Monday, June 12, 2017

The evidence behind the use of plasmapheresis in glomerular diseases

Therapeutic plasma exchange (PLEX) is a type of therapy where a patient’s plasma volume is removed over a period of a few hours, through a process of centrifugation of blood with subsequent separation of its constituents, and replaced by different types of colloid fluids, most commonly Albumin or Fresh Frozen Plasma (FFP). Its most frequent use in nephrology is for certain glomerulopathies such as ANCA-associated vasculitis, anti-GBM disease, recurrence of idiopathic FSGS post-transplant and atypical HUS. The data with regards to PLEX use for these indications is mostly historical and based on mechanistic concepts which led to multiple observational studies and reports claiming efficacy. There are very few randomized trials directly comparing PLEX to other therapies and a review of the data (see table summary of KDIGO recommendations) has really brought me a new perspective on what justifies our use of PLEX for renal disease.
Table: KDIGO recommendations on PLEX use for glomerular disease

KDIGO recommendation
Grade
Anti-GBM disease
We recommend initiating immunosuppression with cyclophosphamide and corticosteroids plus plasmapheresis in all patients with anti-GBM GN except those who are dialysis-dependent at presentation and have 100% crescents in an adequate biopsy sample, and do not have pulmonary hemorrhage
1B
Start treatment for anti-GBM GN without delay once the diagnosis is confirmed. If the diagnosis is highly suspected, it would be appropriate to begin high-dose corticosteroids and plasmapheresis while waiting for confirmation
Not graded
ANCA-associated vasculitis
We recommend the addition of plasmapheresis for patients requiring dialysis or with rapidly increasing SCr
1C
We suggest the addition of plasmapheresis for patients with diffuse pulmonary hemorrhage
2C
We suggest the addition of plasmapheresis for patients with overlap syndrome of ANCA vasculitis and
anti-GBM GN, according to proposed criteria and regimen for anti-GBM GN
2D
Post-transplant FSGS
We suggest plasma exchange if a biopsy shows minimal change disease or FSGS in those with primary FSGS as their primary kidney disease
2D
Atypical HUS
Not mentioned in KDIGO guidelines
Adapted from: Kidney Disease: Improving Global Outcomes (KDIGO) Transplant Work Group. KDIGO clinical practice guideline for the care of kidney transplant recipients. American journal of transplantation: official journal of the American Society of Transplantation and the American Society of Transplant Surgeons. 2009 Nov;9:S1; and Kasiske BL, Zeier MG, Chapman JR, Craig JC, Ekberg H, Garvey CA, Green MD, Jha V, Josephson MA, Kiberd BA, Kreis HA. KDIGO clinical practice guideline for the care of kidney transplant recipients: a summary. Kidney international. 2010 Feb 2;77(4):299-311.


Anti-GBM disease
Anti-GBM disease has always been regarded as a sine qua non indication for PLEX.  I still remember my first teachings on anti-GBM disease in medical school which were were it causes lung hemorrhage, it causes acute kidney injury and it is treated with plasmapheresis. When reviewing the literature, I was surprised to find only a single randomized trial on the use of PLEX in anti-GBM disease. This a study published in 1985 in Medicine by Johnson et al where they randomized 17 (yes, only 17!) patients with biopsy and serology proven anti-GBM disease to either Prednisone + PO Cyclophosphamide (N=9) vs PLEX + Prednisone + PO Cyclophosphamide (N=8). Patients at baseline were not equally matched as the patients in the conventional group had higher serum creatinine (SCr) at start of therapy and had more severe disease on biopsy. Indeed, 5/8 biopsies available had >70% gloms with crescents in the conventional group vs only 1/7 in the PLEX group had > 50% gloms with crescents. They did find that more patients were dialysis dependent at the end of the study in the conventional group (6/9) vs the PLEX group (2/8). Four patients in the PLEX group had improvement in their renal function vs only 1 in the conventional group. These patients who had improvement were the ones with lower SCr at presentation. Eight patients had pulmonary hemorrhage (4 in each group) and all these episodes were treated with IV Methylprednisone pulse and responded promptly. There were only 3 deaths in all, 1 in conventional and 2 in PLEX group. Anti-GBM titers became undetectable much more quickly with PLEX, after about 2 months. Overall, it is somewhat surprising that what we consider such a strong indication for PLEX is supported by only 1 randomized trial showing improved renal survival with PLEX, where the groups were unevenly matched. What is also important to consider with anti-GBM disease is the possible futility of treatment in patients with most severe disease. A review of anti-GBM disease in the UK from 1980-1984 by Savage et al looked at outcomes for 108 patients. There were 69 patients who were dialysis dependant on presentation. At 8 weeks, none were off dialysis (51 on dialysis and 18 dead). Out of 12 who presented with a SCr>600umol/L, only 1 had improvement in renal function (other 11 either on dialysis or dead). Another British study published in 2001 in the Annals of Medicine by Levy et al retrospectively looked at all anti-GBM disease treated at the Hammersmith hospital in since 1975. They had 71 patients, 39 of which were dialysis-dependant on presentation. Only 2 were off dialysis at 1 year follow-up. When looking at outcomes based on biopsies, they found that 23% of patients with >50% crescents survived off dialysis and that 3 patients survived off dialysis despite >70% crescents. However, no patients with 100% crescents recovered renal function. Together, these results certainly seem to suggest that patients with severe anti-GBM disease presenting dialysis dependent have very little chance of recovery and may not benefit at all from immunosuppressive therapy and PLEX. While a trial of treatment may still be indicated, I believe a rapid re-assessment of the patient’s condition and need for continued immunosuppressive therapy is indicated given the high infectious risks with treatment. Patients with 100% crescents on an adequate biopsy are very unlikely to get any benefit and should probably just be managed conservatively.

ANCA-associated vasculitisAnother frequent indication for PLEX in glomerular disease is ANCA-associated vasculitis and thankfully there is a bit more data to guide us here. The best data comes from the MEPEX trial by Jayne et al published in JASN in 2007 where 137 patients with biopsy/serology proven ANCA vasculitis and SCr > 500umol/L were randomized to either PLEX (7 exchanges in 14 days) or IV methylprednisolone (1g IV daily x 3), both in combination with oral Cyclophosphamide and Prednisone. Patients with severe lung hemorrhage requiring mechanical ventilation were excluded. Just over 2/3 of patients were dialysis-dependent on presentation. They found that treatment with PLEX had a better renal recovery (alive and off dialysis and SCr<500umol/L) at 3 months than IV steroids (70% vs 49% respectively, P=0.02). The HR for ESRD at 12 months for PLEX vs IV steroids was 0.47 (0.24-0.91, P=0.03). Survival however was not significantly different (19 deaths in PLEX group vs 16 in IV steroids group) and most deaths were due to infections (19), lung hemorrhage (6) or cardiovascular disease (4) and very few due to vasculitis. A sub-study of the MEPEX trial by Van Wingaarden et al found that for patients requiring dialysis on presentation and with severe tubular atrophy on biopsy, the point at which patients would get more benefit for renal survival from treatment over risk of death from treatment was when they had 18% or more normal glomeruli for IV steroids group as opposed to only 2% normal glomeruli for PLEX. This suggests that patients with most severe disease are more likely to reap renal benefit from treatment when they are given PLEX. In 2013, Walsh et al published the long term follow-up data from the MEPEX trial and found that the short term benefit seen in MEPEX was lost. Indeed, for patients treated with PLEX there was no significant improvement in the composite of ESRD or death (HR 0.81, 0.53-1.23; P=0.32) nor in the outcome of ESRD (0.64, 0.40-1.05; P=0.08). At final follow-up, half the patients died and 2/3 were either dead or on dialysis, reaffirming the poor prognosis of severe ANCA vasculitis. A meta-analysis of all randomized trials looking at PLEX for ANCA vasculitis by Walsh et al in 2011 found PLEX to be associated with a 20% risk reduction in the composite of ESRD-death (HR 0.80, 0.65-0.99) and a 36% reduction in ESRD (HR 0.64, 0.47-0.88) but no effect on death (RR 1.0, 0.71-1.42). The authors did warn though that overall most trials were small, none of them individually found a significant result for the composite of ESRD-death and they had notable methodological flaws such as randomization concealment was only performed in 4/9 trials and the methods of concealment weren’t described in any. Hopefully, the PEXIVAS study which is now nearing completion and should be presented at the upcoming ASN Kidney Week 2017 (hopefully as a late-breaking trial) will help clarify the role for PLEX in ANCA vasculitis. For now, it would seem that PLEX is indicated for patients presenting with severe renal failure due to ANCA-vasculitis as it improves renal survival, without a mortality benefit though.

Post-transplant FSGS
It is sad to say that there are unfortunately no other prospective trials studying the role of PLEX for glomerular disease. While the use of PLEX for recurring idiopathic FSGS post-transplant is recommended (the rationale being the removal of some as of yet unidentified pathogenic plasma permeability factor), the data is purely observational. A review of 77 case-reports and case-series, totalling 423 patients with recurrence of FSGS post renal transplant, published in BMC Nephrology in 2016 showed that overall 71% of patients achieved complete or partial remission. Factors most associated with response were male sex and starting treatment within 2 weeks of recurrence. However, the lack of control group prevents us from establishing a clear benefit from PLEX itself. Also, the treatment regimens were extremely varied and it is unclear exactly how much PLEX, for how long should it be given and which replacement fluid to use. Finally, the frequent use of PLEX in these patients who are already on immune-suppressing drugs may predispose to even more infections given the removal of Immunoglobulins by PLEX.

Atypical HUS
Similar to FSGS, the use of PLEX for atypical HUS is based purely on observational data (case reports and case series). The rationale behind its use for this indication certainly makes sense by removing defective complement factors and replacing properly functioning complement factors to halt the overactivity of the alternative pathway. Patients diagnosed with TMA are often started on PLEX promptly while awaiting results of diagnostic testing (Shiga toxin E. Coli cultures, ADAMSTS13 level and complement pathway factor levels and mutations). If diagnostic tests suggest an alternative complement pathway disease, it will be maintained until Eculizumab is available. Unfortunately, while most observational studies suggest an initial response around 60% to PLEX, this is mostly a hematologic response and patients will often become dialysis dependent.

Conclusion
Overall the evidence to support the use of PLEX for the treatment of glomerular diseases is not great. While anti-GBM disease and lung hemorrhage are considered some of the strongest indications for PLEX, this is not firmly supported by good prospective data. The best evidence is for its use is in severe ANCA-vasculitis and with the upcoming PEXIVAS study results hopefully this will further help us in our decision making. I think it is great that such a large study such as PEXIVAS (over 700 patients) for such a rare disease has been able to come to completion and this highlights the importance of proper collaboration to conduct prospective studies in GN. Hopefully this will inspire us to continue to strive for well-designed studies to guide us in the treatment of our patients.

David Massicotte-Azarniouch
Nephrology Fellow, University of Ottawa

No comments: