Monday, April 25, 2011

Acute Thrombocytopenic Purpura, the MMR and Natural Infection

Twitter's a great tool for rapidly disseminating information. With a large network of followers, a simple message can spread like wildfire. This can be great for getting important facts out to a wide audience, like instilling a bit of rationality around the fear-infused media exaggerations of the events at the Fukushima Daiichi nuclear plant in Japan.

Of course, like all decent tools, there can be a bad side to Twitter, as well. Just as rational facts can be spread quickly, so, too, can misinformation. Given the character limits on tweets, a lot of the nuance and complexities of a given subject are often left out, resulting in messages that, on the surface, may instill readers with a sense of unease or outright anger. The careless may inadvertently scare people about a certain topic, while the nefarious use the limitations of Twitter to their advantage, purposefully spreading partial-truths or even outright lies to promote their agendas.

Such was a tweet I saw just the other day.

A certain California doctor that we all know and love retweeted something from the Twitter bot @VaxCalc. The person who set up that account uses it to spread half-truths and fear about vaccines. Their sole purpose, it seems, is to misinform people about immunizations. To what end, one can only guess. At any rate, the aforementioned pediatrician passed along the following:

That sounds pretty bad. I mean, bleeding into the skin? And less than 1% of that is reported to VAERS (Vaccine Adverse Event Reporting System)? Reading that, it sounds like acute thrombocytopenic purpura is a horrible, horrible reaction and that it actually happens quite frequently after receiving the MMR. Further, our friendly neighborhood pediatrician to the stars added his own comment of "Not great medicine" when he retweeted @VaxCalc's note:

Well, between the two of them, they got me to wondering: first, what the heck is acute thrombocytopenic purpura (ATP)? Second, how often does it occur following MMR? Third, how often does it occur following measles (or other viral illnesses)? Since @VaxCalc is just a bot, and because the good doctor decided to share with his nearly 11,000 followers this tidbit of information, I decided to ask that third question. I mean, since he commented that either ATP after MMR is not good medicine or the reporting rate is not good medicine, then he must know how often it occurs after infection, right? I mean, if he didn't put the information in context, well, that might be considered either intellectually dishonest or, to be charitable, sloppily lazy.

He hasn't answered me yet, other than to ask "what's ATP in that context?" Because, y'know, it's really difficult to look at the tweet to which I was replying.

That had the benefit of forcing me to look for the answer myself. Actually, I started looking for the answer right away, figuring that I wouldn't get a straight answer from the fellow. So off we go on a merry literature-reviewing adventure, but first, a quick note about VAERS.


The Vaccine Adverse Event Reporting System is a passive surveillance system. As its name implies, it records reports of adverse events following the administration of vaccines. It's important to note that VAERS tracks adverse events, not adverse reactions. Though similarly named, these two terms have distinct meanings. An adverse event is any unfavorable or unintended event following administration of a medical product; there is a temporal connection, but not necessarily a causal connection. By contrast, an adverse reaction is an adverse event which has a reasonably causal relationship to the medical product. So that's the a significant limitation of VAERS. There's a lot of noise, lots of events that, in the end, have no causal connection with a vaccine. The biggest limitation, though, is that it is a passive system, meaning that no one is forced to report each and every occurrence of an adverse event (AE). You may wonder, well, why not? Because anyone can submit a report to VAERS - pharmaceutical companies, physicians, researchers, even regular Joes like me. The makers of medical products are required by law to report all AEs to FDA, but it's kinda hard to enforce that for every single person in the U.S. That said, it's not any huge surprise that there is underreporting.

The take away is this: VAERS is a system for raising questions and looking for trends. It is not a definitive source for data. It's a springboard, not a finish line.

Now, onto the meat of this whole mess.

Acute Thrombocytopenic Purpura

Thrombocytopenic purpura is a blood-clotting disorder in which the body makes antibodies against its own platelets. This results in the destruction and elimination of platelets, the parts of your blood that cause clotting and help you stop bleeding if you get cut. So, the "thrombocytopenic" part describes a reduction in the number of platelets and "purpura" refers to the purplish discoloration or bruising that may result. Acute means that it comes on quick and doesn't last very long. Now, just because it doesn't last all that long does not mean that it is not serious. ATP could potentially lead to hemorrhage or other dangerous effects from excessive bleeding.

ATP and the MMR Vaccine

So what does ATP have to do with the MMR vaccine? Well, thrombocytopenic purpura is a known risk of the vaccine. Numerous studies have looked at ATP (also referred to as immune thrombocytopenic purpura, or ITP). To keep matters simple, I'll convert all rates to # per 100,000. France, et al. (2008), found a rate of about 2.5 per 100,000; Peltola, et al. (1994), found 3.3 per 100,000; and Black, et al. (2003), put the risk at around 4 per 100,000. Mantadakis, et al. (2010), performed a systematic review of 12 studies, finding that ATP following MMR occurred from 0.087 per 100,000 to 4 per 100,000, with a median of 2.6 cases per 100,000 vaccine doses. In fact, the CDC even acknowledged in one of their MMWR reports (1998) that it occurs and that it should be considered when deciding whether or not to vaccinate a child with a history of thrombocytopenic purpura:

Children who have a history of thrombocytopenia or thrombocytopenic purpura may be at increased risk for developing clinically significant thrombocytopenia after MMR vaccination (172,175). Although thrombocytopenia can be life threatening, no deaths have been reported as a direct consequence of vaccine-induced thrombocytopenia. The decision to vaccinate with MMR should depend on the benefits of immunity to measles, mumps, and rubella and the risks for recurrence or exacerbation of thrombocytopenia after vaccination or during natural infection with measles or rubella. The benefits of primary immunization are usually greater than the potential risks, and administration of MMR vaccine is justified, particularly with regard to the even greater risk for thrombocytopenia after measles or rubella disease. However, avoiding a subsequent dose of MMR vaccine may be prudent if an episode of thrombocytopenia occurred within approximately 6 weeks after a previous dose of the vaccine. Serologic evidence of measles immunity among such persons may be sought in lieu of MMR vaccination.

Keep in mind that the CDC's statement is from 1998. Lots of research has been done since then. As the Mantadakis review shows, "MMR vaccination of unimmunized patients with ITP and revaccination of patients with prior ITP did not lead to recurrence of thrombocytopenia," and around 93% recovered within 6 months.

We have, then, a known risk of thrombocytopenic purpura following MMR vaccination. It occurs at a rate of roughly 2-4 per 100,000 doses. In the majority of cases, it resolves on its own, with no lasting problems and is generally not life-threatening. This is something that we should strive to improve on, but it isn't necessarily a deal-breaker in the decision to vaccinate or not. We also need to consider how frequently ATP occurs after infection.

ATP and Infection

Now we get to the question that I asked Dr. Jay: what is the rate of acute (or idiopathic) thrombocytopenic purpura following natural infection? Information on the rate of ATP following infection was a bit more difficult to find. In 1951, Fisher and Kraszewski (PDF) described two cases of thrombocytopenic purpura following natural measles infection. Cines, et al. (2009), describe numerous causes of TP, including autoimmune disorders (e.g., lupus) and chronic infections (HIV, Hepatitis C, H. pylori), as well as following natural infection with rubella, varicella zoster virus (chicken pox) and many other viruses. Yenicesu, et al. (2002), found that following viral infection, ITP occurred about 13.3% of the time. Likewise, Rajantie, et al. (2007) found that thrombocytopenic purpura occurs more frequently following natural infection than after immunization, and that vaccien-associated TP is generally mild and resolves within 6 months in about 90% of cases. √únal, et al. (2009), also describe mumps as a cause of ATP. Tucci, et al. (1980) discovered subclinical thrombocytopenic purpura in 55% of children with measles, 25% of children with mumps and 30% of children with rubella, among other viral causes. Finally, in the same CDC report stating that ATP occurs in about 1:30,000-1:40,000 (~2-4:100,000) cases following vaccination, it was also reported that ATP occurs in about 1:3,000 (33:100,000) cases of rubella.

By looking at the collection of the literature on the topic, then, we can see that thrombocytopenic purpura occurs at a much higher clinically significant rate (at least 10x more frequently) following natural infection with the viruses against which the MMR offers protection than from the vaccine itself.

MMR Vaccine vs. Measles, Mumps and Rubella

On the one hand, we have a vaccine which carries a risk of acute thrombocytopenic purpura, which typically resolves within 6 months or less, in about 2-4 for every 100,000 doses administered. By contrast, each separate disease (measles, mumps and rubella) individually carry a higher risk of ATP, with similar outcomes as the vaccine. As an example, rubella occurs around 10 times more often following natural infection (~33 per 100,000 cases) than the rate from the vaccine.

Although thrombocytopenic purpura associated with the MMR vaccine is generally mild and resolves without complications, further study to determine why it occurs and how to minimize the risk is certainly called for. However, though we acknowledge that the risk exists, and that we should strive to make the vaccine safer in this aspect, we also must recognize that natural infection with measles, mumps or rubella carry a greater risk than the vaccine. Looking at the risks of both the vaccine and natural infection, the chances of acute thrombocytopenic purpura are reduced with vaccination.

I have to thank Dr. Gordon. If he had not passed on a biased and incomplete view of vaccines and not bothered to even attempt to answer the question I posed to him, I might not have put in the research to figure out the answer to my question. Thanks to Dr. Gordon, when looking at ATP, I was able to discover not only the rate following vaccination and natural infection respectively, but also that vaccination mitigates the overall risk. As I said before, the rate at which ATP follows vaccination should be improved, but it should not be used to dishonestly sway anyone away from the MMR. To do so is nothing but base manipulation.


  1. Wow, Todd, one of your best ones yet. Well-researched, well-written... You're not an MD, but you just contributed A LOT to good medicine. I'm printing this off and passing it on to my wife. She might be "just a PA", but she's always open to learning about the true risks from vaccines vs. the infectious diseases they cause. Thanks again.

  2. Thanks, Ren. Hopefully Jay'll actually read this and learn. I doubt he'll correct his tweet, but one can always dream.

  3. As a mother to a child that suffered from ITP, I have to say it's a very scary condition. My son had a level of 1 platelets. I was told he could bleed spontaneously into the brain, he had only just started walking so I had to stop him walking and falling (doctors orders). He managed to cut his mouth which meant I had to take him back to hospital for a platelet transfusion. He was set back a long time mentally and physically from ITP.

  4. In the Yenicesu et al (2002) study you cite it appears from the abstract that the viruses were grouped together, rubella was not assessed separately:
    "In viral serological studies, the EBV, CMV, and rubella antibodies were investigated for all patients at diagnosis (ELlSA). The proportion of children whose ITP was associated with documented acute viral infection was 13.3% in this group."
    Furthermore, this study in way showed that 13.3% of those who contract these illnesses also get ITP. Instead it showed that of those who were diagnosed with ITP, 13.3% of those had one of these illnesses. So what about the other 86.7% who were diagnosed with ITP, was their ITP caused by MMR vaccination? I guess we'll never know.

    For the Rajantie et al (2007) study you cite you mention that the authors conclude:
    "We conclude that the incidence of symptomatic thrombocytopenia after vaccinations is much lower than that after respective natural infections and that the outcome in most cases is excellent."
    Yet in the abstract we see no figures, no justification as to how they came to this conclusion, and without such evidence it makes their statement nothing but baseless conjecture that we often see touted.

    For the Tucci et al (1980) study you cite we see in the study that:
    "The Authors performed platelets count in 207 children of age varying 7 months to 12 years from the diagnosis and during the course of the illness. A subclinical thrombocytopenia was revealed in 55% of children affected by measles, in 25% of mumps, in 45% of varicella, in 30% of german measles and in 55% of infectious mononucleosis. In all the cases we didn't observe hemorrhagic manifestations and platelets count has come back to normal and higher levels spontaneously."
    As the authors note the children showed SUBCLINICAL thrombocytopenia, not idiopathic or acute thrombocytopenia purpura that is seen after MMR vaccination. To my knowledge no study of this nature has ever been performed to test subclinical thrombocytopenia after MMR. Maybe you know of one. But until one is done, comparing the results of this study to reports of idiopathic or acute thrombocytopenia purpura is purely illogical and fraudulent.

  5. Next you write that:
    "Finally, in the same CDC report stating that ATP occurs in about 1:30,000-1:40,000 (~2-4:100,000) cases following vaccination, it was also reported that ATP occurs in about 1:3,000 (33:100,000) cases of rubella."
    Yet the CDC document provides no citations for the statement that: "ATP occurs in about 1:3,000 (33:100,000) cases of rubella." Numerous other figures and statements have citations in this document, but not this one. How exactly did they come this figure, what studies are they basing this on? Until a citation is provided, this statement seems defunct aswell.
    Something else is interesting about the CDC document, they state:
    "Surveillance of adverse reactions in the United States and other countries indicates that MMR vaccine can, in rare instances, cause clinically apparent thrombocytopenia within 2 months after vaccination. In prospective studies, the reported frequency of clinically apparent thrombocytopenia after MMR vaccination ranged from 1 case per 30,000 vaccinated children in Finland and Great Britain (167,168) to 1 case per 40,000 in Sweden (169), with a temporal clustering of cases occurring 2-3 weeks after vaccination. Based on passive surveillance, the reported frequency of thrombocytopenia was approximately 1 case per 100,000 vaccine doses distributed in Canada (170) and France (171), and approximately 1 case per 1 million doses distributed in the United States (172). "
    Odd that in other countries the rate of thrombocytopenia was 1 per 30,000 - 40,000, but in good old USA the rate was 1 case per 1 million! My goodness, either the USA's children have some rare ability to ward off this reaction, or more likely the USA's surveillance is terribly inadequate.

    The truth is, as far as I could find, there are no studies that define what the rate of idiopathic or acute thrombocytopenia purpura is during or after infection with measles, mumps or rubella. The studies that report ITP incidence after MMR are severely lacking as well.

    As you also mentioned Black et al. (2003) found that:
    "The relative risk estimate for ITP within 6 weeks after MMR vaccination, compared to the combined group of unvaccinated children and children vaccinated with MMR more than 26 weeks previously was 6.3 (95% CI 1.3–30.1). The attributable risk of developing ITP within 6 weeks after MMR vaccination was estimated to be 1 in 25 000 vaccinations (95% confidence interval 21 300, 89 400). The estimated absolute risk of developing ITP during the 6 weeks after MMR vaccination was 1 in 21 000 (95% CI 10 500, 47 800) vaccine doses. This study confirms the increased risk of ITP within 6 weeks after MMR vaccination."
    Miller et al. (2001) fount that:
    "The absolute risk within six weeks of immunisation was 1 in 22 300 doses, with two of every three cases occurring in the six week post-immunisation period being caused by MMR."
    The authors identified a peak in the number of cases of ITP within 6 weeks after MMR vaccination, with an RR 3.27.
    There are a number of other studies that have been done using the same methodology which come to similar conclusions.
    The downfall of these studies is that they are retrospective and rely on reported incidence, which could miss cases that go unreported. Unless parents are made well aware before MMR vaccination to look for symptoms of ITP, cases may well go unreported. The same can be said for studies that research ITP associated with natural infection.
    In your closing paragraph you state:
    "As I said before, the rate at which ATP follows vaccination should be improved, but it should not be used to dishonestly sway anyone away from the MMR. To do so is nothing but base manipulation."
    This seems ironic to me, being that it seems you manipulated results from studies or documents to try to push your own agenda. If no one had bothered to look into the studies you quoted, they may well have believed your twisted conclusion.

  6. Cherie,

    Thanks for your comments. One problem that you are having is that you seem to be basing your conclusions solely on the abstracts. You really need to read the full studies. For example, you criticize the Rajantie study as having no figures in the abstract, yet these are available in the full study. The authors noted that ITP following immunization was acute (symptoms for less than 2 weeks) and never insidious (symptoms for more than 2 weeks before diagnosis), as after infection. They also noted a lower incidence of ITP after vaccination than after infection. Profound disease requiring treatment was as frequent as other non-immunization-related ITP, but more benign and of shorter duration. Remission occurred in just a few weeks and chronic ITP was rarer than after infection. None of the study subjects experienced ITP after the second dose of MMR. They conclude that whil ITP is observed following vaccination, it occurs at a significantly lower rate than following infection, indicating that vaccination may have a protective effect in preventing ITP. Note that this was a prospective study, not a retrospective study.

    "Odd that in other countries the rate of thrombocytopenia was 1 per 30,000 - 40,000, but in good old USA the rate was 1 case per 1 million! My goodness, either the USA's children have some rare ability to ward off this reaction, or more likely the USA's surveillance is terribly inadequate."

    You need to keep in mind differences in reporting structures. For example, those countries reporting higher rates have more rigorous registries than the U.S., thanks to socialized medical systems. The numbers cited for the U.S. are based on passive surveillance. The basis for that figure is, admittedly, rather old, and since it is based on passive surveillance, is likely an underestimate. The real number is likely closer to the rates seen in places like the U.K., Sweden and Finland.

    "it seems you manipulated results from studies or documents to try to push your own agenda"

    I manipulated nothing. I reported the evidence exactly as it was presented. I have acknowledged that ATP does occur following vaccination, but the evidence (if you actually read the studies and go beyond just the abstracts) suggests that acute thrombocytopenic purpura following vaccination is rarer than after infection, of shorter duration, and of milder severity. If you feel that the risk of ITP/ATP is too high and that it is of sufficiently great risk that it outweighs the benefits of vaccination, by all means, present the evidence supporting your position.

  7. I was born in 1951. I had a bad case of the measles when I was an infant. Later, at the age of 3 yrs old, I contracted a bad case of TP., which resulted in a splenectomy. I improved but my platlet count eventually dropped. I was administered ATCH, but every time they weaned me off, my platelets would drop. Eventually, the doctors tapered the cortisone off at a slower rate and put me on a milk free and chocolate free diet. I am now 62 years old healthy. I never knew there was medical connection between TP and measles until a few years back. I am pretty sure that at one time I may have had the mumps too. My father told me that I could have died during the splenectomy and that I had a large blank bruise on my stomach. My mother told me that they didn't have the MMR vaccine back then and I was never given the vaccine. I hope this information is helpful.

  8. Perhaps it is because of Dr. Gordon's tweets, but lately on another blog there have been comments saying their child at thrombocytopenic purpura after the MMR, therefore don't get vaccinated. Since it is a NVICP table injury I generally ask how their claim is going, and with them luck.

    But I did do a search of "thrombocytopenic purpura" on the US Court website and found a few cases (only forty one hits, less than half were relevant). Some are granted compensation without much detail (a "stipulation"), and others are denied due to being out of the range of the table time period and there was not enough evidence given.

    But there were a couple which were very involved, with fascinating and often disagreeing testimony from expert witnesses. You might find them interesting reading.

    One from 2010 where compensation was granted:

    The poor woman went through a horrible ordeal. An additional document that came up on the search shows that she was awarded over half a million dollars because she "suffered an unusually severe form of thrombocytopenic purpura."

    One from 1998 where compensation was denied:

    Some commentary on the expert testimony: "Tr. at 37-38. Aside from the difficulty in understanding Dr. Levin's interpretation of the article, Dr. Levin also makes an unreasonable assumption. He assumes the one patient who did not recover suffered a chronic course, however, there is no indication of that in the article. In fact, he may have been hit by a bus. The point is, it is pure speculation by Dr. Levin to say what happened to that one patient."

  9. I think this sentence needs rewording:
    "As an example, rubella occurs around 10 times more often following natural infection (~33 per 100,000 cases) than the rate from the vaccine."

    Every time I try to zoom in and out on my phone I get redirected to a post of yours called 'A Limerick'. Not so when scrolling, except when I try to scroll left so I can publish this comment. This is the third time I've tried to write it!


You need a Google, LiveJournal, WordPress, TypePad, AIM or OpenID account to comment. If you need one of those, click on the appropriate link. Spam comments will be deleted. Comments on posts older than 30 days will go into moderation, due to the activities of commercial spammers.