Unfortunately, the cost and availability of HIV diagnostic and monitoring assays aren't likely to bend to the same market pressures that are bringing drug prices down. In contrast to the ease of making and distributing medications, most current HIV laboratory diagnostics require blood samples to be collected, preserved and transported to a central facility for processing by trained technicians using expensive technology. The test results still need to be transmitted back to a medical provider in the field who is qualified to interpret the results and make an informed judgment about what should be done. If weeks have passed and the patient is no longer in the area or if there are no new drugs available to respond to a need for a change in regimen, then the expensive test result has been wasted.
It is now accepted that triple combination therapy, when prescribed and monitored by an expert provider, gives the best results in terms of sustained suppression of HIV with minimal toxicity. Although current generation treatments are not perfect and require a lot of management and laboratory resources to get optimal results, even the simplest 3-drug regimens given by rote are adequate to prevent death for many in the near term. Still, even in the rich countries of the world, optimal results are not universal. Poor adherence and viral resistance frequently send patients back to their doctors for more testing and a new lineup of drugs. Emerging toxicities need to be identified and treated with a change in strategy or medication. Although these problems can often be caught and corrected without serious clinical consequences, the costs of increased utilization of expert and technical resources are significant.
In rural Haiti, a small pilot project by Partners in Health has been demonstrating that antiretroviral drugs, when given with directly observed therapy (DOT), community support and guidance from a trained healthcare provider, can be successfully offered with no special lab monitoring at all. But what are the acceptable limits of unmonitored antiretroviral therapy in less supportive settings? As Joep Lange of the University of Amsterdam said at a recent meeting on affordable diagnostics sponsored by GMHC in Bethesda, Maryland, "we may lose 2 out of 100 patients treated due to drug toxicity, but we will lose 100 out of 100 patients who never receive treatment."
(Notes and slides from the meeting will be available at www.gmhc.org.)
Spokespeople for the pharmaceutical industry have in the past defended the refusal to lower drug prices by citing the absence of monitoring infrastructure in developing regions. Now it is clear that the availability of drugs is driving the formation of support resources, not the other way around. As drug costs continue to drop, barriers to treatment should continue to fall and treatment can be extended to an increasing number of the 40 million people who need it most. But just as lack of diagnostics should not be a reason to withhold treatment, the increasing availability of drugs should not be a reason to stop seeking better and cheaper assays and simpler clinical management plans.
Laboratory tests and assays are used by doctors to evaluate a patient's general state of health, to warn of the need for prophylactic care, to decide if antiretroviral treatment (ART) is warranted, to track the recovery of health after treatment has begun, to monitor the ongoing success of treatment, and to identify treatment toxicity before serious symptoms arise. For the management of HIV disease, two special laboratory tests are commonly ordered in addition to standard blood chemistry tests. CD4 tests count the number of a crucial immune cell circulating in the blood; low CD4 counts tend to predict disease progression. Viral load tests report the number of HIV particles floating freely in the blood. Viral load counts tend to drop after antiretroviral therapy has begun; a rising viral load result may indicate that a treatment is no longer working.
In most Western countries, CD4 counts are performed by flow cytometry, a technology intensive process of marking blood cells with fluorescing antibodies then passing them single file before a tuned laser beam that illuminates each specifically marked cell as it is counted by a computer. Flow cytometers begin at $30,000 and individual tests may cost upwards of $50.
The standard HIV PCR test for measuring viral load involves extracting the viral RNA in a known volume of blood, amplifying the RNA through several cycles of a polymerase chain reaction (PCR) technique, then separating and tagging the amplified genetic sequences for quantification. Basic equipment to perform PCR ranges around $20,000 and individual tests can cost about $100 apiece.
In the absence of laboratory tests, clinical evaluation and syndromic management (as practiced in rural Haiti) can provide guidance for administering and monitoring treatment. A CD4 count can be very useful for evaluating if a new patient should begin treatment, but the decision to treat can also be made based upon a positive HIV antibody test and a finding of weight loss, fatigue and perhaps some minor oral symptoms. While this may be acceptable for one person, another HIV-positive individual may feel fine until she develops symptoms of pneumonia, meaning that she has reached a dangerous state of disease progression.
The value of viral load testing also needs further discussion. Consider an individual on a simple daily regimen of pills prescribed by a regional doctor and dispensed by a respected community member. Because the response to antiretroviral drugs is fairly predictable, if adherence is good and she reports feeling better, a viral load test may not offer much extra useful information. A viral load test, though, is the best way to get early warning that high level drug resistance is developing -- before future treatment options are compromized. Yet without followup drug susceptibility testing, the only strategic option may be to switch every drug in the regimen. Individuals who only have a partial virologic response will probably need several confirmatory viral load tests to establish a trend. Often having one viral load result leads to the need for another, a problem if resources allow for no more than one per year.
For limited resource settings, a question could be posed: If you could only have one test per patient, which would you choose? Baseline CD4 count? The CD4 count after several months of treatment? Viral load after a few months of treatment to see if it's working? More research needs to be done to find out when lab tests are crucial and when they are luxuries.
No country is likely to be without any laboratory support, although distance and cost may render the resource impractical for routine use in the clinic. For example, several countries in Africa host facilities that can perform tests equal to those offered anywhere in the world. Such national centers of excellence (COE) are usually linked in a partnership with a Western university or as part of research collaborations with international health organizations. These high-tech centers have the potential to offer validation and quality control for laboratories at regional and district hospitals but they lack the capacity to provide routine diagnostic support for more than a few elite patients.
Examples of the lower and mid-range of existing diagnostic technologies are increasingly being installed at regional-level hospitals, with funding coming from foundations and committed governments. To sustain these facilities, personnel need to be trained, maintenance provided and reagents and supplies purchased on an ongoing basis. Regional centers may have the capacity to process tests for residents of the immediate urban area as well for blood samples transported to them from district centers. But cost (a single viral load test can cost as much as several months worth of drugs) and the delay involved in receiving samples from remote areas limit these facilities. At best, regional level service may be able to provide only a single key test (such as CD4) per year for patients within its reach.
Outside of the regional-level hospital, existing standard technologies are impractical to implement beyond a few pilot projects. Current technology depends on equipment that is costly to purchase and maintain and must be located near clean water and a reliable electricity supply. Furthermore, existing commercial diagnostics all require skill and interpretation at the point of use. A few lower cost alternative technologies appropriate for district-level hospitals are available and are being used -- yet all have problems.
One simple and practical approach to the problem is to design a dedicated CD4 cell counter using commodity level technology (i.e. lasers similar to those used in compact disk players). The advantages to automated flow counters include high throughput with excellent reproducibility. If equipment costs can be brought down sufficiently, such technology might make cell counting available at district hospitals. Innovations on this front will probably require one of the current makers of cytometry equipment to commit to entering the market in this unexplored niche. Once this commitment is made, however, progress should be rapid since the underlying technology is well understood.
Dynabeads is a method of tagging CD4 cells with antibodies linked to tiny magnetic beads. The tagged cells are separated from the blood in a magnetic field and counted under a microscope. Dynabeads brings down the cost of equipment for counting CD4 cells to under $10,000 -- but the technique is slow to perform and requires a trained human eye for evaluation. Under ideal conditions, the test can correlate well with results from flow cytometry, but in multiple tests this has varied greatly depending on the laboratory. Although the precision and reproducibility of the Dynabeads technique remains uncertain, it is commercially available and is being used in Senegal and elsewhere.
Another diagnostic short cut receiving a lot of attention is the potential for total lymphocyte count (TLC) to stand in for CD4 counts. TLC is much easier and cheaper to determine than CD4-specific T-cell counts and some researchers have suggested that low TLC, especially when combined with clinical staging, may be a good predictor of survival and can be used to help decide when to start treatment. However, while very low TLC correlates fairly well with low CD4 counts, higher TLC does not -- even though absolute CD4 counts may be dangerously low. TLC also fails to correlate with viral load and short-term response to treatment. Other researchers are investigating the value of hemoglobin levels when combined with symptoms as a predictor of clinical progression.
On the virus counting front, a number of alternatives to PCR appear promising although no single product stands out as ideal. It may be possible to modify existing viral load technology to simply give a readout within clinically relevant ranges instead of reporting absolute quantification. For example, to tell if therapy is failing it may be enough to have a reading fall into one of three bins: say, under 5,000, 5,000 to 50,000, or over 50,000 copies/mL.
Per test cost improvements may be realizable with existing commercial products such as real-time nucleic acid assays for HIV (NASBA). Optimized p24 assays that don't rely on PCR are also in development but these tests need continuing comparative studies with established assays to standardize and validate their use.
As in developed nation settings, blood samples collected in rural areas or primary care clinics have to be preserved and transported to the nearest laboratory. A breakthrough in low cost point-of-use diagnostics will undoubtedly find a market in healthcare facilities in all parts of the world. Until then, researchers in regions without access to dry ice and FedEx are experimenting with techniques such as dried blood spots for preserving and shipping samples under adverse conditions of heat, humidity and long delay.
For diagnostic monitoring performed at the district level to become feasible, new assays will have to meet very demanding criteria for accuracy, cost and ease of use. But if assays are to become useful at the point of care, then radically different criteria need to be imagined and met.
The cost of existing diagnostic technologies can be marginally improved by finding alternative sources of antibodies and test reagents, removing taxes and tariffs, and negotiating price breaks. But existing systems are not amenable to the kind of price pressure that has brought drug costs down. The only HIV diagnostic tool that has met the criteria for a cheap, easy-to-use commodity for field use are certain versions of the HIV antibody test, in particular those which use simple dipsticks to give rapid results from urine or saliva.
Disposable dipstick tests are also a promising candidate to achieve commodity status for use in HIV treatment management. Mass-produced dipsticks could report one or possibly two test results at an affordable though not negligible cost (several dollars per test). Unfortunately, developing these products requires going through an extensive and expensive multidisciplinary industrial process. Hopefully the projected size of the market for easy-to-use technologies -- in both the developed and developing world -- will attract sufficient investment to realize this potential. Dedicated dipstick assays for HIV management are not likely to appear within the next five years.
In the same way that the cell-phone has allowed many less-developed regions to leapfrog to world-class communications without the investment in costly land-based telephone infrastructure, the criteria for universal, low-cost, easy-to-use testing may only be met after the most technically advanced product candidates are fully developed.
Though still in its infancy, microarray chip assay technology might ultimately offer a solution to the most difficult-to-meet criteria. Imagine a small hand held device like a Palm Pilot. Place a drop of blood onto a window and insert a blank card. The sample is mixed with chemicals and distributed into hundreds of tiny wells where various chemical reactions take place. Within a few moments, an optical device scans the wells, the computer analyzes the output into a readable form and the results are displayed or transmitted to the user. A single HIV multitest chip might tell a patient's CD4 range, viral load range, detect abnormal blood chemistry, report liver enzyme levels, and even tell if he has a genetic predisposition for having mitochondrial toxicity from one of the drugs he is taking. A palmtop reader will cost a few hundred dollars, and because the quantities used are so small, the reagents and chip itself will cost only pennies per test. This technology is still on the horizon but in fifteen years it may bring excellent standards of diagnostic support to the even most remote rural areas.