LUST but do not fall in Love with POCUS yet?

More information
The LUST Trial in Hemodialysis: Can Lung Ultrasound Guide Decongestive Therapy?
The LUST Trial in Hemodialysis: Can Lung Ultrasound Guide Decongestive Therapy?

In this commentary, three POCUS experts provide some background and interpretation for the LUST trial.

LUST but do not fall in Love with POCUS yet?
LUST but do not fall in Love with POCUS yet?

This week, we will discuss the LUST trial, to see if use of a POCUS changes your focus enough to help improve clinical outcomes

LUST trial.png
LUST Visual Abstract

For additional reading about POCUS, check out this special issue of ACKD. All articles are free for 50 days (about January end 2022). Follow this link for full text access.

Introduction

Fluid overload is an important indicator of mortality and morbidity in end-stage kidney disease (ESKD), but the ideal body weight is a dialysis holy grail that we have never clearly pinpointed. One approach is to ‘probe’ and gradually decrease the dry weight until symptoms appear (as was done in the DRIP trial, Agarwal et al Hypertension 2009). Other fancier technologies have been tried: bioimpedance, blood volume monitoring, biochemical measures (eg BNP) and measurement of volume with other tests such as IVC diameter. None of them have progressed beyond being interesting toys, sadly enough. Another aspect that makes reducing dry weight difficult is the development of adverse effects. Cramps and hypotension doesn’t make anyone happy - and their presence has nothing to do with the achievement of dry weight, rather with the rate and speed of ultrafiltration. Along that line, rate of ultrafiltration also has become a dichotomized quality measure, with misguided attempts to keep it at less than 13 ml/kg/hour. Is there a way out of this, to safely lower dry weight and improve volume status? Longer dialysis and more frequent dialysis aficionados aside, what can we do within the confines of conventional hemodialysis?

Lung ultrasound is considered a promising tool for assessment of pulmonary edema.  An A-line pattern is termed as normal lung parenchyma, which consists of horizontally oriented reflections of the hyperechoic pleural line pleural depth. With  progressive pulmonary edema, there is a shift of A-line pattern to a vertically oriented B-line pattern, which is a discrete laser-like hyperechoic line that radiates to the edge of the ultrasound field and moves with respiration. Thus, an ultrasound lung scan can be a very fast and simple technique to measure extravascular lung water. 

In the emergency settings, use of an ultrasound, with POCUS or point-of-care ultrasound, has been reported to improve diagnosis of acute pulmonary edema (Al Deeb et al, Acad Emerg Med, 2014). In a study of 75 dialysis patients, use of an ultrasound was helpful in detecting signs of volume overload in symptomatic and asymptomatic dialysis patients (Mallamaci et al, JACC imaging 2010).  

But if a patient is asymptomatic, what is the utility of knowing there is excessive fluid in their lungs? Can it improve any outcomes? This week’s randomized controlled trial, LUST (a randomized multicenter trial on a Lung UltraSound-guided Treatment strategy in patients on chronic hemodialysis with high cardiovascular risk) was aimed to evaluate the treatment guided by extravascular lung water measurements by an ultrasound scan to decrease all cause mortality, decompensated heart failure, and non-fatal myocardial infarction in patients on hemodialysis compared to standard care based on clinical signs and symptoms. 

The Study

Study design

Open label, randomized controlled clinical trial (Trial reg: NCT02310061).  An Internet-based platform was created and an open call was made to the European Cardiovascular and Renal Medicine (EURECAm) working group of the European Renal Association and the European Society of Dialysis and Transplantation (ERA-EDTA). A total of 24 units expressed their interest and finally, 18 centers entered the trial. 

Interventions

Patients were randomly assigned to a lung ultrasound-guided treatment protocol or standard clinical care. Lung water assessment was done by a chest US performed by a nephrologist before and after hemodialysis session and the predialysis lung scan was used to titrate the ultrafiltration (UF). Patients with moderate to severe lung congestion (>15 US- B lines) were followed with once a week lung US until the treatment goal (<15 US-B lines) was achieved and then continued the assessment once a month throughout the study.  

The treatment goal was achieved by using intensification of ultrafiltration either by lengthening of dialysis sessions or extra dialysis sessions. If unable to achieve treatment goal within the first 3-4 weeks or developed intolerance to UF, drug treatment (carvedilol or fosinopril addition) was adjusted according to Kidney Disease Improving Global Outcome (KDIGO) guideline. 

The control group was managed with standard care including optimization of fluid volume control based on clinical criteria and the use of carvedilol, angiotensinogen-converting enzyme inhibitors, or -sartans. Patients in both groups were followed up at baseline visits, at 6, 12, and 24months by a cardiologist blind to this study with lung ultrasound. Patients were also evaluated with echocardiography during the same visit. 

Inclusion criteria:

Patients older than 18 years, who were on hemodialysis for more than 3 months were included in the trial who have a high-risk cardiovascular profile:

  1. A history of myocardial infarction with or without ST elevation or unstable angina

  2. An acute coronary syndrome is documented by ECG and cardiac troponin

  3. Stable angina pectoris with documented coronary artery disease by coronary angiography or echocardiography

  4. Heart failure with dyspnea class III-IV according to New York Heart Association classification

Exclusion criteria:

  1. Cancer 

  2. Advanced non cardiac disease or comorbidity (e.g., end-stage liver failure)

  3. Active infections

  4. Relevant intercurrent illness

  5. Inadequate lung scanning and echocardiographic studies

Outcome measures

The primary outcome of the study was a MACE outcome, a composite of all cause mortality, nonfatal myocardial infarction, or decompensated heart failure. 

The secondary outcomes were:

  1. All-cause and cardiovascular hospitalization

  2. Changes in echocardiographic parameters including LV mass index, left atrial volume index, ejection fraction, and the early diastolic transmitral flow velocity to early diastolic mitral annular tissue velocity 

Statistical analysis

All the patients were followed up for 24 months after randomization. A Cox regression hazard ratio (HR) and Kaplan-Meier method was used to analyze the effect of the intervention on the primary, composite endpoint and all-cause and cardiovascular hospitalizations (secondary endpoints). The effect of the allocation arm on the number of US-B lines and on echocardiographic parameters was investigated by linear mixed models. A secondary (post hoc) analysis was done to see the effect of the intervention on heart failure was analyzed by the zero-inflated binomial regression, which is a method for modeling count variables with excessive zeros and for overdispersed count based variables (Jahn-Aimermacher 2008). A total of 500 patients were planned to recruit to provide 80% power to detect a 33% risk reduction (45% in intervention and 30% in control) in the primary endpoint with an assumed type 1 error rate of 0.05, 2-sided. 


Funding:  The LUST trial was funded by a grant from the ERA-EDTA

Results 

Though the plan was to enroll 500, only a total of 363 patients could be enrolled into the trial. The authors blame the nephrologists, who thought the technique was complex and time-consuming, and despite a longer recruitment period (4.5 years) this was all they could do. 

This left 183 in the lung US-guided therapy and 180 in the standard care. Interestingly enough, the figure 1 doesn’t follow CONSORT guidelines, and we don’t know the number of patients screened, eligible with reasons for exclusion, we only see the randomized N, as shown below. 

Most of the study participants were receiving pharmacologic therapy for hypertension and cardiovascular comorbidities. Demographics and clinical characteristics of the study participants were depicted in table 1. About 70% were men, with 40% diabetes, and about 15% active smokers. The median dialysis vintage was about 4-5 years. Since it was part of the eligibility, about 70-75% had a history of ACS or angina and about 38-47% (in each arm) had a history of heart failure. 

Change in Volume

The study participants in the lung US group had a total of 4103 predialysis and equal number of post dialysis scans by the attending nephrologist with an average of 24±17 for each patient in either pre or post hemodialysis measurements. In the 4 prefixed visits, patient in the intervention arm had a decline in the number of US-B line (baseline 15; 95% CI 12-19, study end: 9, 95% CI 5-12) compared to control arm [baseline 16; (95% CI 13-20) to 30 (95% CI 20-39)]. Similarly, the number of patients who achieved the treatment target (<15 US-B lines) was higher in the active arm (78% vs 56%; p= <0.001), see figure below. However, blood pressure and weight of the study participants from both arms did not change over the period of trial. There was no difference in the echocardiographic parameters as well (left atrial volume, LV mass, EF) all in Supplementary table S3.

Safety related to intervention

The lung US based therapy was safe and risk of dialysis hypotension was less in the active arm and vascular access related problems, intradialysis and extra-dialysis arrhythmia did not differ between two groups (see some adverse events reported in table 2). 

Study outcomes

The composite endpoint occurring in 62 patients in the active arm and 71 patients in the control arm during a mean follow-up of 1.5 years years was not statistically significant (HR 0.88; 95%CI 0.63-1.24, p= 0.47), see figure 3 below. Death occurred in 28% of patients in the lung US-guided therapy group compared to 33% in the usual care group which was also not statistically significant. 

The subgroup analysis did not demonstrate any notable differences (though note the difference between men and women). 

Post hoc outcomes

The total number of repeated episodes of decompensated HF and repeated cardiovascular events in both arms showed a significant decrease in the lung US arm compared to the usual care arm (Table 3). 

Other Outcomes

Secondary analysis of patient-reported outcome (depression and Standard Form 36 Quality of Life Questionnaire, SF 36) and two additional questionnaires (Berlin Questionnaire and Karnofsky score) also did not differ.

Discussion

In summary, the use of lung ultrasound did improve the subsequent appearance on lung ultrasound, though without a change in weight or BP. It also  did not have a significant reduction in MACE, but there was a decrease in recurrent hospitalization for decompensated HF in a post hoc analysis. 

The major limitation was the limited enrolment: if nephrologists who wanted to participate in this clinical trial found it too onerous, what hope do we have of translating this into clinical practice?

Secondly, the change in a post hoc decompensated HF recurrent event outcome is difficult to fathom in the absence of change in weight, change inBP, or change in any echocardiographic parameters. 

We also have a very sparse adverse event table, which does not even include overall hospitalization rates. Dialysis hypotension decreased in the intervention arm (though we do not know how this was defined), but we do not know granular details about BP, or about other parameters such as cramps.  

How were these ‘recurrent decompensated HF’ events decided? They were not adjudicated by an independent committee - but were entered by the local investigator. Keep in mind that this was an open label trial as well. 

Should we now embrace POCUS, as POCUS enthusiasts would have us do? It does look like an interesting tool, but this trial fails to convince us that universal usage of POCUS will change clinical outcomes. A simple probe method to decrease dry weight meticulously may be better than the latest shiny toy. 

Conclusion

In hemodialysis patients who are at high risk of cardiovascular risk, a lung US guided therapy may safely reduce lung congestion but was not effective compared to usual care in improving the primary endpoint of the trial. This trial highlights that the road for universal incorporation of POCUS in clinical practice is bumpy. 

Summary prepared by Abdul Qader, MD 

Pediatric Nephrologist, Dhaka, Bangladesh

 NSMC Intern, Class of 2021