Thread by @roby_bhatt

At the risk of shouting into the void, 🧵 on papers from 2020 that most changed how I think abt COVID, as an ID physician-scientist. 280-char summaries + URLs for each. Thx to all authors & apologies for any omissions; this list is unofficial, personal, idiosyncratic, & LONG. 1/

2/ Early summary of 72,314 (hospitalized) cases from Chinese CDC, broken down by mild vs severe vs critical, early hint at CFR (overestimated b/c mild cases undersampled), & sharply age-dependent mortality. Also, risk to HCWs. Fig 1 (epidemic curve) key.

jamanetwork.com/journals/jama/…

3/ Another inpt obs cohort study from China 🙏🏽. Fig 2 shows prognostic biomarkers (lymphs, D-dimer, IL-6 – not CRP, despite my false memory). Watched (helplessly) a lot of these rise in worsening patients. Still wonder why we ordered so many, so often.

thelancet.com/journals/lance…

4/ Too much quibbling when this one came out over semantics (asympt vs presympt), imo, & not enough attention to the devastating potential impact of transmission from ppl who don’t (yet) know they’re infected. This is so key to COVID’s rampant spread.

nejm.org/doi/full/10.10…

5/ This one I’m including for Fig 1, the now-oft-used schematic showing an early virological and a late hyperinflammatory phase of COVID. This may not be the earliest to frame the disease course like this, but it’s the first I saw.

jhltonline.org/article/S1053-…

6/ First report I saw confirming that SARS-CoV-2, like SARS Classic, uses ACE2 as the viral receptor for cell entry, and TMPRSS2 and cathepsins as candidate processing proteases. Rapid, beautiful, and important molecular virology & cell biology.

sciencedirect.com/science/articl…

7/ Reports of the outbreak devastating a Seattle-area SNF in Feb-Mar 2020 presaged the havoc COVID can wreak in congregate living facilities & surrounding community. 167 linked cases, 33% fatality rate for residents.

nejm.org/doi/full/10.10… &
cdc.gov/mmwr/volumes/6…

8/ Reports from Diamond Princess cruise ship: more early evidence of both facile transmission & wide range of possible symptoms, incl many folks without symptoms while shedding virus (some went on to develop symptoms later).

nejm.org/doi/full/10.10… &
thelancet.com/journals/lanin…

9/ More than any other, this paper shaped how I view viral shedding & transmission kinetics. Fig 1 is key (corrected here: nature.com/articles/s4159…).

Infectiousness peaks early; far more presymptomatic spread than flu. Huge implications, as we've seen.

nature.com/articles/s4159…

10/ N only 9 inpts, but a detailed early look at viral burdens from various sites, esp respiratory tract, over time, incl culture & the now-infamous subgenomic RNA, with paired serology (anti-spike IgG/M and neutralization). VL & culturability peak early.

nature.com/articles/s4158…

11/ Prescient modeling study informed by seasonal CoVs, foretelling the shitshow that would ensue “post” lockdown. @yhgrad is a close friend; when he texted me a working version of Fig 5 in early March, I got my first inklings of what 2020 might become.

science.sciencemag.org/content/368/64…

12/ Of three reports of outbreaks in homeless shelters in rapid succession, this was the first I saw and the closest to my home. The extent of asymptomatic carriage puzzled me at the time, and underscored many unique challenges of COVID.

jamanetwork.com/journals/jama/…

13/ No one paper can fully convey the many ways in which COVID-19 has laid bare society’s inequities, but this is one example: Fig 2 summarizes: 250% greater risk of death from COVID among Blacks, & 88% increase among Latinx. SDOH, structural racism.

link.springer.com/article/10.100…

14/ Preprint, now paper, that opened my mind to the (alas still untested) possibilities of high-freq rapid testing, even if lower sensitivity than PCR. Clear, compelling logic. Lots of noise on all sides of this issue, but to me, this is peak modeling.

advances.sciencemag.org/content/early/…

15/ Another modeling paper on testing, this one from my soon-to-be-former boss @RWalensky (maybe you've heard of her?). Incorporates cost-effectiveness to argue for frequent testing in college settings, even if low sensitivity - now widely adopted.

jamanetwork.com/journals/jaman…

16/ Lots of great work to clarify IFR, but somehow this speculative one stuck with me & met my brain where it was already going: age-dependent IFR means COVID ~doubles actuarial death rates (Fig 2), & thus may tell us more abt humans than abt SARS-CoV-2.

bmj.com/node/1033603.f…

17/ Superspreading, now appreciated to play a large role in driving spread, was well demonstrated in this contact-tracing study in Hong Kong. Fig 3 shows how most cases did not spread, while Fig 2 graphically illustrates how some spread a LOT.

nature.com/articles/s4159…

18/ A more visceral demonstration of superspreading: the infamous choir paper, with 32 confirmed and 20 probably secondary cases in a 61-person choir practice. Singing generates aerosols. And COVID is a menace.

cdc.gov/mmwr/volumes/6…

19/ More superspreading: the MMWR report from the GA summer camp was nuts: 1 camp counselor tested positive on day 4 of an overnight camp; everyone was sent home within 3 days. 344 of 597 campers had tests available, and 260 of these were positive! Yikes.

cdc.gov/mmwr/volumes/6…

20/ Mol epi: gargantuan sequencing effort from many of my colleagues at MGH and the Broad as well as MADPH showing the downstream effects of the first known superspreader event in the US.

science.sciencemag.org/content/early/…

21/ A systematic analysis of respiratory mechanics in COVID found similarities to other causes of ARDS. Also rewarding to see favorable survival in intubated pts at my hospital during my service block. Proud to call these folks colleagues.

atsjournals.org/doi/10.1164/rc…

22/ What more can I say about this one? ACTT-1 (double-blind, placebo-controlled RCT) showed shorter hospitalizations with remdesivir (Fig 2,3) & thisclose to stat sig morality benefit (Table 2). The first big practice-changing RCT…

nejm.org/doi/full/10.10…

23/ …until this (sigh): pragmatic, open-label (non-blinded) randomized WHO Solidarity trial showed no benefit to remdesivir (Fig 2a)! Go w gold standard (ACTT-1), or larger N but inferior design (Solidarity), or parse by subgroup?

¯\_(ツ)_/¯ I’m pro-rem

nejm.org/doi/full/10.10…

24/ Biggest surprise for me (pre-vax) was easily the Recovery trial: steroids work!? In the critically ill, anyway; trend twd harm if not yet on O2. Fig 2 says it all. My takeaway: antivirals early? steroids late!

Kudos to UKNHS for the trial platform 👏🏽

nejm.org/doi/full/10.10…

25/ My synthesis on rem/dex: meh antiviral fx may be due in part to late Sx; once ppl need admission, viral phase mostly over. But variable progression risk (& IV only) = early admin hard. Steroids may help here but not in other ARDS b/c COVID more uniform; ARDS is a grab-bag.

26/ Nerdy, but I study transcription (mostly in bacteria) & thus loved this comprehensive overview of the transcriptome of SARS-CoV-2 grown in Vero cells. Along with a Susan Weiss seminar, this opened my eyes to subgenomic RNAs before they were cool.

cell.com/cell/fulltext/…

27/ For vaccines to work, we needed to know if infection would protect against re-infection. This work (from my former MSTP colleague @GreningerLab) found a ray of hope: neutralizing Ab’s protected 3 ppl from a superspreader event aboard a fishing vessel.

jcm.asm.org/content/58/11/…

28/ First heard about this from @DrJLi on a division COVID call. Masterful work showing intra-host immune evasion (non-synonymous mutations enriched in spike) in a heavily immunosuppressed patient with a protracted COVID course. Presage of UK variant?

nejm.org/doi/full/10.10…

29/ Forgive me, but this paper I’m on shaped my thoughts on VL: cross-sectional study of ~2000 COVID pts in MA SNFs, no relationship b/w VL & Sx at time of testing. Plenty of ppl felt well with 1 billion copies/mL. Still winding its way thru peer review.

medrxiv.org/content/10.110…

30/ Cool preprint (not peer reviewed & I’m no immunologist): pts who died of COVID got neutralizing Ab’s later & clear virus slower than survivors. VL doesn’t predict symptoms, but VL on admission predicts survival. Slow clearance may square that circle.

medrxiv.org/content/10.110…

31/ From the gee whiz files: identical twins got COVID, thought to be from the same exposure. One was discharged after a brief hospitalization, the other nearly died (ICU, prolonged intubation). Randomness affects more than we want to accept, imo.

acpjournals.org/doi/10.7326/L2…

32/ 2 recent reports (UK, SA) of CoV2 variants that may incr transmission. Key ?: any reinfections? No word yet; if not, hopefully natural & vaccine-induced immunity will hold. Either way: better masking, distancing, ventilation.

cmmid.github.io/topics/covid19… &
medrxiv.org/content/10.110…

33/ Finally: VACCINES! Tbh, the thunder was stolen a bit by press releases and text alerts, but this was still a thing of beauty. Look at Fig 3 and try not to weep for joy. I still hope it can change the trajectory of human history.

nejm.org/doi/full/10.10…

34/ On that note, honorary mention of the 2017 paper that laid the foundation for the vaccine successes by demonstrating a stabilized variant of the pre-fusion spike from MERS. Amazing feat of protein engineering: check out This American Life ep. 727.

pnas.org/content/114/35…

35/ List is very much unofficial: most chosen b/c of how they hit me at the time, in turn affected by what I did when. I was on service for April surge in Boston = early clinical studies had outsized impact. Later, more basic/translational science. I’d love to hear your faves!

36/ I omitted serosurveys; they’re important but imo they’ve too often added confusion, not clarity. I also skipped school articles, despite having thought and read a TON about this (globalhealth.massgeneral.org/covidlibrary.p…) - no one article clarified that hornet’s nest for me.

37/ Also, I’m an ID physician & molecular (micro)biologist. So while I’ve learned a TON abt new fields this year through COVID, I don’t feel as qualified to select from papers on, say, hard-core epi or immunology or droplet physics. Tell me your favorites, experts of Twitter!

38/fin Debated, but didn’t tag authors I don’t know. Hopefully paper links give credit where it’s due, & I didn’t want to distract folks (or usurp followers). As a clinician, my boundless gratitude for the efforts of so many to guide understanding & management of our patients.