class: RWH_bg_title white # .black[To PLEX or not to PLEX?] ### Management of acute vasculitis after PEXIVAS .footnote[ .RWH_footer_bold[ SpR Teaching ] ] .RWH_footnote_right[ .RWH_footer_bold[ 25th Nov 2021 ] ] ??? Created with [xaringan](https://github.com/yihui/xaringan). --- class: center, middle, inverse *NB some of the slides in this presentation have been redacted to remove patient identifiable information* --- # Overview - a case and then straw poll - background to PLEX - PEXIVAS - meta-analyses and guidelines - summary and another straw poll ??? Session inspired by 1) this case and was always worried about this coming up after PEXIVAS; 2) being asked to review the Walsh BMJ guidelines. --- # Hypothetical case 34 year-old male. Previously healthy. Presents with one week of malaise: purpuric rash; crampy, central abdominal pain, symmetrical large-joint arthralgias, stuffy nose, red eyes, hot spellsand a small volume haemoptysis (< 1 tsp). Non-smoker. No cocaine / travel / other weird stuff. You review him on day 3 of hospital admission. HR 110; ABP 120/70; SaO2 99% air; T37.5. He looks flushed; classic vasculitic rash on lower limbs,red eyes, crepitations in mid-zones. -- ### Investigations Hb 105 (fallen from 130); WCC 15; plt 187. Cr 61 mcM (baseline 55). CRP 215. Blood +++; protein +; uPCR 70. PR3 ANCA 109. A few blood and urine cultures sent so far; none positive. ??? Urgent CT shows reasonably impressive dense, bilateral consolidation – consistent with diffuse alveolar haemorrhage. How would you manage? Would you give PLEX? --- # Images < redacted: CXR and CT images > --- class: center, middle, inverse # History of PLEX --- # ANCA are pathogenic   .RWH_footnote_right[Xiao (JCI, 2002)] --- # ANCA are pathogenic  .RWH_footnote_right[Xiao (JCI, 2002)] --- # History of PLEX - animal data and [case-report](https://pubmed.ncbi.nlm.nih.gov/15806479/) in pregnancy support pathogenic role for (MPO) ANCA - small trials in [anti-GBM disease](https://pubmed.ncbi.nlm.nih.gov/56532/) (n = 7) and also "idiopathic" [RPGN](https://pubmed.ncbi.nlm.nih.gov/1745027/) (n = 48; 1991) or ["crescentic GN"](https://pubmed.ncbi.nlm.nih.gov/1519607/) (n = 32; 1992) - mainly AAV <br> - MEPEX: - SCr > 500 mcM (n = 137) - 3g IVMP *vs.* PLEX - 25 % reduction in ESKD at 12 m; no change in mortality <br> - guidelines: SCr > 500 mcM or DAH - in practice: used more liberally, often informed by biopsy etc. ??? --- class: center, middle, inverse # PEXIVAS ---  ??? - Positive MPO or PR3 antibodies (by ELISA) - Kidney involvement (with an estimated glomerular filtration rate of less than 50 ml/min/1.73m^2) - either Bx or urinary sediment - OR pulmonary involvement (with diffuse alveolar hemorrhage) ---  ---  ---  --- # Criticisms of PEXIVAS - open-label (proabably not a big deal) - IVMP for all - no biopsy (but isn't that realistic?) - no ANCA titres / confirmation of "dose" of PLEX - alb replacement - very hard endpoints - equipoise (NB lower than expected event rates) - powerwered to detect large effect (80% power to detect HR 0.64) - ...and small numbers in subgroups (n = 61 for severe DAH) ??? A fantastic trial! Arguably set up as an "effectiveness" rather than "efficacy" trail with real-world inclusion criteria. --- class: center, middle, inverse # Meta-analyses and guidelines --- # Meta-analyses <br> <br> <table> <thead> <tr> <th style="text-align:left;"> outcome (1 yr) </th> <th style="text-align:left;"> all </th> <th style="text-align:left;"> GN </th> <th style="text-align:left;"> DAH </th> </tr> </thead> <tbody> <tr> <td style="text-align:left;"> death </td> <td style="text-align:left;"> ~0.9 (0.7 -- 1.3) </td> <td style="text-align:left;"> ~1.0 (0.6 -- 1.8) </td> <td style="text-align:left;"> ~0.9 (0.7 -- 1.3) </td> </tr> <tr> <td style="text-align:left;"> ESKD </td> <td style="text-align:left;"> ~0.6 (0.4 -- 0.9) </td> <td style="text-align:left;"> ~0.6 (0.3 -- 0.8) </td> <td style="text-align:left;"> NA </td> </tr> <tr> <td style="text-align:left;"> serious infection </td> <td style="text-align:left;"> ~1.3 (1.0 -- 1.6) </td> <td style="text-align:left;"> ~1.2 (1.0 -- 1.4) </td> <td style="text-align:left;"> ~1.3 (1.0 -- 1.6) </td> </tr> </tbody> </table> .footnote[From [Cochrane review](https://doi.org/10.1002%2F14651858.CD003232.pub4), [ACR guidelines](https://onlinelibrary.wiley.com/doi/full/10.1002/acr.24634) and BMJ guidelines (in revision)] ??? C = cochrane B = BMJ guidelines (in review) A = American guidelines renal vasc C mort 12 m 1.04 (0.57-1.92) C ESKD 12 m 0.45 (0.29-0.72) C serious infection 1.26 (1.03-1.54) A mort 1.15(0.77-1.70) A ESKD 0.61 (0.42-0.90) A serious infection 1.19 (0.99-1.42) all vasc B mort 12 m 0.93 (0.66-1.31) B ESKD 12 m 0.58 (0.38-0.89) B serious infection 12 m 1.27 (1.03-1.56) DAH without kidney B mort 12 m 0.93 (0.66-1.31) B serious infection 12 m 1.27 (1.03-1.56) A mort 0.95(0.70-1.30) --- # Guidelines **Cochrane**: - PLEX effective in renal vasculitis at reducing ESKD at 3 and 12 months (NNT = 5; low-quality evidence) <br> **ACR**: - do NOT use PLEX in renal AAV (conditional - consider if "high risk" of ESKD) - do NOT use PLEX in DAH (conditional - consider as salvage / rescue Rx) <br> **BMJ CPG** (in review): - use PLEX if SCr > 300 mcM (weak recommendation) - do NOT use in DAH (weak recommendation) ??? Cochrane is review rather than guidelines --- # Crunching the numbers .RWH_bold[$$NNT = 1/(PEER \times RRR)$$] *Assume a RRR of 0.4 (for RR of 0.6) that is constant across range of baseline risk*\* *Assume baseline risk correlated with Cr*\* <br> <br> .grey[$$Cr < 200 \hspace{0.5em}| \hspace{2em} 1/(0.002 \times 0.40) = 1250$$] .green[$$Cr > 200 \hspace{0.5em}| \hspace{2em} 1/(0.05 \times 0.40) = 50$$] .orange[$$Cr > 300 \hspace{0.5em}| \hspace{2em} 1/(0.10 \times 0.4) = 25$$] .red[$$Cr > 500 \hspace{0.5em}| \hspace{2em} 1/(0.40 \times 0.4) = 6$$] .footnote[\**unlikely to be correct*] ??? Where does the Cr of 300 mcM come from? - see [KI reports abstract](https://www.sciencedirect.com/science/article/pii/S2468024921003235?via%3Dihub). https://www.bmj.com/content/312/7028/426 - PEER = patients estimated (i.e. baseline) event rate - ARR = CER - AER - AER = CER * RR - RRR = 1 - RR - NNT = 1/ARR Therefore: - ARR = PEER - PEER.RR = PEER(1-RR) = PEER.RRR - NTT = 1/(PEER*RRR) https://www.ncbi.nlm.nih.gov/pmc/articles/PMC81587/ https://www.ncbi.nlm.nih.gov/books/NBK426103/ RRR likely to be broadly consistent between groups for therapies designed to prevent / slow progression of disease. RRR likely to be higher as baseline risk increases for therapies designed to reverse disease. --- class: center, middle, inverse # Summary --- class: split-two white .column.RWH_bg1[.content[ ### .black[.center[For PLEX]] - biological plausibility - if it works for the kidney it should work for the lung - PEXIVAS imperfect / not representative (IVMP etc.) - practicalities: are we ever sure this is not anti-GBM? ]] .column.RWH_bg2[.content[ ### .black[.center[Against PLEX]] - effect likely to be small - line, infections, hospital stay, faff... - kidney biopsy ]] --- # When to use PLEX? - default position is NOT to use unless... - pulmonary-renal syndrome - severe RPGN but minimal chronic damage - high risk of ESKD (more than just SCr) - severe PH (see [NEJM](https://www.nejm.org/doi/full/10.1056/NEJMe1917490?query=recirc_curatedRelated_article) & [AJKD](https://pubmed.ncbi.nlm.nih.gov/32277949/) editorialists) - higher ANCA titres - double-positive for anti-GBM --- # Hypothetical case Return to case: would you use PLEX? ??? So who would give PLEX now?