Venous thromboembolism (VTE), including deep vein thrombosis (DVT), superficial venous thrombosis, and pulmonary embolism (PE), has been considered an important medical and social problem. One of the significant risk factors for the development of VTE is recent hospital treatment, especially accompanied by major surgery [1-7].
Therefore, stratification of inpatients according to their individual risk of VTE, and provision of appropriate preventive measures, are crucial during the management of inpatients [8,9].
Currently, the Caprini score (2005 version) [10] is considered the most validated individual risk assessment model for postoperative VTE. It was verified in about 3 million surgical patients worldwide, and showed a strong correlation between the total score and the frequency of symptomatic VTE events [11,12].
In a patient with a score of 9 or greater, the frequency of symptomatic VTE was as high as 11%, and those with a score of 11 or greater had symptomatic VTE events in 59%, despite standard prophylaxis [13 ]. These patients may be classified as an extremely high-risk group requiring improvements in the VTE preventive protocol.
Intermittent pneumatic compression (IPC) after major surgery is as effective as pharmacoprophylaxis in reducing the risk of VTE, but does not affect the rate of bleeding complications [14], and is generally suggested as an alternative to anticoagulants in those patients. who have an increased bleeding risk [15].
However, the combination of CNI and prophylactic anticoagulation, also known as pharmaco-mechanical modality, is more effective than mechanical and pharmacological approaches alone, providing an additional risk reduction of DVT and PE of 48%–61% [16]. .
The CNI in Surgical Patients at Extremely High Risk of Venous Thromboembolism (IPC SUPER) trial was a 2-center, randomized, open-label, investigator-initiated clinical trial with one evaluator blinded to the results. of efficacy, which was based on the hypothesis that combining IPC with standard preventive measures [above-knee anti-embolic elastic compression stockings with an ankle pressure of 18 to 21 mmHg (ECM) and standard doses of low-molecular-weight heparin (LMWH)], could reduce the incidence of postoperative VTE in extremely high-risk patients.
| Methods |
> Study design
The study was carried out at 2 sites in Moscow (Russian Federation). The trial protocol was designed by the authors, approved by the institutional review board of the Pirogov Russian National Research Medical University, and registered at ClinicalTrials.gov (NCT03044574). Data monitoring and quality checks were performed by the institutional review board.
> Patients
Patients were eligible for inclusion in the trial if they were over 40 years of age, required major surgery, had a high risk of postoperative VTE, Caprini score ≥ 11, and provided informed consent.
The Caprini score (2005 version) [10] was evaluated twice: at baseline, before inclusion, and before discharge or after death. The maximum score was considered for the analysis.
Patients with an assumed score ≥ 11 at baseline, but no less than 9 actual points, were included. It was allowed to assume the presence of some additional risk factors such as malignancy, congestive heart failure, chronic venous disorder, prolonged bed rest, among others, which should have been confirmed within the hospital treatment.
> Randomization and masking
Patients were randomly assigned in a 1:1 ratio to receive CNI plus MCE and LMWH (CNI group), or standard prophylaxis only (control group). Randomization was performed by the investigator and controlled by the Data Monitoring Committee independently for both clinical centers. It was based on the number of hospital medical records: those with the last even digit were assigned to the experimental group, and
those with the last odd digit to the control group; if the last digit was zero, the penultimate digit was used. In order to reduce the bias of deliberate non-inclusion of patients with the highest risk of VTE in the CNI group, all ineligible subjects were repeatedly evaluated, according to the exclusion criteria, by the Data Monitoring Committee. . After group assignment, the patient was assigned an individual code that was used in all study documentation.
The allocation list was kept by the researcher. The blinded expert performing the Doppler ultrasound (DUS) did not have access to the original medical record or allocation list; He only used the patients’ individual code for identification. Furthermore, to achieve unawareness, most EDs were performed in a separate room, away from the patient’s bed. If it was impossible to transfer the patient, the ED was performed in bed, but the IPC device was removed before the expert visit.
> Procedures
All randomized patients received standard VTE prophylaxis according to the National Guidelines [17], which are complementary to those of the American College of Chest Physicians (ACCP) [15], using CSM and LMWH injections.
The use of stockings began before the operation and were used throughout the hospital treatment period. After discharge, patients were recommended to use ECM during the night and during long periods of bed rest for 1 month.
Subcutaneous enoxaparin 40 mg was started once daily, 12 hours before the operation (elective surgery without high risk of bleeding), within 12 hours after the intervention (emergency surgery, procedure with high risk of bleeding). , or within 5 days (highest risk of bleeding), and continued until discharge, but for no less than 7 days. Prolonged prophylaxis with LMWH in the outpatient setting was not prespecified and depended on surgeon preference.
CNI was performed with the Kendall SCD 700 Series Controller device , and with the Kendall SCD thigh-length Comfort Sleevesk (Cardinal Health). The cuffs contain 3 chambers in the lower calf, upper calf and thigh, which sequentially compress the leg, starting with a pressure of 45 mm Hg at the ankle.
Using the vascular refill detection system, the pump measures the venous refill time and adjusts the interval between compressions, providing a better hemodynamic response [18,19].
CNI was started in the operating room before the incision, or in the intensive care unit within 12 hours after surgery, and was applied throughout the period of the patient’s hospitalization: throughout the time inside of the intensive care unit and at the time of bed rest, except for a nightly interval of 6 hours (0 am to 6 am) in the surgical service, which corresponded to a duration of 18 hours per day, excluding the period of wandering. CNI was discontinued when DVT or PE was suspected or confirmed, or when other contraindications appeared.
> Measurements
Regular clinical and ultrasound evaluation was performed during the inpatient treatment period. Two certified expert sonographers performed all EDs in a blinded manner (outside the patient’s bed and/or after removing the compressor and stockings for IPC), with evaluation of the veins in the calves and thighs being mandatory, and optional of the iliac veins and the inferior vena cava, if there was suspicion of thrombosis.
The ED study was performed at baseline (the day before elective surgery, or within 2 hours before/12 hours after emergency surgery), and then every 3-5 days until discharge or lethal outcome. It could be performed urgently if any clinical suspicion of VTE appeared. The main criterion for venous patency was its compressibility; In addition, the color flow mode could be used to reveal the blood flow in the target vein.
Venous thrombosis was defined as the partial or complete impossibility of compressibility of a venous segment at any location, and/or the absence of blood flow in the proximal deep veins. PD verification was performed emergently if a clinical suspicion arose. An autopsy was performed on all deceased patients. After discharge, patients were evaluated at 30 and 180 days after surgery with ED and/or clinical examination and/or telephone call, for new VTE events.
No specific compliance measures were designed for the use of ECM and LMWH injections. Compliance with the IPC device was measured as follows: each day, the investigator (weekdays) or on-call surgeon (weekends and holidays) visited the patient 5 times and documented whether he was confined to his bed, whether or not the cuffs were applied to his legs, and that the pump was working.
These assessments were recorded on a specifically designed compliance sheet, and compliance was calculated as the proportion of all visits documented with CNI applied and functioning, compared to the total number of visits in which the patient was found in bed. .
> Results
The primary outcome was asymptomatic lower limb venous thrombosis, detected by repeated ED every 3-5 days after surgery, until hospital discharge or death.
Secondary outcomes included those obtained during the inpatient period (i.e., isolated calf muscle DVT; proximal DVT; symptomatic PE; fatal PE; total VTE events; postoperative mortality; leg skin injury; combination of major bleeding [20] and clinically relevant non-major bleeding; and compliance with CNI), and those obtained on days 30 and 180 after surgery, during outpatient follow-up (i.e., combination of symptomatic and asymptomatic venous thrombosis lower extremity and symptomatic PE; VTE-related mortality; and non-VTE mortality).
> Statistical analysis
The prevalence of postoperative asymptomatic venous thrombosis at the height of standard prophylaxis with CSM and LMWH was assumed to be 30% [13,21]. Taking into account the fact that pharmaco-mechanical prophylaxis should provide an additional VTE risk reduction of 60% [16], an absolute difference of 18% was assumed.
Using Fisher’s exact test for the primary statistical analysis, with a power of 80% and a type I error of 0.05, the sample size was estimated at 180 subjects (90 + 90 in each group). Assuming 10% dropout, the sample size was calculated at 200 subjects. Considering a prespecified subgroup analysis for patients with early and late onset of LMWH, the sample was doubled.
The primary statistical analysis was designed on an intention-to-treat basis and a 2-tailed Fisher’s exact test was used to compare the primary outcome between the 2 groups. The primary outcome was reported as an absolute number, percentage, absolute risk difference, and relative risk with 95% confidence interval (95% CI). A secondary statistical analysis of the primary outcome was performed using an unadjusted Cox proportional hazards model.
Data were censored at the time of VTE confirmation, death, or hospital discharge, and reported as a hazard rate with 95% CI. Kaplan-Meier curves were used to compare the distribution of time to event.
Cox proportional hazards models were adjusted for care center, surgery profile, type of surgery, emergency of surgery, type of anesthesia, surgical radicality, malignant disorder, Caprini score, and time of start of injections. LMWH.
A similar analysis for the per-protocol population, excluding patients with gross protocol violations and premature discontinuation of at least 1 component of the combination prophylaxis (ie, MCE, CNI, or LMWH), was performed for the primary outcome.
Analyzes of secondary outcomes, subgroup analyses, and sensitivity analyzes were performed according to the statistical analysis plan. The loss of primary outcome data was not anticipated in the hospitalization scenario; therefore, a missing data analysis was not planned, and no adjustments were made for multiple comparisons and imputation of missing values.
The 95% confidence intervals were not adjusted for multiplicity and, therefore, inferences drawn from these intervals may not be reproducible. All analyzes were 2-sided and were carried out with the IBM SPSS Statistics program, v.26.
| Results |
> Patients
From February 2017 to October 2018, a total of 812 patients were evaluated for selection. Of these, 407 patients underwent randomization and were included in the intention-to-treat analysis (204 patients were assigned to the CNI group and 203 to the control group). The groups were comparable with respect to the main demographic and clinical characteristics, and well balanced according to the Caprini score.
> Interventions
The ECMs were applied in all patients before surgery. LMWH was started before surgery in 205 patients (50.4%), within 12 hours after surgery in 168 (41.3%), and within 5 days after surgery in 34 patients ( 5.3%). IPC was applied before surgery in 202 patients (49.6%), and within 12 hours after surgery in 205 (50.4%).
Protocol deviations occurred in 7 patients in the CNI group, who prematurely discontinued LMWH injections due to bleeding events (of these, 4 patients discontinued CNI and ECM due to concomitant skin injury). and in 11 patients in the control group, who discontinued LMWH due to bleeding events (of them, 1 subject also discontinued CSM due to skin injury).
No other deviations from the protocol were reported. After discharge, 117 (57.4%) of 204 patients with CNI, and 92 (45.3%) of 203 patients in the control group, reported continued use of CNI. Long-term use of enoxaparin was recommended in 57 (27.9%) of 204 patients in the CNI group, and in 48 (23.6%) of 203 patients in the control group. None of the observed subjects continued VTE prophylaxis for more than 1 month.
Primary result
Scheduled EDs revealed asymptomatic lower extremity venous thrombosis in 1 (0.5%) of 204 patients who received CNI in addition to CEM and LMWH, compared with 34 (16.7%) of 203 patients who received CNI alone. standard prophylaxis (relative risk 0.03; 95% CI: 0.01-0.21; P < 0.001).
No additional EDs were performed due to clinical suspicion of venous thrombosis. Findings from the secondary analyzes of the primary outcome in the intention-to-treat and per-protocol populations were consistent with those from the primary analysis. High heterogeneity in outcome was observed in the subgroup analysis.
Adjunctive CNI provided maximum benefits in patients with a Caprini score of 11 or more, who underwent elective abdominal surgery under general anesthesia, which was intended to cure a malignant disease, and which was radical.
> Secondary results
The significant difference in secondary outcomes was observed only in calf muscle DVT and total number of VTE events, in favor of CNI. The risk of proximal DVT and symptomatic PE decreased in the CNI group, but not significantly.
Five of 203 patients (2.5%) in the control group developed PE, and 3 of them (1.5%) died. However, no difference was observed in PE-related mortality or total postoperative mortality.
Leg skin injury was reported in 25 (12.3%) of 204 patients who received CNI, and in 15 (7.4%) of 203 patients who received standard prophylaxis only, with no significant difference between the groups.
Major and clinically relevant non-major bleeding were observed in 7 (3.4%) patients in the CNI group, and in 11 (5.4%) patients in the control group, without any difference between the groups. Individual compliance with the CNI varied between 45.5% and 100%, with a mean value of 95.2 ± 9.9.
At 30 and 180 days after surgery, no new VTE events were detected and no difference was observed in VTE-related or non-VTE-related mortality.
> Sensitivity analysis
The sensitivity analysis did not show significant differences compared to the primary statistical analysis.
| Discussion |
The IPC SUPER trial was the first study to evaluate the efficacy and safety of CNI adjunctive to standard VTE prophylaxis in extremely high-risk patients, taking into account the Caprini score. Patients with a score of 11 or more received maximum benefit with the pharmacomechanical modality.
The authors believe that this approach can be used to identify patients who will gain the most benefit from adjunctive CNI. Conversely, ignoring individual VTE risk assessment can lead to controversial results using the pharmacomechanical approach, as in the recently published PREVENT trial, which found no benefit with adjunctive CNI in critically ill patients [22].
The other condition to achieve better results with CNI is to exclude calf DVT at baseline. Omitting venous evaluation of the calf in the ED before applying CNI may lead to extension of the thrombus into the proximal veins, which was possibly observed in the PREVENT trial. Compliance with the CNI is another crucial issue.
By providing 6-hour overnight compression-free intervals and removing the cuffs before any attempt to get out of bed, while maintaining 18 hours of compression daily, excluding the period of ambulation, a very high compliance with CNI was achieved, exceeding previously published thresholds of 40% [23-26].
It is important that not only the application of the cuffs but also the confirmed operation of the pump was recorded, which is not always the same [25]. All of these factors, accompanied by the use of a single, consistent and well-established CNI device, formed the basis for the favorable results of this trial.
The danger associated with CNI was well balanced with the benefits. The incidence of skin lesions (12.3%) in this study is slightly higher than that in the CLOTS 3 (3.1%) and PREVENT (2.9%) trials [22,27], without any significant difference. between the groups. This may be related to the concomitant use of CNI and ECM. However, only skin necrosis (2.0%) required discontinuation of ECM and IPC, while in cases of hyperemia and blisters (10.3%), mechanical prophylaxis was continued without worsening of symptoms. .
There is no robust evidence that the complementary use of compression stockings can increase the clinical or hemodynamic efficacy of IPC [28]. However, individual data suggest a further reduction in the risk of VTE after major surgery [29].
In the present study, the use of ECM was crucial due to the presence of a 6-hour CNI-free nighttime interval, when patients were still protected against VTE by elastic compression. Furthermore, this combination may be essential in situations of low compliance with the CNI.
This study is limited by its open design, absence of centralized randomization, absence of placebo, high heterogeneity between surgical procedures, types of intervention, types of anesthesia, and lack of non-abdominal surgery and non-malignant disease. Furthermore, it failed to include enough patients with delayed initiation of LMWH injection.
A decrease in the preliminary Caprini score assessed at baseline compared to the final score calculated at the end of hospitalization was observed in some patients, mainly due to early ambulation, and could result in a lower incidence of venous thrombosis. than expected.
In conclusion, among patients with a Caprini score of 11 or greater, receiving standard prophylaxis with CSM and LMWH, adjunctive CNI resulted in a significantly lower incidence of asymptomatic lower limb venous thrombosis, relative to standard thromboprophylaxis alone.
