Jason Napodano, CFA & Aafia Chaudhry, MD, recently wrote a nice article discussing some of the flaws in the current bear thesis on Trius Therapeutics (TSRX). One of the bear theses that they mention is the looming patent expiration of linezolid (brand name Zyvox and marketed by JNJ). While they are accurate in that this fear is overblown as a competitive threat to tedizolid, there are very specific reasons why this is the case that were not detailed. In particular, an unwritten assumption of this bear thesis is that resistance to linezolid will not become a problem even if its use increases dramatically. This makes little sense. It is true that hospital and insurance companies may want to rely on a cheaper generic linezolid over a less effective vancomycin and a more expensive tedizolid but they cannot in the face of resistance. So the question then is whether linezolid is resistant to resistance.
Given enough time bacteria can become resistant to any antibiotic. For instance, staphylococcus aureus developed resistance to each of the main antibiotics used to treat it: penicillin, methicillin, vancomycin, and now linezolid. While the resistance mechanism varies, what does not vary is that the increasing use of an antibiotic leads to resistance and the proliferation of that resistance. Vancomycin, for example, saw rapid uptake in the early 1990s as the drug of choice to treat methicillin-resistant staphylococcus aureus (MRSA). This put selective pressure on the bacteria and led to the development of vancomycin-intermediate Staphylococcus aureus (VISA) and heterogeneous vancomycin-intermediate Staphylococcus aureus (hVISA). While cases of complete resistance remain rare, Howden et al (2010) are quite clear in showing that it is steadily growing and unlikely to stop. Some estimates of vancomycin resistant MRSA have it increasing from 2.8% in 2005 to 11.1% in 2008 (slide 6). While this trend is certainly problematic, these strains often remain sensitive to linezolid.
Early in its development linezolid showed efficacy in a large number of resistant bacteria and led to it becoming an antibiotic of choice in treating resistant infections. Linezolid is interesting in that it was both the first in a new class of antibiotics (oxazolidone) and a fully synthetic compound (unlike other antibiotics that are developed from natural molecules). The synthetic aspect of linezolid was unique and "it was initially expected that there is no natural pool of resistance gene which could facilitate the development of clinical resistance" (Nian et al 2012: 5043). Despite being a fully synthetic antibiotic and the first in a new class, however, resistance developed. While in 2008 the rate of resistance in the United States was low (about 0.4%), this will (as was seen with vancomycin) only increase with time and use of linezolid.
Of course, resistance to linezolid is worrying but what is also critical is understanding how quickly these mutations may spread. For any given antibiotic, there are a number of mutations that can generate degrees of resistance. The most dangerous are those that can be passed with relatively ease between bacteria and do not result in a large cost to fitness (the bacteria need to reproduce in a hostile and highly competitive environment). Linezolid inhibits bacteria by binding in a deep cleft in the 50S ribosomal subunit surrounded by 23s rRNA nucleotides. Given the binding site, mutations in 23s rRNA could generate resistance and "the most frequently reported mutation in linezolid-resistant clinical isolates is 23S RNA G2576U … [where] this mutation has been reported in both staphylococci and enterococci" (Long et al 2012: 606). In addition, Long et al (2012) notes resistance being related to mutations in ribosomal proteins L3 and L4.
While these mutations are important, the most problematic is resistance caused by multiresistance gene cfr as it is transferable between bacteria. Long et al (2012:609) highlights the import of this mutation as:
The methylation confers combined resistance to five different classes of antibiotics that bind at overlapping nonidentical sites at the PTC. This resistance is substantial and functions in both Gram-positive and Gram-negative bacteria. The phenotype is call PhLOPSa, for resistance to phenicols, lincosamides, oxazolidinones, pleuromutilins, and streptogramin A antibiotics. In addition, Cfr confers significant resistance to selected 16-membered ring macrolide antibiotics such as josamycin and spiramycin, but not tylosin.
The key here is not just the breadth of resistance but the transferable nature of the mutation. In addition, the cfr mutation does not come with a fitness cost, which "is troubling, as it suggests that cells can maintain the gene even in absence of antibiotic selection" (Long et al 2012: 609). The question then is whether these linezolid resistant strains remain sensitive to a second generation oxazolidinone like tedizolid.
At the 2012 ICAAC meeting, Cercenado et al presented data on the sensitivity of linezolid resistant strains to tedizolid and radezolid (another second generation oxazolidinone being developed by Rib-X Pharmaceuticals). They measured the MIC50 (the minimum inhibitory concentration (mg/L) needed to inhibit the growth of 50% of the bacteria) of linezolid, tedizolid, and radezolid in 298 linezolid resistant strains. The mean MIC50 for linezolid was highest at 16 followed by radezolid at 2 and then tedizolid at 0.5 (keep in mind that the lower the MIC50 is better and under 8 is considered sensitive to a drug). Obviously, linezolid would have the highest as there are strains that are known resistant to it but what is critical is that both second generation oxazolidinones were effective. They also looked at particular mutations. For instance, they had 141 strains that gained linezolid resistance through cfr and the mean MIC50 of linezolid was 8 compared to 1 for radezolid and 0.25 for tedizolid. The most difficult mutation were strains with a combination of cfr and G2576T mutations and in those 53 strains the mean MIC50 for linezolid was 128 compared to 8 for radezolid and 2 for tedizolid. They also looked at linezolid resistant MRSA and in those 73 strains the mean MIC50 for linezolid was 8 compared to 2 for radezolid and 0.25 for tedizolid. When all was said and done, it looks like both second generation oxazolidinones are effective in linezolid resistant strains with tedizolid consistently being the most effective.
So why does any of this matter? As noted at the beginning, one of the biggest questions with Trius is the commercial viability of tedizolid, especially with a generic linezolid coming on the horizon (although precise timing is still not clear). The bears are basically asking whether a market with an effective and generic linezolid and payers sensitive to costs has enough room for tedizolid. This line of argument has a critical flaw and that is linezolid will not remain as effective in the future as it is today. In fact, a market with high volume linezolid use will only increase the speed at which resistance spreads. As such, tedizolid at a bare minimum has a place in linezolid resistant strains, which will only be growing over time and this is a segment where the generic nature of linezolid would have absolutely no import.
Of course, tedizolid should also find a place outside of linezolid resistant strains. Deane et al (ICAAC 2012) found that "TZD [tedizolid] consistently demonstrated more potent in vitro activity than LZD [linezolid] against all the S. aureus and CNS [Coagulase-negative staphylococci] strains studied, and the difference in potency was usually by several doubling dilutions." In other words, tedizolid was significantly more effective than linezolid. In another ICAAC presentation, Deane et al found that "TZD consistently demonstrated more potent in vitro activity than LZD, DAP [daptomycin], or VAN [vancomycin] against E. faecalis and E. faecium; and was consistently more potent in vitro than LZD against all the organism groups studied." These two studies essentially show tedizolid as being one of the most effective antibiotics against some of the most common bacterial strains. While oxazolidinones are mainly used to treat Staphylococci and Enterococci infections, Cynamon and Sklaney (ICAAC 2012) also showed that tedizolid was more effective than linezolid in treating M. fortuitum.
Overall, the data seem fairly consistent and clear: tedizolid is an active antibacterial agent and effective against some of the most resistant strains. These characteristics open two markets for tedizolid. First, the growing linezolid resistant strains will have few options outside of tedizolid and will be immune to any effect from the arrival of a generic linezolid. While this might be a small market in terms of prescriptions, it will only grow over time. Second, tedizolid is a highly effective drug with both an ease of use and clean safety profile and it will take market share from current drugs. It is difficult to completely model this but some market research from Trius indicates significant interest from doctors (see slide 11). This market will be impacted by the arrival of a generic linezolid but perhaps not as significantly as many think (see slide 12). While this does not mean that tedizolid will have the same sort of sales success as linezolid, it has a quite favorable comparison and between the two potential markets it is set to be a commercially successful drug.