An ongoing dialogue on HIV/AIDS, infectious diseases,
June 29th, 2012
HIV Testing Roundup, and a Brief Rant
I’ve written so many times about HIV testing that a complete list of the headlines fills two full web pages.
But since the last entry on the topic was more than a month ago, one might think I’ve lost interest in the topic.
Never!
Three items on the HIV Testing radar, two national, one local.
First, for a classic good news/bad news report, read these latest data from CDC on previous HIV testing among newly diagnosed people with HIV in the United States. Among the highlights/lowlights:
- Good news: 60% of more than 50,000 newly diagnosed individuals reported having a previous negative test, many within the previous couple of years so that their HIV was diagnosed early.
- Prior negative testing was correlated with factors associated with access to care and high awareness of HIV risk: whites, individuals aged 13 to 29, and males reporting male-to-male sexual contact as their sole risk factor.
- Bad news: blacks, those aged ≥50, heterosexual contact as the sole risk factor, and males reporting injection-drug use were least likely to have a prior negative HIV test — and also far more likely to progress to AIDS at or shortly after diagnosis.
Since people with no history of negative tests are at risk for being diagnosed later in the course of the disease, it’s got to become easier for them to get tested.
Second, could rapid HIV testing in pharmacies be part of the solution?
This sounds like a great initiative, as there is little doubt it’s easier for many people to go to their pharmacies than their health care providers. I’ll be particularly interested to see how this is implemented on a practical level, including feasibility, cost, acceptance rate, and most importantly, how people with positive tests get linked to care.
Third, we’re only about a month away from when it becomes legal in Massachusetts to obtain an HIV test without written informed consent — law signed in April, “official” July 31. (We’ll be the last state to drop the requirement, for the record.) It’s not an “opt out” policy — verbal consent is still required — but it’s substantial progress nonetheless.
All good, right? Not so fast.
The privacy part of the HIV testing law — which states that HIV testing results cannot be disclosed without a patient’s written consent — has not changed. Some (I hope a minority) have interpreted this to mean that written consent is still required before the test result can be reported by the lab — even if that means simply putting the result into the patient’s medical record (a record already protected by HIPAA). To quote/paraphrase an email forwarded to me from another institution:
… although the law now permits oral consent, we must still obtain written consent to share the test or results with any provider – meaning we still need written consent to include this information in the medical record.
It goes without saying — but I’ll say it anyway — that this bass ackwords interpretation obviously leaves us back where we were before the law changed. Thank goodness it’s not an interpretation shared by all institutions and practices.
And, for the record, it’s an intepretation that is utterly at odds with the intent of the change in the law, which was to expand HIV testing by removing written consent — so that people with this potentially fatal infection can access lifesaving treatment and stop spreading the virus to others.
So glad I didn’t go to law school. I’d spend hours a day with smoke coming out of my ears.
June 24th, 2012
ID Learning Unit — Choosing a Quinolone
We love quinolones on medical services, and it’s easy to understand why. Advantages:
- Ideal spectrum for several common infections, including community-acquired pneumonia, UTIs, and more complex infections when combined with other drugs
- Great oral absorption
- Few drug-drug interactions
- Once- or twice-daily dosing
- Generally well tolerated
- Reasonable cost
But how do you choose between them? Below, in a concise and admittedly somewhat oversimplified table, is all the medical intern/resident/hospitalist needs to know about our friends the quinolones:
| Drug | Activity vs gram – rods (rank) | Activity vs gram + cocci (rank) | Activity vs Anaerobes (rank) | Metabolism | Comments/Fun facts |
| Ciprofloxacin | 1 | 3 | 3 | Kidney | Don’t use for Strep pneumo; no association with increased cardiac death in recent study (unlike levo, azithro) |
| Levofloxacin | 2 | 2 | 3 | Kidney | Active (left) enantomer of ofloxacin – hence its name. How cool is that? |
| Moxifloxacin | 3 | 1 | 1 | Liver | Not active vs. Pseudomonas; generally the preferred quinolone for mycobacterial infections |
But it’s not all sunshine and happiness with these drugs. Problems:
- That nasty tendon issue
- Can make people (especially the elderly) kind of crazy – “like 5 cups of Starbucks, and not in a good way”, was one patient’s memorable description
- Photosensitivity
- QT prolongation
- Oral absorption blocked by concomitant iron, magnesium, calcium, aluminum, and probably also Krugerrands if you happen to be eating them
- C diff risk, something we really didn’t see until the emergence of the more virulent strain
- Ever rising rates of resistance with gram negative rods, Staph aureus
Bottom line? These are great antibiotics – but they are not interchangeable, and not 100% safe.
Extra credit: Can you name the five quinolones that have been pulled by the FDA for safety reasons? Double points if you get the reason they are no longer available, triple points if you also get the brand names and match them to the actual drugs.
Here’s a hint to get you started.
June 18th, 2012
ID Learning Unit — Serologic Tests for Syphilis
Diagnosing syphilis is tricky for lots of reasons, including:
- The protean disease manifestations, many of which were best described in older medical literature — and hence not known to people who don’t read words on paper (vs a screen) very often.
- You can’t visualize the bug (Treponema pallidum), unless you happen to have a darkfield microscope nearby — which, if you’re like 99.99% of clinicians, you don’t have at all.
- You can’t culture the bug, unless you use a rabbit — and all I can say about the various techniques of xenodiagnosis is yuck.
- There is a seemingly endless list of available serologic tests, each having a potentially different purpose and all of them carrying an indecipherable acronym.
It’s this last one that is the topic of today’s ID Learning Unit. All you need to know (not everything you could know) about blood tests for syphilis.
There are the screening tests:
- Rapid Plasma Reagin (RPR) — detects antibodies not to T pallidum, but to cardiolipin-cholesterol-lecithin antigens, which are present in active syphilis; which begs the question, what the heck is a “reagin”, and how do you pronounce it? Lots of false positives.
- Venereal Disease Research Laboratory (VDRL) — detects antibodies to the same antigens, just a different technique, and seems to be used much less often; test of choice in the CSF, where it’s believed to be specific (but not sensitive) for neurosyphilis. Of course, how we know that is very mysterious, just like everything about neurosyphilis.
- Treponema pallidum enzyme immunoassay (TP-EIA) — the new kid on the block and, finally, a screening test that actually looks at specific treponemal antigens — hence more specific, fewer false positives. The TP-EIA is replacing the RPR and VDRL as a screening test in many high-volume settings (hospitals, STD clinics, etc); we switched to it at our hospital a few years ago.
Next, there are the confirmatory tests:
- Fluorescent treponemal antibody absorption (FTA-ABS)
- Microhemagglutination test for antibodies to Treponema pallidum (MHA-TP)
- Treponema pallidum particle agglutination assay (TP-PA)
All positive screening tests must be confirmed with one of these confirmatory tests (our state lab does the TP-PA). It’s particularly important for the RPR and the VDRL because, as noted above, these tests measure antibodies to non-specific antigens and have lots of false positives. Hey, weren’t you listening?
And finally, we bring back a couple of the screening tests to measure disease activity:
- RPR (mostly)
- VDRL
Both of these rise shortly after the lesion of primary syphilis (the chancre), then peak during secondary syphilis. Treatment hastens their decline, but some people will spontaneously revert to negative even without treatment (see figure, which is from an ancient but excellent paper, Ann Intern Med 1986;104:368). Which is why the above screening tests (RPR and VDRL) may be unreliable in evaluation of that elderly patient with dementia, and why they don’t really measure disease activity for late manifestations of syphilis. So if you really think it might be neurosyphilis, you should order TP-EIA or one of the confirmatory tests, listed above.
Two small digressions about the RPR and VDRL. First, the units. When positive, these tests are reported in dilutions — e.g., 1:4 is a one-dilution lower (and a two-fold lower) titer than 1:8. Not many of our tests get reported this way (ANA comes to mind as another); in general, it’s only considered a significant change if it’s two dilutions or more — 1:16 goes down to 1:4 after treatment, for example.
Second, they may be falsely negative due to the “prozone” reaction, which does not refer to place to buy auto parts. The prozone reaction occurs when the the antigen and antibody bind together forming a complex, blocking the clumping that labs look for when doing an RPR or a VDRL. If labs are alerted to the possibility of a prozone effect, they can do dilutions to reveal the expected agglutination.
All very simple, right? Hey, it’s topics like this that keep ID doctors in business!
June 17th, 2012
For Inpatients, HIV Medication Errors Common — Then Promptly Corrected
Several papers have shown that antiretrovirals may be incorrectly prescribed for hospitalized patients with HIV. How do they do at Johns Hopkins — the site of one of the best comprehensive HIV programs in the country (and perennial US News and World Report #1 Hospital in the Universe)?
As described in a new CID paper, investigators reviewed ART medication orders from all hospital admissions among HIV-infected patients in 2009 (702 admissions among 388 patients). In 380 admissions, ART was prescribed on the first day of hospitalization, and in 29% of these, a medication error occurred (145 total). By the second day of hospitalization, however, the error rate had dropped to 7%.
The most common errors were incomplete regimens (for example, prescribing only the PI component and leaving off the NRTIs), but incorrect doses, incorrect dosing frequency, and significant drug-drug interactions were also noted.
Not surprisingly, errors were more common with PI-based components, as these have a wider range of recommended doses, generally require ritonavir, and have more drug-drug interactions. (Plus, let’s face it it’s easier to get right single pill regimens — Atripla one PO QHS — than salvage regimens, pretty much all of which include boosted PIs.) There were significantly more errors on surgical services. The authors attribute the rapid correction of medication errors to their using clinical pharmacy review of medication orders.
These results are certainly consistent with our experience as well: Patients with HIV who require hospitalization often report that on the first day in the hospital, the regimens that they have been taking so meticulously are given incorrectly.
Since hospitalizations among HIV patients are steadily declining, and the vast majority of antiretrovirals are prescribed by a relatively small number of clinicians, we can’t expect hospitalists and medical housestaff to be familiar with these medications. Having them reviewed by a clinical specialist in HIV is critical — a practice we’ve adopted at our hospital as well.
June 15th, 2012
ID Learning Unit — The D Test
I suppose it’s not surprising that we’d follow-up the Etest with the D test, though perhaps if I were being alphabetical, the order would have been reversed.
The D test is important, because it screens for a form of clindamycin resistance in MRSA that might otherwise not be detected — the “inducible” kind, which can be associated with treatment failures. About half of MRSA isolates have this form of resistance.
Here’s how it works:
- Take some erythromycin-resistant, clindamycin-“sensitive” MRSA, spread it on a culture plate
- Drop an erythromycin disk on the left side of the plate, and a clindamycin disk on the right
- If there’s a flattening of the zone of inhibition between the two disks, then the test is positive, confirming inducible clindamycin resistance
- Report that bug as clindamycin resistant
First, critical thinkers will wonder why this is important, since we obviously don’t give erythromycin with clindamycin to an actual patient. My big-picture explanation is that it’s a marker for easily inducible resistance even without the erythro being present.
And second, note the critical step of putting the clindamycin disk on the right — otherwise the shape of the zone of inhibition won’t be a “D”, and everyone will be confused because who knows what to call that shape.
(That second part was a joke.)
June 13th, 2012
Questions About HIV Cure, and a Very Funny Quote
The single case of HIV cure following allogeneic bone marrow transplant is in the news again, this time because of data just presented at “The International Workshop on HIV and Hepatitis Virus Drug Resistance and Curative Strategies” (formerly known as the “HIV Resistance Workshop” — how’s that for rebranding?).
I’m not at the meeting, which is too bad since they often have it in splendid locations.
But from what I gather based on the report, here are the key findings in the study:
- The patient remains off antiretroviral therapy, with a normal CD4 cell count
- He has generously submitted multiple specimens for research analyses at multiple different time points
- Highly sensitive assays of various sorts have been performed at several labs
- HIV RNA has been detected in plasma in 2 (of 4) labs from 3 different time points; levels are lower than those typically seen in virologically suppressed patients on ART
- HIV DNA was detected in a rectal biopsy sample by one lab
- No HIV has been detected in CSF or peripheral blood mononuclear cells
- No replication-competent HIV has been isolated
A more detailed summary of the presentation is over here at NATAP, and some of the controversy it has sparked is summarized at Science and, um, POZ.
My gut feeling is that the findings are potentially real, but unlikely to be of much clinical significance if virus can only be detected intermittently by special assays, especially since he’s been off all HIV treatment for more than 5 years.
But they may not be real — we need to remember that a false positive test result is more likely when the test’s sensitivity is cranked up and it’s performed multiple times.
Or as more colorfully put by Doug Richman:
If you do enough cycles of PCR, you can get a signal in water for pink elephants.
And if this interesting presentation tells us anything, it’s that defining success in any study of an HIV cure strategy is going to be very, very difficult.
June 12th, 2012
ID Learning Unit — The Etest
Every year I attend on the general medical service, so it gives me a chance to work directly with the medical residents — and to brush up on my non-ID-related Internal Medicine.
In exchange for what they teach me, each day on rounds I try to tell them about at least one ID-related thing that they may not know. Since I do this in the spring, they’re awfully sharp. Fortunately, I’ve been doing this ID stuff a long time, so can usually find something.
Today’s learning unit was the “Epsilometer Test, or “Etest” — that brilliantly simple way of estimating the MIC of an antibiotic to an organism, by using a drug-impregnated strip that has a gradient of concentration, from high to low. Find where the “elipse” of inhibition crosses the strip, and presto! There’s your MIC.
We take for granted that this method is readily available and, for the most part, clinically valid. It’s worth remembering, however, that when it debuted in the late 1980s/early 1990s, it was considered pretty slick and not entirely trustworthy.
And there’s still some debate about how reliable it is, especially with MRSA.
If you want to learn how to do it, watch this movie.
June 8th, 2012
SPARTAN: Two-Drug, NRTI-Sparing Strategies Continue to Disappoint
Just published is the cleverly named “SPARTAN” study — spartan because it leaves out both NRTIs and ritonavir — and the results are very interesting.
Ninety-three treatment-naive HIV-positive study subjects were randomized 2:1 to receive either a two-drug regimen of raltegravir 400 mg BID + atazanavir 300 mg BID, or a standard regimen of TDF/FTC + boosted atazanavir. (The higher ATV dose in the two-drug arm targeted ATV exposures comparable to those achieved by ritonavir-boosting.)
At week 24, virologic suppression rates numerically favored the two-drug regimen (75% vs 63%); as a small pilot study, the trial was not powered for a statistical comparison between the two arms. As has been observed in other studies of integrase inhibitors, the non-integrase group had a slower initial response and appeared to be catching up.
Despite these early favorable results, there were 6 virologic failures in the RAL + ATV group — vs only 1 in the standard arm — and 4 of these failures showed evidence of raltegravir resistance. Notably, all 4 of these patients with resistance had baseline HIV RNA > 100,000, a similar finding to ACTG 5162 (which examined RAL + darunavir/r). Rates of grade 3-4 and grade 4 hyperbilirubinemia were also higher in the ATV + RAL arm. Based on these efficacy and safety results, the sponsor elected to terminate the study.
As I’ve noted before, initial two-drug HIV therapy without NRTIs hasn’t fared well, even when it includes our best drugs. In addition, we still don’t know why these two-drug regimens don’t do as well, especially in patients with high viral loads.
Furthermore, the protective effect that boosted PIs have on the development of NRTI resistance doesn’t apply either to NNRTIs (as shown in ACTG 5142) or to raltegravir (again, ACTG 5162). And though in SPARTAN ritonavir-boosting wasn’t used, it doesn’t appear we can blame PK, as ATV exposures indeed were comparable to those seen with ATV/r dosed at 300/100 mg daily. Do the NRTIs provide some key antiviral component mechanistically? Do we just not have the right two active drugs? Or is there something magic about using 3 rather than 2 drugs, regardless of mechanism of action?
Suffice to say, the results of the fully powered NEAT study — which compares RAL to TDF/FTC, with all study subjects receiving boosted darunavir — will be of great interest, as will a similar study that uses maraviroc instead of raltegravir.
June 6th, 2012
A Fun Internet Poll for ID Nerds
Over on Medscape, one of my ID heroes, John Bartlett, has a new series called, “The Medscape Awards in Infectious Diseases” and it looks like a winner.
Here’s how it works:
The Medscape Awards in Infectious Diseases is a new series that will honor the greatest achievements in the field of infectious diseases during 1980-2012. John G. Bartlett, MD, Professor of Medicine, Johns Hopkins University School of Medicine, identified 8 key categories for these infectious disease awards… Readers will be asked to select the candidate that they believe is most worthy of this title, and then Dr. Bartlett will reveal his personal choice and the reasons for that choice.
OK, I’m game — for two reasons. First, practically anything that has John’s distinctive combination of scholarliness and practicality behind it has got to be good; and second, I am a towering nerd when it comes to Infectious Diseases. I admit it.
So bring on the first category — “Bacterium”.
The question is, “What do you believe was the most important discovery of a new bacterium during the time period 1980-2012?” Here are the choices, and my thoughts about each one:
- 1980: Borrelia burgdorferi (Lyme disease) — has spawned an entire parallel universe deftly described here.
- 1982: Escherichia coli O157:H7 (hemorrhagic colitis) — posits the eternal question, can a person really learn to prefer a hamburger well-done? Here’s my opinion.
- 1983: Helicobacter pylori (peptic ulcer disease) — one of the true oddities of helicobacter is that ID doctors know nothing about it; it’s like asking an MD about teeth.
- 1986: Chlamydia pneumoniae (atypical pneumonia) — also the cause of coronary artery disease … NOT.
- 1999: Bartonella henselae (cat-scratch disease) — this is my first Ted Nugent citation on this blog.
- 2000: Tropheryma whipplei (Whipple disease) — number of cases most ID doctors personally have diagnosed = zero. Which equals the number of Whipple Procedures they have done, too.
- 2000: Methicillin-resistant Staphylococcus aureus (USA 300 strain) — USA 300 sounds like a motor race, and not surprisingly, it actually is one too.
- 2000: Clostridium difficile NAP-1 strain (C difficile epidemic) — probably the best reason out there to avoid unnecessary antibiotics; I’d be shocked if we’re still using antibiotics as our primary treatment for C diff in 5 years.
- If not listed here, tell us about your choice — I would consider Klebsiella pneumoniae carbapenemase bacteria (KPCs), since they’re a glimpse of a post-antibiotic era, or Streptococcus gallolyticus, because it’s the new name of Strep bovis that I can never remember, or if we’re talking new names in the past 20 years, Stenotrophomonas maltophilia, since it’s such a wonderful mouthful to say.
So my vote?
Paul E. Sax, MD
Contributing Editor
NEJM Journal Watch
Infectious Diseases
Biography | Disclosures | Summaries
Learn more about HIV and ID Observations.
NEJM Journal Watch — Recent Infectious Disease Articles- Observations from ID and Beyond: Musings About a Bruising and an ID Link-o-Rama
- The Bad with the Good: Donor-to-Recipient Transmission of Bartonella quintana
- Largest Recorded Blastomycosis Outbreak in the U.S. Occurred in Michigan
- Fecal Microbiota Transplantation for Recurrent C. difficile Infection: The Devil in the Details
- Observations from ID and Beyond: The Riveting Conclusion of How PCP Became PJP