The Tweetup has become an yearly event at the American College of Rheumatology Annual Meeting, and the use of Twitter at medical conferences is now so ubiquitous that it is indexed and searchable on Symplur, which includes the meeting hashtag for this year: #ACR15.
”We’ve reached a point where social media is now part of the professional workflow. While it’s a minority that understand and leverage these tools, the ones who are onboard are helping reshape the image of our organizations and our profession. Those of us creating, curating and conversing in the great wide open will continue to benefit from our public presence.”
The goal of the Tweetup is always to give people a chance to connect on the topic of social media in rheumatology and medicine. Every year, it leads me to new projects such as #RheumJC, the recently formed Rheumatology Twitter-Based Journal Club. I hope others find similar opportunities from attending.
Looking forward to seeing everyone at the Tweetup!
The project points out that nephrology has the fewest randomized controlled clinical trials of any subspecialty in medicine, and hopes to bring interest to areas where gaps in knowledge are most lacking.
In the area where nephrology intersects with rheumatology, we have actually been quite fortunate. The two conditions in this area which carry the highest burden are ANCA-associated vasculitis and lupus nephritis. For ANCA-associated vasculitis, we have been fortunate to have recently well performed RCTs regarding the use of rituximab for induction of remission and maintenance as well as treatment of relapses.
We have also made recent advances in our understanding of therapies for induction of lupus nephritis, with studies looking at the role of tacrolimus as monotherapy or as part of a multitarget therapy regimen along with MMF.
Despite this, one of the biggest questions rheumatologists and nephrologists have regarding lupus nephritis is: why the heck doesn’t rituximab seem to work in bigger RCTs?
My entry for DreamRCT: RoRo-LuN (Role of Rituximab only in Lupus Nephritis):
Previous studies looking at the role of rituximab for the treatment of lupus nephritis have been highly criticized for poor design. Initial data from the RITUXILUP group (rituximab and IV methylprednisolone on days 1 and 15 with background MMF but no oral steroids) have been extremely promising, but many patients cannot tolerate MMF, and the role of rituximab as monotherapy given over 6 month intervals will remain uncertain. RoRo-LuN would randomize patients with biopsy proven class III or IV lupus nephritis to one of three arms to be followed over 2 years, with the primary endpoint to be renal remission defined as normal creatine or return to baseline creatinine, inactive urinary sediment, and urine protein/creatinine ≤0.5. Group 1: rituximab without oral steroids (rituximab 1 g on weeks 0 & 2, 26 & 28, 52 & 54, 78 & 80, IV methylprednisolone 1 g on weeks 0 and 2); group 2: same as group 1 but with the addition of tapering oral steroids over 6 months; group 3: standard therapy (initial pulse steroids, MMF, tapering oral prednisone).
Unfortunately, despite clinical experience by clinicians and promising reports in many smaller studies, larger RCTs have not shown effectiveness of rituximab against lupus and lupus nephritis. However, these studies have been extensively criticized for their trial design as the reason for failures.
The first of the larger RCTs evaluating the role of rituximab in lupus was the EXPLORER trial, which looked at patients that did not have renal involvement. This trial randomized 257 patients with moderate-severe SLE (on one background immunosuppresive [methotrexate, azathioprine, or MMF], with 57% of patients corticosteroid deponent) to rituximab infusions or placebo at a ratio of 2:1 on days 1, 15, 168, and 182. The primary endpoint was the effect of achieving and maintaining clinical response at week 52, assessed using BILAG, was not met. The EXPLORER trial was criticized for having a small number of participants, confounding background immunosuppressives, and for questions regarding the ability of BILAG to detect a meaningful clinical response.
Following this, the LUNAR trial looked at 144 patients with class III or IV lupus nephritis being treated with MMF and corticosteroids, and randomized them 1:1 to receive rituximab or placebo on days 1, 15, 168, and 182. The primary endpoint was a 20% superior renal response in the rituximab group at week 52. Again, the primary endpoint was not met, although overall response rates were 56.9% in the rituximab group compared to 45.8% in placebo. Failure to meet the primary endpoint was thought due to faulty design due to background immunosuppressives confounding any benefit of rituximab, as well as being underpowered.
Interestingly, 78-week follow up data to the LUNAR trial did suggest that rituximab had a longer term effect, with improved proportion of patients who had remission of proteinuria and fewer patients who required additional immunosuppression:
”In LUNAR, the exploratory data demonstrated that at week 52, the difference (10%) in the proportion of patients with 50% reduction in proteinuria favored rituximab treatment; the difference increased to 17% at week 78 (P = 0.04).”
”The other compelling suggestion of a benefit is the finding that significantly (P < 0.01) fewer patients in the rituximab group required cyclophosphamide for worsening disease, and more achieved a renal domain BILAG C score, and this was sustained up to 78 weeks.”
The currently ongoing RITUXILUP trial hopes to avoid oral steroids entirely in patients with class III/IV or V lupus nephritis while determining whether rituximab is an effective therapy when added to maintenance MMF. In this regimen, patients are given two doses of rituximab (1 g) and methylprednisolone (500 mg) on days 1 and 15, and maintenance treatment with MMF, compared to standard therapy using initial IV methylprednisolone, MMF, and tapering oral steroids. This trial is powered to show superiority non-inferiority of the Rituxilup regimen, with patients followed for at least 2 years, and should be completed in 2018.
“Life can be much broader, once you discover one simple fact, and that is that everything around you that you call life was made up by people that were no smarter than you. And you can change it, you can influence it, you can build your own things that other people can use. Once you learn that, you’ll never be the same again.” — Steve Jobs
Here’s a quick roundup of things I’ve been up to around the web, and my other active projects.
More details to come in future posts, but a short summary of what’s coming for ACR 2015:
I’ll be speaking at on the new TechMed track at ACR15 in San Francisco (along with Dr. Suleman Bhana (@DrBhana) on Sunday, November 8, 2015 from 2:30–4:00 PM on Tech Tools for Rheumatologists: Introduction to Automation and Workflows to Save Time and Increase Office Productivity.
This will be followed shortly after by the annual #ACR15 Tweetup from 4:30–6:00 PM in the West Building Rm 2000–2002. If you’re unfamiliar with the Tweetup, check out my post from last year.
Upcoming Experimental stuff:
I’m in the beginning (soft-beta-pre-launch) phase of a new site I’m calling ArthritisProject.com, which I hope to turn into a more structured educational resource for patients to learn about their arthritis. If you’re a patient interested in this, sign up for the mailing list on the site, where I’ll push out updates as they’re available.
In hopes of keeping myself on a low information diet, I had long avoided using Twitter lists by trying to carefully curate who I followed in my main timeline.
Over time, the task of limiting the number of users I followed became impossible while working on projects such as The Rheumatology Podcast and #RheumJC, and connecting with related groups such as #NephJC.
Recently, I discovered a service called Nuzzel, which curates the most frequently shared stories from your timeline. In their words:
“We created Nuzzel to solve the problems of social overload. You can use Nuzzel to discover the best news stories shared by your friends on Facebook and Twitter without being overwhelmed or missing anything.”
Because of Nuzzel, I have actually found it useful to follow more people in my areas of interest (healthcare/rheumatology, technology, fitness), and let the service automatically curate the most frequently tweeted stories.
After following more users on Twitter, I finally started using Twitter lists. Since the methods built into Twitter (and my beloved Tweetbot) are not extremely useful for reorganizing significant numbers of users into Twitter lists, I found twitlistmanager.com extremely helpful for managing this task.
The advantage of using Twitter lists because of Nuzzel is that it easily allows me to keep the number of people in any of my lists well under 100 (maintaining my monkeyshere), while still allowing me to find the shared stories of highest importance.
”I don’t have big ideas. I sometimes have small ideas, which seem to work out.” – Matt Mullenweg
I previously detailed my WordPress setup, but over the last few weeks, I have given my site (another) overhaul worthy of an updated blog post.
New WordPress Theme
I recently switched to the X WordPress Theme. This theme comes with four different minimalist style layouts that help focus on blog readability over everything else, yet remain easily and highly customizable through the built-in WordPress customizer. I’ve played with a ton of different WordPress themes, and for a premium theme, this one is as easy as they come. If you want to make any very specific modifications, you may need to learn a little bit of CSS and/or PHP.
Strong Focus on Increasing Site Speed
The speed that your site loads is highly important for two reasons:
Improved experience for readers (better chance that readers will actually read your writing). According to KISSmetrics:
”47 percent of visitors expect a website to load in less than 2 seconds, and 40 percent of visitors will leave the website if the loading process takes more than 3 seconds.”
Unfortunately, many of the tips offered by Google are difficult to know exactly how to optimize. Besides having your site hosted on a fast server, the lowest hanging fruit are easily solved with a few WordPress plugins:
WP-Optimize. This plugin actually increased my PageSpeed score by 10–14 points, and works by removing any excess junk such as comment metadata in your WordPress pages.
Smush.It. This optimizes images in several lossless ways, such as stripping JPEG metadata and optimizing compression. The free version can “smush” image files under 1MB.
BJ Lazy Load. This plugin uses a jQuery script to improve site loading by delaying loading of any images files until just as the reader has scrolled down to that image.
VaultPress. Premium service from Automattic (the makers of WordPress) that offers real-time backup and has options for security protection. I switched to this service given the ease of restoring backups compared to WordPress Backup to Dropbox.
“It sounds plausible enough tonight, but wait until tomorrow. Wait for the common sense of the morning.” ― H.G. Wells
Humans don’t wake up ready to give 100%.
Our daily startup sequence has more similarities to older PCs that run best if restarted daily than modern devices that are always ready to go. Most of us benefit from going through some type of morning routine to get ourselves set for optimal output.
A well defined morning routine is important because it decreases the number of choices that you make in the beginning of your day, which reduces decision fatigue.
“Imagine you have 100 points for making decisions every 24 hours. The more decisions you rack up in one area, the fewer decisions you can effectively make elsewhere.
If you make a ton of unnecessary decisions: which email to check, what to do first in the morning, what breakfast to have, you are going to deplete your hit points, and that will lead to poor decision making later, because you’re going to run out of your 100 points.”
An effective morning routine should essentially be a mental do-confirm checklist with the goal of getting minimizing or eliminating unnecessary decisions as you get ready for the day.
My typical morning has five things that rarely deviate:
Water. The first thing I do is drink a large glass of water. As much as I wish I could start with coffee, water is what wakes me up in the morning.
High protein breakfast. For at least the last 10 years, I have eaten two eggs almost every morning, typically with some sausage or ham. Rare exceptions are when I’m traveling or eat at a restaurant. (I’d love to eat bacon every day, but don’t typically have time to do the full bacon method). Sometimes I’ll have a small handful of berries, depending on the season.
Coffee. Usually French press with freshly ground beans. Caffeine is a wonderful thing.
Walk the dog. Although I typically only have time for a 5-10 minute walk on work days, I always find it refreshing to get outside and move around. Bonus: it keeps the dog from plotting against me.
Shower. I always feel mentally fresher after a shower, and given a chronically busy schedule, I used to say, “a shower is worth 2 hours of sleep.” We now know that sleep is much too important to skimp on, but my patients probably prefer that I take a morning shower anyway.
Over time, I’ve essentially designed my work wardrobe in the style of Steve Jobs’ daily uniform, so that I could get dressed in the dark (which occasionally happens) and still match.
For my rheumatology colleagues that are not familiar, #NephJC is a thriving nephrology journal club on Twitter and NephJC.com. If you’re relatively new to Twitter and using hashtags, but would like to participate, the simplest way would be to follow the instructions on their site.
I’m excited to see what we’re able to learn from our nephrology colleagues, and hope us rheumatologists have much to share as well.
Background on establishment of the use of rituximab for ANCA-associated vasculitis
Until recently, the only standard therapy for treatment of ANCA-associated vasculitis had been induction with cyclophosphamide plus glucocorticoids. In 2010, the RAVE and RITUXVAS trials respectively established rituximab as non-inferior therapy for the induction of systemic and renal ANCA-associated vasculitis, with similar adverse events.
The RAVE trial compared rituximab to oral cyclophosphamide, and had a primary outcome of complete remission and tapering of glucocorticoids at 6 months, with results as follows:
Rituximab group: 63/99 patients (64%)
Cyclophosphamide group: 52/98 patients (53%)
Despite a trend toward improved outcomes in the rituximab group, the RAVE study was powered as a non-inferiority trial, and the differences between the groups were not statistically significant (p=0.09, 95% CI -3.2 to 24.3).
Cyclophosphamide/azathioprine: 39% at 12 months, 33% at 18 months.
Again, despite the numerical trend, this data was only able to support rituximab as non-inferior to cyclophosphamide. Additional analysis showed that in patients with relapsing disease, rituximab was superior at 6 months (p=0.01) and 12 months (p=0.009), but this effect did not extend out to 18 months (p=0.06).
The difficulty of frequent relapse in ANCA-associated vasculitis
Clearly, maintenance of remission for patients with ANCA-associated vasculitis remains a challenge and patients continue to relapse frequently despite current management strategies. Prior studies looking at remission rates with maintenance immunosuppression show significant relapse rates:
What is the role of rituximab in maintenance therapy for ANCA-associated vasculitis?
Until now, only retrospectivedata has suggested that rituximab might be useful as a maintenance strategy for ANCA-associated vasculitis. This question was addressed in our current journal club article:
This study recruited patients aged 18-75 with newly diagnosed ANCA-associated vasculitis who were in complete remission after treatment with glucocorticoids plus pulsed cyclophosphamide. Disease activity was measured using BVAS, a standard vasculitis activity measure. Patients who previously received rituximab were excluded.
Induction of remission in both groups was:
Prednisone starting at 1 mg/kg with gradual taper, with some patients initiated with pulse methylprednisolone 500-1000 mg for 1-3 days.
Pulse cyclophosphamide 0.6 g/m² on days 0, 14, and 28, followed by 0.7 g/m² every 3 weeks for 3-6 additional pulses until remission was attained.
After 4-6 months, patients were randomly assigned 1:1 to either:
Rituximab 500 mg on days 0 and 14, then at months 6, 12, and 18.
Azathioprine 2 mg/kg for 12 months, 1.5 mg/kg for 6 months, then 1 mg/kg for 4 months.
In both groups, prednisone was further tapered to approximately 5 mg/day for at least 18 months, after which prednisone could be further tapered at investigator discretion.
“The primary end point was the percentage of patients with major relapse (reappearance or worsening of disease with a BVAS >0 and involvement of at least one major organ, a life-threatening manifestation, or both) at month 28.”
Secondary end points: Rates of minor relapse, adverse events, mortality.
Under the hypothesis that rituximab would decrease major relapses by 25% at month 28, assuming 5% exclusion or dropout, to have 80% statistical power and a 2-sided alpha risk of 0.05, 118 patients had to be enrolled in the trial.
115 patients were enrolled (87 with GPA, 23 with MPA, and 5 with renal-limited ANCA-associated vasculitis). 58 received azathioprine and 57 received rituximab. Patient demographics were not significantly different in either group. 80% of patients had newly diagnosed disease, and 20% had relapsing disease.
Primary end point
At month 28, major relapse rates in each group were:
Azathioprine 17/58 patients (29%)
Rituximab 3/57 patients (5%)
Which was statistically significant, with hazard ratio for relapse, 6.61; 95% confidence interval, 1.56 to 27.96; P=0.002.
To put it another way, in the authors words, “hence, to avoid one major relapse, 4 patients (95% CI, 3 to 9) had to be treated with systematic rituximab infusions rather than with azathioprine.“
Secondary end points
Minor relapses were not statistically different between each group, with 9/58 (16%) in the azathioprine group and 6/57 in the rituximab group (11%) (p=0.43).
Rates of infection were similar in both groups, with 8/58 (14%) in the azathioprine group and 11/57 (19%) in the rituximab group.
There were two deaths in the azathioprine group, one with vasculitis involvement of the aortic valve who had a major relapse and received pulse steroids but died of E. coli bacteremia. The second patient had developed pancreatic cancer.
The authors concluded that “in the present study, rituximab was superior to azathioprine at maintaining remission of ANCA-associated vasculitis; this was especially true for granulomatosis with polyangiitis, which was the condition seen in most of the study population.”
“Our data also show that successive 500-mg infusions of rituximab, given every 6 months up to month 18 after remission, were not associated with more frequent severe adverse events than azathioprine.”
Interestingly, the authors note that the 6-month interval between rituximab infusions was chosen based on expected B-cell reconstitution and relapses after a median of 1 year.
They also state that a single 500 mg rituximab dose was chosen because patients were already in remission (i.e. B-cell deplete) and in hopes of decreasing risk of infection.
Unable to generalize to all patient types with ANCA-associated vasculitis since most had PR3-ANCA/GPA.
Leaving prednisone tapering up to discretion of investigators after 18-months (although authors note that only 2/20 relapsed patients had stopped prednisone).
Azathioprine maintenance until month 22, with a gradual taper (although 8 relapses occurred in the first 12 months at a dose of 2 mg/kg, and only 2 relapses occurred between months 12 and 22).
Another important point from the authors is that “several major relapses (7 of 17 in the azathioprine group and 2 of 3 in the rituximab group) occurred after treatment with the trial maintenance drugs was stopped.”
Having personally used rituximab in the clinical setting for both remission-induction and treatment of relapse for ANCA-associated vasculitis, and knowing that B-cells play a significant role in disease activity, I was not surprised that rituximab would be an effective therapy. This being said, I certainly did not expect a NNT of only 4 patients. I was also impressed were able to achieve these results using only single 500 mg doses of rituximab every 6 months, which may be important not only for potential decreased infectious risks, but may have financial implications as well.
One concerning observation in this study is that after stopping maintenance therapy in either group led to several major relapses. I have often wondered if we can ever safely stop all immunosuppressive therapy in these patients, and if so, how to identify which patients this can be safely achieved.
[Update 12/12/2014: Added slideshow version of the Storify below by @nephondemand. This gives additional background on management of ANCA-associated vasculitis and the trial covered in this journal club.]
Despite ways to manage many of these challenges, certain decisions in medicine are inherently going to remain extremely difficult.
Features of extremely difficult decisions in medicine:
An incorrect decision has a high risk of adverse consequences.
Limited information is available to make the decision.
Of these two features, we typically have no way to control the amount of risk involved in our decision. We might also feel that we have no control over having limited information, but this is incorrect.
We are actually able to manage information limitations in medicine.
The 40-70 Rule in Decision Making
Former U.S Secretary of State and General Colin Powell is known for the 40-70 rule in decision making, stating that a leader should make a decision when they have between 40% and 70% of the information available.
”If they make the decision with less than 40% of the information, they are shooting from the hip. But waiting for more than 70% of the information delays the decision unnecessarily.”
Applying this to medical decision making:
By the time you have gathered 40% of necessary information, a skilled clinician should have narrowed diagnostic and treatment possibilities enough to make an effective decision.
By waiting for more than 70% of information to become available, we may actually increase risk by delaying a decision.
The surprising part here is that by delaying our decision beyond a certain amount of information, we may end up with worse outcomes.