Written evidence submitted by Phil Robinson and Jon Curtis (CLL0055)
About the authors
Mr Robinson is a former founder and Managing Director of KBiosciences ltd. Previous to this he was employed as head of genetic technology development at GSK. He has also wide experience of liquid handling, automation and drug discovery whilst also at GSK and Amersham International. He has over 30 years experience of working in and leading laboratory teams. Additionally he has a number of patents and involvement in many thousands of scientific publications predominantly in the field of mass genetic testing.
Mr Curtis is the second founder and chief Automation Officer of KBiosciences ltd. Previous to this role his role at GSK was one of implementation of technologies across the gamete of disciplines in the Drug Development world of GSK. Mr Curtis was one of the first engineers to bring modern automation to the Genetics world where he was the pioneer of ultra-high throughput PCR at levels of up to 1 million reactions per day. Additionally he has a number of patents and involvement in many thousands of scientific publications again in the area of mass genetic testing.
Mr Curtis and Mr Robinson are both retired individuals who founded KBiosciences and grew it into a world leading genetic testing business that 10 years ago was performing 1 million tests per day with a team of less than 50 members of staff. The testing process required the invention and development of a number of industrial robotic and biological solutions creating the worlds highest throughput genetic testing pipelines. These were so successful that a significant number of laboratories worldwide wished to use the same setup and thereby purchased these robots and reagents. They are still being used worldwide today. At last count over 4900 publications cite these technologies. The business was purchased by LGC genomics in 2011 and Mr Robinson and latterly Mr Curtis retired. It should be noted that both Mr Curtis and Mr Robinson have no commercial affiliation. They are both experts in the mass PCR testing field who volunteered their services (for free) solely for the good of the UK.
When the Covid-19 Pandemic broke it became immediately obvious that their skill set was desperately needed and they volunteered to help. The Lessons Learnt Feedback herewith comes from direct experience of the Lighthouse Labs.
Opening Statement
The Government, PHE, HHSI and the DHSC have made a number of serious mistakes that has cost the lives of many people. They and the scientists put in charge have consistently ignored constant advice from the experts in mass genetic testing. If they had accepted this advice as it was given in April / May and followed the guidelines set out in our proposal for testing the entire population (www.popmaxcovid.com) on a regular basis the UK would not be in the complete mess it is in currently. It is a statement of fact that only now 8 months after our proposal and offer of help (on repeated occasions) that mass testing is being deployed with inferior lateral flow testing and the new Lighthouse labs are being built in Leamington Spa and Scotland finally utilizing the techniques we proposed in April. These should have been deployed in the original Lighthouse labs in May / June, but yet again we were ignored. Finally it should be stated that the recent impressive testing regime in Wuhan where they tested the entire population in under 10 days was using our proposal which we sent to them. We have the evidence to support this statement by way of the audio from the Zoom call that was carried out with their testing heads. It therefore cannot be over stated that the UK Government and testing heads must be held accountable for missing the biggest and best opportunity for keeping Covid-19 infection rates very low. It is unforgiveable.
Evidence for Lessons to be Learnt
All statements in this section are supported by documented evidence (either, email, audio or phone texts). Should the enquiry wish to see this or to have us speak then feel free to request it.
The Covid-19 PCR test is the gold standard for detection of the virus in nasopharyngeal swabs. It would be correct to therefore employ the leader of the Lighthouse Labs to have some knowledge or at least to have someone in their line that had some experience of this type of ultra high through put testing (there are not many people with this experience but they do exist). The appointment of Chris Molloy and the subsequent appointments that he made are wholly inappropriate. Prof Molloy is a great scientist but from the completely wrong discipline. He is an expert in the field of Drug Discovery not Genetic testing. When directly questioned he freely admitted that he had no experience of mass PCR testing (in fact he admitted he had not even carried out a single PCR). This was the first huge lesson to learn. Appoint a scientist who has experience of the testing process, or at least appoint someone in the line structure who knows the process inside out. The appointment of Prof Molloy caused the cascade of appointments of drug discovery scientists all the way across the testing lab management. It is not surprising that Prof Molloy appointed people he trusts but again they had no experience of high throughput genetic testing. This is forgivable if the mistake is realized and rectified early on, but even though this was pointed out a number of times it was not acted on. Indeed, Mr Curtis and Mr Robinson extended multiple offers to work for free to help improve the testing labs to the standard and throughput they should be.
Testing large numbers of samples (millions) is perfectly possible and has been demonstrated by us in a commercial setting for many years. It is fair to say that we are one of a handful of world experts in genetic testing at this level of throughput. There are many scientists in the UK who have dealt with 1000’s of samples a day such as Alan McNally who designed the lighthouse lab setup. It is a very different proposition to set up a lab to process 100,000 – 1million samples per day at fast turnaround rates. The technology choice makes a huge difference to what a lab can achieve. Constant technological advice and offers of implementation from us were consistently ignored when we already had the experience of running a lab at ultra high throughputs.
The testing process can be broken down into a number of steps. Sample receipt, Sample transfer, RNA extraction, PCR amplification, Detection and Analysis. The inexperience of the management team led to poor technical choices as follows. Again, it should be noted on the public record that these poor choices and the improvements that should have been made were communicated to the Lighthouse management team and not acted on.
Sample receipt and Sample Transfer.
This is the most difficult part of the process. When visiting the Lighthouse lab at Milton Keynes in April it was apparent that there are over 20 types of tube being received by the labs. In a high throughput environment this is the biggest lesson to be learnt of all (mistake). Laboratory robots require the same type of tube to be used at all times. In the early days of testing, tube supply was short, but this should have been the first problem to be resolved. It required strong management to refuse to test samples in the incorrect tube to reinforce this behavior from the sampling centres. You can go further than this and suggest that the management should have forced the suppliers to all standardize their tube formats thus ensuring supply going forward.
The second lesson to learn for this stage is that automation is the key. Still to this day the lab scientists are manually interacting with tubes, in that they are unpacking, de-capping and placing them on the deck of the liquid handling robot. We identified this problem in April and sent recommendations to the light house labs for how to solve this. Indeed we even offered to set it up for them. Again we were ignored. The importance of this stage should not be under-estimated
Plate Density Choice
Another of the fundamental problems in the testing labs was the reticence to switch to a 384 well plate density. The current testing is still predominantly in a 96 well format, with the exception of Leamington Spa which will use 384 well plates when it comes on line. The use of 384 well plate technology improves throughput fourfold with little extra effort. It also has a concomitant saving in reagent cost of fourfold. Current published costs for the PCR phase of the testing process are about £5 per reaction. This would have dropped to £1.25 per reaction x 500,000 / day (current quoted testing level) = £1.375 million saving per day!
Additionally the RNA extraction phase is subject to the same argument. Current RNA extraction is carried out using the Thermo Kingfisher system. There are industrial alternatives that are scaleable into 384 well plates that again have been ignored. The same throughput and cost argument applies.
PCR amplification and detection
Initially the use of Real-time PCR was mandated, but again in April it was suggested to Lighthouse management that this part of the process be switched to end-point PCR. Briefly, end-point PCR reads the exact same reaction at the end rather than at every stage. The reaction and reagents are the same. It was madness to not make this small switch in April. Indeed we took it upon ourselves to prove that this would work well, in fact (using our 20 years of experience) we were able to improve on the MK data significantly in three days experiments. The data was submitted and again ignored. This was yet again unforgivable, especially as end-point PCR diagnosis has been used in an industrial setting for at least 20 years. The advantages of using end-point are many, but the two headlines are that the data analysis is simple and the throughput gain is massive (with a concomitant reduction in required manpower and costly instrumentation). The Real-Time PCR data analysis is still being carried out by a third party at a significant cost to the Taxpayer. Switching to end-point would have negated this need and sped up the reporting of results to the individual.
Informatics
When visiting the MK labs it was clear that data was being manipulated and reported using MS-Excel. The lab should have had a commercial Laboratory Information management System (LIMS) in place from the start. This was not the case. Again for a certified lab this is unforgivable. The mistake made with results not being reported back due to MS-Excel having a 65k entry limit would not have happened if an industrial strength LIMS was in place. In fact Dr Cox, head of the MK lab should have been aware of this as his previous role was head of Bionformatics at the Sanger centre. It should never have happened.
Summary for Technology Choices
The Leamington Spa and Scottish labs that are due to come on line shortly have finally taken notice of the points raised above but it is too late to impact the Covid-19 second wave. The overarching lesson to be learnt here is to listen to the experts and have a highly active process improvement team. The testing labs could have been so much more effective at a significantly lower cost to the taxpayer.
The government testing strategy must be called into question. Covid-19 is the perfect virus for wreaking havoc in a human population. It only kills a small subset of people infected, and in a large percentage of people it shows no-symptoms. These asymptomatic people unknowingly spread the virus, hence making it extremely difficult to control once in a population. Without a vaccine, social distancing and testing are the only real tools available to control its spread. To date the testing capability of the country has not met the demand.
It must be strongly argued that mass testing is the only real tool capable of finding the asymptomatic spreaders of the virus. Again it must be argued that PCR testing is the most sensitive and gold standard method for mass testing. The Antigen based lateral flow testing is widely reported to only find about 60-70% of people carrying the virus, and thus not ideal although some success in controlling viral spreading in Liverpool has been reported.
The only way to control the spread of Covid-19 is to test the entire population regularly, (we proposed by household in a pooled strategy). This is possible and was proposed in April (www.popmaxcovid.com). Indeed the Chinese have used this strategy to successfully test and control Covid-19 in the huge population of Wuhan and other cities. The question must therefore go on public record as to why this strategy was actively repeatedly ignored. It was sent to every minister in power, our local MP’s, and all the heads of testing in the UK on multiple occasions.
The lesson to be learnt from the ignoring of the proposal to test the entire population is to understand why it was so totally ignored. The entire proposal was submitted into the DHSC NHS-I portal for ideas to improve the Covid-19 response and seven months later is still sitting there showing under consideration. Please remember the authors of the proposal have carried out testing at population scales on many billions of samples for many years and are world leaders in the field, trying to help the country in its hour of need.
Summary and Closing statement
The Government and heads of the testing labs in the UK have fallen dramatically short in their implementation of a National Covid-19 Testing Service, in terms of strategy, management and technological choices when constant advice on improvement was given from experts in the field. The dire consequence of this is a larger number of deaths and a wrecked economy than if a population wide testing strategy was implemented. The authors of this document had the knowledge and experience to ameliorate or even prevent the huge spread of Covid-19 in the country, if they were given the mandate to deploy their proposal. It must never happen again that people who have the ability and vision to effect national disasters are ignored.
P Robinson
J Curtis
Nov 2020