Chemical antibiotics were one of the great health successes of the 20th century. Antibiotics both naturally-derived and synthetic have resulted in huge decreases in both morbidity and mortality from bacterial infections As a consequence, the "antibiotic age" has changed public expectations about the results of infectious disease. However, this has led to high levels of inappropriate prescribing, where antibiotics may be given to fulfil patient expectations rather than for clinical benefit. Along with unwise uses in agriculture and elsewhere, this has contributed to recent rises in numbers of antibiotic-resistant bacteria.

As a result, many commentators have described the end of the antibiotic age, and the term "superbug" has entered the common vocabulary for mutidrug-resistant bacteria such as vancomycin-resistant Enterococcus (VRE), multidrug-resistant Staphylococcus aureus (MRSA), and multidrug-resistant Pseudomonas aeruginosa (MRPA). More recently, the term "XDR" has been introduced to describe bacteria that are beyond all effective antibiotic therapy.

New chemical antibiotics do not seem to be the answer, and Zyvox provides a worrying example. It is the first of the oxazolidinones, hailed as the first new class of antibiotics for thirty years. Development took fourteen years from the first laboratory work to licensing for use.

Superbugs are the subject of a high level of media attention, and are widely perceived to be resistant to all antibiotics. This is rarely the case. MRSA, for example, is usually vulnerable to a range of antibiotics in laboratory testing. However, it is important to discriminate between susceptibility to a particular antibiotic on a test plate in the laboratory and the efficacy of that antibiotic in the patient.

A good example of this is Tobramycin. This is the drug of choice for Pseudomonas infection of the lung in cystic fibrosis (CF) patients. Resistance rates are higher in CF than in the general population, but in the laboratory Tobramycin is still relatively effective, with resistance apparent in only 10% of CF strains in one major study (Pitt et al, 2003). However, in that study another 35% showed reduced efficacy, and the authors concluded that resistance to existing antibiotics was "disturbingly high". In the clinic Tobramycin does not show even this level of efficacy; it reduces rather than eliminates Pseudomonas infection of the CF lung, and Pseudomonas remains the biggest killer of CF sufferers worldwide. Thus in this setting even a "90% effective" drug is unable to resolve clinical disease.

Gales et al, reporting in 2001 on the worldwide SENTRY Antimicrobial Survelliance Programme noted that "P.aeruginosa shows a particular propensity for the development of resistance, and this situation is associated with increased risks of morbidity and mortality and higher costs". They also noted that "a decrease in the susceptibility of P.aeruginosa was seen over the 3-year period in all the geographic regions studied". Resistance ranged from 10 to 40% to individual drugs, with multidrug-resistant forms "increasingly being isolated".

The most worrying aspect of this study is that of the 13 most valuable antibiotics for the control of Pseudomonas, seven are likely to be affected by the transmissible carbapenemase determinants (TCD's) which are now spreading. These are encoded by transmissible genetic elements which produce a potent and wide-ranging beta lactamase that can confer resistance to all remaining beta lactam antibiotics. If these are eliminated, only six remain; two aminoglycosides and four quinolones. Worryingly, neither of these classes are free of concerns over significant levels of side effects.

There are also considerations which limit the use of antibiotics, notably the issue of toxic side effects. These can be severe, especially for those rarely used antibiotics which, as result, remain relatively effective. These "drugs of last resort" may also require intravenous administration, necessitating in-patient hospital care.

One example is the polymyxin antibiotic Colistin, noted by Gales et al (2001) in their authoritative report as a possible treatment for multidrug resistant Pseudomonas. However, they also described it as "considerably toxic", while the British National Formulary (the UK prescribers handbook) describes it as follows: "It is not absorbed by mouth and thus needs to be given by injection to obtain a systemic effect; however, it is toxic and has few, if any, indications for systemic use".

By contrast, microbiological agents used for biological control typically show high specificity, which minimises side effects while maximising efficacy.

Bacteriophages (often known simply as "phages") are viruses that grow within bacteria. The name translates as "eaters of bacteria" and reflects the fact that as they grow, the majority of bacteriophages kill the bacterial host in order to release the next generation of bacteriophages. Bacteriophages are incapable of infecting anything other than specific strains of the target bacteria, underlying their potential for use as control agents. 

Bacteriophages were first discovered in 1915, and were then shown to kill bacteria taken from patients suffering from dysentery. Felix d'Herelle, a pioneer of this work, noted that bacteriophage numbers rose as patients recovered and suggested that the two were linked. Following this, d'Herelle was an early pioneer of what became known as bacteriophage therapy.

Up until the discovery of effective antibiotics, bacteriophages remained a real alternative to existing therapies. When broad spectrum antibiotics came into common use, bacteriophages were seen as unnecessary, with antibiotics being seen for many years as "the answer" to bacterial disease. This attitude persisted until the development of the wide-ranging (and in some cases total) resistance to antibiotics seen within the last ten years. In many cases it is necessary to use expensive new antibiotics (such as Zyvox for Staphylococcus aureus) or "drugs of last resort" (such as Vancomycin, again for Staphylococcus aureus) which often require complex routes of administration and can show toxic side effects, necessitating prolonged hospital treatment. Even to these drugs, resistance is reaching worrying levels.

In the light of current knowledge, it is apparent that early work with bacteriophages was hindered by many factors. One of the most significant was the widespread belief that there was only one type of bacteriophage, a non-specific virus that killed all bacteria. This was wrong. In fact, the exquisite specificity of bacteriophages is one of their main strengths. In addition, poor understanding of the causes of disease led to exaggerated claims that damaged the reputation of bacteriophage therapy.

With the far greater understanding of bacteriophages and their function that is now available, it is possible to identify the bacteria which are causing disease and then target them with bacteriophages that will kill those bacteria, and only those bacteria. This specificity has other benefits. Whereas antibiotics can kill a wide range of bacteria, leading to recolonisation of the body by inappropriate and often harmful bacteria, bacteriophages selectively eliminate only the target.