Understanding possible differences between haemodialysis and haemodiafiltration which affect survival in end-stage kidney disease dialysis patients: the STITCHED study

Background

The mortality of end-stage kidney dialysis patients remains high, with the 5-year survival rate being lower than that of more common solid organ cancers. Cardiovascular diseases are the most common cause of mortality, followed by infections. In addition to a 10-fold increase in the risk of stroke, elderly dialysis patients who start dialysis are reported to have an increased risk of progressive vascular dementia. Standard dialysis treatments are based on the diffusive clearance of uraemic toxins, with high-flux haemodialysis (HD) being the current standard of treatment. Recently, online haemodiafiltration (OL-HDF) has been introduced into clinical practice. This adds convective clearance, thus increasing the spectrum of uraemic solutes that can be removed from the blood. However, three major clinical trials were completed in 2017/2018 and failed to show a difference in patient outcomes between these two dialysis modalities. However, these trials did not target a specific amount of convective clearance. Another trial targeting a high dose of convective clearance reported better patient survival with high-volume OL-HDF. To investigate the amount of convective clearance required to improve patient outcomes, the UK High-Volume Haemodiafiltration vs. High-Flux Haemodialysis Registry Trial (H4RT) study to compare high-volume OL-HDF with standard HD was proposed.

The HR4T trial enrolled end-stage kidney dialysis patients in the UK into a randomised prospective trial comparing high-volume OL-HDF and HD treatments.¹ The trial collected data on patient demographics, comorbidities, and medications. Longitudinal laboratory results and dialysis session data were also collected by linking to the UK Renal Registry. The H4RT study endpoints included a comparison of hospital admissions due to cardiovascular and infection-related events, patient survival and mortality data, patient-reported outcomes, and health care economics. The primary outcome was a composite of non-cancer mortality or hospital admissions with a cardiovascular event or infection during a 3-year follow-up period.

Study design

The H4RT study was designed to investigate the amount of convective clearance required to improve patient outcomes. However, it was not set up to determine why HDF could provide improved patient outcomes. The STITCHED study (Study To Investigate the Change in Hypotensive Episodes during Dialysis) was designed as a mechanistic study nested within the main H4RT study. Recruiting patients already entered into the H4RT study, randomised to either standard high-flux HD or high-volume OL-HDF, and followed for 2 years.

Previous studies comparing high-flux HD and OL-HDF would suggest that mortality, particularly cardiovascular mortality and stroke, is more likely to be reduced with OL-HDF.^2–8 The H4RT study was designed to deliver high-volume OL-HDF convective exchanges, and these greater exchange volumes are expected to increase the thermal losses during treatments. Previous studies have demonstrated that negative thermal balance increases sympathetic nervous system activity and reduces the risk of intra-dialytic hypotension.^9–11 Fewer hypotensive episodes during treatment coupled with more stable systolic blood pressure control would be expected to reduce intradialytic hypotension and potential ischaemic damage to the heart and brain. A more stable blood pressure would also allow more reliable fluid removal, so by reducing the risks of volume overload and leading to reduced patient morbidity and mortality.^12–15 Thus, one possible mechanism that may explain the potential benefit of OL-HDF is more stable blood pressure. The European Dialysis and Transplant Association recommended ambulatory blood pressure monitoring for patients with end-stage kidney disease treated with dialysis.¹⁶ As such, we planned to monitor the blood pressure during both dialysis sessions and the interval between dialysis sessions. Additionally, both observational data and experimental evidence suggest that OL-HDF may reduce or prevent an increase in arterial vascular stiffness. Advances in ambulatory blood pressure monitoring also permit the measurement of vascular stiffness using pulse wave velocity.¹⁷ Thus, we planned to determine whether there was any improvement or reduction in vascular stiffness over time with OL-HDF compared with standard HD. As some of the uraemic toxins that accumulate in patients with end-stage kidney disease have been implicated in reducing arterial compliance and reactivity, we planned to determine whether there are changes in uraemic toxins with OL-HDF treatments.

The risk of stroke, particularly ischaemic stroke, has been reported to increase in end-stage kidney dialysis patients. Similarly, elderly HD patients have been reported to be at a greater risk of developing vascular dementia. As episodes of intradialytic hypotension are reported to reduce cerebral blood flow during dialysis sessions, we planned to assess whether differences in blood pressure stability between HD and OL-HDF led to differences in cognitive function over time using the Montreal Cognitive Assessment (MoCA) test.¹⁸ Similarly, dialysis patients gain fluid between dialysis sessions. Hypotensive episodes during dialysis stop fluid removal, thus increasing the risk of patients leaving the dialysis volume overloaded. To investigate whether more stable intradialytic blood pressure with OL-HDF has beneficial effects on the heart, we were to use echocardiography to assess structural and functional changes^19,20 and measure volume status with bioelectrical impedance measurements.^21,22

Study objectives²³

Primary objective:

Secondary objectives:

Ethics

We planned to recruit adult kidney dialysis patients who had provided written informed consent and were enrolled into the H4RT trial. These patients were then given STITCHED study information and subsequently then asked to provide written informed consent to then be recruited into the STITCHED sub-study. In keeping with the Helsinki accord only patients who could provide written informed consent were recruited into STITCHED. The STITCHED study was approved by the East Midlands – Leicester Central Research Ethics committee 18/EM/0213.

Study methods

We planned to capture the following data, had the study not been halted.

Equality, diversity and inclusion

Patient groups included the patient advisory board for the UK Renal Registry and the Royal Free Hospital Kidney Patients’ Association. These committees comprised patients of different sexes, religions, and ethnicities. Patient recruitment at the main site, the Royal Free Hospital, included patients of different sexes, religions, and ethnic backgrounds.

Patient and public involvement in the study

Results

The study was halted before any relevant data was obtained and as such statistical analysis could not be undertaken.

Discussion

The study was halted before any relevant findings were obtained due to the withdrawal of funding from the NIHR. The study began in the summer of 2020, immediately after COVID-19 reached London. The Royal Free Hospital London was the primary site for STITCHED, and the hospital initially stopped all clinical research. This meant that the research team could not visit kidney dialysis centres, and consequently, all patient recruitment was halted. This practice of closing down clinical research activities was mirrored in our other participating centres. Our renal research team was seconded to other posts within the hospital or to COVID-19 studies. When clinical research resumed, COVID-19 and cancer studies were prioritised. Research activities were restarted in a stepwise manner, with renal research in the last tier to be reactivated. We and our other centres that had agreed to participate in STITCHED found that when we were allowed to recommence renal clinical research studies, we had fewer renal research staff. The reduction in staffing was due to a combination of staff leaving or moving to other posts, as there had been no clinical renal research, long-term staff sickness due to COVID-19, and no new staff recruitment to replace the staff members who had left.

As STITCHED fell behind in the study milestones, a review was conducted by the NIHR. A number of problems were highlighted and solutions proposed by the review committee, which included more local ownership of the development of STITCHED-specific clinical research forms. The University College London surgical trial group was subcontracted to produce and test a new series of study case forms. The costs for this work were paid independently from internal sources and not from the STITCHED monies. New milestone targets were set by the trial review committee, but despite meeting these targets, the NIHR decided to withdraw funding for the SITCHED study, and the study was halted.

When funding was withdrawn for STITCHED, 98 patients had consented to participate in the study. As requested in the study closure notification, all blood samples which had been collected for batch analysis were destroyed and not analysed. Only a minority of patients had completed baseline ambulatory blood pressure and pulse wave velocity measurements at that time, so there were insufficient data available for any informative analysis.

Conclusion and future directions

The STITCHED study was halted before meaningful results were obtained to draw any conclusions. The H4RT study subsequently completed recruitment and is currently in the final analysis stage. Recently, a European trial comparing high-volume OL-HDF and standard high-flux HD (CONVINCE) reported some of their initial findings, with higher survival for those treated with high-volume OL-HDF.²⁴ Although both trials appear to compare the same dialysis modes, there are differences in patient recruitment, target convective exchange volumes, and primary outcomes. However, neither trial was designed to answer the question: “if high volume OL-HDF does improve survival for kidney dialysis patients, what is the reason?” This was the rationale behind the STITCHED study.

The STITCHED study was designed to compare changes in blood pressure during dialysis treatments, the inter-dialytic interval, and the clearance of uraemic toxins and biomarkers. By comparing these changes in arterial stiffness, cardiac modelling, and cognitive function, STITCHED was designed to determine whether the improvement in survival was due to improved blood pressure control, removal of uraemic toxins, and changes in the thermal or sodium and calcium balance between diffusive and convective clearance. The reason for improved survival is important in deciding the future of dialysis treatments. For example, there is now a new design of dialysers that allows for a similar spectrum of uraemic toxins to be removed by internal convection compared to HDF with greater clearance compared to diffusion.

The economic healthcare costs of high-volume OL-HDF compared to high-flux haemodialysis were reported to be greater in the CONVINCE study group. Although this was partly due to longer patient survival, the sessional cost of an individual dialysis treatment was greater with OL-HDF; however, this cost was considered acceptable when considering the benefits to patient healthcare. Although dialysis machines for OL-HDF require three pumps, a conventional haemodialysis machine for diffusive clearance only requires two pumps; therefore, it is less complex, has fewer parts, and may potentially have lower capital and running costs. On the other hand, if survival is related to improved blood pressure control during treatment, this may reflect the differences in thermal balance between convective and diffusive modes, and dialysers and dialysis machines need to be developed to maximise convective exchange so that all patients could benefit from high-volume OL-HDF. Again, if OL-HDF leads to better blood pressure control during the inter-dialytic interval, bioimpedance-measured volume control, and sodium and calcium balance between the different modalities, this may alter the choice and composition of dialysis fluids.

Although H4RT and CONVINCE should provide more evidence as to whether OL-HDF improves patient outcomes and is cost-effective compared to high-flux HD, an understanding of the mechanisms underlying these outcomes will remain unexplored without further research.

Role of the sponsor

The sponsor provided funding for the study but did not materially influence the conduct of the study until it was closed down, or influence any interpretation of study findings.

Data sharing statement

As the study was stopped prematurely, no significant data were collected for sharing. Data protection was provided by the UCL, according to the GDPR guidelines.

Disclaimer

This study was funded by the National Institute for Health and Care Research (NIHR) under the EME program (NETSCC 17/61/04). The views expressed are those of the author and not necessarily those of the NIHR or Department of Health and Social Care. This project was carried out between May 2019, when regulatory approval was obtained, and November 2021. The program examined the report at the time of submission to assess completeness against the stated aims. These reports did not involve a peer-review process https://doi.org/10.1186/ISRCTN74893205.