The emergence of previously unreported infectious diseases, the re-emergence of infectious disease thought to have been on the way to elimination, and the rapid evolution of infectious pathogens exhibiting antimicrobial resistance (“AMR”) and in particular, multiple-drug resistance (“MDR”) have created a major clinical and public health threat of global dimensions.

The idea that AMR can be delayed or even prevented by combining drugs with different targets as so-called “free” combinations or “fixed-dose” combinations (FDCs) has been shown in animal models of malaria and circumstantially in field trials of tuberculosis drugs but is difficult to rigorously test in the field. Microorganisms have several strategies (some used simultaneously) to resist being killed by chemotherapeutic agents. These include lack of, or a decrease in, transport of drugs into the cell, the operation of pumps to remove the agent from the cell, the production of drug-inactivating enzymes, and the mutations in the genes encoding drug targets. Microbes may be inherently resistant to an anti-infective agent and can acquire resistance to anti-infective agents. Most antimicrobials cause the selection of preexisting mutations, not the emergence of new mutants and acquired resistance is driven by mutation and selection - sometimes referred to as vertical evolution. This is particularly important for Mycobacterium sp., protozoans (Plasmodium sp., possibly Leishmania sp.) and viruses (HIV).

There are several ways that the different components of free or fixed-dose combinations produce their antimicrobial effect. The different drugs may attack the same biochemical target by different mechanisms (e.g., cotrimoxazole). Alternately, combination therapy may use drugs with completely different modes of action (e.g., artemether-mefloquine for malaria) and which in theory do not share the same resistance mechanism.

FDCs may be better than free combinations in slowing or even eliminating AMR. Multiple interruptions when using free dose combinations of pills creates the risk of monotherapy on some drugs and not in others. This fact, coupled with the in vivo mutation rates of the genome, rapidly leads to drug resistance to one or more of the free combination drugs. Fixed-dose combinations make the possibility of monotherapy even more remote. Effectiveness of FDCs, however, depends on detailed knowledge of the epidemiology and microbial ecology of the particular pathogen. Since in HIV, malaria, or TB, development of AMR commonly occurs by rapid genetic alterations, if evolution of AMR is occurring within a host during course of therapy (which in the case of HIV or TB is quite long), then FDCs would theoretically be effective if more than one drug is present in therapeutic concentration at any one time. If one in 10⁹ microbes are resistant to drug A and one in 10¹³ are resistant to drug B, and the genetic mutations that confer resistance are not linked, only 1 in 10²² will be simultaneously resistant to both A and B. If correctly given, combination drug treatments should in theory retard emergence of resistance compared with sequential use of single drugs.

The literature directed to determining if FDCs or free combinations are more effective in slowing or eliminating the development of AMR is weak. In this regard, we summarize our conclusions below:

• Most head to head comparisons/trials of monotherapy versus fixed combination versus free combinations are safety and efficacy studies;

• Only relatively recently has individual resistance to anti-TB and anti-malarials been measured at the molecular level;

• Responses of TB, malaria and HIV pathogens to combination drugs are very complex, particularly for malaria and HIV and the more and different combinations that will be used, the more complex will be the interactions;

• Free combination drugs are generally more prone than FDCs to dispensing and patient error. No studies of which we are aware have systematically looked at the effect of blister packs compared to FDCs and/or free combinations with regard to development of resistant pathogens. In this regard there seem few studies on health outcomes generally;

• Some studies suggest that decreasing overall antibiotic use may reverse bacterial resistance in human populations. One cannot assume from this that combination therapy will have the same effect. It is thus critical to know if using FDCs will prevent the appearance of drug resistance and/or reverse existing rates of drug resistance at both individual and population levels. The primary difficulty in assessing the evidence will be to actually measure developing/ongoing antimicrobial resistance in populations in field situations;

• Recent uses of molecular biology techniques might allow for easier tracking of clinical resistance markers although this genotyping must be correlated in the field with clinical outcomes. Larger longitudinal and community based studies are needed;

• Head to head comparisons of blister packs compared to FDCs and/or free combinations are needed with regard to health outcomes, including development of AMR;

• For HIV and malaria, it is critical to increase the pace of our understanding of genetic resistance pathways and mutations, since this understanding has not kept up with the increasing number of therapeutic options.