected.

The inclusion criteria were as follows: (1) unilateral entire-limb DVT; (2) first episode of DVT; (3) a symptomatic period less than 14 days; (4) receiving CDT treatment; (5) a minimum follow-up period of 12 months. The exclusion criteria were: (1) symptom duration longer than 14 days or history of DVT; (2) bilateral DVT; (3) combined PMT; (4) CDT with both antegrade and retrograde accesses. Finally, 121 patients were included in the study.

Once diagnosed, patients were administered low-molecular-weight heparin at a dosage of 100 IU/kg per 12 h. The interventional procedures were conducted by two radiologists, each with 2–5 years of practice experience. It was a standard practice to implant a retrievable inferior vena cava filter (IVCF) before starting CDT (17).

For the antegrade approach (AGA), the ipsilateral popliteal vein was selected. After IVCF placement, the patient was placed in the prone position. Then, under local anesthesia and venographic guidance, a 5 French sheath was inserted. Subsequently, a multiple-holes infusion catheter (4 French Uni*Fuse Infusion Catheter, AngioDynamics, USA, or 5 French Infusion Catheter, COOK, USA) was inserted, with its tip placed into the inferior vena cava. The retrograde approach (RGA) was performed through the contralateral common femoral vein. An Omini, Cobra, or Simon catheter was used to explore the entrance of the ipsilateral common iliac vein, and then the infusion catheter was introduced to the distal popliteal vein.

The selection of the access approach was mainly based on the following clinical factors: (1)

Vascular anatomical conditions: The anatomical structure of the patient's blood vessels was evaluated by preoperative imaging examinations. If there were anatomical abnormalities in the ipsilateral popliteal vein (such as stenosis, tortuosity, or variation) that made it difficult to puncture and insert the catheter, the RGA approach was selected.

Treatment efficacy was evaluated using a modified SVS scoring system (10). Scores were defined as follows: 0 for a patent vein without thrombus; 1 for partial occlusion with thrombus present; 2 for complete occlusion. The clot burden reduction rate was calculated as (pre-CDT limb scores—post-CDT scores)/pre-CDT score × 100%. Thrombus burden reduction was categorized into grades (10): Grade I for a reduction of less than 50%; Grade II for partial thrombolysis, corresponding to a 50%–95% reduction; Grade III for a 95%–100% reduction or complete lysis. Inflow patency was classified as follows (11): Inflow was classified as “poor” when the inflow vessels small caliber inflow vessels with many collateral vessels or even the trunk was undetected. “Great” indicated no post-thrombotic changes in adequate caliber inflow vessels. “Fair” indicated mild post-thrombotic changes. Specifically, the caliber of the blood vessel was reduced but the blood flow was still basically smooth, and there was no obvious blood flow stasis. “Great” and “fair” patency inflows were regarded as “good” inflow, while “poor” patency inflow was categorized as “bad” inflow. Venogram image interpretation was performed by three radiologists, each with over 5 years of experience, who were blinded to the intervention.

All patients were evaluated at 1, 3, 6, and 12 months post-lysis, with annual follow-up visits thereafter. Endpoint events included patient mortality and loss to follow-up at the last recorded visit. PTS was assessed using the Villalta scale (18).

All continuous variables were expressed as means and standard deviations (SDs), while categorical variables were presented as frequencies or percentages. Categorical variables were compared using the Chi-square test, and continuous variables were compared using Student's t-test. Statistically significant differences were defined at P < 0.05 (two-tailed). Kaplan–Meier analysis was used to calculate the free-PTS time, the log-rank test was used to assess the association between relevant variables and free-PTS. The Cox's hazard multivariate regression model was employed to explore potential independent prognostic factors. Propensity-score matched (PSM) analysis, adjusted for “gender” and “further treatment”, was performed to balance the baseline characteristics between the two inflow groups. The two groups were matched at a 1:2 ratio using the nearest-neighbor method with a caliper width of 0.1. All statistical analyses were conducted using SPSS software (version 17). Univariate and multivariate analyses of PTS incidence after PSM were performed using R (4.3.1).

The demographics and risk factors were summarized in Table 1. The mean age of the patients was 56.1 ± 12.5 years, 38.8% (47/121) were male. Seventy-four cases (61.2%) involved the left limb. The leading risk factor was May-Thurner syndrome, present in 59.5% (72/121).

Thirty-four patients underwent CDT via the AGA approach, while 87 received the RGA method (Table 2). The AGA group had significantly better inflow patency than the RGA group. For the popliteal vein, the AGA group had a much higher rate of Grade III patency (76.5%) compared to the RGA group (4.6%). For the femoral vein, the AGA group also showed superior patency rates (Grade III: 67.6%, Grade II: 23.5%, Grade I: 8.8%) compared to the RGA group (Grade III: 1.1%, Grade II: 44.8%, Grade I: 54.0%) (P < 0.0001). Overall, the AGA group showed a higher rate of great inflow patency (67.6%) and lower rates of fair (23.5%) and poor (8.8%) inflow patency compared to the RGA group (1.1% in great, 43.7% in fair, and 55.2% in poor) (P < 0.0001).

Table 3 showed the differences in variuos characteristics between the “good” and “bad” inflow groups. Most patients with “bad” inflow (94.1%) received the retrograde approach, while only three patients (5.9%) with “bad” inflow received the antegrade approach. The “good” inflow group had a lower incidence and less severity PTS compared to the “bad” inflow group (P < 0.0001). Since there were differences in gender and further treatment between the two groups, PSM analysis was performed to balance these two potential confounding factors. As a result, 97 well-matched pairs were obtained, with 59 patients in the “good” inflow group and 38 in the “bad” inflow group. After PSM, the “good” inflow group still had a lower incidence and less severe PTS compared to the “bad” inflow group (P < 0.0001).

After PSM, the impact of the access approach and inflow patency on the incidence and severity of PTS were assessed (Table 4). There was no significant difference in the incidence, severity and two00 year cumulative incidence of PTS between the two access groups (Figure 1A). However, the “good” inflow patency group had a lower incidence of PTS (35.6%) than the “bad” inflow patency group (78.9%) (P < 0.0001). Moreover, the severity of PTS in the “good” inflow patency group was milder than that in the “bad” inflow patency group. A significant difference was observed in the two- year cumulative incidence of PTS between the two groups (P < 0.0001) (Figure 1B).

After PSM, a total of 51 patients (52.6%) developed PTS during the two-year follow-up. Univariate analysis showed that further treatment had a protective effect [odd ratio (OR) 0.16, 95% CI: 0.09–0.29, P < 0.0001]. Further analysis revealed that balloon dilatation alone (OR: 0.21, 95% CI: 0.03–1.53, P = 0.123) and balloon dilatation + stent implantation (OR: 0.16, 95% CI: 0.09–0.29, P < 0.0001) were associated with a reduced risk of PTS. In contrast, “bad” inflow patency was identified as a risk factor for PTS occurrence (OR: 3.55, 95% CI: 2.02–6.25, P < 0.0001). Additionally, poor inflow (OR: 6.6, 95% CI: 2.31–18.83, P = 0.0001) was associated with an increased PTS risk (Table 5).

Multivariate regression analysis also demonstrated that further treatment had a protective effect (OR: 0.17, 95% CI: 0.1–0.3, P < 0.0001), with balloon dilatation alone (OR: 0.17, 95% CI: 0.02–1.3, P = 0.088) and balloon dilatation + stent implantation (OR: 0.16, 95% CI: 0.09–0.29, P < 0.0001). Conversely, “bad” inflow patency increased the risk of PTS (OR: 3.41, 95% CI: 1.94–6, P < 0.0001), and poor patency (OR: 6.88, 95% CI: 2.4–19.72, P < 0.0001) (Table 6).