In the real world getting 100% is really hard. The contraceptive pill is very effective but results are 97%. There are many issues.
GT2 is probably the easiest to clear - 97% with Sof+Riba beat all the other genotypes. GT1b is easier than GT1a. GT3 is much harder than GT1.
I have a GT2 patient who is a SVR4 but he took 4 weeks Sof+Dac, had 2 weeks off, took another 4 weeks then stopped and still has the last 4 weeks. He was just lucky.
I have another patient who went on holiday, forgot his tablets, did not take for 2 weeks in the middle but completed 12 weeks over 14 weeks. He has relapsed, and probably did himself no favours. I have another with the same profile and pending results.
When I ask patients about this I don't say "did you take all the pills" I ask "how many pills did you miss" because most people do miss a few.
Absorption and Dosing
Some patients taking a small dose of drug have high blood levels - take warfarin as an example - for some patients 1-2 mg a day is enough for others 10 mg is required. We use fixed doses, with no weight adjustment - big people get less mg/kg of drug than small people. At least one relapse I know if was well over 100 kg and perhaps under dosed.
With daclatasvir we seem to have identified Taurine in Red Bull and other energy drinks as a problem by inducing CYP3A4 and lowering the daclatasvir levels. Some supplements people take may not be helping.
Although it is not talked about in the english literature a lot coinfection exists. It makes sense that someone exposed once may have been exposed twice, and while one genotype may dominate you could still have low levels of another. Say you have GT1 as the main type and GT3 at low level and take Harvoni - this is not so good for GT3 and there are reported cases of GT1 relapsing as GT3 - one explanation is re-infection - another is coinfection inadequately treated
Response and Resistance
We have seen on treatment virological breakthrough in 2 patients and both behaved like patients on NS5A monotherapy with a slow viral load decline then relapse. It was like the sofosbuvir was not working. This is entirely possible and could be due to NS5B resistance, but my theory is that the sofosbuvir pro drug (which requires 2 phosphorylation steps to become active) was not phosphorylated by CatA and/or CES1 enzymes - deficiencies of both these enzymes are reported.
NS5A RAVs also exist at baseline and are more common in patients that relapse.
For GT1 fibrosis does not seem to make a huge difference but for GT3 it definitely does. Some people's fibroscans are out of date and they may have not been given sufficient duration.
With GT1 past failure reduces the 97% SVR to 94% with 12 weeks treatment. This increases to 96% with ribavirin or 99% with 24 weeks treatment.
Notice that the extension by 12 weeks only gets an extra 5% in this case.
Viral Load and Innate Immunity
With GT1 we know that with low fibrosis and treatment naive the SVR rate is excellent with 8 weeks treatment (98%) if the viral load is < 6 million, but over 6 million and 8 week SVR rate is 90%. More virus probably means more mutations to deal or more to kill or that the patient needs more help (as their immune system is basically ignoring the infection). Whatever the reason we need longer, at least for GT1, with a high viral load.
As well as the possibility of CatA and CES1 defficiency there are probably many other factors.
IL28 is not commonly considered but may have a mild impact with CC people more sensitive to the body's natural interferon than CT or TT people.
What is required
If a researcher inject 10 virus particles into a monkey they might get an infection. If it is 100 virus particles they WILL get an infection so we really need to kill just about every last virus on treatment.
So there are many possible reasons for relapse.
Longer, stronger and different drugs are the suggested approach to retreatment.