HIV Dissidents, continued
A while back I wrote about the case of HIV-dissident Christine Maggiore and the death of her three-year-old daughter. To recap: mother is HIV positive, she refuses anti-retrovirals while pregnant and after daughter's birth, she breastfeeds daughter (a known risk for HIV transmission), daughter dies at 3-and-a-half-years-old two days after starting amoxicillin for a proposed ear infection, medical examiner reports "AIDS-related pneumonia", family is livid, LA Times chronicles all in this article. The crew at Dean's World, who deny the HIV-AIDS hypothesis, support the family, who has the report reviewed by a pathologist. You can read his report here.
This is where it gets interesting. The pathologist who reviewed the report is Dr. Mohammed Ali Al-Bayati, PhD, DABT, DABVT, and he finds that the medical examiner in LA is in error. He is being passed off as an expert by Dean and by others. Writes David Crowe:
Dr. Mohammed Al-Bayati is a respected pathologist (PhD) and a dual board certified toxicologist with over twenty-five years experience and over forty articles published in the scientific and medical literature.
All true, but extremely misleading. Dr. Al-Bayati may be perfectly competent at what he does, I don't know. But he has no particular expertise that makes him qualified to refute the medical examiner's (ME's) report. What an autopsy report includes is a summary of the clinical events and medical history, a gross description of the body and internal organs, and a description of microscopic histologic findings. The last two are the job of a pathologist, something Dr. Al-Bayati proclaims to be. However, it is helpful to know that the world of pathology is split into two: clinical pathology and anatomic pathology. When a pathology resident sits for board exams, he actually takes two different tests, one for clinical and one for anatomic. It is possible to become board certified in one and not the other. In fact, it is possible to become board certified in one with zero experience in the other; and it is possible to become board certified in clinical pathology without any training in skills required to create or understand an autopsy report that addresses mainly anatomic findings.
This is important to know - because Dr. Al-Bayati is not a board certified anatomic pathologist! This matters, because an autopsy report is the purview of an anatomic pathologist, someone who is trained and educated in the recognition and interpretation of anatomic data, mainly microscopic histology. Dr. Al-Bayati could have reached his current station without ever once taking a single anatomy, histology, or anatomic pathology course. Indeed, there is no indication from his credentials that he is the least bit qualified to claim expertise in reviewing the medical examiner's findings. So how could he be considered an expert in anatomic pathology? (He isn't a board certified clinical pathologist either; however, he does have education and experience in clinical chemistry and toxicology, so any expertise in these areas will be granted).
This is not an ad hominem attack. He (and others) is passing himself off as someone with special expertise in evaluating these anatomic pathological findings, and he simply does not; the American Board of Pathology would agree with me. It certainly does not mean he can't have an opinion or that his arguments don’t deserve refutation (if I were saying that, then that would be ad hominem). It just means it is very misleading to say his opinion has any special sway over any other average person's. I should disclose that as a pathology resident, I am not board certified in anatomic pathology, either. However, one day soon I will be, and I am currently learning and training in basic skills that allow me to evaluate Dr. Al-Bayati's report. The opinions that I render here are based on my education as a pathology resident and the materials available to me in pathology texts and the medical literature. I reserve the right to change my mind based on more information - I am not an expert, but I am competent and knowledgeable of anatomic pathology.
At this point, if you are interested, you should probably read the report. Then go read this rebuttal by Orac at Respectful Insolence. Much of what I say overlaps with his post (however, his is much more thorough and encompassing than mine could ever be). I hope to add to a few points he has brought up.
To summarize, the ME found the following (via Orac):
- Pneumocystis carinii was found in Eliza Jane's lungs by Gomori methenamine silver staining in association with pink foamy casts in the alveoli. The lungs were also edematous (water-logged).
- Eliza Jane was mildly neutropenic (low neutrophil--a type of white blood cell--count) and profoundly anemic (low red blood cell count)
- Eliza Jane's brain contained throughout its white matter with relative sparing of cortex a number of variable-sized microglial nodules characterized by multinucleate giant cells associated with moderate pallor and myelination, occasional macrophages, and angiocentric pattern. These lesions stained positive by immunohistochemistry (IHC) for the HIV core p24 protein, a finding consistent with HIV encephalitis.
- There was atrophy of the spleen and thymus
- There was enlargement of the liver with fatty infiltrate of the cells (steatosis) and ascites
(note: immunohistochemistry is staining based on the presence or absence of a specific substance with antibodies. Normal histologic staining is just pink and purple color based on the acidity or basicity of the tissue in question.)
Taking these finding at face value for the time being, Dr. Al-Bayati has two tasks: 1) to discredit the findings from the ME and draw into doubt her conclusions, and 2) to come up with an alternative explanation for the child's death. #1 is the most important; the ME basis her conclusions on the probability that the findings can be explained by an AIDS diagnosis (relative to other possibilities). Dr. Al-Bayati can hypothesize (#2) all he likes, but it's a waste of time if the likelihood of these hypotheses are extremely low and the likelihood of the ME's explanation much higher. So he tries to refute each of these findings as being consistent with an AIDS diagnosis where he can; where he can not, he pushes an alternative:
- P. carinii pneumonia was not present and the findings were consistent with pulmonary edema caused by anaphylactic shock
- The blood abnormalities are a classic symptom of parvovirus B19, which he will later hypothesize as the culprit in this whole saga
- Most of the brain findings were non-specific, and the specific findings are false positives.
- The atrophy of the spleen and thymus is caused by the previous three-week illness
- The liver changes were due to liver toxicity from the amoxicillin (penicillin) the child was taking for two days for an ear infection.
Lets start with the first one, since this is by far the most important. The fungus P. carinii is a very rare human pathogen that is seen only in hosts with compromised immune systems. (Dean Esamy disputes this fact, but he is grvely mistaken; expanded thoughts on this here.) Historically it was seen in patient on long course steroid treatment and treated cancer patients with destroyed bone marrows. It was exceedingly rare until the advent of AIDS, where it became much more common, afflicting most of these patients at first (note: this was how AIDS was first recognized, by the sudden appearance of many cases of P. carinii pneumonia - called PCP - in the early 80's). A finding of PCP equals an immunocompromised host; and it means an AIDS diagnosis if no congenital or acquired cause of immune system compromise can be found (based on statistical likelihood). While the organism is fairly ubiquitous, it is almost never found in normal histologic examination, never in the alveoli (small air sacs that make up the lung), and never in the quantities or with the same microscopic background seen in AIDS patients and other immunocompromised individuals. The ME found P carinii on special stains (GMS) and "pink foamy alveolar casts" on normal stains (H&E).
As a matter of fact, if the presence of PCP can't be refuted, Dr. Al-Bayati's whole case goes down the drain. So what does he claim? That the organisms were there but there was no pneumonia. No PCP, no AIDS:
Pneumonia is a term that refers to inflammation and consolidation of the pulmonary parenchyma. The microscopic examination of Eliza Jane's lungs revealed no inflammation. The ME did not observe any inflammatory response in the alveoli or in the interstitial tissue to justify a diagnosis of...PCP or any other form of pneumonia.
The lesions of PCP usually comprise an interstitial infiltrate of plasma cells and lymphocytes; an interstitial fibrosis; an interstitial diffuse alveolar damage; and hyperplasia of type II pneumocytes; the alveoli are filled with characteristic foamy exudates...
So, got that? PC, but no P, because it doesn't fit the (strict) textbook definition of pneumonia. Well let's just see what the textbooks have to say. From Pediatric Pathology (Stocker & Dehner):
The organism has an elliptical shape with sharply demarcated borders and a pale center containing a distinctive black dot. A foamy transudate often totally lacking in inflammatory cells fills the alveoli.
From Spencer's Pathology of the Lung (Hastleton - 1996):
In typical cases of PCP, many alveoli and alveolar ducts are filled with characteristic amorphous, foamy eosinophilic material... There is an associated interstitial pneumonia with an infiltrate that consists mainly of lymphocytes and plasma cells... Alveolar walls may appear thickened due to the cellular infiltrate and edema, but fibrosis is not a feature in the early stages of the disease. The interstitial infiltrate is very variable and may even be minimal.
From Surgical Pathology of Diffuse Infiltrating Lung Diseases (Flint & Colby - 1987):
The histologic changes associated with PCP vary enormously from nearly normal histology to the classic foamy intra-alveolar exudate... Histologic patterns that may be appreciated include: nearly normal histology with few inflammatory cells scattered around aggregates of organisms plastered against the alveolar walls; an intense interstitial pneumonia with interstitial infiltrates...; diffuse alveolar damage...; the classic pattern with eosinophilic foamy intra-alveolar exudates and edematous alveolar walls with a mild chronic inflammatory infiltrate and prominent type II cells.
From Pulmonary Pathology (Dunnill - 1987):
Histologically alveolar septa are increased in thickness and infiltrated with plasma cells and lymphocytes... In immunosuppressed patients plasma cells and lymphocytes may be scanty. The characteristic feature is the intra-alveolar exudate... The intra-alveolar cellular reaction to the exudate is notable for its absence...
Finally, fromPractical Pulmonary Pathology (Leslie & Wick - 2005):
(PCP) can mimic any lung injury pattern... The histology of pneumocystis infection is that of frothy intra-alveolar exudates (so-called "alveolar casts") with many organisms.
So an exhaustive search of my pathology texts shows that Dr Al-Bayati's assertion, that inflammation must be present, is not at all supported by the whole of the pathology profession. In fact, it seems logical that diseases which depress the body’s abilities to mount inflammatory responses would not have very much inflammation around these organisms. The ME's findings, organisms and foamy alveolar casts with little inflammation, is entirely consistent with the diagnosis if PCP. Not to put too fine a point on it: show 100 pathologists lung tissue that matches the description in the ME's report, and you'll get 100 diagnoses of PCP.
This, of course, does not equal AIDS; it just makes it very likely. Dr. Al-Bayati then goes on to site many studies of PCP in HIV-negative and non-AIDS patients (but there's no PCP, right? Well, just in case, I guess). However, every one of the patients in the studies he cites is severely immunocompromised from various medical treatments or malignancies, neither of which describes this child. His lame attempt to account for an immunodeficiency brought upon by thymic and bone marrow atrophy from a chronic (3 weeks maximum) illness. This doesn't pass the clinical smell test, and no such patients are represented in the studies he cited. People don't develop PCP after short duration of an upper respiratory tract infection and an ear infection. I challenge Dr. Al-Bayati to find a single documented case.
At this point, short some plausible explanation for immunosuppression, in a child of an HIV+ mother who breastfed, AIDS is high on the list of probable diagnoses; in fact, it’s the only probable diagnosis.
Moving on to #2, that the child was anemic (low number of red cells) and neutropenic (low number of specific white cells, called granulocytes). From Pediatric Pathology:
Red and white cell elements may be increased, normal, or suppressed.
The presence of these abnormalities in this case neither proves, nor disproves, the presence of AIDS. As we say, it neither sensitive, nor specific. The more interesting fact is how these relate to the alternative hypothesis. For this, you should read Orac. The upshot is that B19 usually causes pure red cell abnormalities (but not always), and it only causes that in patients with a predisposing problem in red cell production or immunosuppression – and AIDS is one of these predispositions. If B19 was involved, (and there is no specific evidence that it was) it doesn't preclude a diagnosis of AIDS - it may actually support it.
Now for #3, that the HIV p24 positive staining seen in sections of the brain are not specific and could be false positives. This is the only point made by Mr. Al-Bayati that has any real plausibility. He cites a study done in 1992 that calls into question the specificity of the p24 immunostain. Many previous studies, especially studies involving CNS tissue, had expressed confidence that the p24 stain would stain positively only in cases where HIV was present. This study found that some kidney tissues with similar-looking pathology stained positive with p24 antibodies.
A few points: In terms of immunostains, 1992 is ancient history. I have contacted an author from that study to determine if any further studies were done, and if improvements have been made to the stain. I will update if he and I are able to communicate.
Moreover, positive-staining with a non-specific stain just means that the information yielded from this particular study is not-conclusive. The other non-specific pathological findings (giant cells, etc.) are consistent with a diagnosis of HIV-encephalitis; nothing Al-Bayati quoted from the ME does the slightest to rule this out – we must rely on the other information elsewhere in the study that more or less points specifically to an AIDS diagnosis.
In other words, this simply means that a positive p24 stain is no the be-all-end-all test for HIV/AIDS; there was evidence given (when a lot was needed) that this was a false positive.
#4: the thymic and splenic atrophy. From Pediatric Pathology:
Marked thymic atrophy... is seen in children with AIDS. It may be impossible to visualize the thymus at the autopsy table... Most children with AIDS have massive splenomegaly. There is obliteration of the white pulp with hyperplasia of the splenic cords.
This is consistent with what the ME saw. Dr. Al-Bayati would argue that he saw a small spleen, not a large one. However, he claims the 40 g spleen was "85% of the expected normal average weight for age." However, a large or small spleen is not determined on its relation to the average, but it's the relationship to the normal range (2 standard deviation around the mean). Furthermore, the normal range for a three-year-old is 36 to 45 g according to pediatric pathology texts; so there was no underweight spleen in any meaningful sense of the word. And histologically, she saw atrophy of certain elements consistent with an AIDS spleen.
Again, neither of these findings is sensitive nor specific; yet they are consistent with an AIDS diagnosis in a child.
Lastly, #5, that the changes seen in the liver are consistent with his alternative hypothesis.
What the ME saw was called steatosis, or accumulation of fats inside the liver cells. Also of note: the liver weighed 500g and there was no mention of liver inflammation or necrosis (cell death) in Al-Bayati's relayed report. He claims that this weight represents a rapid accumulation of water due to shock and that the histologic changes are consistent with those seen in amoxicillin toxicity (the drug she received for two days before death). He even provides several citations of the literature that support this; however, a careful reader will notice that these cases all describe hepatic drug toxicity, which is accompanied by destruction of the liver cells themselves.
But you will recall that the ME found no evidence of such destruction - just fat accumulation (which also takes more than two days to occur). So if the picture seen by the ME is inconsistent with Dr. Al-Bayati's hypothesis, what is it consistent with?
Well, isolated steatosis is usually seen in alcoholics, pregnancy, and obesity. None of these describe this child. However, Albisetti et al. described hepatic steatosis present frequently in a study of HIV+ children using ultrasound, biopsy, and autopsy. Where previously this change had been associates with HIV drugs (AZT, etc. - this has been described quite a bit in the literature), this study found isolated steatosis in many children before they had been treated. In fact, every autopsy or biopsy occurrence of steatosis had been untreated except for one. And this makes sense, if you realize that most children are treated in this country, studies would tend to make it seem as if it were due to the drugs themselves (it would be hard to differentiate). Since this study found many children yet untreated, it was able to show this change in isolation – associated only with AIDS.
And, again, the weight is above the weight he cited, but not outside the normal range; also, it was actually right at the mean according to a different citation in a pediatric pathology text.
This is in no way sensitive or specific for an AIDS diagnosis. But it is consistent with it, and inconsistent with the alternative hypothesis.
So, in summary, all five main pathologic findings, as presented by Dr. Al-Bayati himself, are consistent with an AIDS/HIV diagnosis. The PCP pneumonia in patient lacking the other risk factors (and three weeks of a mild upper respiratory illness is not a risk factor) is fairly specific for AIDS. All the other findings do not constitute strong evidence, but such is already represented by the PCP. As far as the alternative hypothesis goes, the PCP and steatosis are inconsistent with Dr. Al-Bayati’s characterization. Their presence does not rule out his hypothesis, but it does require an explanation. And the literature supports AIDS as by far the most likely suspect.
(You’ll note I confined myself to the pathologic aspects of this case, and not the clinical features. I feel I am only qualified to adequately analyze the former and not the latter. For this, see Orac.)
As a postscript to this analysis, let me comment on the case as it stands now. Rumor has it that the child’s parents are being investigated with the intention of possibly charging them with a crime for exposing their child to HIV and failing to treat. Most people who have written about this case who also agree with my conclusions support this course of action.
But I think I can, at the very least, make a good argument why this should not happen. In all I have read regarding this case, and from all I have read of Christine Maggiore and her activism, I have seen no evidence that she has an agenda that exceeds advocating for what she finds to be the truth. Also, I choose to believe, until presented with evidence otherwise, that she was a loving mother who is very saddened by the loss of her daughter. Though her beliefs, and the actions that extended from them, may have directly led to the infection of her daughter with HIV, its progression to AIDS, and her death from its complications, I believe the pain that Ms. Maggiore now suffers from the consequences of those beliefs is more than any punishment that she should have to endure.
She shares responsibility in this child’s death. But she was acting in good faith in what she thought was the best interests of her child. I don’t see any crime in that. (I have significantly expanded and qualified these thoughts here.)
Update: Follow-up here.