False diagnosis

A detailed explanation of why the genetic diagnosis given by the hospital was fabricated and false.

The CARS2 gene is known as autosomal recessive, it is a gene expressing for a protein, one copy from each parent is translated to RNA and then transcribed for two copies of the same/similar protein to perform a single function. In this case to carry or “charging” an amino acid “cysteine” within the mitochondria.

Another example of a protein might be POLG as mentioned by Dr Sampaio, again manufactured in the nuclear genome and serves to repair errors in copying when the mitochondria divides and replicates. Mitochondria need this protein as they have an error rate 100x that of nuclear DNA.

The Mitochondria are the “powerhouse of the cell” or engine where the energy is produced (energy transportation in the body is done via ATP), a symbiotic relationship in Eucaryotic cells (cells found in animals).

Mitochondrial DNA is a separate enclosed ring away from the Nuclear (Genome) of DNA. DNA is the blueprint and makes RNA. RNA is a transportable near copy of this, and carries the instructions of those blueprints to be made into proteins.
The small mitochondrial DNA ring is not able to independently produce all of the proteins needed for functionality; so, mitochondria rely heavily on imported nuclear gene products

You inherit your mitochondria from your mother – it only has 37 genes (to make 13 proteins). It is served by proteins mostly from the nuclear genome (a copy comes from each parents) to produce the energy for the entire body

To summarize, the CARS2 gene is the DNA code to make a string of amino acids, assembled into a protein that is a ‘Brick delivery truck’ to carry a particular “building block or brick” (amino acid – Cysteine) for when the mitochondria divides and builds itself a duplicate version of itself. As a form of redundancy, the nuclear genome produces two of these delivery trucks. Only one needs to be fully functional. In fact it only needs to be partially functional (this is called missense), depending on the specificity and frequency for that particular “brick”. A missense function is often a benign, but with a noted difference to the normal type (called the wild type)

POLG is a gene to make the protein that resolves many of the mistakes made in this division and rebuilds the mistakes in Mitochondria.

The genetic report from the Hospital informed the parents they both carried a mutation in the CARS2 gene. The mothers mutation passed to Alina was confirmed from a previous example (from comparison of offspring from a related marriage where the same protein is expressed identically – a family in Turkey, this is called homozygous)

Another example was given of unrelated parents in Sweden, but the proteins were slightly different from each parent (this is called heterozygous – one was the same as the Turkish/Maternal protein, the other was unknown), but still causing a similar illness type and both differed from the wild type

However, Alina’s fathers protein was an unknown variation seen only 7 times before in a database of 230,000. In order to establish this was a genetic illness, the hospital needed to prove this as deleterious like Alinas mother (i.e. both delivery trucks were faulty, and beyond functional) to qualify this as a Mitochondrial Gene defect. Since there was no demonstrated association of this missense gene with illness, the Hospital Genetics department chose to model it using computer software (in-silica)

So, to concentrate on an explanation with the paternal gene – since only one copy needs to be functional. Each variation is given a letter to represent it below). Also, a comparison to typical an obvious trait of this gene defect, the lifespan of the patient to demonstrate how essential this particular “brick” is. You are born with your nuclear and mitochondrial genome, they don’t change throughout your life.

To run a genome test, a full set of genes (both copies) was taken from Alina. This is then compared by overlap with the genome of each parent who contributed half each. This determines the source of the DNA error.

If W = Wild type protein (no consequence), Q = Turkish/Maternal fatal protein, J = Scandinavian fatal protein, P = Paternal protein. X = unknown, could be W, could be anything.

WW = good = both trucks are working

WQ= good = only one truck is working

QQ = 20 year life span

QJ = 10+ year life span

QP = unknown. Alina’s variation.

In the database of 230,000 people the following results were noted:

Q is listed 3 times and has 2 fatal submissions associated (QQ and QJ) – so its pretty accurate database, out of 3 submissions of a suspected pathogenic protein variation from 230,000+ entries, its confirmed 2/3rds of them (1/3 must be like mum; QX, and she is alive and healthy of course).

Given the varying abilities and education of the Medical staff around the world and such a small set of data, that’s very accurate!

P has 7 submissions, yet none are known to cause issues. Its unlisted as a recognised and known issue (variable insignificance).

The only evidence for P being fatal is the pictorial heuristics of a smaller neutral Amino Acid is replaced by larger charged Amino Acid (missense) in the protein chain. And…

It happens to be different to the normal wild type (W)

Since 2/3rds of the Q variations are fatal proteins, you would guess therefore that the data would offer roughly 1/3rd of 7 examples of P with associated fatality by combination with itself, like PP? Even in a tiny dataset maybe at least one out of 230,000+ samples The linear expectation would be about 2

The answer is none of course.

Maybe from an alternative angle. Since Q has been associated with other protein variations as fatal (itself Q and also when combined with J) then maybe P should exhibit the same with a fatal outcome not only by itself, the equivalent such as PQ or PJ… out of 230,000+ examples

The answer again is none.

I’m also going to assume given the expense, that a larger majority of the gene testing is for investigation of an illness rather than not, then its very safe to deduce that there are slightly more other people with the P+ gene combination than this ratio as well. The chances are that the P gene features on this database is more likely because it’s a casual observation from another unrelated disease, or from random sample testing, but most certainly not associated with Mitochondrial defect. There are possibly thousands of PP, PQ or PX variations walking around and enjoying their day not knowing on computer they should be dead

This is a phenomenon known as the “Birthday Paradox”, the chances of two people in a room at a party sharing a birthday. With each additional person arriving, the odds of this happening are not linear, but exponential (since they can share a birthday not only with the person hosting the party, but all others who are already there). If after 230,000 samples to compare against and the P gene variation is not linked to a fatal outcome, it is highly unlikely that it ever will be. As there were just two variations of fatal CARS2 genes, then you can assume there are many more also to be discovered, so increasing further the possibility of both genes being faulty when linked with P

The database seems effectively tiny enough for the Hospital to use in confirmation of a disease in-silica, but too small to be used to disprove it

I will go a little further into the accuracy of in-silica predictive models as I’m sure it will become a focus, since the Hospital chose to use this diagnostic method as they could not use known examples of P being pathogenic/deleterious, because there aren’t any

ACMG guidelines, the governing resource on in-silica predictions of proteins “…In general, most algorithms for missense variant prediction are 65–80% accurate when examining known disease variants. Most tools also tend to have low specificity, resulting in overprediction of missense changes as deleterious, and are not as reliable at predicting missense variants with a milder effect….” This is posted on the ACMG website as an overview of in-silica modelling.

I trust the points are noted, firstly, its not a percentage to consider as a valid diagnosis from a computer, merely a suggestion of missense. Proteins have a great deal of plasticity, they are not rigid

Secondly, this is for known disease variants, the reverse is not necessarily true as subject to a lack of individual reference data (hence overprediction), experimental bias, as well as experimental error.

Plus of course the overall effect may have been as mentioned, “milder” and something a specialist would not be able to distinguish without some supporting evidence. A mild effect might be considered closer to a strong trait. Which in this case, means it takes many years to develop (as is the case on even the case examples given).

https://www.nature.com/articles/gim201530

The genetic report develops to link this missense suggestion close to an area of ATP binding by vicinity, but not by a known interaction. This is a demonstration of confirmation bias. It is a speculative & supportive opinion, not a computer result.

Mentioned in the report of this gene also is comparison to the CARS2 gene as being highly conservative references to several animal studies.
You can ask how many animals were tested for this gene defect comparison? Surely not 230,000 to confidently pick out a missense mutation present in 0.003% of the sampled population of humans. You might consider this inclusion passed off as a matter of fact, is really academic padding of a very weak argument.

To quote the Coughlin 2015 report on CARS2, one of the reports of a child born with this error “…His early clinical course was marked for a delay. His highest level of develop-mental function included ambulation with the assistance of a walker, hitting a toy and babbling…”

Alina used to play on her electronic drums everyday, she understood two languages, and was well ahead of her peers at nursery in playing tricks against them, ones that even fooled adults. There are plenty of videos to show this; She was thriving as a child, other than a documented repeating viral infection

The clinical aspects of the disease progression can be explained more clearly by an alternative disease, a chronic Enterovirus, which is entirely in line with all observations, responses, progression and medication.

To expand further, the Boolean aspect of deduction: to review not the overlapping clinical features, but the observations that do not align. The first and most obvious is the disease rate progression, namely Alina seemed to deteriorate to beyond recovery in little under 6 weeks from admission, or 12 weeks from the first seizure in the throat and lip. The 2 examples given by the hospital progressed over several years, with a distinct feature, microcephaly (which Alina did not have – her head circumference was only 7mm less than average). Once you remove the clinical symptoms of an Enterovirus from the report, the only overlap of the one example of a child suffering this genetic defect is “mild gross motor delay”, a trait shared by both Alina’s parents in their youth and not considered by them or any medical professionals prior to admission as being a concern, but something the Hospital chose to amplify and rely on in support of its very desperately needed diagnosis to support a Doctor who failed to properly review her medical history, withheld significant observations, had very little knowledge of the testing equipment, and especially after the trial of the autoimmune drug Anakinra had made very clear improvements

A diagnosis that co-incidentally was the original pursuit of proof by Dr Sampaio, but then abandoned by Dr Cardamone as Alina was so responsive to the ketogenic diet

Since the parents were presented this diagnosis, an additional report has been compiled of 4 further cases of CARS2 gene defect (2020). Unsurprisingly all the patients are alive in their 20’s. The paternal defect does not feature in any of these missense variations either

Incidentally, cysteine (the amino acid “brick”) is required for Keratin production, and hair is made up of 14% Cysteine. An observation was made with these additional cases was the presence of “tiger tails” and ‘brittle hair”

https://onlinelibrary.wiley.com/doi/full/10.1002/fsn3.1818#fsn31818-bib-0038

American Journal of Human Genetics report (PDF, 2Mb)

The parents, being sentimental happened to keep a lock of Alina’s “crazy hair”. Under 50x magnification (with thanks to a local optician’s microscope), it of course does not show the distinct “tiger tails” – the final proof that the CARS2 gene defect was valid

Alina’s hair x50 magnification (below)

The summary of this genetic report is very confidently proven as false. Fabricated and contrived to fit the lack of diagnosis and total mismanagement, and something that should have been very clear to the Genetics team led by Prof. Edwin Kirk, who was ‘very likely’ aware, so included the words “quite cautious about the reporting” in his summary in the Hospital Notes. However, he still wrote a letter to the parents informing them that in order to consider children again, they would need IVF to exclude this gene error. A procedure that was unsuccessful and ultimately caused the failure of the relationship.

Sydney Children’s Hospital maneuvered “its problem, to make it the parents’ problem”, and shamelessly told them when they were incredibly vulnerable and defenceless facing the death of their only child after a 5 month battle for her life in hospital, both parents exhausted beyond measure as they made it their sole promise to never let Alina be alone.

The false genetic diagnosis was more beneficial to the Hospital than the parents; excusing Dr. Sampaio and the Neurological “team”, their poor diagnostic techniques, the peer review consultations, the unnecessary tests and medications, suffering and cruelty, as well as any legal complaint. They had invested and committed to Dr Sampaio’s belief this was genetic Mitochondrial, that they had little option but to make it so, regardless of the lack of ethics, honesty and disclosure. By forcing a “genetic” diagnosis, the Hospital absolves itself from any responsibility for the outcome of Alina, as well as the emotional trauma experienced by the parents, implying “it simply wasn’t their fault!”