After analysing the Ebola crisis in a previous post, I
reasoned that a single drug would be unlikely to control the crisis due to its
lack of supply, the lack of infrastructure to transport and distribute the drug
in a controlled setting and the long periods of testing that have to be endured
by potential drugs. In this post, I will analyse the ZMapp drug (the so called “secret
serum”) in greater detail although I still maintain that the lack of supply and
wide spread testing of the drug will limit its immediate use in Ebola stricken
countries.
The Drug
The ZMapp drug consists of a “combination of monoclonal antibodies
“(originating from a single immune cell which has been cloned to produce
identical antibodies). This unites two previous cocktails of antibodies, MB-003
and ZMAb. According to a research article published in the journal Nature, this
combination was shown to save the lives of all of the rhesus macaques monkeys
tested in the trial. Yet more promisingly, the drug showed that late symptoms
of “elevated liver enzymes, mucosal haemorrhages and … pertechia (red blood
spots beneath the skin) could be reversed”. This is particularly useful as it
allows symptoms to be seen which means that individuals could be reported for
treatment by communities and non-healthcare professionals. This is important in
the general dispersed living conditions of Western Africa where screening for
the virus is not capable so detection in earlier stages is not an option.
MB-003 was described as “partially efficacious” by the
research team, yet this was only administered therapeutically after the
detection of disease “triggers” which delays the treatment and also limits its
widespread use. Similarly ZMAb added to an adenovirus adjuvant therapy provides
full protection in rhesus macaques yet only up to 72 hours after infection. The
ZMapp drug however, uses a combination of these treatment drugs whilst adapting
the murine based antibodies in the ZMAb drug so that they would last longer in
humans by slowing breakdown in the body.
The Research
Initially, the research group tested on guinea pigs,
comparing the efficacy of the adapted ZMAb drug, MB-003, the individual antibodies
that comprise MB-003 and a control (Phosphate Buffered Saline). Survival and
weight loss were tracked over the period of the trial which was 28 days. During
this test, it was shown that c13C6 and h13F6 (individual antibodies) gave a
survival rate of 17% (as shown above). These drugs tests were repeated in monkeys yet ZMAb wasn’t
tested. Here c13C6 again showed the highest individual success rate with 1/3 of
monkeys surviving the virus. Thus c13C6 was incorporated as a central antibody
in the new ZMapp cocktail. However, this can be questioned as the success rate
of MB-003 (which included c13C6) in guinea pigs was significantly less than in
primates, the MB-003 drug saving 1 in 3 primates yet none of the guinea pigs.
This implies that the success of c13C6 in guinea pigs may have been cancelled
out and so this would imply that certain combinations including the antibody
may limit the therapeutic effect. Furthermore, the wide variation in the
success of MB-003 across guinea pigs and primates implies that the Ebola virus
may be more susceptible to mutation and hence resistance in guinea pigs. This
undermines the potential of such drugs in humans. On the other hand, the
primate data is arguably more valid as humans are more genetically similar to
primates and hence this data is more applicable. Therefore the success of c13C6
in primates when in combination with other antibodies, suggests that the virus
that primates are susceptible to responds to the drug and hence a similar
result may be expected in humans.
Another slight criticism I have of this selection is that it
ignores the efficacy of h13F6 which showed equal success in guinea pigs. While
I understand that the lack of apparent effectiveness in primates is more
important, the interpretation of the data is subjective as if we were to select
based on guinea pig data then in situations where survival was similar we may
look to those cases of death and compare which gave the longest survival. Alternatively
other researchers may argue that the least weight loss signifies a potential
palliative effect. Thus based on “time to death” h13F6 had a time of 10.2 days
in guinea pigs, whereas c13C6 had a time of 8.4 days. Yet based on weight loss,
h13F6 fared worse with a loss of 21% in weight compared to only 9% with c13C6.
Thus this research could be improved by a clearer definitive statement over
which results they prioritise to explain the selection process of the drugs.
Ideally however, two branches of trials should have been conducted to assess
this discrepancy further with c13C6 as the starting antibody of one branch and h13F6
the starting point of another.
Hence, three new drugs were created, ZMapp1, ZMapp2 and
Zmapp3. This is useful in trials such as these because it allows subtle
variations to be made to an initial drug. This ensures that the maximum
efficacy can be achieved which is more cost effective as it means that drugs
companies don’t miss out on a slight variation of a drug that could have
offered a larger advantage. This means that the most ‘efficient’ drug can be
found. From the guinea pig data collected on these three variations, the
researchers concluded that ZMapp1 gave the greatest effect saving 67% of the
guinea pigs, followed by ZMapp2 which saved 50% and then ZMapp3 which had no
statistical difference to the existing drug, ZMAb. Thus ZMapp1 and 2 were
carried forward to be tested on monkeys. Arguably, the reasons for the distinct
lack of effectiveness of ZMapp3 over the other two forms could have been
identified which due to the cocktail variations must have arose due to
conflicting effects between the drugs. More investigative research into this
could have explained it further and hence allowed the researchers to consider
whether the maximum effect was being achieved in the other two forms. Consequently,
this would have enabled potential further production of varieties of the ZMapp1
form for example to achieve the most efficacious response.
Yet the researchers changed the cocktail of ZMapp1 by
swapping on version of an antibody with another similar yet not identical
version. This is important as it shows the inconsistency in the chemicals across
the guinea pigs and primates and so this means that the results seen in the
guinea pig data for ZMapp1 may not be representative of the drug seen in the
primate trials. Although this is only a slight change, it deserves attention as
it undermines the reliability of the data and therefore this means that the
initial tests of ZMapp1 lose, in part, their validity. Furthermore, the
apparent repeatability across both tests is questioned and so suggests that we
cannot assess the repeatability.
ZMapp NHP results coloured based on when they received the drug |
Non-human primate (NHP) trials showed that 100% of the
monkeys treated with ZMapp1 survived whereas 5 out of the 6 monkeys tested with
ZMapp2 survived. Both of these drugs were better than the control monkey
treated with Phosphate based saline and also the control monkey treated with a
control antibody. Both of these controls died. Initially this would seem to
suggest that ZMapp1 was the more effective than ZMapp2. However, the monkey
that died in the ZMapp2 treated group may have had a natural susceptibility to
the virus. Hence this would suggest that the death was due to heterogeneity as
opposed to the inadequacy of the treatment. Adding to this, the clinical score
for the NHP that died was higher than the others. This is a scoring method
which determines the severity of the symptoms and hence could be used to
predict the death of organisms. This suggests that this monkey was already
vulnerable to the virus due to the large degree of difference between it and the
other NHPs treated with ZMapp2. Furthermore, only 6 NHPs were tested in each
group. Hence the natural level of this susceptibility couldn’t be identified. This
means that the results do not take into account this level of susceptibility in
the population which may mean that the NHPs treated with ZMapp1 were
misrepresentative of the population, this means that it undermines the validity
of this conclusion because the conclusion assumes that the development of
symptoms were solely dependent on the drug as opposed to other factors. This
means that it may not have been a fair test. The number of NHPs tested is also
an issue because it means that the results are less reliable as the result as a
whole is more likely to be influenced by chance or genetic misrepresentation as
I have already discussed. In addition, the animal that died had elevated levels
of “alanine aminotransferase, total bilirubin, blood urea nitrogen and creatinine,
in addition to decreased levels of glucose, suggesting multiple organ failure”.
When the data for these levels are plotted on a graph, this particular individual
follows the path of the control groups closely which again infers that this
animal was ‘resistant’ to the drug due to a lack in any meaningful change.
Despite these issues, ZMapp1 was considered the most
efficacious drug and hence was carried through to the next stage of testing.
ZMapp1 is now trademarked as ZMapp. In this final test stage, the drug was
given to sets of NHPs at different times, e.g. one primate group received it on
days 3, 6 and 9 whereas another group received it on days 4, 7 and 11 etc. Yet
it was found that all animals survived the infection (shown to the left). Nevertheless, in two primates (named E4
and E6) who were treated last, their clinical score peaked close to the limit
that would mandate euthanasia. This again suggests a level of natural
resistance and hence supports my view that ZMapp2 should not have been
discarded purely based on its survival rate. In addition, E4 incurred a severe
rash on 40% of its body surface between days 5 and 8.
On the other hand, this genetic susceptibility is
overcome even in E4 and E6 and although the NHP that begun treatment first
shows the most positive result (due to a lower amount of viraemia and less
advanced symptoms) by the end of the trial there is no discernible difference
graphically in the clinical scores of the NHPs hence suggesting normalisation.
This is significant as it could infer that the drug was able to reverse even
advanced symptoms in later test groups. This can be seen in the results for the
F group which was treated last. The peak in the clinical score for this group
was one of the highest which can explained by their later treatment. On the
other hand the length of time this level took to descend is a useful measure of
responsiveness to a treatment (this can be seen in the partially resistant E4
and E6 which took a longer time to decline. This supports the correlation as I
believe that these animals were resistant and so their longer decline could be
linked to this. From this, one can conclude that perhaps time taken for decline
represents responsiveness). Therefore the rapid decline of the F group (as shown in the graphs above) shows
that the drug remains effective at later stages. This cannot fully be assigned
to the drug though as this may be part of a natural decline yet we cannot test
this as control groups died before this stage.
Will it be Useful?
Even so, the form of Ebola currently spreading in West
African is a form called EBOV-G (the Guinean variety) whereas this trial uses
EBOV-K as the virus to be treated. This poses a problem in the application of
these results to the current crisis. However, the researchers explain that the “epitopes
(targeted areas of the virus) targeted by ZMapp were not mutated between the two
virus variants” and hence the antibodies should “retain their specificity”.
Nevertheless the group admits that the susceptibility of survivors to
re-infection is unknown yet I don’t believe that this is a substantial issue as
the aim of this particular drug is to treat those infected with the virus and
hence has an immediate benefit. Therefore, the susceptibility after treatment,
although advantageous if reduced, is not a massive drawback if it is not.
My Verdict
Overall, these results are positive and the method used is
reliable and valid. The only issue I have with the method is the limited aspect
of the testing which is an issue as it means that genetic variability and
natural resistance are not taken into account. Furthermore, it means that the
results are vulnerable to chance which may mean that the result obtained is not
reliable. This is an issue because, when considering that this drug could be
used as a forefront method of treatment of Ebola it means that the small test
group size may misrepresent a wider response to the drug and hence may cause
health to be threatened due to adverse reaction for example which may not have
been seen in prior tests. In addition I maintain my previous viewpoint despite
the high potential of the drug. This is because the drug is not yet being
considered for passionate use and instead wider trials in human volunteers will
occur. This, the researchers estimate could take months (a significantly
shortened timescale due to the urgency of the situation yet still long-enough
to avoid depending on a drug based control system). To conclude, the drug in my
view shows massive potential particularly over the current treatments and the
control group yet is not yet to be considered vital for the fight against Ebola.