I’m back. True, I’m back a little later than I had planned, having hoped to have resumed dishing out my usual Respectful—and not-so-Respectful—Insolence on Monday, but let’s just say that my return from hiatus and recharging was…not smooth…and leave it at that. In any event, I found myself greeted soon after resuming my usual monitoring of misinformation and disinformation with hysterical Tweets about mRNA from mRNA-based COVID-19 vaccines supposedly having been found in breast milk. In fact, there are a number of topics that bubbled up during my absence vying for my attention, but I decided that this fear mongering about COVID-19 vaccines supposedly contaminating breast milk was the best one to ease my way back into the blogging routine because it lets me look at an actual study and allows me to revisit my expertise in PCR technology. Also, I noticed that antivaxxers kept Tweeting just the main results table from the study without actually citing the study and asking me to tell them what it meant, which is annoying.
But first, let’s see how the antivax crankosphere is spinning the study.:
I could go on and on and on, as there are thousands of Tweets like this, but I do want to include a Tweet by an antivax legislator:
And I’m not even going into the hysterical reactions coming from antivaxxers with Substacks with large readerships, such as Jessica Rose, Byram Bridle, and, of course, Steve Kirsch who titled his reaction The data says you should never breast feed your baby after you’ve been vaccinated. (Unsurprisingly, the data, even if you take this study entirely at face value, say no such thing.) If you really want to see the bonkers, you could also always look at Naomi Wolf’s video about the study, but, to be honest, I couldn’t even make it through the full 12 minutes. You’re welcome to try if you wish, though. If you do, I’ll quote an old Yardbirds song Mr., You’re a Better Man Than I.
I could list even more antivax amplification of this study as slam-dunk evidence that COVID-19 vaccines are killing babies, but, as I like to say, instead let’s go to the tape (I mean the actual study). It’s a Research Letter, which is the shortest form of study, and, unfortunately, this particular study is not open access, which means that I had to use my university access to read it.
I’ll dismiss one thing that I noticed about the study right away that was also pointed out by Kyle Sheldrick in comments; specifically, the study was listed as having been carried out from February to October 2020, which was before any vaccine was available to the general public. True, there were people who received the vaccine as part of a clinical trial, but the Pfizer and Moderna clinical trials specifically excluded lactating individuals. Apparently that was an error, and the study was conducted in 2021, which leads me to a second question. Why were there so few subjects recruited during an eight month period. Only eleven subjects were recruited. That’s not exactly a robust accrual. Similarly, why did it take so long to publish the results if accrual ended in October 2021? That was nearly a year ago, and it doesn’t take that long to do RT-PCR on breast milk specimens, analyze the data, and write it up for publication as a research letter.
I was also irritated that the methods section was consigned to a Supplemental Online Content file. I hate that. The methods should be in the paper, and if a research letter is too short a format to include it in the actual paper published in the paper journal then the findings should have been published as an actual paper. Finally, I was annoyed that the actual sequences of all of the primers used to measure both the spike protein mRNA used in the Pfizer and Moderna vaccines and the housekeeping gene controls. There is no excuse for not including the exact sequences or referencing a paper that does if the same primer sequences have already been published.
All the paper said was:
Based on the putative sequences of vaccine BNT162b2 (Pfizer) and mRNA1273 (Moderna) by Dae-Eun Jeong et al.22 two sets of primers targeted two regions of each vaccine mRNA were designed, and the primers and probes were synthesized by Integrated DNA Technologies (Coralville, IA). Expression of RNAse P and beta Actin transcripts were assayed as an internal control to verify the sample quality throughout the RNA isolation, reverse transcription, and PCR processes. Real-time PCR was performed using Lightcycle 480 Probes Master (cat#04707494001, Roche LifeScience, Indianapolis, IN) on the QuantStudio3 (Applied Biosystems, Foster City, CA). Cycle threshold (Ct) was calculated with Quantstudio Design and Analysis software version 15.1 (Applied Biosystems, Foster City, CA). BNT162b2 (Pfizer) and mRNA-1273 (Moderna) leftover vaccines immediately after clinical use in our hospital (that were designated to be discarded) were used to validate the real-time PCR assay. For assay validation, a known amount of vaccine in the range of 100,000 to 0.1 pg/mL was spiked in whole milk collected from individuals prior to June 2019.
It is, of course, reasonable to construct a standard curve using breast milk collected and stored before the pandemic spiked with known amounts of vaccine to cover a concentration range from 1 to 100,000 pg/mL, as the authors did, but, again, as a longtime PCR maven who wrote about misadventures in PCR by antivaxxers long before the pandemic through to the claims about how too-sensitive PCR assays for SARS-CoV-2, the coronavirus that causes COVID-19, were producing a “casedemic” (although who also, admittedly, hasn’t done a lot of PCR in the last few years but before that had run literally thousands of assays during the preceding couple of decades), I want to know the sequence. True, the authors list a reference (#22), but there are only six references listed in the paper; so I remain puzzled whether the cited reference lists the sequences. Indeed, I’m not the only person irritated by the lack of explicit listing of the primer sequences. In a comment, Kevin McKernan notes that “it would be helpful to know the primer sequences to rule out off-target effects” and that the “vaccines have diff GC content and sequence due to codon optimization but some homology to C19 may still exist.” That first comment refers to effects that are nonspecific to the mRNA sequence being subjected to PCR, and the second refers to the changes in the sequence of the spike protein gene that do not change the protein produced used by Pfizer and Moderna to decrease immunogenicity and increase stability of the mRNAs.
However, let’s assume that the primers were well-designed, that the investigators did all the correct controls to rule out off-target effects and ran the PCR products out on gels to make sure that products of expected size were generated and that there were no primer-dimers (for example). What did the authors do, and what did their study find?
Here’s the study population:
This cohort study was conducted from February to October 2020 and included 11 healthy lactating mothers who received either the BNT162b2 (Pfizer) or mRNA-1273 (Moderna) COVID vaccine within six months after delivery. Demographic data were collected through in- person interviews (Table 1). NYU Institutional Review Board approval was obtained, and all study volunteers signed written informed consent.Strike 2020, as this study was carried out in 2021. Also, as you can see, this study was carried out at NYU.
And here is Table 1, with the demographics:
As for the breast milk, it was treated thusly:
Extracellular vesicles (EVs) were isolated from 3 ml of BM by sequential ultra-centrifugation method. After removing cell and bulky debris by serial centrifugation at 2,000- and 17,000 x g for 10 min at 4°C, the supernatant was collected and subjected to 100,000 g centrifugation in a TLA 110 rotor for 18 hrs at 4 °C. The EVs pellet was suspended in PBS, and EVs concentration was determined by ZetaView by Particle Metrix ZetaView Nanoparticle Tracking Analysis ((Particle Metrix, Germany).
To translate, the breast milk was subjected to centrifugation to remove “bulky debris” (particulate matter, etc.) and then the supernatant (the layer that didn’t end up at the bottom of the tube) harvested and subjected to a much faster centrifugation for 18 hours at refrigerator temperature, which resulted in pelleting the extracellular vesicles from the breast milk at the bottom of the tube, which was then suspended in phosphate-buffered saline. The whole breast milk and EV pellets were then subjected to PCR, and the following results obtained:
These results were summarized thusly:
Of 11 lactating individuals enrolled, trace amounts of BNT162b2 and mRNA-1273 COVID-19 mRNA vaccines were detected in 7 samples from 5 different participants at various times up to 45 hours postvaccination (Table 2). The mean (SD) yield of EVs isolated from EBM was 9.110 (5.010) particles/mL, and the mean (SD) particle size was 110.0 (3.0) nm. The vaccine mRNA appears in higher concentrations in the EVs than in whole milk (Table 2). No vaccine mRNA was detected in prevaccination or postvaccination EBM samples beyond 48 hours of collection. Also, no COVID-19 vaccine mRNA was detected in the EBM fat fraction or the EBM cell pellets.
Interestingly, none of the antivaxxers touting the results of this study as proof positive that the vaccine mRNA is finding its way into mothers’ breast milk and causing horrific harm to our innocent babies mention the finding that no vaccine mRNA was detected in the breast milk beyond 48 hours after vaccination, which should tell you that, even if you are frightened by these results, that there’s nothing to be concerned about beyond 48 hours after vaccination. Funny how that message seems to have been lost among those fear mongering about this study. Similarly, nowhere in the study is it shown that the mRNA detected is actually intact and functional. What was detected could easily have been fragments of partially broken down spike mRNA, given that PCR used in this fashion. generally amplifies only relatively short (<200 base pair) fragments, as the longer the fragment amplified, the worse the efficiency of amplification.
The authors themselves state:
The sporadic presence and trace quantities of COVID-19 vaccine mRNA detected in EBM suggest that breastfeeding after COVID-19 mRNA vaccination is safe, particularly beyond 48 hours after vaccination. These data demonstrate for the first time to our knowledge the biodistribution of COVID-19 vaccine mRNA to mammary cells and the potential ability of tissue EVs to package the vaccine mRNA that can be transported to distant cells. Little has been reported on lipid nanoparticle biodistribution and localization in human tissues after COVID-19 mRNA vaccination. In rats, up to 3 days following intramuscular administration, low vaccine mRNA levels were detected in the heart, lung, testis, and brain tissues, indicating tissue biodistribution.4 We speculate that, following the vaccine administration, lipid nanoparticles containing the vaccine mRNA are carried to mammary glands via hematogenous and/or lymphatic routes.5,6 Furthermore, we speculate that vaccine mRNA released into mammary cell cytosol can be recruited into developing EVs that are later secreted in EBM.
I’ve discussed that biodistribution study before in depth. If you want the details, here they are, but one key thing to note is that the biodistribution study used a dose of lipid nanoparticle ~18-35x higher by weight than the typical adult human dose from vaccine—as biodistribution studies tend to do in order to map where even the tiniest amount of drug or vaccine go in the body.
I find it rather odd, though, that the authors failed to mention a previous study that failed to find vaccine mRNA in human breast milk, such as one published nearly a year ago in JAMA Pediatrics, which found no vaccine-specific mRNA in the breast milk of individuals vaccinated less than 48 hours before. Similarly, they were not the first to find vaccine mRNA in breast milk. This study from 2021 found vaccine mRNA in the breast milk of a small proportion of subjects:
Five breastmilk samples from 4 mothers had detectable vaccine mRNA, out of 309 samples from 31 mothers tested (Supplemental Table 2 ). All positive samples were collected within 3 days of the vaccine doses – two samples from days 1 and 3 of dose 1 ( Figure 6B ) and another three from days 1 and 3 post dose 2. One mother had detectable vaccine mRNA in both breastmilk and serum samples. The median vaccine mRNA amount in both sample types were comparable: 14ng/100ml (IQR 8-23) in serum compared to 7ng/100ml (IQR 6-7) in breastmilk (p=0.2).
None of the serum samples from the five infants tested had detectable vaccine mRNA. Of the five, one infant was from a mother with detectable vaccine mRNA in the both breastmilk and serum and another three were from mothers with vaccine mRNA in the serum.
In other words, the the findings of the current study are not novel, and they didn’t even do anywhere near the largest study addressing this question, which makes me wonder how this study was worthy of publication in such a high impact factor journal. Sure, it’s worth publishing, if only to document the results given that the concentrations of mRNA detected were in line with the previous study, but in JAMA Pediatrics, even as a Research Letter? I suspect that the editors knew this study would be good clickbait particularly given how they cited it:
I would also mention that the amount of mRNA detected in breast milk was incredibly tiny. I was about to do the calculations, when I saw that an immunologist already did them for me and put them in a form that I now wish I had thought of:
To address the concern that somehow the evil lipid nanoparticles getting into breast milk must be poisoning our babies, there’s this:
Of course, where the rubber really hits the road is whether the finding of COVID-19 mRNA in breast milk is clinically relevant; i.e., whether it does any detectable harm. I found multiple papers in my searches for other studies measuring mRNA levels in human breast milk that looked at this very question, for example this study from Singapore that looked at 88 mother-child dyads and found no evidence of problems. However, one of the more comprehensive reviews can be found here, which notes:
Only a small percentage of milk samples from women who received an mRNA vaccine contained trace amounts of mRNA. Thirty-six of 40 milk samples in one study and 5 of 309 milk samples in another had detectable mRNA levels; the highest concentration found was 2 mcg/L in one study and the median concentration was 70 ng/L in another; mRNA has not been detected in the serum of any breastfed infants.[18–20] mRNA has an estimated serum half-life of 8 to 10 hours.[18,19] The tiny amount of polyethylene glycol-2000 in Pfizer-BioNTech vaccine is not found in breastmilk or absorbed orally, so breastmilk PEG exposure from maternal immunization is not a concern. Neither of the mRNA vaccines available in the US contains a preservative or adjuvant.
Also, Ed Nirenberg also led me to a large study:
Basically, we have copious evidence that the finding of vaccine-derived mRNA in human breast milk is uncommon, transient, and without evidence of harm, as well as evidence that none of the other components of the vaccines cause harm. This result is, of course, unsurprising, given that we ingest RNA and DNA all the time from our foods. Every cell in every plant and animal contains RNA and DNA. Moreover, RNA is notoriously unstable, even the modified RNAs used in the vaccines, which are only less unstable than unmodified RNAs, not particularly stable. DNA is more stable, but even DNA has a problem tolerating the conditions in the digestive tract. Acids from the stomach will attack the nucleic acids, and there are a number of digestive enzymes such as pepsin and nucleases that will break RNA down very quickly. As Ed Nirenberg also notes, there are exceptions, such as vaccines for viruses that primarily infect the digestive tract (e.g., poliovirus) and are therefore resistant to the harsh conditions there, but COVID-19 mRNA vaccines are not those exceptions.
Finally, one can’t help but be amused at the inconsistency of antivaxxers. Does anyone remember back when they were castigating PCR-based COVID-19 tests because they involved going out to 40 cycles, which, antivaxxers claimed, produced many, many false positives because that many cycles made the test too sensitive? That was the origin of the “casedemic” conspiracy theory, which falsely claimed that there really wasn’t a pandemic, just a “casedemic” of diagnosed cases of COVID-19 that were meaningless because there were so many false positives. I just can’t help but point out that antivaxxers hate very sensitive PCR assays and blame them for a “fake pandemic”; that is, until very sensitive PCR assays find tiny, clinically irrelevant amounts of vaccine mRNA (probably fragments of it) in the breast milk of vaccinated mothers that persist no more than at most three days after vaccination. Then they decide that PCR is great because it lets them spread the lie that vaccinating lactating individuals is poisoning or babies with the deadly spike protein.