Earlier this week I blogged on the potential (yet poorly proven) effects of bacteriophages as salvage therapy for infections caused by AMR, and stated: “Phages and their active enzymes are proteins that evoke an immunological host response when injected, and up till now all attempts to circumvene those unwanted effects have failed.” Two recent case reports challenge part of that statement.

The first one from San Diego is in AAC, and describes a 68-year old man with gall-stone induced necrotizing pancreatitis complicated by MDR Acinetobacter baumannii infection. Due to clinical deterioration (despite multiple antibiotic courses) the patient was intubated at day 51, after which deterioration continued. On day 106 he was on multiple pressors with hepatic and renal failure. Phage therapy through percutaneous draining of pseudocyst  cavities, gall bladder and intra-abdominal cavity was started on day 109 and IV phage treatment was started on day 112. He abruptly deteriorated after 2 days and therapies were stopped, and then resumed after another 2 days until day 167. During this period and thereafter, in which antibiotics were also prescribed based on changing susceptibility profiles, patient fully recovered.

The second one from Belgium is in Crit Care, and describes a 61-year old woman, initially hospitalized with septic shock and Diffuse Intravascular Coagulation, due to E. cloacae peritonitis, complicated by gangrene of limbs. The remaining pressure sores were colonized with P. aeruginosa only susceptible to colistin, which caused septicaemia 4 months after initial admission. Ten days into colistin treatment she developed kidney failure, after which antibiotics were discontinued, which was then followed by a new septic episode with the same pathogen. After start of IV phage therapy, blood cultures turned negative and further deterioration (and hemofiltration) was avoided. Four months later she died from K. pneumoniae infection, despite full sensitivity of the strain to antibiotics of treatment.

Importantly, in both cases phages were apparently applied only after approval by national regulatory agencies: “under the umbrella of Article 37 (Unproven Interventions in Clinical Practice) of the Declaration of Helsinki in the Belgium patient and as an Emergency Investigational New Drug authorized by the FDA in the American patient.

More information on (very old) experiences with IV administration of phages is in this review, which also nicely summarizes 5 potential arguments against using phages intravenously (with my response added):

• Rapid removal of phages from the bloodstream (partly because there are so few bacteria to replicate).

Few if any “PK/PD” data available (but if it works, it works).

• Severe adverse immune responses, such as anaphylaxis.

No reports published, although this has always been a strongly voiced argument.

• Rapid development of antibodies might preventing long-term efficacious use.

Few if any data (but if it works, it works)

• Treatment delay due to time needed for phage selection.

Serious limitation for acute episodes with unidentified pathogen, which represent most episodes of sepsis. Yet, not such a problem if the pathogen is identified (as in both cases). Alternative is to administer broad cocktails (that might increase the risk of point 2?)

• Phage-mediated lysis of bacteria triggering a dangerous release of endotoxins.

Sounds irrelevant, as that is what expect antibiotics to do anyway.

So, within a week, I have learned more on bateriophages. Monday I wrote: Nevertheless, the concept is highly interesting, needs further studies, but clinical use is not expected soon.

Today, I remain with that conclusion. There is an urgent need to carefully report all human experiments (as that still is what it is) with intravenous use of bacteriophages, including the unsuccessful attempts. Clinical use may then be expected sooner, and the evidence may not have to come from Tbilisi.