Safety of exercise therapy after acute pulmonary embolism

Abstract

Objective:

The role of exercise therapy after acute pulmonary embolism (PE) is unknown. Exercise therapy is safely used after myocardial infarction and chronic obstructive pulmonary disease. The aim of this study was to investigate the safety of exercise therapy after acute PE.

Methods:

We implemented a 3-month exercise program after acute PE. Outcomes were death, bleeding, readmissions, recurrent events, changes in peak VO2 and quality of life (QoL).

Results:

A total of 23 patients were enrolled and received anticoagulation; no adverse events were reported during the exercise period. One death, 1 DVT and 5 readmissions were reported due to non-exercise related reasons. Functional capacity improved as evidenced by an increased peak VO2 at 3 months (+3.9 ± 5.6 mL/kg/min; p = 0.05). Improvement in QoL was observed at 6-months on the functional (+17.0 ± 22.6, p = 0.03) and physical health factor scales (+0.9 ± 4.6, p = 0.03).

Conclusion:

Exercise therapy is feasible and safe in appropriately anticoagulated patients after PE.

Introduction

Pulmonary embolism (PE) causes between 50,000–100,000 deaths per year in the United States.¹ Among patients who survive, 4–5% develop chronic thromboembolic pulmonary hypertension (CTEPH), a severe complication involving persistent dyspnea, increased pulmonary artery pressure and right ventricular (RV) dysfunction, with a high mortality of 40–50% if left untreated.² Among patients who do not develop CTEPH, 40–50% develop post-PE syndrome, defined as persistent dyspnea, impaired exercise capacity and reduced quality of life (QoL).^3,4 Exercise training improves physical capacity in patients with a myocardial infarction, chronic obstructive pulmonary disease, or untreatable CTEPH.^5–7 Recent studies suggest that post-PE syndrome may not be associated with residual clot burden or RV dysfunction,⁸ but rather it may be due to physical deconditioning after acute PE.⁹ Therefore, it is possible that patients with acute PE may also benefit from an exercise intervention. However, information about the safety of such an approach after acute PE is limited. Significant safety concerns regarding exercising patients with acute PE exists among physicians. The aims of this study were to investigate the safety of exercising patients after acute PE, and the role of exercise therapy in recovery of functional capacity.

Methods

Study design

This was a prospective pilot study to evaluate the safety of an exercise intervention in patients with acute PE. The study was approved by the University of Maryland Institutional Review Board, and all patients provided written informed consent. Patients with a diagnosis of acute PE and DVT during the last 28 days were invited to participate. Patients were excluded if they had severe peripheral arterial disease or inferior vena cava thrombosis, had a contraindication to anticoagulation or exercise, were pregnant, had a life expectancy less than 2 years, or had hemodynamic instability at the time of enrollment. Pulmonary embolism was defined by computed tomography-pulmonary angiography (CTPA) as a filling defect in the main, segmental or subsegmental branches of the pulmonary artery; or by a ventilation perfusion (VQ) scan demonstrating a new mismatch defect in the perfusion images. Deep vein thrombosis was defined by venous ultrasonography as the inability to occlude vein walls by compression during B-mode images or lack of flow on Doppler scans.¹⁰ Clinical features, comorbidities, and risk factors for venous thromboembolic disease (VTE) were collected at the baseline visit. In addition, patients underwent exercise testing to evaluate functional capacity, and the 36-item Short Form survey (SF-36) to evaluate quality of life. Then patients started a 3-month program of aerobic exercise training. Exercise testing was repeated at 3 months and the SF-36- was repeated at 6 months. A venous duplex ultrasound of the lower extremities was obtained at each visit. Echocardiography and ventilation perfusion scans were ordered by treating physicians as part of standard clinical care when indicated, and results were extracted by chart review. Adverse events were recorded at each study visit.

PE classification and treatment

Each patient was classified according to the American Heart Association classification into massive, sub-massive and low-risk PE.¹¹ The PE severity index (PESI score) was also calculated at the time of admission.^12,13 Diagnosis and treatment for PE was instituted per standard clinical care and was determined based on preferences of the treating physician. Treatment consisted of anticoagulation alone or in combination with one or more advanced therapies such as systemic thrombolysis, catheter directed therapies, surgical pulmonary embolectomy, extracorporeal membrane oxygenation (ECMO), or the use of an inferior vena cava (IVC) filter.

Assessment of functional capacity

Functional capacity was assessed by the Long-Distance Corridor Walk test¹⁴ conducted in a 20-meter long unobstructed hallway. The test included a 2-minute walk warm-up, where heart rate (HR) and blood pressure were recorded before and after. The pre-determined maximum HR was set at 135 beats per minute (bpm) for patients aged 70 years or older, and otherwise estimated based on age (0.85*(220 – age)).³³ If the patient’s HR exceeded the pre-determined maximum at any point during the warm-up, the 400-m walk was not performed. After the warm-up, patients rested for 60 seconds before starting the 400-m test. Patients were asked to walk 10 forty-meter laps as quickly as possible, and their HR and blood pressure were recorded before and after the test. Peak oxygen consumption (VO_2peak in mL/kg/min) was estimated using the following formula:

$$\begin{gathered} V{O}_{\text{2}\mathit{\text{peak}}}=\text{39}.431-\left(\text{0}.054\times se{c}_{\text{400}m}\right) \\ -\left(\text{0}.031\times SB{P}_{end}\right)+\left(2.832\times LS\right) \\ -\left(0.064\times CF\right) \end{gathered}$$

(sec_400m = patient’s time to walk 400-m in seconds; SBP_end = systolic blood pressure at the end of the test; LS = Long Stride; and CF = Correction Factor, 0 if walking time ≥240 seconds, sec_400m −240 if walking time <240 seconds).¹⁵ Heart-rate reserve (HRR) was estimated using the peak HR in the 400-meter test and the resting HR prior to the start of the warm-up and was calculated using the following formula: HRR = HR_peak − HR_resting.

Exercise training

To monitor compliance, patients were invited to exercise in our hospital gym facility. If travel was not feasible, patients were allowed to exercise at home, for which they were provided a portable HR monitor (Polar^®, Ann Arbor, MI, USA) and instructed to log details of their exercise in a journal. Information from the HR monitor was downloaded and evaluated for compliance at each follow-up visit. Later in the study, updated monitors (Fitbit^®, San Francisco, CA, USA) allowed remote access to the exercise information; a weekly report was downloaded and reviewed to ensure compliance. Exercise training was conducted in 3 phases. Phase 1 was conducted while the subject was hospitalized and lasted up to 7 days or until they were ambulatory. It consisted of three 30-second bouts of upper extremity cycling ergometry (approximately 5W at 75–85 rpm), each followed by a 10-minute rest period performed three times a day.¹⁶ Phase 2 was initiated when patients became ambulatory or when they were discharged, and consisted of weight-bearing aerobic exercise. They were asked to walk 3 times a week, starting at 40–50% of their age predicted HRR for 20 minutes, and progressing to 60–70% for 30 minutes over the subsequent 3 weeks. Phase 3 was initiated 1 month after study-enrollment, concluded at 3 months, and consisted of 30 minutes of aerobic exercise at 60–70% of HRR followed by lower extremity stretching exercises 5 days a week.

Quality of life assessment

The 36-Item Short Form (SF-36) QoL questionnaire¹⁷ was administered at baseline and at 6 months. In addition to the summary score for overall QoL, domain scores were computed from the SF-36 according to the Research and Development corporation (RAND) scoring method.^18,19 Scores were calculated by first converting all items to a 0-to-100 scale, where a higher score denotes a favorable health state. Then, items within the same domain were averaged to create 8 scales: Physical Functioning, Physical Health Problems, Energy/Fatigue, Social Functioning, Emotional Health Problems, Emotional Well-being, Pain, and General Health Perceptions. Composite Physical and Mental Health Factor Summary T-scores were obtained by standardizing the domain scores using age- and gender-adjusted means and standard deviations for the 1988 US general population. A summary T-score of 50 would be the mean domain score for the US general population, and a summary T-score of 60 or 40 in our own cohort would be one standard deviation away from this mean.²⁰

Outcomes

The primary safety outcome was death during or after exercise therapy. Secondary outcomes included any recurrent symptomatic PE diagnosed by CTPA, or a new mismatch defect in the VQ scan in patients with clinical suspicion of a recurrent PE or recurrent acute DVT. Other adverse events recorded were major cardiovascular events and hospital readmissions over the study time period. The effectiveness of exercise therapy was evaluated based on changes from baseline in the 400-m walking time, 400-m peak HR, and estimated peak VO₂ and 3 months; and the difference between baseline and 6-month in QoL RAND and summary T-scores scales.

Statistical analysis

Statistical analyses were performed using SAS 9.3 (Statistical Analysis System, SAS Institute Inc., Cary, NC, USA). Baseline characteristics were reported in percentage, mean and standard deviation (SD) for parametric variables and median and interquartile range (IQR) for non-parametric variables. The median and interquartile range PESI total scores were reported as well as percentages of patients in each sub-category. Right Ventricular function, use of advanced therapies, type of anticoagulation and safety outcomes were reported in percentages. Effectiveness outcomes were reported as follows: for functional capacity and for assessment of the QoL the change from baseline in each metric was calculated by subtracting the baseline value from the value at follow-up. Box-and-whisker plots were used to summarize the data distribution. Subjects who were unable to complete baseline or follow-up visit testing did not contribute data. Paired t-tests were used to assess the significance of changes observed over time. A two-tailed P-value<0.05 was taken as indicating a statistically significant difference.

Results

Patient characteristics

A total of 23 patients with PE were enrolled in the study. The mean age ± SD was 49.2 ± 12.9 years, 43% were female and 52% were Caucasian. The mean BMI was 33.0 ± 9.3 kg/m², hypertension was present in 50%, 27.2% had recent surgery and 17.9% had a diagnosed cancer. The median hospital length of stay was 5 days, with an interquartile range (IQR) of 3–8 days, and a range of 0–20 days. The median time between diagnosis and baseline visit was 18 days, IQR 8–33. (Table I).

Table 1.

Demographic, clinical characteristics and comorbidities of patients.

Patient characteristics (N=23)	N (%)
Age at admission (years, mean ± SD)	49.3 ± 12.9
Female sex	10 (43.5)
Caucasian	12 (52.2)
BMI (kg/m2, mean ± SD)*	33.8 ± 8.8
Hypertension	12 (52.1)
History of Cancer	4 (17.4)
Surgery within 90 days of PE	6 (27.3)
Length of hospital stay (days, median IQR)	5 (3–8)
Time between PE and exercise therapy (days, median IQR)	18 (8–33)

^*BMI n = 22.

SD= Standard Deviation; BMI= Body mass index; PE= Pulmonary Embolism; IQR= Inter-Quartile Range.

Pulmonary embolism characteristics

Five patients were diagnosed with massive, 16 with sub-massive and 2 with low-risk PE. The median PESI score was 77 (IQR 57–109), including 9 (39.1%) patients in class I, 4 (17.3%) in class II, 4 (17.3%) in class III, 3 (13%) in class IV, and 3 (13%) in class V. (Table 2).

Table 2.

Baseline characteristics of pulmonary embolism.

AHA PE Clinical classification, N (%)
Massive PE	5 (21.7)
Sub-Massive PE	16 (69.6)
Low risk PE	2 (8.7)
PESI score, median (IQR)†	77 (57–109)
PESI Category I	9 (39.1)
PESI Category II	4 (17.3)
PESI Category III	4 (17.3)
PESI Category IV	3 (13.0)
PESI Category V	3 (13.0)
Echocardiographic findings, N (%)‡
Normal Right Ventricle	6 (31.6)
Dilated Right Ventricle	5 (26.3)
Dysfunction of right ventricle	1 (5.3)
Dilatation and dysfunction of right ventricle	7 (36.8)
Treatment of Pulmonary Embolism, N (%)
Anticoagulation alone	11 (47.8)
ECMO	2 (8.6)
Catheter directed thrombolysis	6 (26.0)
Surgical pulmonary embolectomy	2 (8.6)
Systemic thrombolysis	1 (4.3)
Surgical pulmonary embolectomy + ECMO	1 (4.3)
IVC filter	9 (39.1)

AHA= American Heart Association; PE= Pulmonary Embolism; PESI= Pulmonary Embolism Severity Index; ECMO= Extracorporeal Membrane Oxygenation; IQR= Inter-Quartile Range; IVC= Inferior Vena Cava Filter.

^†PESI score indicates 30-day mortality risk: Category I (0–65 points, 0–1.6 %), Category II (66–85 points, 1.7–3.5%), Category III (86–105 points, 3.2–7.1%), Category IV (106–125 points, 4–11.4%), Category (V ≥ 125 points, 10–24.5%).

^‡4 patients did not get an echocardiogram at baseline

Treatment of pulmonary embolism

Eleven patients (47.8%) were treated with anticoagulation alone: 20 received unfractionated heparin, 1 argatroban, 1 rivaroxaban and 1 enoxaparin as initial therapy. Additional advanced therapies were used in the remaining patients: 6 had catheter directed thrombolysis, 2 surgical pulmonary embolectomy, 2 ECMO, 1 systemic thrombolysis, and 1 a combination of ECMO and surgical pulmonary embolectomy. Inferior vena cava filters were implanted in 9 patients. Sixteen patients were discharged on a direct oral anticoagulant, 5 received enoxaparin bridging to warfarin, 1 had enoxaparin monotherapy, and 1 received fondaparinux bridging to warfarin.

Safety outcomes

There were no deaths, recurrent PEs, or major adverse cardiovascular events during the 3 months of exercise therapy. There were 7 adverse events in 5 patients subsequent to the exercise intervention, but none were related to exercise. One patient died from sepsis, 1 had a recurrent DVT, 1 required an emergent cholecystectomy, 1 was diagnosed with prostate cancer, 1 had a minor gastrointestinal bleed due to a supratherapeutic anticoagulation, and 1 was admitted with acute pancreatitis.

Effectiveness outcomes

Exercise testing was completed by 15 patients at baseline and by 12 patients at 3 months (10 patients had both baseline and 3 months exercise testing). The reasons why patients were not able to complete exercise testing were: test aborted based on warm-up HR or patient’s discomfort, a temporary contraindication to walk (e.g. ankle fracture) or a missed study visit. In addition, 3 patients withdrew from the study mid-way through the exercise program.

Baseline exercise capacity of participants is described in Table 3. Figure 1 describes the changes in exercise capacity and at 3 months. The time required to walk 400-m improved (i.e. reduced) by 1.1 ± 1.7 minutes; the estimated peak VO₂ improved (increased) by 3.9 ± 5.6 mL/kg/min (p = 0.05); and the peak HR during the 400-m walk changed by +5.2 ± 14. bpm.

Table 3.

Baseline Functional Capacity (Exercise testing and Quality of Life Questionnaire).

Baseline Walking tests (N=17)	Mean ± SD
Time required to walk 400-m (min)	5.9 ± 1.9
Peak VO2 achieved during 400-m walk (mL/kg/min)	16.2 ± 6.2
Peak heart rate during 400-m walk (beats/min)	107.5 ± 18.2
Baseline SF-36 Health Survey RAND Scales (N=20)
Physical Functioning	65.4 ± 30.1
Role Limitations due to Physical Health Problems	12.8 ± 11.4
Energy/Fatigue	67.5 ± 25.7
Social Functioning	69.4 ± 23.5
Role Limitations due to Emotional Problems	18.3 ± 9.6
Emotional Well-Being	73.9 ± 21.8
Pain	65.5 ± 27.5
General Health Perceptions	69.5 ± 18.6
Baseline SF-36 Health Survey Summary T-scores (N=20)
Physical Health factor T-score	41.6 ± 8.5
Mental Health factor T-score	46.0 ± 9.2

RAND= Research and Development corporation; SF-36= 36-Item Short Form Health Survey.

RAND scales: a higher score denotes a more favorable health state, maximum score 100.

Summary T-Scores: have a mean of 50 and a standard deviation of 10, compared to the 1988 US general population.

Figure 1.

Change in 400-m walking time, peak oxygen consumption and peak heart rate from baseline to 3 months.

The middle line of each box represents the median; the x represents the mean; the bottom line of the box represents the 1st quartile; the top line of the box represents the 3rd quartile; the vertical lines extending from the ends of the box represent the minimum value at the bottom and the maximum value at the top; the dots represent outliers, more than 1.5 times the interquartile range (IQR) away from the median.

The QoL questionnaire was completed in 20 of the 23 patients at their baseline visit, 13 patients at 6 months (11 patients had both baseline and 6 months SF-36 questionnaire). The mean RAND scales at baseline are described in Table III. Figure 2 describes changes in QoL questionnaire. All the RAND scales and summary t-scores improved but were statistically significant for the RAND physical functioning scale (+17.0 ± 22.6, p = 0.03) and for the Physical health factor summary T-score (+ 0.9 ± 4.6, p = 0.03).

Figure 2.

Change in Quality of Life measures from baseline to 6-months

Plot (a) shows differences (baseline vs 6-month visit) in the 8 SF-36 RAND Scales, while Plot (b) shows differences in the 2 Summary T-Scores.

The middle line of each box represents the median; the x represents the mean; the bottom line of the box represents the 1^st quartile; the top line of the box represents the 3^rd quartile; the vertical lines extending from the ends of the box represent the minimum value at the bottom and the maximum value at the top; the dots represent outliers, more than 1.5 times the interquartile range (IQR) away from the median.

On baseline echocardiography, 68.4% of patients had an abnormal RV (26.3%, dilated; 5.3%, dysfunction, and 36.8%, both dilation and dysfunction of the RV). Thirteen patients had follow-up echocardiograms, and all showed normal RV systolic function, though 1 patient at 6 months, and 3 patients at 1 year had a persistently dilated RV. Five patients had a VQ scan 1 year after the PE, and only 1 of those had residual thrombosis and changes consistent with CTEPH.

Discussion

This is one of the first prospective studies assessing the safety of exercise therapy in patients with acute PE. The monitored and personalized exercise training program combined with serial testing provided an objective assessment of safety and effectiveness. The results suggest that low-to-moderate intensity exercise therapy can be safely implemented as early as 4 weeks after the acute event. Although almost half the patients had severe PE with a high PESI score (III or more), no exercise-related adverse events were reported during the intervention. No patient developed heart failure or acute cardiopulmonary decompensation secondary to exercise or recurrent PE. No new VTE occurred during the exercise period.

Our study included 21.7% patients with massive and 69.6% with sub-massive PE but did not encounter exercise- or VTE-related death. Other studies have reported an in-hospital mortality of 32.6% and 4.5% in patients with massive and sub-massive PE respectively. At 90 days, mortality has been reported to be even higher, at 41.3% in massive and 12.3% in sub-massive PE. In that report, the cause of death was directly related to the PE in 45% of the massive and 31.6% of the sub-massive PE cases.²¹ Our results provide evidence exercise programs can be safely implemented in patients early after an acute PE.

Our results may also imply a salutary effect of the exercise and/or a general improvement in care of PE patients. Although we did exclude patients that were unable to exercise, so it is possible our patients were in better physical condition than the those included in other studies. Recurrence of VTE is anticipated at a rate of 3% at 3 months²² up to 20% at 12 months.²³ Given the absence of recurrent VTE in our study, it is possible that exercise therapy may have a protective effect for VTE, though larger randomized-controlled trials will be required to support these findings.

Exercise therapy is routinely used to rehabilitate patients with myocardial infarction or COPD.^24,25 Exercise therapy also improves exercise capacity and QoL in patients with established CTEPH,.⁶ Grünig et al. found improvements after implementing an exercise program in patients with pulmonary hypertension, noting a survival of 95% at 2 years.²⁶ Ehlken et al. used exercise therapy in patients with severe pulmonary hypertension and CTEPH, and reported improvements in their oxygen consumption, right heart function and pulmonary resistance.²⁷ Patients with PE present with temporary or permanent pulmonary hypertension and RV dysfunction, therefore it is plausible to hypothesize that they may also benefit from exercise therapy.

Only one study has tested exercise therapy in 5 patients after acute PE, reporting improvement of physical activity without adverse events.²⁸ Their exercise protocol was based on cardiac rehabilitation and weight loss programs²⁹ contrary to our study where the exercise intervention was specifically designed to improve cardio-pulmonary function, based on the individual patient’s functional capacity, and guided by objective measures of patient peak VO₂ values. Our exercise program was derived from the experience of prior studies in related patient populations but was more focused on PE pathophysiology and individualized for each patient, with the goal of maximizing safety of and benefit to the patient.

Our study showed significant improvements in peak VO₂ but, regardless of this improvement, most of the peak VO₂ values were <80% of age-sex predicted values, reflecting the general deconditioning of our patients. Other studies have reported similar deconditioning in patients with acute PE.^8,9 Although our exercise intervention did not increase the peak VO₂ levels to above 80%-predicted values, this was a relatively small cohort that included a large proportion of patients with high PESI scores. The improvements in peak VO₂ observed in our study argue for a larger prospective randomized evaluation of exercise intervention in acute PE.

As can be expected, the mean baseline physical and mental summary T-scores of our patients were lower than in healthy adults. The mean physical health summary T-score improved while the mean mental health summary T- score, which was close to normal at baseline, did not change at follow up. The RAND scales also demonstrated improvements after the exercise intervention (physical functioning, physical health problems, and pain). Exercise has been shown to improve QoL in Parkinson’s disease,³⁰ Alzheimer’s disease,³¹ cancer,³² diabetes mellitus³³, knee osteoarthritis,³⁴ among other chronic medical conditions. One study demonstrate an improvement in QoL after exercise in patients with CTEPH.⁶ There is an increased risk of VTE among cancer patients³⁵ and other chronic diseases, therefore exercise therapy may play an important role in the physical recovery of patients with these co-morbidities after an acute PE.

Limitations

The improvements in physical function and QoL are encouraging in this pilot study, but more reassuring is the fact that exercise intervention after acute PE appears to be safe. An important limitation is that not all the patients were able to complete their baseline and follow-up exercise testing and questionnaires, making the assessment of effectiveness challenging. Our study does not assess for long-term outcomes and echocardiographic evaluation was not performed routinely; therefore, we cannot report on the incidence of CTEPH or post-PE syndrome. Other limitations include a small sample and challenges regarding patient retention. Our patients had limited availability of time due to work-related activities, or had multiple comorbidities making it difficult for them to integrate time for exercise with their medical appointments and other activities.

Conclusion

Initiation of exercise therapy as early as four weeks after acute PE is feasible and safe in appropriately anticoagulated patients. Exercise intervention may be associated with significant improvements in physical function and QoL measures. Whether these improvements translate into a reduction in subsequent post-PE syndrome and CTPEH will require a randomized-controlled study.