1648

Trevor Bedford has a good analysis taking the 22 day lag between testing positive and death that shows cfr 1.7

1649

Here is a good explanation on testing and why tests need to target the strains in your geo.

1650

That’s pretty interesting. The paper link is here, for people who don’t want to dig through the tweets to get to the pre-print:

Having errors at the 3’ end of primers is a well known problem in the field (most of my molecular biology skills are well out of date, and I still remember this from my bench days) - it’s real unlucky that there’s been enough random drift that there are single nucleotide changes that might have an impact.

Also, a 30-fold change is (as noted) about a change of 5 on the CT scale - most actually infectious people are scoring around 17-25, and the PCR test is generally flagged as positive up to cycle 34/35, so this will only make a difference for the qualitative result for people with a much lower viral load (people that are close to being outside the infectious range.)

This is SUPER interesting, though, given some data we’ve seen at Broad. We have this really weird bimodal distribution of qPCR scores [not sure if I can post the plot without being somewhere between a jerk and actually doing something “wrong”, but it looks bimodal around CT=17 and CT=35], and Broad is running ~ 100,000 tests a day with large geographic spread. This would be an interesting source of variation in the data, which folks who run the clinical lab didn’t seem to have an answer for when I saw their talk on thursday.

1651

So, from Vincent Racaniello’s (TWIV owner) video on when he tested positive, he discussed that a CT of 35 has a very good chance of a false positive - and was retested negative the next time.

Am I interpreting your statement that the clinical lab didn’t have an answer implies they aren’t acknowledging this issue of fine tuning tests for geo discrepancy?

It does seem that positive testing %'s are going way way up. I wonder when states report their positive tests if they adjust these later when someone retests and gets a negative?

God, just so many unknowns and variables right now. Doctorate candidates in a few years are going to have such a target rich environment in picking what to go study instead of digging / scraping the bottom of the barrels.

1652

Just to make sure we’re on the same page:

When we talk about cycle threshold result for quantitative PCR, it’s worth going through what’s going on under the hood. When a sample is taken, some number of viral copies of the genome is taken. Assuming 100% efficiency of the primers, every cycle of PCR doubles the amount of viral DNA present at the locations bracketed by the primers - that is, the assay makes a copy of that portion of the genome. Each cycle doubles the amount of cDNA present. This amplification happens for many cycles, and when the total amount of viral material reaches a certain concentration, you’ve reached the threshold of detection.

Because of this, the lower the CT value, the fewer doublings needed to reach the threshold, thus you started with more material.

In VR’s case, he was borderline. (Btw, he had his test run by institution, which currently runs about 100,000 tests per day.) That’s why VR got in contact for his positive test score and found out the CT value, which isn’t normally reported.

What would happen if VR had a mutation that was in the 3’ part of one of the primers as mentioned in the manuscript? The PCR reaction would be less efficient. You can think of this like 2 cogs in a machine that are slightly mismatched, so the teeth don’t always match up and force isn’t transferred. Since some small subset of the molecules generated wouldn’t double, instead of doubling each cycle, the total number of molecules would be slightly less than expected. This means the cycle at which you reached the threshold would be higher than expected.

To be clear, under Chan’s hypothesis the opposite would happen - there would be more false negatives if you have mismatched primers. Thus VR would have a CT value ~40 (under the case they cite), which would have been negative result.

I’m unaware of the inner workings of the clinical lab, but considering Alica Chan and Ben Deverman both work at the same place I do, I’m sure there are internal discussions.

Hopefully this is useful, but let me know if I can elaborate, and sorry if i went too basic.

1653

SUPER helpful. Thanks! I like your explanation of how CT values are derived. I had no idea.

1654

The race for a coronavirus vaccine has received another shot in the arm with the US biotech firm Moderna becoming the latest to reveal impressive results from phase 3 trials of its jab.

An interim analysis released on Monday, and based on 95 patients with confirmed Covid infections, found the candidate vaccine has an efficacy of 94.5%. The company said it now plans to apply to the US regulator, the Food and Drug Administration, for emergency-use authorisation in the coming weeks. In the trial, 90 of the patients received the placebo with the remaining five the vaccine.

The results are the latest encouraging news to emerge from the breakneck effort to develop a vaccine against coronavirus and follow a similar interim analysis earlier this month from a collaboration between Pfizer and the German firm BioNTech, which suggest its vaccine is 90% effective at preventing illness.

The Moderna vaccine is not expected to be available outside the US until next year. The biotech said it would have 20 million doses ready to ship in the US before the end of 2020 and hoped to manufacture 500 million to one billion doses globally next year.
The Moderna vaccine, which is based on similar mRNA technology as BioNTech’s, is expected to be assessed by the FDA on a final analysis of 151 Covid cases among trial participants who will be followed on average for more than two months.

If the results remain as impressive as the trial goes on, the Moderna vaccine could potentially provide a major advantage over the Pfizer vaccine. While Pfizer’s vaccine requires ultracold freezing between -70C and -80C from production facility to patient, Moderna said it had improved the shelf life and stability of its own vaccine, meaning that it can be stored at standard refrigeration temperatures of 2C to 8C for 30 days. It can be stored for six months at -20C for shipping and long-term storage, the company said.

1655

If we can roll this vaccine out quick enough, we might be able to get thing under control even in the US in six months time.

At this point, I’d support a hard lockdown until we get to that point, it could be justified as a short-term sacrifice.

That percentage was a lot higher than I was anticipating- I thought it would be about 50-70%.

Seriously, this feels like the equivalent of the battle of Trafalgar or Gettysburg in terms of significance.

1656

Does that seem like a rather small trial? 90 patients got a placebo and only 5 got the actual vaccine?

1657

I think the designation of this as an interim analysis suggests that it’s a slice of the total data in a larger trial of 30,000 people. If the full data supports the interim analysis…then yahtzee.

1658

I suspect it’s 95 positives out of the folks they trialed, 90 got placebo, 5 got vaccine.

That’s statistically significant.

1659

That is correct.

This excerpt from the WP is a little more clear.

Moderna’s vaccine, co-developed with Fauci’s institute, is being tested in 30,000 people. Half received two doses of the vaccine, and half received a placebo. To test how well the vaccine works, physicians closely monitored cases of covid-19 to see whether they predominantly occurred in people who received the placebo group.

Of the 95 cases of covid-19, the disease caused by the virus, 90 were in the group that received the placebo. There were 11 severe cases reported — all in people who received the placebo. With cases of covid-19 confined almost exclusively to trial participants receiving a placebo, that sends a strong signal that the vaccine is effective at thwarting the virus

1660

Yup. A study with 30,000 is pretty damned good. Of course the really important N is being driven by the number of people getting sick, and so far they’re doing amazingly well. They’re aiming to have 151 sick, hopefully 95% of those will stay in the placebo group. I find it both sad and hilarious that they are getting to this number well in advance of when they’d originally predicted because of the horrific US spike. It’s exactly as @espressojim predicted. Trump basically created a nationwide Petri dish.

Also really positive: the Moderna vaccine is stable at normal refrigeration temperature for 30 days.

1661

Sorry, does 95% here refer to that 95 percent of recipients won’t spread the virus to anyone else but 5 percent will, or that symptoms will be 95 percent less damaging to recipients, or some combination of those, or…?

1662

The relevant quote from the WaPo article:

Of the 95 cases of covid-19, the disease caused by the virus, 90 were in the group that received the placebo. There were 11 severe cases reported — all in people who received the placebo. With cases of covid-19 confined almost exclusively to trial participants receiving a placebo, that sends a strong signal that the vaccine is effective at thwarting the virus.

90/95 = 94.7%, and there’s no mention yet if it prevented transmission of the novel coronavirus. All we know for sure right now from this report is that it prevented development of the disease (COVID-19).

1663

Yeah, the portability (for lack of a better term) of the Moderna vaccine is really going to help distribution, and I’m especially thinking of rural areas in that.

1664

On Twiv this weekend someone mentioned that the Pfizer vaccine may have better stability than originally advertised but they hadn’t completed stability testing yet so were using the default “guaranteed” temps for now, and that might change in the next few weeks/months, before full FDA approval.

1665

Were half of the study participants given the vaccine and half the placebo? That’s an important factor in understanding the study’s significance.

1666

Yep! Exactly that:

Half received two doses of the vaccine, and half received a placebo.

1667

And the 1/2 that got the vaccine, were locked in individual Fallout style vaults for the last 6 months.

Just anticipating the anti-vaxer’s next argument.