Thursday, July 7, 2011

Forensic tests for the presence of blood

Part 30 in the Knox/Sollecito case
Updated three times (see below)

Blood has a high proportion of red blood cells that are packed with a protein called hemoglobin but are without DNA. Hemoglobin has a helper molecule called heme, which contains iron, and the iron binds and releases oxygen. Roughly 1 in 800 blood cells is a white blood cell that does not have hemoglobin but does have DNA. The detection of blood is an ongoing problem in forensic science, and advances are continually being made. A previous entry in this blog also treated luminol.

Presumptive tests
Two major kinds of presumptive tests are chemiluminescent and chemical. Presumptive tests rely upon the pseudoperoxidase acivity of hemoglobin. Hemoglobin is not an enzyme (catalyst) in its role carrying oxygen, but it can often speed up the reaction of hydrogen peroxide with a reduced molecule. In these tests hemoglobin is mimicking the action of a class of enzymes called peroxidases, thus one refers to the pseudoperoxidase activity of hemoglobin. Peroxidases oxidize organic molecules using peroxides (such as hydrogen peroxide) as the oxidant. Luminol is chemiluminescent, giving off a bluish light in the presence of dilute blood when it is oxidized. Tetramethylbenzidine (TMB) is a chemical test, relying upon a change in color upon oxidation.

The murder of Meredith Kercher may have featured a misuse of the Kastle-Meyer test, a presumptive test for blood. The Kastle-Meyer chemical test relies upon hydrogen peroxide oxidizing phenolphthalin, which is colorless, to phenolphthalein, which is pink.

The police released photos (see above) to the press that some observers, such as Judy Bachrach of Vanity Fair, thought was blood. Based on the discrepancy between Ms. Knox’s description of the bathroom and its seemingly bloody appearance, some observers lost trust in Ms. Knox.***

The actual appearance of the bathroom was unremarkable, except for a small number of droplets of blood and a partial footprint that may have been made in bloody water.

However the pink color may have been from using the Kastle-Meyer reagent over a large surface area and allowing air to oxidize the phenolphthalin. However, the color may have instead been the result of reagents used for latent fingerprint analysis. The pink photo at the top of this page was apparently not entered into evidence at the trial.

Confirmatory tests
Confirmatory tests are typically run after presumptive tests have given a positive result. Their purpose is to distinguish blood from substances that can give false positives for blood. Forensic scientists Dr. Kelly Virkler and Dr. Igor Lednev identify five classes of confirmatory tests: microscope tests, crystal tests, spectroscopic methods, immunological tests, and chromatographic methods. Crystal tests are based on the formation of crystals of heme (or a pyridine-based derivative of heme), and they are not performed as much as in previous years. Immunological tests that are based upon antibodies (immunoglobulins) that bind to hemoglobin, lactate dehydrogenase (more specifically, the distribution of the isozymes of lactate dehydrogenase), glycophorin A, or other biomolecules are more recent innovations. There are many methods which employ antibodies, such as the early double diffusion (Ouchterlony) experiments and more recent immunochemical techniques, such as enzyme-linked immunosorbent assays (ELISAs).

False Positives and the language of forensic reports
Presumptive tests will often give false positives in presence of metal ions or plant material containing peroxidases. Drano and bleach are two of several household products that can yield a false positive with luminol, as noted by Lt. Robin Bratton. Some of the putative bloodstains in the Lindy Chamberlain case (in which she was wrongfully convicted of murdering her infant) may have been the result of copper dust from the atmosphere, discussed by Dr. R. V. Winchester. The Chamberlains lived in Mt Isa, which is home to copper mines.

The case of Greg Taylor in North Carolina illustrates how one’s fate can turn on the choice of which information is revealed and the choice of words the forensic scientists use to convey the results. Joseph Neff and Mandy Locke wrote, “In Taylor's case, an alleged blood stain was the only physical evidence tying him to the murder. The presumptive test was positive; the confirmatory test was negative. The lab report made no mention of the negative confirmatory test…The FBI's written policy directed the analyst first to report the positive presumptive test results. If the confirmatory test is negative, the analyst would write, ‘Further testing could not confirm the presence of human blood.’” Some problems that can arise when a confirmatory test is negative are discussed in Appendix A.

The luminol-positive areas in the murder of Meredith Kercher
There were two luminol-positive areas in Filomena’s room, and both contained mixtures of DNA (the DNA mixtures will be treated in a separate blog article). There were three footprints in the hallway, and all tested negative for DNA. There was also a shoe print that contained both Meredith’s and Amanda’s DNA.

It emerged during the trial that the luminol-positive areas that contained DNA were subjected to a second presumptive blood test, tetramethylbenzidine (TMB). Dr. Sarah Gino noted in her testimony that TMB is negative about 50% of the time when luminol-positive areas are tested (Massei Motivations Report, p. 258, English Translation). Although luminol is more sensitive than TMB, TMB can detect blood that has been diluted up to 10,000-fold.

It is sometimes argued that negative TMB results in the Kercher case can be ascribed to this difference in sensitivity. There are several reasons to reject this explanation. First, if it were the only explanation for TMB giving a negative result, no forensic personnel would ever use TMB after using luminol: A negative result would not rule out the presence of blood, yet a positive result would still require a confirmatory test afterwards. Second, the window of dilution factors for which one would expect a positive luminol reaction but a negative TMB reaction is relatively small. Third, if one did have a sample which fell into this range, the luminol response would be weak, whereas Colonel Garofano remarked upon the sheer luminosity* of the footprints in the book Darkness Descending. A study by Bilous and coworkers showed that the maximum intensity of light emitted fell with decreasing concentration of blood (see Table 1).

With respect to the luminol-positive/TMB-negative/DNA-negative areas, I asked the authors of a recent study on the forensics of body fluid identification for their interpretation. Drs. Virkler and Lednev wrote, “So, there was either no blood and the luminol was wrong, or there was blood and the TMB had interference and the luminol damaged the DNA. We think it is more likely that there was no blood, and that the luminol was reacting with something else, possibly plant matter from the bottom of the shoes causing the footprints (the intensity of the luminol reaction might give some more insight). The prosecution should have used much more convincing evidence to prove the presence of blood.”

Were the luminol-positive areas related to the crime?
The footprints in the hallway are all right feet images** and do not form a trail. No reference footprints were taken from anyone except Amanda, Raffaele, and Rudi, nor can the prints be dated. Yet Judge Massei regards the prints as being made in blood. He said (p. 284 in the English translation of the Massei Motivations Report), “In this regard, one cannot simply disregard the fact that the bloodstains were undeniably abundant in Meredith’s room, from which easily, or indeed inevitably, they must have been exported to other parts of the house by anyone who, coming out of Meredith’s room, went into these other parts.” In some respects this line of reasoning is similar to Dr. Stefanoni’s argument in front of Judge Micheli during the pretrial, as reported in Candace Dempsey’s blog, Let's Talk About True Crime. This argument is extremely poor. It suggests that the footprints should form a continuous trail of right and left footprints away from Meredith’s room, contrary to fact. It treats luminol as if it were a confirmatory test for blood, and it ignores the negative TMB testing that was done on at least some of the luminol-positive areas.

Finally, two other facts lead one to question whether the luminol-positive spots are related to the crime. One is that the luminol data were collected on December 18, not in early November right after the crime but rather after the police had tossed the crime scene. This means that law enforcement personnel may have tracked luminol-positive material into Filomena’s room, for example. Two is that the police found many luminol-positive areas in Sollecito’s flat. There is no reason to associate any of these regions with the murder, and Sarah Gino’s testimony suggests that luminol-positive areas are not uncommon in forensic investigations.

The relationship between DNA testing and confirmatory blood testing
Although DNA typing is a powerful tool in forensics, it is not a test for blood. The National Forensic Science Technology Center said, “For example, while examining the clothing of a suspect, a forensic biologist might visually locate a brown stain that presumptively tested positive for blood and was then DNA typed. The DNA type is found to match the victim. Knowing that the loci tested are higher primate specific, what conclusions can be drawn? The only unqualified conclusion that can be offered is that the stain contains DNA that matches the victim. It has not been proven to be blood.” In response to a question of mine, Dr. Virkler and Dr. Lednev concurred: “It is correct to assume that DNA profiling is not a confirmatory test for blood because it can be found in so many other things. Just confirming the presence of the victim or suspect's DNA has absolutely no bearing on what type of tissue or fluid it is. There could have been skin cells scattered in a pile of ketchup that would match a person's DNA, but that doesn't make it blood.”

In "An Independent of the SBI forensic laboratory" in North Carolina, Chris Swecker and Michael Wolf stated, “It should be noted that the confirmatory ‘Takayama’ blood test that was at issue in the Taylor Innocence Commission proceedings was discontinued in 2003 and replaced with DNA and rapid Stain identification tests.” The Rapid Stain kit manufactured by Independent Forensics of Hillside, IL “uses two mouse monoclonal antibodies specific for human glycophorin A.” Glycophorin A is a protein found on the membranes of red blood cells.

The lack of DNA in a sample suggests that a substance is not blood, but that relationship is not absolute. The sample might contain an inhibitor of the polymerase chain reaction needed to amplify DNA in present-day DNA forensics or the presumptive test itself may have an effect on the DNA profiling. The exact formulation of luminol affects how much DNA is recovered.

Update 2, 6 PM 7/8/11
The luminol-positive areas can be subdivided into those that did and those that did not have DNA. The amorphous regions in Filomena’s room fall into the former category, and the footprints in the hallway fall into the latter category. However, none was subjected to a confirmatory test for blood, and none should be concluded to be blood. The presence of DNA in some of the areas does not constitute a confirmatory test; additionally, the areas with DNA were negative in the TMB tests. The other luminol-positive areas were not confirmed to be blood. Ms. Comodi asked for the jury to decide whether or not the areas were blood. The jury should not have concluded that any luminol-positive area was blood, and that is one reason that the notion that these areas were mixed blood is a fallacy. We will explore this erroneous notion in a subsequent entry.

***A previous version of this sentence was incomplete. It read, "Based on the discrepancy between Ms. Knox’s description of the bathroom and its seemingly bloody appearance." I completed the sentence and added a link to a discussion board.

Update 1 7/7/11, 2 PM EDT
*On page 377 in the book Darkness Descending, Colonel Garofano discussed the luminol-positive prints in Amanda's room and the prints attributed to her in the hallway: "FIrst of all, from their sheer luminosity they are blood. The DNA test showed Meredith's blood in all cases except for two places in which we have a mixed Amanda and Meredith sample." Colonel Garofano's statement implies that the luminol-positive areas all had Meredith's DNA, which is false. They also seem to equate a hypothetically presence of DNA as meaning that blood is also present, and this is also false, as discussed above. Thanks to Rose Montague for asking for the exact quote.
**A reader suggested a different wording, such as "each footprint in the hallway is a right-foot image," would be clearer. I am grateful for this suggestion.

Update 3 6 PM 7/12/11
I added two links on Drs. Virkler and Lednev.

Appendix A, Forensic bias and confirmatory blood tests
The Raleigh News and Observer reported, “Jed Taub, a 30-year veteran of the SBI, said, ’We didn't report the negative result of a confirmatory test because, really, it's misleading,’ said Taub, who now works as a forensic investigator for the Pitt County Sheriff's Office. ‘We couldn't be sure it wasn't blood, so those tests really didn't matter.’’ Reporters Mandy Locke and Joseph Neff continued, “Taub said that the only times he reported the absence of blood was when he got a negative result on that first, presumptive test. Any negative results after that were irrelevant, he said.”

In another article in the series on North Carolina’s SBI forensic laboratory Neff and Locke reported, “Perhaps the biggest challenge facing new SBI Director Greg McLeod is changing the culture of the SBI and the laboratory. Analysts have worked to support the theories of prosecutors, instead of rendering detached scientific analysis. Training manuals, some which have been withdrawn, have coached analysts to support prosecutors and distrust defense attorneys. This bias extended down to the very way analysts reported test results. ‘They were writing reports to law enforcement,’ said Chris Swecker, the former FBI supervisor who audited the blood cases. ‘They were trying not to write any negative test results.’”

These stories suggest that some aspects of forensic bias can be traced to the close relationship between law enforcement and forensic science laboratories. With respect to the forensics there is nothing that happened in Perugia that could not happen in the United States.

Appendix B, Some other presumptive and confirmatory tests

Leucomalachite green: Hemoglobin catalyzes the oxidation of the reduced form of leucomalachite green to the oxidized form, much as in the Kastle-Meyer test.

Leucocrystal violet: This test is very similar to the leucomalachite green test.

Takayama crystal tests: The sensitivity is about 0.001 mL of blood or 0.1 mg of haemoglobin. The crystals are pink in color.