Saturday, February 6, 2016

O1010 - Active Immunization Against Tetanus

This study looked at antibody levels in subjects immunized with various kinds of tetanus vaccine. So nothing that new. But the discussion was pretty interesting:
"The first dose of alum precipitated toxoid affects the antitoxin producing cells of the body in such a manner that following a second injection there is a rather prompt increase in the antitoxin content of the blood. A certain period of time must elapse between these two injections before this release of antitoxin into the blood stream takes place...The effect of the first dose of alum toxoid is rather unique in that it cannot be replaced by the natural occurrence of tetanus and recovery from it. We have had occasion to confirm Cowles' finding that a patient who had recovered from tetanus 10 years before still required two doses of alum toxoid before any demonstrable antitoxin appeared in the blood. Following the second dose, a few days elapse before a protective titer develops. Hence, if an injury occurs during the interval between the two injections of alum toxoid and for a week or two following the second injection, passive immunization may be necessary in order to get full protection against tetanus."
"The 'repeat' or stimulating dose of plain or alum precipitated toxoid produces within one week a remarkable increase in the antitoxin content of the blood. In most cases about five days elapse after the injection of the 'repeat' dose before the antitoxin titer is brought up or above the protective level. It is stated that the period of incubation of tetanus in man is usually from six to 14 days and is directly proportional to the amount of toxin and the severity of the disease. With a short period of incubation, six days or less, the disease is almost invariably fatal. Whether the mobilization of antitoxin that occurs after the injection of the 'repeat' dose is fast enough to prevent all cases of acute tetanus is not definitely known, but more than likely it will be able to do so, since our minimum protective value of 0.1 unit is a conservative one."
"From our studies to date, it would appear that active immunization against tetanus by means of the injection of two doses of alum precipitated toxoid followed by a 'repeat' injection upon the occurrence of an injury, will prove to be of value in military practice, in civil life where frequent injuries make the repeated injection of tetanus antitoxin impossible, and in the handling of allergic individuals who are sensitive to horse serum."
So that's interesting, and explains how things pretty much still work today, except that Td is given upon injury only if one hasn't received it for a number years.

Reference:
Gold, H. Active Immunization Against Tetanus. Ann Intern Med 13, 768–782 (1939).

Saturday, January 16, 2016

101 - Prophylaxis and treatment of whooping cough with a pertussis antigen

This was an unusual study on a vaccine-like method of protecting children from whooping cough. Instead of using whole killed bacterial cells, they grew the cells and filtered them out, leaving only soluble factors the bacteria had produced externally, which they called "antigen." Then they inactivated this antigen so it wouldn't cause any problems but would still induce an immune response, treating it with formaldehyde and mild heat. This was sorta like how tetanus and diphtheria toxoid vaccines were produced.

Then they tested this antigen for how well it could a) treat whooping cough cases, b) prevent catching the disease when given when not at risk, and c) prevent catching the disease when given after being exposed.

Antigen Therapy
They had two main groups: children with whooping cough treated with antigen, and control children with whooping cough. The controls were treated with either vaccine or vaccine + antigen. They don't specify what they mean by "vaccine;" details are pretty sparse in general.

The average time before each group was similar, about 1.5 weeks, but the duration of disease after treatment was about 1.6 weeks for the antigen-treated and 3 weeks for controls, so a pretty big difference.

The number of complications in the groups was different too: About 0.7% of the antigen-treated had complications from the disease (bronchitis, conjunctival hemorrhage, otitis media), while 18% of controls had complications, including some deaths

Antigen Prophylaxis
Of those immunized without first being exposed, about 81% of those who did become exposed didn't get the disease, so they said it was 81% effective. That's not exactly how such trials should work; there should be controls.

They claimed this immunity lasted at least 5 months and up to 2.5 years, but I'm not sure how they determined this.

The same was true of those treated after being exposed: 76% didn't get the disease, but again there were no controls reported.

In terms of adverse reactions, some few patients had a short, mild fever, arm redness, or hives, but nothing worse.

Conclusions
So considering the lack of details, controls, and blinding, and how this approach didn't seem to be considered worthwhile either before or after this study, I'm not convinced it's very valuable.

Reference:
Joslin, C. L. & Christensen, T. A. Prophylaxis and treatment of whooping cough with a pertussis antigen: Report of results. Am J Dis Child 60, 1269–1276 (1940).

Saturday, December 12, 2015

100 - A Comprehensive Study of Influenza in a Rural Community

Ok, here's another study that doesn't look at vaccines specifically, but rather the background of the disease they prevent, and how our bodies respond to it. In this case, influenza.

The researchers followed a small community of people in New York over the course of a flu season, looking at their antibody levels. The tests they used were neutralization (mixing serum with virus sample and seeing if the serum neutralized the virus so it didn't make mice sick) and complement fixation, though of the two tests the former seemed more useful.

In the period before the flu season, teenagers had the highest titers in the neutralization test, which were lower at younger and older ages. For complement fixation, the highest levels were in those 40 to 60 years old. Both tests observed a decrease above 60, in the elderly.

Then in the flu season, they made sure to distinguish clinical flu from the common cold or other flu-like illnesses: the flu was defined by a fever over 99 degrees F, headache, pains, general malaise, and respiratory symptoms. About 85% of the disease in the community was actually colds, which peaked in December, while the flu didn't peak until February. About 62% of the flu tested was found to be Influenza A, the rest other kinds.

So regarding the neutralization test results, having a cold didn't increase one's titers at all, but having the flu did (makes sense). Some had an increase without observing symptoms, possibly subclinical infections, too mild to be noticed. There was a definite correlation between having lower antibodies before the flu season and being likely to get the flu during the season, so titers correlated with protection. Though even those with the highest titers occasionally got sick.

One year after the season, some people's titers had returned to the same level they had been before the season (even if they got the flu), some stayed the same, the rest fell partway back. The average was 5 times more 1 year after than before.

So overall, it seems like higher titers correlate with more protection, but there's no really safe level that's 100% protective. And correlation isn't causation, so it's not clear the antibodies are doing the protecting, though neutralizing the virus seems like a good sign. Also I'm not sure how specific these tests are for specific antibodies, so they could show high levels of antibodies that aren't actually that useful, or something like that. Needs further study.

Reference:
Rickard, E. R., Lennette, E. H. & Horsfall, F. L., Jr. A Comprehensive Study of Influenza in a Rural Community. Public Health Reports (1896-1970) 55, 2146–2167 (1940).

Saturday, October 24, 2015

099 - Secondary Familial Attack Rates from Pertussis in Vaccinated and Unvaccinated Children

This study is a follow-up analysis of previous results by Kendrick and Eldering of a clinical trial in Grand Rapids, Michigan, of a whole-cell pertussis vaccine.

Rather than the whole study population, this study focused only on 165 families that were definitely exposed to pertussis from one of their own members. The other members' immunity status was known from history of vaccination or the disease itself.

What they saw was that, of all 78 vaccinated subjects, only 28 got whooping cough (36%). All these were under age 7. In contrast, of those without history of vaccination or disease, 79% got sick. Of those unvaccinated under age 7, 92% were attacked (the rate decreased with increasing age of subject). 36% vs. 92% for those most vulnerable? Not bad.

Most of the primary cases, that brought the disease to a family, were in the older category, confirming that older children often bring it to their younger siblings. Interestingly, with the subjects who had the disease before, 5 of the 6 that got sick were over 7 years old, so it seems there's a period after which the immunity is not so great, even after natural infection. The study was set up to analyze that in depth though.

Also noteworthy is that, of 172 primary cases that brought the disease to the families, 157 were unvaccinated (91%), 12 were vaccinated (7%), and 3 had the disease previously. There's some herd immunity in action, perhaps, but it's hard to tell.

This wasn't a very rigorous study (lack of blinding or placebo), but considering that an intimate exposure, such as in the same household, is probably the most difficult to have immunity against, there seems to be some effect. And apparently whooping cough is so contagious that 80-90% get sick when exposed to it this way, if not immune somehow.

References:
Kendrick, P. L. Secondary Familial Attack Rates from Pertussis in Vaccinated and Unvaccinated Children. Am. J. Epidemiol. 32-SectionA, 89–91 (1940).

Saturday, October 10, 2015

O984 - Tetanus Immunization

The author of this study (Philip B. Cowles) seemed skeptical of some claims made about the tetanus toxoid vaccine, so set out to test them.1

The first was the claim that 0.01 unit of antitoxin per milliliter of serum (the level of antibodies against the tetanus toxin in the blood) was enough to protect against the symptoms of tetanus. He used guinea pigs and mice to figure out how much antitoxin was necessary to protect. He seemed to find that 0.1 unit per mL was usually enough, though not always. As another author summarizes,
"This author felt that though 0.10 unit is not sufficiently great to protect all animals against a maximum infection, it is probably much larger than is necessary to care for many infections resulting from wounds judged to be too slight for surgical attention, and probably large enough to care for the majority of injuries that receive surgical treatment."2

The second question was how long does it take for a booster shot to raise the level of antitoxin. The common practice was to give a booster when a person presented with a wound of some sort that might expose them to tetanus infection, thinking that the vaccine would raise their antibodies enough to protect. But if that doesn't happen quickly enough, it wouldn't be very useful.

Cowles tested a number of people before and after a booster. Most had about 0.02 unit or less before, and it took 4-5 days to rise up to 0.1 or higher. Is this fast enough? Tetanus seems to develop in 7-10 days, but it's possible the toxin has started working before then, before symptoms show up. So more study would be needed to say a booster is definitely helpful, but it's pretty definitely better than nothing if an antitoxin treatment is unavailable.

As a final addendum, the author mentions the question of whether having tetanus makes one immune to further tetanus. There had been reports of people have tetanus repeatedly, so obviously there isn't always natural immunity, but Cowles describes a man who had recovered from a mild tetanus attack at some point, but didn't seem to have any antitoxin in his blood. So it seems like the tetanus toxin is toxic at lower concentrations than it is antigenic (capable of producing antibodies and immunity), so it's rare to become immune naturally without dying or something. Small sample size though.

References:
1. Cowles, P. B. Tetanus Immunization. Yale J Biol Med 9, 409–416 (1937).
2. Gold, H. Active Immunization Against Tetanus. Ann Surg 114, 1060–1068 (1941).

Saturday, September 19, 2015

O983 - Active Immunization Against Tetanus Infection with Refined Tetanus Toxoid

This study compared unrefined tetanus toxoid and also a refined version, precipitated with alum (aluminum potassium sulfate) in guinea pigs. The former gave 0.5 units antitoxin after 6 months, and the latter 1 unit after 3 months. They conclude the latter is better (though it doesn't quite seem like a good comparison to me).

Then they used this refined toxoid to immunize 30 people. None had reactions to it, local or general, which is good. Their antitoxin went from as much as 0.0005 to up to 0.04 units in 90 days. After a 2nd dose, levels got up to 9 units, but were widely variable between people. The highest levels 1 and 2 years after the 2nd dose were 0.4 and 0.35 units, respectively; high enough for protection, they judged.
"The amount of antitoxin found in the serum of human beings thirty days after receiving a second dose of the toxoid was considered to be sufficient to protect them by giving a third dose of toxoid, on injury, instead of a prophylactic dose of tetanus antitoxin."
There was some variation too on whether it lasted a whole 2 years protectively. It seemed to depend more on the individual's general health than on age.

The authors state that the Societe de Chirurgie of Paris committee recommends:
"vaccination against tetanus [is] the best method to decrease the mortality. The vaccine (anatoxin of the Pasteur Institute) is harmless, not being followed by any local or general reaction. The vaccination should be repeated every year or two...In war, vaccination should be made obligatory."
The authors recommend this vaccination for soldiers and others who regularly get injured (farm workers, etc), and also for pregnant mothers:
"The antibodies contained in the placental circulation before birth and the antibodies contained in the mother's milk, afford protection to the new-born child."
And the final recommendation is that if a person has been vaccinated with toxoid and gets injured (more than 1-2 years after the vaccination) they should receive another dose of toxoid to rapidly stimulate a response, just in case. But if they haven't received any toxoid doses, they should get toxoid and also a protective amount of antitoxin serum, because the toxoid won't stimulate a response in time, if necessary.

So that's what they say.

Reference:
Bergey, D. H. & Etris, S. Active Immunization Against Tetanus Infection with Refined Tetanus Toxoid. J Immunol 31, 363–371 (1936).

Saturday, September 12, 2015

098 - Anaphylaxis after Injection of Tetanus Toxoid

I once read an article questioning the safety and usefulness of the tetanus vaccine. This vaccine is made of the tetanus toxin itself, denatured such that it doesn't cause problems, only induces an immune response. Preparing it requires growing large quantities of the organism, Clostridium tetani, so components of the bacterial medium are involved in the production process. This is relevant, because some of the problems this anti-tetanus-vaccine article mentioned were anaphylactic shock from the vaccine, and cited several case studies to demonstrate this.1-4

So is anaphylaxis a real risk of the tetanus shot? Let's look at the data.

These four studies pretty much all say the same thing, with slight variations. The first, by Parish and Oakley, is a case report of a person who experienced an anaphylactic reaction after the second in the two-shot series of tetanus toxoid, though she hadn't reacted at all to the first injection, or to previous injections of diphtheria toxoid, typhoid, or Staphylococcus toxoid. It didn't seem too serious and she didn't have any breathing problems, but they gave her some adrenaline and the symptoms mostly went away. Everything was better after two days. This person was the only one to experience such a reaction out of 300 others that received the series from the authors.

The second study, by Whittingham, discusses how toxoid is a much safer way to prevent tetanus than the other option at the time, antitoxin serum, because 10-15% of subjects got something called serum sickness (an immune reaction to the serum, not as serious as anaphylaxis, though some got anaphylaxis too). And immunity was better from toxoid. But he reports that, out of more than 61,000 soldiers who got the tetanus toxoid series, 651 (about 1%) had local reactions, 14 (0.023%) had headache and body pains and such lasting up to 8 hours, and 2 (0.0033%) had anaphylaxis. So anaphylaxis occurred in 1 out of 30000. These two cases were fine after adrenaline injections.

The third study, by Cooke, Hamptom, Sherman, and Stull, was another case report of a reaction soon after the second injection of toxoid: the patient had hives, itching, and swelling, but was fine after epinephrine injection.

The fourth, by Cunningham, was a report of a nurse who had had a reaction to a diphtheria toxoid before: hives and malaise. This person had a reaction 3 weeks after the 1st tetanus injection, so it wasn't clear that it was connected at all; she was playing tennis and got hives and irritation. Adrenaline solved it. The second injection was given, along with some adrenaline, and the patient had more problems, but recovered after 24 hours. The author recommends keeping patients under observation for 30 minutes after injection.

So it seems that reactions happened, but didn't seem serious and were pretty rare, based on these studies.
"Anaphylactic reactions following tetanus toxoid injection have also been reported in the medical literature. Fortunately, sensitivity reactions to tetanus toxoid are very rare."5
"Systemic anaphylaxis has been reported following administration of tetanus toxoid also, but this is extremely rare."6

Relevant Component Removed
The other thing that all these studies talk about is that when the patients mentioned were tested for allergy to components of the vaccine, they all showed reactions to a particular component of the bacterial medium, a kind of peptone (digestion of protein). So it's this component that seemed to be the culprit. Fortunately, it seems that since these reports, steps have been taken to remove this component from production, thus greatly reducing the risk of anaphylaxis:
"From data collected early in the program, it appeared that the proportion of reactions of sensitivity following the injection of these toxoids was at least six times that following the use of toxoids free from these particular peptones.
"As a result of these findings, the procurement and distribution of tetanus toxoid containing Witte or Berna peptones were discontinued. As indicated above, the incidence of reactions of sensitivity has fallen sharply since that time."7

"The production of tetanus toxin on a medium free from peptone has been accomplished with one strain of Cl. tetani. As good, or somewhat better, toxin titers have been obtained as this strain repeatedly yields on peptone-infusion media."8
"Employed on a tremendous scale for military immunization, its initial preparation on the traditional media containing Witte's peptone has had to be modified because of the presence, in that brand of peptone, at any rate, of a substance inducing severe anaphylaxis in certain individuals. Other varieties of peptone are still considered admissible, largely because no serious accidents have thus far been reported following their use."9
"A formula is given for large-scale production of tetanal toxin on a medium initially free from antigenic components."10
"This observation, considered in the light of the experience of Cooke and Whittingham and associates with the sensitizing properties of Witte's and Berna's peptones in toxoids, led to cessation of the use of toxoids containing these peptones with a resultant much lower rate of reactions to tetanus toxoid. In the early immunization toxoids of these types had been used for about 75% of the injections reported and were responsible for 90% of the sensitivity reactions observed."11
"It should be noted, however, that such proteoses were eliminated from all toxoids by 1942."12
So overall, the reaction was rare to begin with, but much rarer after the chief offending component was removed. The original article I mentioned did talk about the peptones, but didn't mention how rare the reaction was to begin with (or how easily treated), and didn't mention that the peptones were removed from vaccines in the future. So I have corrected this misinformation.

Note that this is not a thorough review of the subject; there are other reports of reactions like this, some of which are probably more serious than these four, but I haven't gotten to those reports yet, except insofar as they address these four.

References:
1. Parish, H. J. & Oakley, C. L. Anaphylaxis after Injection of Tetanus Toxoid. Br Med J 1, 294–295 (1940). 
2. Whittingham, H. E. Anaphylaxis following Administration of Tetanus Toxoid. Br Med J 1, 292–293 (1940). 
3. Cooke, R. A., Hampton, S., Sherman, W. B. & Stull, A. Allergy induced by immunization with tetanus toxoid. JAMA 114, 1854–1858 (1940). 
4. Cunningham, A. A. Anaphylaxis after Injection of Tetanus Toxoid. Br Med J 2, 522–523 (1940). 
5. Buff, B. Fatal anaphylactic shock: Following intradermal skin test with dilute horse serum tetanus antitoxin. JAMA 174, 1200–1201 (1960).
6. Vessal, S. & Kravis, L. P. Immunologic Mechanisms Responsible for Adverse Reactions to Routine Immunizations in Children. Clin Pediatr 15, 688–696 (1976).
7. Long, A. P. Tetanus Toxoid, Its Use in the United States Army. Am J Public Health Nations Health 33, 53–57 (1943). 
8. Mueller, J. H., Schoenbach, E. B., Jezukawicz, J. J. & Miller, P. A. Production of Tetanus Toxin on Peptone-Free Media. J Clin Invest 22, 315–318 (1943). 
9. Mueller, J. H., Seidman, L. R. & Miller, P. A. A Comparison of Antigenicities of Hydrolysate and Peptone Tetanus Toxoids in the Guinea Pig. J Clin Invest 22, 321–324 (1943). 
10. Mueller, J. H. & Miller, P. A. Large-Scale Production of Tetanal Toxin on a Peptone-Free Medium. J Immunol 47, 15–22 (1943). 

Saturday, August 1, 2015

097 - An Experiment in Immunization Against Influenza with a Formaldehyde-Inactivated Virus

There had been some good results in animals with vaccines against influenza, and possibly some good trials in humans too, though not in others; influenza is tricky though, due to antigenic drift.

So this was another study, in Hungary during an epidemic in 1937. They used formaldehyde-inactivated virus taken from infected mouse lungs, and vaccinated 306 nurses and children in Budapest institutions with a single dose. 336 were controls.

They tested some subjects before and after the vaccination, and found that antibody levels rose a decent amount after the vaccination, 24x on average. One subject got tested before, after the vaccine, and then again after getting the flu, and the disease hadn't increased the levels any higher than the vaccine had (though obviously it wasn't a protective level somehow).

The epidemic was pretty small, so there were only 34 cases total in the study: 20 in the unvaccinated, and 14 in the vaccinated. This wasn't a significant difference, so it was negative.

But they isolated virus from some of the cases and found that it was a different serotype, so the vaccine might not've been good at targeting it anyway. Oh well.

Reference:
Taylor, R. M. & Dreguss, M. An Experiment in Immunization Against Influenza with a Formaldehyde-Inactivated Virus. Am. J. Epidemiol. 31-SectionB, 31–35 (1940).

Sunday, July 26, 2015

O953 - Prophylactic immunization against measles, scarlet fever, diphtheria, whooping cough, and influenza

This paper was a speech given by Sir Ledingham, director of the Lister Institute in London. He says some interesting things. Speaking on the safety of vaccines:
"The safety of any method to be applied to man must always be a prime consideration. There must be no risk to life, and even the minor inconveniences, local and general, incidental to the introduction of a prophylactic through the skin must also, so far as possible, be reduced to a minimum consonant with the effectiveness of the product."
The first topic is diphtheria, a bacterial disease that can be prevented simply by inducing an immune response toward the toxin the bacteria produce (diphtheria toxin).
"Diphtheria undoubtedly merits chief consideration because we now know quite definitely from field trials...that intelligent and persistent immunization of the child population, starting with the pre-school child one year old, can bring the diphtheria morbidity down to zero."
The best approach to this is using something called toxoid, which is toxin treated such that it produces an immune response but doesn't cause problems.

"By the end of 1938 some 800,000 troops had been so vaccinated [with diphtheria toxoid], and Ramon was able to state that clinical diphtheria was disappearing from the French army."
"It is only when the proportion of vaccinated children reaches 70 to 80 per cent. that, as Ramon says, one sees the real fruits of one's work in a great reduction of the diphtheria incidence, perhaps to zero."
"Among the immunized children there have been since March, 1935, seventy-three notifications of diphtheria, giving an attack rate of 26 per 10,000 children, while among the unimmunized, including some 3,000 known Schick-negatives, the attack rate was 251 per 10,000, or ten times that among the immunized."
The Schick test was a way to test immunity in a person (though not perfect): if there was no reaction (Schick-negative), that indicated some immunity.

Considering the success of immunization, Ledingham makes a statement that seems relevant still today:
"When liberty becomes anti-social and impedes the application of scientific truth to the improvement of human well-being then I think some form of compulsion is the only remedy, though one might hope it would be merely temporary, pending the wider diffusion of scientific knowledge in the community."
It was a bit more complicated than just toxoid though, for diphtheria vaccine options. There was formol-toxoid, toxin-antitoxin floccules, alum-precipitated toxoid, and toxoid-antitoxin mixture, each with different properties in terms of safety or immune stimulation. Ledingham has some opinions:
"T.A.F. and T.A.M. are relatively innocuous at all ages, while A.P.T. is well borne in young children. In older children and adults it may give rise to painless nodes, which take some time to disappear. A.P.T., by virtue of the slowness of absorption of the antigen and consequent longer stimulus to the antibody-forming mechanism, can give rise...to considerably higher antitoxin titre, given in a single dose, than any of the other antigens. For this reason it may well become the antigen of choice for young children, the group we wish above all to see in course of time fully immunized"
In terms of stimulation, alum-precipitated toxoid seemed the best (probably the alum was an adjuvant); two doses of it was equivalent to three of formol-toxoid. Ledingham also stressed that trials of the vaccine should have prevention of diphtheria as their endpoint, not Schick test results; with a good vaccine, the test shouldn't be necessary. One dose with APT followed by another of FT might be good.

Then he moved on to whooping cough:
"As a killing disease of infancy, therefore, whooping-cough closely rivals diphtheria"
He cites Kendrick and Eldering's study that showed only 3.8% of vaccinated subjects with severe disease, compared to 13.1% in the unvaccinated, and other studies giving similar results.

"The age at which a pre-school child, say 9 months, may be submitted to immunization against diphtheria is also the age at which this same child might very desirably be immunized against whooping-cough. What is to be done? I am not aware that diphtheria toxoid has so far been administered in combination with a pertussis vaccine, but I see no reason why this should be attempted if Ramon's claim holds that the potency of diphtheria toxoid in mixture with T.A.B. [typhoid] vaccine is in now way interfered with and indeed appears to be enhanced. Such mixtures would first require to be tested on animals in which, at any rate, the rise in antitoxin titre can be accurately evaluated."
Then moving on to scarlet fever, he thinks it isn't worth trying to vaccinate most people, even if there were an effective vaccine, because it wasn't a serious disease anymore. Also it seemed that there was a patent preventing good vaccine development.

He thought treating measles with immune serum seemed to help, but there hadn't been much effort to acquire or store it. For active immunization, not enough was known yet about the disease.

With influenza though, more was known and a vaccine was closer, with some animal trials.
"One finding, however, which is likely to complicate the problem of active immunization as applied to man is the multiplicity of antigenic types among the strains recovered from different outbreaks of human influenza."
But he was frustrated by public health policy in general to some extent:
"When I reflect that, owing to ignorance, vested interest, or complacency in high places, a municipality is still prevented from ordering the pasteurization of all milk that comes into it, I sometimes despair of getting preventative science across."
One final quote:
"It may sound a totalitarian policy, if you like, but in essence it seems to me quite democratic, for it involves a negligible sacrifice on the part of the individual for his own and the common good."

Reference:
Ledingham, J. C. G. Prophylactic immunization against measles, scarlet fever, diphtheria, whooping cough, and influenza. British Medical Journal 2, 841–846 (1939).

Sunday, July 12, 2015

096 - Active immunization against pertussis: Final report on the cleveland immunizations of 1934-1935

This is a follow-up to an interim report on a big trial of whooping cough vaccine (065). That report didn't show much difference between vaccinated and unvaccinated, and this final report doesn't improve things much. This is weird, because other big trials at the time that showed pretty positive results (072). There was some suspicion that it was the preparation of the vaccine that made the difference, but that didn't seem to be true (093), so it's pretty mysterious.

So in this final report, the proportions of subjects in each group that got sick were about 15% in the vaccinated and 19% in the unvaccinated controls. Allowing an extra three months before starting to count cases (to let immunity develop after vaccination), the numbers were 16% and 18%, so no better.

The average severity between groups might've been different though. For one thing, there was one death in the control group and none in the vaccinated, but that's not enough to make any conclusions. But apparently multiple people reviewed whatever records they had about severity and came to the same result, that the proportion of mild cases vs. severe cases was higher in the vaccinated group: about 59% mild in vaccinated vs. 35% in controls, and 0% severe vs. 3.5% in controls. But it could've been more rigorous in record-keeping.

So overall, not promising results, but not nothing. Why it differed so much is still mysterious. One interesting thing to note is from this other study that compared several large trials at the time, the attack rate of unvaccinated subjects in Doull's study seemed a lot lower than in others (about half) for some reason.2

Though other reviews seemed to think that this was one of the best studies of the time:
"Doull...took particular care to avoid the numerous pitfalls that attend trials carried out in a clinic population...The same cannot be said of many of the apparently successful trials."3
It'll be interesting to see why the positive studies overwhelmed this kind of result, such that now pertussis vaccine is so common.


References:
1.
2.
Perkins, J. E., Stebbins, E. L., Silverman, H. F., Lembcke, P. A. & Blum, B. M. Field Study of the Prophylactic Value of Pertussis Vaccine. Am J Public Health Nations Health 32, 63–72 (1942).
3.
Vaccination against Whooping-cough. BMJ 2, 222–223 (1945).

Saturday, June 27, 2015

095 - The Agglutinative Reaction in Relation to Pertussis and to Prophylactic Vaccination against Pertussis with Description of a New Technic

One big question is whether you can tell if someone is immune to a disease (with methods other than exposing them and seeing if they get sick, of course). Often antibody levels are used as a proxy for immunity in the absence of an epidemic or something. This study looks at antibodies to whooping cough that cause agglutination (clumping together of cellular material) and seeing if that correlates with immunity and such. It would make the test easier than the ones that were done at the time, at least, though those might've been more reliable.

So they tested 101 children that had never had pertussis or vaccination against it. Ten of them showed some agglutination.

164 others were vaccinated with killed bacteria. All but 3 showed high levels of agglutination, much higher than the 10 negative controls.

Finally, 71 children during or after an infection with pertussis were tested. The titers were lower than after vaccination, but of those tested during, 15 out of 17 had agglutination, and 36 of 67 had it after their coughing stopped. Titers went down over the several months following the disease.

So it seemed like agglutination correlates well with vaccination status, but not so well with actual infection history. So it's not clear how useful it actually is, as far as I can tell. Maybe having high levels is indicative of immunity, but having low levels doesn't always mean lack of immunity, but it's not conclusive yet.

Reference:

Saturday, June 20, 2015

094 - Active Immunization Against Whooping-Cough

This study has two parts, both relating to whooping cough vaccination. The first seems like a brief summary of a clinical trial, and the second is an animal trial testing which method of killing pertussis bacteria makes the best vaccine.

Part I
This was more exciting, and thus unfortunately lacking in details. There were 288 vaccinated and 1007 control children, followed by questionnaires to parents and physicians. There were 52 known exposures of controls to disease, and 97 exposures in the vaccinated group. From these exposures, 43 resulted in disease in controls, and 10 in vaccinated. This means 82% infection rate in controls, and 10% in vaccinated, meaning 72% protection from the vaccine. That's pretty good.

Also worth mentioning was that 23 of the 97 vaccinated exposures were intimate, meaning siblings that shared a bed or played together or whatever, but none of these resulted in disease in the vaccinated child.

So pretty good, but lacking in necessary details.

Part II
This was a study in mice using vaccines made of killed bacteria, but killed in different ways: phenol, formalin (formaldehyde), merthiolate (thimerosal), or heat. They tried increasing doses of each in mice before challenging with pertussis to see how many died. They found that phenol-killed vaccine required the lowest dose to start showing protection, followed by heat-killed. The others were worse.

The dose required to show any protection, relative to the weight of the mice, was very high, but this could be explained by the requirement that mice be able to resist a serious bacterial infection (the bacteria are injected into their body cavity), rather than a simple respiratory exposure, so it's probably not comparable to humans, in dose at least.

Reference:
Silverthorne, N. Active Immunization Against Whooping-Cough. Can Med Assoc J 41, 263–265 (1939).

Saturday, June 13, 2015

093 - H. pertussis Vaccines: The Effect of Washing and the Use of Mouse Protection Tests

In previous studies of whooping cough vaccines, most showed protection but a few did not, and it wasn't really clear why these studies were negative. One hypothesis was that while preparing the vaccine for these negative studies, a washing step was included that made the vaccine ineffective. So this study investigated this question by trying different preparation methods and testing the results on mice.

The three preparations were unwashed, washed with distilled water, or washed with saline. The protein observed in the washing liquid was similar between the two kinds, and didn't seem to have much toxicity or antigenicity, at least not for mice.

So then they tested the vaccines themselves with mice. It took a few trials before they got a really good protocol for mouse protection trials, with the unwashed vaccine giving good protection and the infection challenge being strong enough to kill most of the controls. Eventually they succeeded though. But none of the trials seemed to consistently show a significant difference between types of washing, or washed vs. unwashed.

So the reason the negative trials were negative is still a mystery, sadly. But at least future studies had a decent protocol for testing pertussis vaccines with mice, which seems to be the main reason they cite this study.

Reference:
Miller, Jr., J. J. & Silverberg, R. J. H. pertussis Vaccines: The Effect of Washing and the Use of Mouse Protection Tests. The Journal of Infectious Diseases 65, 16–23 (1939).

Sunday, May 24, 2015

092 - Immunization against pertussis

This is another trial of a pertussis vaccine, similar to the others. The justification was that whooping cough is contagious before it is easily recognizable, so it's hard to prevent by anything but immunization.

The subjects were children of 6 to 30 months old, most under 1 year. They were selected from those attending the Stanford Well Baby Clinic randomly, and controls were chosen randomly from the same group, so it was pretty well randomized (except that presumably the parents in the vaccine group had to agree to the vaccine).

The vaccine was whole killed bacteria, a normal dose like Sauer's, 3 doses each a week apart. They started counting after the last dose was given, to make sure everyone was fully immunized. From these shots, they saw 4 systemic reactions, which were just a day of fever.

Then they got the parents to inform them of any known cases of pertussis, and any exposures to others with pertussis. These were diagnosed either by the typical coughing symptom or by culturing the bacteria on cough plates. Some had neither but were counted as "probable" because of history of exposure.

Results
Over the 2.5 years of the study, the vaccinated group had 26 exposures resulting in 7 cases (2 of which were questionable). So the infection rate was 26.9%.

The controls had 25 exposures resulting in 22 cases (1 questionable), so 88% infection rate. The difference between the two groups was statistically significant, both in infection rate and attack rate (the proportion of all subjects infected vs. not), so it seemed like the vaccine was protective.

The control group had 11 cases from unknown exposures, and the vaccinated group had only one, so it's possible to assume the vaccinated group actually had 10 more unknown exposures that would otherwise have resulted in disease, lowering the infection rate to 23%.

In terms of severity, the vaccinated cases were never severe; a few were typical, some mild or questionable. The control cases were mostly typical, 3 severe, 5 mild or questionable. So the vaccine seemed partially protective even in those that got sick.

In terms of immunity duration, of those that got sick in the vaccinated group, most had been vaccinated over 1.5 years before, so the authors recommend small yearly boosters.

Overall, not as rigorous as possible, but pretty good, and the vaccine was helpful but not super-great.

Reference:
Miller, Jr., J. J. & Faber, H. K. Immunization against pertussis. JAMA 112, 1145–1148 (1939).

Saturday, May 16, 2015

091 - Active immunization against whooping cough with various specific vaccines

Since various different trials of whooping cough vaccines had given confusing positive and negative results, Morris Siegel wanted to settle the question of whether any version was effective. Sadly, he failed.

This study vaccinated 1324 children in Brooklyn with different versions: Sauer's version from Eli Lilly, a similar version from the NY Department of Health, some versions from Lederle Labs or Povitzky, and then the NY Dept Health version delivered partially subcutaneously instead of all intradermally.

The subjects were under 6, with no history of pertussis, recruited or volunteers from the community. Controls were selected from the same neighborhoods and families. So it wasn't really randomized or placebo-controlled at all.

They followed up once a month with the subjects to get histories of pertussis. Cases were considered cases if whooping was present, or probable cases if there were other symptoms and exposure to a typical case. And cases that happened within a month of vaccination weren't counted either, since immunity hadn't kicked in yet.

So overall there were 1270 vaccinated subjects and 1016 controls that were observed throughout the study. about 80-90% were 1 to 4 years old, and 7-15% were less than a year.

Results
Overall, 3.6% of the vaccinated subjects got whooping cough, vs. 4.2% of the controls. Of the vaccinated, 76% of the cases had received what was considered a full course, three doses.

Not great. However, looking at the breakdown of vaccine versions, no one who got Sauer's vaccine or the similar NY Dept Health one got sick; most of the cases came from the Lederle Labs versions or similar. The intradermal route seemed best. So considering only the good versions, only 1.7% of subjects got sick, vs. 5.4% for all the other versions; worse than the controls. The biggest difference between them is the concentration: there were a lot more dead pertussis bacteria in the seemingly more effective versions.

One other positive thing though was that the age distribution of vaccinated cases was centered on children 2-3 years old, with fewer young infants getting sick than controls, which had a center around 1-2 years. So it's possible the vaccine shifted the age distribution toward older children, who can deal with it better.

Most of the cases happened after four months had passed since vaccination, possibly an indication of waning immunity; or possibly not, since diseases have some seasonality.

Overall, since it wasn't blinded or randomized, and there were so many different versions used in the vaccinated, it's hard to say much about the results. Though some versions seemed to do better than others, it's unclear whether all groups were exposed to the disease equally. So, almost worthless.

Reference:
Siegel, M. Active immunization against whooping cough with various specific vaccines. Am J Dis Child 56, 1294–1303 (1938).

Saturday, May 9, 2015

O850 - Tetanus Bacillus Recovered from Scar Ten Years after Attack

When active, the bacteria that cause tetanus can only survive in conditions completely lacking in oxygen, but they can form very tough spores that can survive oxygen, heat, drying, whatever.

In this case report, a woman had some surgery on her uterus and showed some signs of tetanus paralysis afterward; the doctors treated her with anti-tetanus serum and she recovered.

Ten years later, she needed more surgery in the same place. The doctors feared a tetanus recurrence so they gave her anti-serum prophylactically, and tried isolating bacteria from the scar tissue they removed from her body. They successfully grew some typical Clostridium tetani, which produced tetanus toxin capable of killing mice. I don't know how they tried to avoid contamination though.

They tested the patient's serum to see if there were any sign of immunity against the tetanus that seemingly had been in her body for ten years, aside from the immunity from the anti-serum they had given. But they didn't find any indication of immunity.

They cite other reports of tetanus remaining dormant in the body for long periods, even up to 14 years, and advise that surgeons treat for tetanus prophylactically when a patient has had tetanus before. Bacteria can be pretty tough.

Reference:
Bonney, V., Box, C. & MacLennan, J. Tetanus Bacillus Recovered from Scar Ten Years after Attack. BMJ 2, 10–11 (1938).

Saturday, May 2, 2015

090 - Serologic Studies in Epidemic Influenza: With Particular Reference to the Persistence of Antibodies After Infection

Many virus infections seem to lead to long-term immunity, so you get it only once; at least, that's how it seemed. But one obvious exception is influenza, which you can get every year almost. Part of the reason for this is its ability to mutate frequently, becoming different enough each year that our immune system doesn't recognize it as well; but Fairbrother and Martin wondered whether this were really the whole story.

So they studied the antibody levels of people before and after an epidemic of flu and for about a year after. There were two kinds of tests they used: complement fixation, and neutralization where they mixed antibodies with virus and injected it into mice to see if they died.

As we saw before (O860), antibody levels seem to correlate with immunity, and definitely increase after infection. They saw the same thing here: after being infected in the epidemic, people's levels were much higher. For those that didn't get infected, levels were pretty variable, some high, some low.

But after about a year, antibodies in people who had been infected were much lower than they had been, nearly back to where they had been before the people were infected.

So it's possible that natural immunity to influenza wanes after a year, though that can't be concluded here because they didn't actually test immunity, only antibody levels.

Reference:

Saturday, April 25, 2015

O951 - Does an attack of acute anterior poliomyelitis confer adequate immunity? Report of four second attacks in New York City in 1935

There've been other reports of reinfection with polio after an initial infection (077), so it seems immunity is not always lifelong or perfect from natural infection.

Fischer and Stillerman collected a number of reports of such second attacks, 13 from others and four of their own in New York City in 1935. These four were in children 3-9 years old, and the attacks all came 2-5 years apart.1

The first had some paralysis in the first attack, recovered, and then had some more in the second and took years longer to recover from that. The second had paralysis the first time but none the second; the infection was nonparalytic. The third had some mild paralysis both times. The fourth had a nonparalytic infection first (it was questionable whether it was even polio), but the second attack was fatal.

So from this and other statistics, the authors calculated that second attacks happen around 2 per thousand first attacks. Considering the low attack rate in the first place, they wonder whether there's actually any immunity to polio at all. But others disagree:
"Fischer and Stillerman have raised the question as to whether the low morbidity rate in poliomyelitis would not make the incidence of second attacks rare even if no immunity occurred following the disease. In the 1935 New York City epidemic they observed four second attacks, a rate of 2 per thousand, which was within the limits of expectancy if no immunity resulted from a previous attack. However, these figures were not based on age specific rates and cannot therefore be taken as final."2

References:
2. Horstmann, D. M. Clinical aspects of acute poliomyelitis. The American Journal of Medicine 6, 592–605 (1949).

Friday, April 24, 2015

O860 - Investigation on Volunteers Infected with the Influenza Virus

Apparently back in the 30s, the Soviet Union was doing research on influenza, some of which used human volunteers.1 Well, better flu research than smallpox or something, and it seems to have worked out alright.

Here, they tried infecting 72 volunteers with virus taken from infected mice, by aerosolizing it and letting them inhale it. It seemed to work in some of the volunteers, though their illness was mild; the authors speculate that the virus might have been attenuated by passage through animals.

The most interesting result was the antibody levels in the volunteers before and after inoculation. They measured antibodies by drawing blood, mixing it with live virus, and injecting it into mice. If the mice survived, the virus had been neutralized by the antibodies.

What they saw was that volunteers who did get sick had pretty low antibodies to begin with, almost none in some cases; afterward their levels were 25-100 times higher. In those that didn't get sick, the levels had started out high and risen only a little (to about the same level as the others). So antibody levels correlated well with immunity to influenza.

The authors thought this infection with somewhat-attenuated virus method might be a good strategy for immunization. The editor of this study's journal disagreed with that statement, as did later researchers:
"A paper published from the Soviet Union by Smorodintsev[sic] et al. in 1937 - frequently cited as the first paper on live virus vaccine - described the administration of a mouse-lethal strain of the...virus by protracted inhalation of atomized virus. Typical febrile influenza developed in 20% of volunteers, hardly an acceptable vaccine by present standards, and certainly not attenuated, as claimed by the authors. Remarkably, they claimed, as well, that the virus appeared not to multiply in men, but the study was a landmark in establishing unequivocally the role of the virus in the development of the disease and in demonstrating antibody response to the virus during convalescence."4
Others questioned whether this study could really be compared to natural infections:
"These results suggested that the level of neutralizing antibodies was of significance in determining susceptibility or immunity to influenza A in man. But the conditions under which the experimental disease was produced and the relatively large quantities of virus suspensions used seem so different from conditions encountered in the natural epidemic disease as to make comparisons between these two conditions hazardous."2
"Whether experimentally induced influenza A in human beings is entirely analogous to the naturally occurring epidemic disease may be open to some question."3
Seems like a reasonable question, but at least it's easier to control this kind of study for unwanted variables.

References:
1. Smorodintseff, A. A., Tushinsky, M. D., Drobyshevskaya, A. I., Korovin, A. A. & Osetroff, A. I. Investigation on Volunteers Infected with the Influenza Virus. The American Journal of the Medical Sciences 194, 159–170 (1937).
2. Rickard, E. R., Horsfall, F. L., Jr., Hirst, G. K. & Lennette, E. H. The Correlation between Neutralizing Antibodies in Serum against Influenza Viruses and Susceptibility to Influenza in Man. Public Health Reports (1896-1970) 56, 1819–1834 (1941).
3. Horsfall, Jr., F. L. Recent Studies in Influenza. Am J Public Health Nations Health 31, 1275–1280 (1941).
4. Kilbourne, E. D. in History of Vaccine Development (ed. Plotkin, S. A.) 137–144 (Springer New York, 2011).

Saturday, April 11, 2015

089 - Treatment of pertussis with the New York State pertussis vaccine

Some investigators in 1936 questioned the effectiveness of any whooping cough vaccine developed thus far, though most thought they were helpful. So Thomas Bumbalo studied a vaccine produced by the New York Department of Health, made of killed bacteria.

The study was in children with siblings, half of whom got the vaccine and the other half didn't. All subjects had been exposed to pertussis though. There were four groups of subjects:
  • those who were sick with whooping cough and were treated by vaccination
  • those who were sick but were not treated with the vaccine
  • those who had been exposed but hadn't yet shown symptoms, and were vaccinated
  • those who had been exposed but weren't vaccinated
The first two groups both had 152 subjects, of similar ages and sexes and in the same families if possible. The latter two had 29 and 58 subjects, respectively. Bumbalo explained that mothers in group 3 were hesitant to have their children injected for no apparent reason, and:
"To quote the sentiment of many mothers, whooping cough is 'a mild disease which all kids get sooner or later and they might just as well get it over now.'"
Some things don't change.

So, the results: the average difference between the first two groups in how long they were sick was 32 hours, about 1.3 days. Not too great.

Of the second two, there were 5 cases (17%) in the vaccinated and 6 (10%) in the unvaccinated, though the average severity was milder in the treated. Still, not great.

Still, it wasn't a great study either. Vaccine therapy had already been pretty well demonstrated to be useless, and vaccinating right after the children had been exposed wouldn't help much either; there wouldn't be enough time to form a good immunity. So not much can be concluded.

Reference:
Bumbalo, T. S. Treatment of pertussis with the New York State pertussis vaccine. Am J Dis Child 52, 1390–1396 (1936).

Saturday, April 4, 2015

088 - Studies in whooping cough: Diagnosis and immunization

First, the scary numbers: in 1936, there were more than 300,000 cases of whooping cough in the US, and 15% of infected infants died from it.

Leila Daughtry-Denmark wanted to help make this situation better, so this study was partially to figure out how to diagnose it early, partially to figure out how to determine who was susceptible, and partially to see if it was possible to make people less susceptible (by vaccination).

The cultures and vaccine used in this study were provided by Eli Lilly.

First Daughtry-Denmark tested white blood cell counts and agglutination to see if they correlated with infection, recovery, or immunity. Infection or vaccine seemed to increase the former but didn't seem to affect the latter.

Complement fixation (relating to antibody levels) seemed a much better indicator. Even young children (under 6 months) showed a good response to vaccination. Letting the vaccine age to 4 months didn't reduce its effectiveness. And the Georgia Public Health Lab repeated the test on their own and saw similar results. And they never observed fixation without either the vaccine or the actual disease.

Complement fixation also seemed to correlate with immunity, at least in a small sample (2 brothers), where when they were both exposed later, the brother with good fixation was protected and the other got sick.

Then there were some bigger tests of the vaccine, both with Sauer's version and a more concentrated kind that required fewer injections, making it more convenient. Apparently Sauer's sometimes took more than 8 injections to work, while this double-strength one only took three. They sure had some tolerance to shots back then; though I guess the current recommendation is four shots of DTaP before 1.5 years of age, so that's not so much better.

Anyway, there were 240 subjects who got the vaccine, but only 73 actually got exposed to the disease. Of these, 10 got the disease. So you could say it was 86% effective. But there was no control group, so it isn't possible to make firm conclusions. Also, only those who got Sauer's vaccine got exposed, so there wasn't really a test of the double-strength version.

So it seemed to work, but it's hard to know how effective it really was; exposure doesn't always mean disease, even in the unvaccinated.

A final comment mentioned in the article: there was a test of 50 college students who claimed to have had pertussis as children, and it found that only one had very good complement fixation, while 46 had no detectable fixation. So natural immunity doesn't seem that great for pertussis.

Reference:
Daughtry-Denmark, L. Studies in whooping cough: Diagnosis and immunization. Am J Dis Child 52, 587–598 (1936).

Saturday, March 21, 2015

O936 - Complications and Sequelæ of Measles

In 1904, smallpox and diphtheria were fairly under control, so fever-causing diseases and malnutrition were the most important destroyers of human lives. This review by J. Mayer discussed measles, the complications it caused, and ways to treat it back then.

The article doesn't discuss much the frequency of each of these complications; some of them are probably quite rare. But it does demonstrate the wide variety of things that could possibly go wrong in certain (unknown) circumstances when infected with measles.

Mayer claims that measles generally is not dangerous unless complications arise, but these can arise in almost any type of tissue and can cause permanent, serious problems.
"It is hard to combat the old notion that measles is something akin to a common cold with a rash, and that there is nothing to be done but keep the patient warm. It is also difficult, as you know, to have people accept a new medical idea, which is not to be wondered at. The truth is that we, ourselves, too often encourage indifference by some such remark as 'It is only measles.'"
The most common, and thus most dangerous, is pneumonia, often caused by secondary bacterial infection. This happened in at least 5% and up to more than 20% of cases, sometimes with no clear way to prevent it from happening. Even today, in recent outbreaks in the US, 11-25% of cases needed hospitalization.

Next is laryngitis, which can also be caused by secondary bacterial infections. And otitis media, or ear infection, caused a great amount of suffering; if it led to infection, it could result in permanent hearing impairment or even death. And conjunctivitis is fairly common, sometimes leading to eye damage.

Mayer also claimed measles could cause gut problems—diarrhea, hemorrhage, colitis—but seemed like these might've been a result of mistreatment of the disease; the article wasn't clear.

These were the relatively common complications. Others were more rare but serious: meningitis, kidney problems, heart problems, etc. Skin problems such as eczema could occur.

Finally, Mayer quoted from a pamphlet distributed in Glasgow:
"Measles is a dangerous disease—one of the most dangerous with which a child under five yerars of age can be attacked. It is especially apt to be fatal to teething children. It tends to kill by producing inflammation of the lungs...It tends to maim by producing inflammations of the ears and eyes. Measles has carried off more than four times as many persons as enteric fever. It is therefore a great mistake to look upon measles as a trifling disease...It is therefore a great mistake—because as a rule children sooner or later have measles—to say, 'The sooner the better,' and to take no measures to protect them, or even deliberately to expose them to infection."
Seems like good advice.

Reference:
Mayer, J. Complications and Sequelæ of Measles. Cal State J Med 2, 221–224 (1904).

Saturday, March 14, 2015

087 - Schick Immunity and Diphtheria Infection

Diphtheria is a bacterial infection that can be pretty serious and deadly, especially in children. Fortunately we can form very good immunity against it, either by infection or vaccination, with antibodies targeting only the toxin that the pathogen produces (diphtheria toxin). 

This means the vaccine is fairly simple to make, the only requirement being that it induces an immune response against this toxin. So people were trying different methods to accomplish this: injecting whole toxin (not a good idea), or toxin mixed with antitoxin (antibodies from someone/something else) to neutralize the toxin, or, the best, toxin inactivated just enough that it didn't cause problems but still induced immunity.

One thing I've noticed is that back when the diphtheria vaccine was being developed, some studies seemed to measure its effectiveness solely by something called the Schick test. This involved injecting a small amount of toxin into a patient's skin, then observing the spot for a reaction. Counterintuitively, a negative reaction (lack of inflammation) indicated immunity diphtheria, the idea being that the patient's immune system could neutralize the toxin before it caused problems. So they took "Schick-negative" to mean "immune."

But as with any test, it needs to be validated, to make sure that the results correlate well with the actual state of immunity being tested. Are Schick-negatives actually better protected from diphtheria? The studies I'm reviewing today look into that question.

087a - Schick Immunity and Diphtheria Infection
Not surprisingly, immunity against diphtheria is never perfect in everyone. E. Ashworth Underwood observed a number of subjects in Leeds to see if Schick-negatives ever caught the disease.1

Underwood had been overseeing a vaccination program, using toxin-antitoxin mixtures, toxoid, or toxoid-antitoxin mixtures, that started in 1928. Between then and 1935, he observed 2197 Schick-negative patients. Some of these were naturally immune rather than vaccinated.

Of these, 20 came down with diphtheria. Two of them were just naturally immune, while the others were natural plus vaccinated. Except for 2, all were children between 3 and 10. In 4 cases, the disease was subclinical, but 13 needed antitoxin treatment. So there are two possibilities: either the Schick test can be negative in non-immune people sometimes, or the cases were infected with some type of pathogen that overcame their immunity.

Addressing the first possibility: at the time, C. diphtheriae strains were characterized as one of three types (or as atypical): mitis, intermediate, or gravis. This roughly seemed to correlate with how virulent they were, how serious a disease they could cause. Of the 20 cases, 19 were infected with gravis type; so it seemed possible that this type could overcome a higher level of immunity.

How does the case rate in Schick-negatives compare to the rate in Schick-positives? Underwood looked at about 85,000 Schick-positive individuals in Leeds, and saw a case rate of about 1.1%. At best with the Schick-negative cases, not counting the subclinical cases or those with a questionable Schick result, the case rate is at best half as much. So immunity could be said to be less than 50% effective. Not great.

Still, that doesn't mean the vaccine was useless. Leeds seemed to be experiencing an outbreak of a virulent strain that could overcome immunity, for one thing; the vaccine might be more helpful in other places with other strains. Also it's important to note that none of the 20 Schick-negative cases was fatal, whereas the case-fatality rate in Leeds in general was at least 8%, so the vaccine might still have had a benefit.

087b - The Schick Test and Active Immunisation in Relation to Epidemic Diphtheria
In this similar study, Parish and Joyce Wright observed diphtheria cases in London.2 Usually immunized people didn't get it, or if they did, it was almost too mild to recognize. They did notice a ward with 27 Schick-negative children, 11 of whom seemed to be carrying gravis diphtheria without symptoms.

In a school near London, there were 248 children and 33 staff. All of these got immunized if they were Schick-positive. An outbreak occurred in 1934, started by newcomers (when vaccination was interrupted by measles, chickenpox, and scarlet fever outbreaks), resulting in 4 cases in Schick-positives and 8 in Schick-negatives (one naturally immune). Half the positives were moderate severity, and 5 of the negatives were called "mild or moderate." They also seemed to be carrying the organism for a while, even two months later.

Again, it seemed like gravis strains were the cause of the outbreak. Unfortunately, it wasn't clear how high an antibody might protect against it.

087c - Alteration in the Incidence of the Gravis, Mitis, and Intermediate Types of C. diphtheriæ in Manchester: And Their Clinical Correlation in a Further Series of 940 Cases
According to Robinson and Marshall in this study, it's difficult to distinguish the three types of diphtheria clinically.3 Though they do cause more or less severe cases on average, and have other distinguishing characteristics, such as in lab cultures. The case-fatality rates were about 15% for gravis, 7.7% for intermediate, and 0.5% for mitis, though other researchers had different results.

They report 13 cases in Schick-negative people (12 of them naturally immune, one immunized). None had a mitis strain by itself, though one had a triple-type co-infection. Three had intermediate, the rest gravis. Two of the gravis cases were fatal, but most were mild.

In people Schick-negative due to toxoid-antitoxin vaccination, 11 got diphtheria. All but 2 cases were gravis, and 5 were severe. One died. So overall, it's not possible to conclude that Schick-negatives will only have mild cases, if any. The authors recommend periodic boosters, though nothing specific.

087d - Review of the Observations which have Accumulated with regard to the Significance of Diphtheria Types in the Last Four Years (1931-1935)4
In this paper, K.E. Cooper and colleagues talk about the types of diphtheria again: how types can be distinguished, the severity of disease they cause, etc. Apparently virulence doesn't always correlate with toxin production, they claim.

In Cork, more than 70 people died of diphtheria for every 100,000 in the 1920s, higher than anywhere else. The problem declined after many were immunized. After that, there were 81 cases in 2 years, and 18% of them died in the general population. (22% of non-immunized cases died, while no vaccinated child died.)

In other places, there was a lot of deadly gravis, some more than in others, where other strains predominated. In Leeds, they broke down cases by age, and found that in children under 5, case-fatality was 9-13%; it was 3-11% in ages 5-10 (the high end being gravis epidemics), and 1-4% in ages 10-15. So it's pretty dangerous in young children, unfortunately.

In Cork, Leeds, Manchester, and Stafford together, 5% of all gravis cases were vaccinated, 2.5% of intermediate cases, and 1% of mitis. I'm not sure if this takes into account the proportions of each population (as in, if 95% of the population is vaccinated but only make up 5% of the cases, that's different from if 5% of the population is vaccinated but makes up 5% of the cases). But not all of the cases in vaccinated people were mild; 3 gravis cases were fatal. The conclusion was vaccination is generally helpful though.

087e - Diphtheria in Liverpool with special reference to type incidence and severity5
H.R. Shone and colleagues looked at the different diphtheria types again. They found that of all cases for all types, the highest rates were in ages 5-9 (about 50% of cases), another 30% in ages 0-4, 14% in 10-14, and the rest in over 14. There was maybe slightly more mitis in ages 0-4 and more gravis in 5-9 and over 14, if the numbers were accurate.

In contrast to others, they found that intermediate infections were most severe, but the difference wasn't really significant between intermediate and gravis. This was reflected in case-fatality rates: 2.4% for mitis, 10.7% for intermediate, and 6.6% for gravis. And in terms of cause of death, it seemed like mitis caused more laryngeal complications while the others were just super-toxic.

Case-fatality rates went down as age increased, so the youngest children had the highest proportion of deaths, unfortunately.

108 vaccinated patients got diphtheria, though their Schick status wasn't tested (so they might not have been Schick-negative). 50% of the cases were gravis, 34.3% intermediate. Two of the intermediate cases died, two sisters 6 and 9 years old. Another 21 were severe.

Conclusions
None of these studies really made a rigorous investigation of the connection of Schick status and immunity, so I'm not sure what can be said, except that Schick-negative status is not always indicative of total immunity. It's important to note that it didn't seem to matter whether the immunity came from a vaccine or from previous infection, especially in the face of gravis.

I don't think Schick tests are done today, though I'm not yet sure why, so maybe it was decided they weren't that useful. We'll see, hopefully.

References:
1. Underwood, E. A. Schick Immunity and Diphtheria Infection. The Lancet 225, 364–369 (1935). 
2. Parish, H. J. & Wright, J. The Schick Test and Active Immunisation in Relation to Epidemic Diphtheria. The Lancet 225, 600–604 (1935). 
4. Cooper, K. E., Happold, F. C., McLeod, J. W. & Woodcock, H. E. de C. Review of the Observations which have Accumulated with regard to the Significance of Diphtheria Types in the Last Four Years (1931-1935). Proc R Soc Med 29, 1029–1054 (1936). 
5. Shone, H. R., Tucker, J. R., Glass, V. & Wright, H. D. Diphtheria in Liverpool with special reference to type incidence and severity. J. Pathol. 48, 139–154 (1939).

Saturday, January 24, 2015

Measles and Disneyland: Just the Facts

This is a bit different from my usual style, but a friend of mine on Facebook posted a link to a blog called "Measles Shmeasles Goes to Disneyland" by someone named Jessica Gianelloni, and asked for my input/opinion about what it said, and I put a lot of work into putting together a response, so I figured I might as well post it here too. Note: The original blog no longer exists, as the owner shut it down for some reason, but the content can probably still be found online if you desire it.

Overall I think in this article Jessica gets a bunch of stuff wrong, and at least some of the things she gets right are badly out of context. An interesting thing to note is that I searched for the headline she cites ("Disneyland Measles Outbreak Linked To Anti-Vaccine Movement") and could only find it on The Onion. Make of that what you will.

One thing that seems correct is that the vaccine is not as effective as expected in the 1960s; one dose is not adequate. I don't know why Jessica says 3 or more doses are recommended now though; all the recommendations I could find said only two. But that's not super-important at this point.

HERD IMMUNITY
The next thing is a claim that at least 80% of people being immune is required for herd immunity, which seems accurate based on her "citation." And levels are even higher than that, around 90% vaccinated; I doubt this is incorrect. The question is that if we have such high coverage, which is predicted to prevent spread of measles, why do we have measles outbreaks? And the implication is that the vaccine must not be capable of preventing the spread.

This is a very unsophisticated analysis though. Measles is considered eradicated in the US, which doesn't mean there are no cases, but it means that any outbreaks that start are imported from other countries; once a given outbreak ends, the virus is not present in that area anymore to start any more outbreaks.

And when outbreaks do happen, it's pretty consistent that a majority of the cases are people that never had a measles vaccine; there are areas where the coverage is well below 80% of people. For examples: there was a study on outbreaks from 1989 to 1991, and the risk of catching measles was 35 times higher for unvaccinated people. In 1990 for example, unvaccinated people made up 0.5% of the population but had 17% of the cases of measles in the outbreak. That's a lot higher than would be expected if the vaccine weren't very good. This number was worse in some years, better in others.

In the current Disneyland outbreak, of 34 cases in which the vaccination status of the people was known, 28 hadn't been vaccinated (82%). Similarly, in outbreaks in the first half of 2013 and 2014, of those cases with known vaccination status, 91% and 87% (respectively) were known to be unvaccinated. Only 2% of cases in the 2013 data had received both recommended doses. So it's not really possible to say these were outbreaks in highly vaccinated populations, since it spreads mostly only between unvaccinated. This is something Jessica seems to get wrong.

One thing to note is that of those unvaccinated cases, some of them were too young to be vaccinated. This is important in light of Jessica's Palevsky quote: the reason people who vaccinate are upset with those who don't is that when outbreaks occur, it's mostly the unvaccinated that spread disease to those too young to be vaccinated.

NATURAL MEASLES IS NICE
Next is claims about how before there was a vaccine, measles was a one-time thing, a normal part of growing up, and even contributed to a person's health in many ways other than just the disease itself. But now the vaccine makes it so that instead of measles being found mostly in older children (where it is mildest), it's more common in young children (too young to be vaccinated). The numbers Jessica gives are from less than 0.5% of cases in infants before the vaccine, to 30% now. In addition, implications are that the vaccine A) does not allow mothers to pass protective antibodies to their infants, at least not as well; B) does not offer life-long protection; and C) does not provide the same alleged general health benefits as actual measles infection.

First, about epidemiology before and after the vaccine: I don't know where Jessica got these numbers, but they aren't nearly in line with what I could find. First, from a couple of studies in the US in the 1930s, that I've blogged about before: One in Detroit in 1935 (081) found that in children 0-9 years old, 6% of cases were in children under 1 year old (and 66% in 1- to 4-year-olds). The other, in 1930 in Baltimore (060) found that of cases in children 0-14 years old, 4% were in under 1 year, 8% in 1 year, and most in 1-8 years old. So that's a lot higher than under 0.5%, long before the vaccine.

In more recent outbreaks: In the Disneyland outbreak so far, 6% of the cases have been infants under 1 year old. Again in the first half of 2013, 11% of the cases were under 1 year old. So that's a bit higher than the 1930s numbers, but nowhere near the 30% that Jessica claims.

It's also worth mentioning that in the population overall, in the 1930 Baltimore study, 0.8% of children under 1 year old got sick with measles; so 8 per 1000. Compare that to today, in the 1st half of 2013, there were 18 cases in children under 1 year out of a nationwide population of about 4 million in that age range; so that's 0.0009% of infants got measles. That's 9 per million, almost 1000 times less than before the vaccine. And Jessica acknowledges that the seeming 98% decline in measles was the death rate, not the incidence rate ("Does the incidence rate when the vaccine was introduced even matter?"), so we can attribute this 1000-fold decrease to the vaccine. That seems pretty impressive. I would say, is it important that a slightly higher proportion (2x) of outbreak cases are in younger children, if younger children are much less likely (1000x) to catch it overall?

Next, implication A: the vaccine does not allow mothers to pass protective antibodies to their infants, at least not as well the actual disease. Mothers who had the natural infection pass antibodies to their infants that generally protect them for 12-15 months (as Jessica says). But studies show that mothers who only had the vaccine can also pass protective antibodies to their infants. In this 2010 study, the protection passed from vaccinated mothers to infants was similar to that from naturally immune mothers, though it faded a bit more quickly (1-3 months less time). Here is a nice graph from this study:
Leuridan 2010, Figure 2
Is this significant? Probably somewhat, but not nearly as big a difference as Jessica makes it sound, and measles is a lot more rare now too.

Next, implication B: the vaccine does not offer life-long protection like the actual disease. I didn't look too hard into the claim that the wild virus provides life-long immunity (though one of the first detailed accounts of measles did include an observation of someone who seemed to still be immune to measles after having caught it 60 years earlier. No idea if that is a common thing though). As for the vaccine, a 2012 study found that after 20 years, only 10-15% of people who had received 2 doses had no antibodies. There was also a 1998 study that found that after 12 years, about 98% of people who got the vaccine seemed to have adequate antibody levels to protect them.

At this point in my writing, Jessica shut down her blog. Not sure why. Luckily I found another copy online so I can continue to go back and see what she said.

Next, implication C: the vaccine does not provide the same alleged general health benefits as actual measles infection. Immune diseases, tumors, allergies? I think I know what she's talking about here, something I had heard of before: there was a 1985 study that compared children who got measles and either had a rash or didn't have a rash. There wasn't any comparison with vaccinated or anything, just measles infections. And it seemed to show that children who got the rash had fewer health issues later in life than those who didn't. The lack of rash was explained by children having some sort of passive immunity, either from maternal antibodies or from injections of antibodies; there was no discussion about how the vaccine might affect things. But the hypothesis was that if the body didn't completely deal with the virus all at once, the virus might lurk around and cause health problems later (the ones Jessica claims it prevents). But I don't know if there was any follow-up to this study to clarify anything.

As it stands, it seems like the vaccine might be just as helpful as full-blown measles in preventing these health issues. Other research in Africa found that "vaccine efficacy against death was much greater than the proportion of deaths attributed to acute measles disease...These observations suggest that standard titre measles vaccine may confer a beneficial effect which is unrelated to the specific protection against measles disease." And another study found maybe a slightly increased risk of allergy for those who got wild measles infections.

Overall, it doesn't seem like the data supports Jessica's claims and implications.

MEASLES IS NOT NECESSARILY DANGEROUS
Jessica then claims that measles is not something to be feared, at least not in developed countries such as the US, so a vaccine is not necessary. And even in developing countries such as Africa, the vaccine is not as helpful as sanitation and nutrition would be, especially vitamin A. The implication, I think, is that the costs and risks from the vaccine are greater than the benefits, compared to other treatments or the disease itself.

I'll address the costs and risks of measles first. The CDC in 1998 claimed that measles kills 1 or 2 people for every 1000 it infects, and this is about the same rate as for the encephalitis it causes, a serious brain inflammation. In developing countries, it can kill as many as 1 out of 4 people it infects. I don't know where these numbers come from though.

For encephalitis, I didn't find any other good numbers on that, but it seems like we haven't seen any for a while. However, you can see in the outbreaks I've cited above (Disneyland, 2013, 2014), at least 11% of those who caught it needed to be hospitalized; up to 25% sometimes. Seems pretty serious. I guess I can appreciate Jessica's faith in modern medicine, though, if she thinks being hospitalized is no big deal.

As for death rates, two people died from measles in 2003, and considering the number of cases since 2000, about 1500, that's right in the 1-2 per 1000 range. The two deaths weren't exactly in the healthiest people, but unhealthy people do exist (often through no fault of their own) and should be protected; also, it's not always possible to know who is particularly susceptible to the disease. It could be you!

Another confirmation of the death rate: in the 1999 study mentioned above, there were 26672 cases and 89 deaths. This works out to 3 per 1000 cases; right on target, unfortunately.

But risk of death is not the only factor to consider; there's also cost burden, both to individuals and to the healthcare system overall. Being hospitalized is not cheap, I'm pretty sure, especially for those who have no insurance. A study estimated that each case of measles costs about $20,000 (and that's a conservative estimate!). This is costs for treatment and also public health efforts to track and control the outbreaks. Definitely a cost worth avoiding if possible.

What about vitamin A? That's a fairly cheap and low-risk treatment, right? I don't know what African study Jessica is referring to in the blog, of course, but the World Health Organization agrees that it's worth giving vitamin A to people with measles in developing countries, and agrees with Jessica's 50% figure. A review of other studies concludes that vitamin A might help reduce severity, at least in hospitalized cases. So that's nice, though I don't know how much of an impact it could really have on the risks overall. It seems better to avoid being hospitalized and needing treatment in the first place.

So overall, a fairly high risk of hospitalization (and associated costs), a fairly low (but not negligible) risk of death, perhaps less risk with vitamin A treatment, at least in developing countries. The obvious next question: is the vaccine any better, or is it worse?

THE MEASLES VACCINE IS MORE DANGEROUS
Jessica claims that the vaccine is associated with "seizures, encephalitis, blood disorders, sensory impairments, learning disabilities, immune system suppression, inflammatory bowel disease, inflammation of the brain, and many other severe allergic reactions." Some of those are redundant or too vague for me to figure out what she's talking about, but the Institute of Medicine released a report a few years ago reviewing vaccines and the evidence for their risks, so I'll summarize some of that.

Keep in mind that, for measles itself, if we didn't have a vaccine, there would be about 500,000 to 5 million cases in the US each year, so with a 1 in 1000 to 1 in 10000 death rate (I'm being conservative, giving a range), that means 50 to 5000 deaths and 55000 to 750000 hospitalizations. Are vaccines worse than that?

Regarding encephalitis/brain inflammation, there were studies looking at more than 500,000 children. In that sample, 199 got encephalitis, which overall is about 4 in 10000, except that only 9 of those cases happened within 3 months of vaccination; 80 of them were actually before the children got vaccinated, and the rest were more than 3 months after. So hardly any were likely associated with the vaccine. Another study found no association either.

Regarding febrile seizures, the report concluded evidence was pretty good that they were associated with the vaccine, but they don't seem to cause any permanent harm or learning disability.

Regarding autism, there have been a lot of studies of that with the MMR, some better and some worse. Even of the better ones, they're consistent with their reporting of a lack of association of the vaccine with autism.

Not sure what Jessica means by "blood disorders," but studies consistently report a lack of association of the vaccine with type 1 diabetes at least.

For sensory impairments and other things, the report authors considered the link between measles itself and the vaccine as some evidence for an association, but other than that there wasn't good evidence for a link with the vaccine in particular.

For allergic reactions, there is good evidence that the vaccine is associated with anaphylaxis, but this is rare, happens right after the shot, and is treatable.

Considering all that, it seems pretty clear that we aren't aware of any reason why the vaccine should be considered more risky than measles itself for most people. And it's pretty clear from recent outbreaks that we need to choose one or the other. So those are my thoughts on the issue, based on the data I could find. Just the facts.