TOPIC:

There has been quite a bit of talk about relapse recently across all forums. To me this was entirely expected as with Indian generics appearing in mid December, Jan-Mar was a busy treatment time with SVR12 success and failures starting to appear in bulk in July. Online conversations have a selection bias in that cured patients move on, leaving patients with relapse behind. You can prove that to yourself here by using the forum search tool and selecting a term like "treatment" and then "posts older than 6 months ago" and sorted by views.

fixhepc.com/forum/search.html?query=trea...s&childforums=1&ids=

Here you will see a lot of names you don't recognise, and some that you will. Many of these patients don't post anymore because they have been cured of HCV so our forum is no longer relevant. It's sad to see people go, but I'm happy to know HCV forums have become irrelevant in their lives.

We are active following 448 consecutive patients for reporting purposes. We are following more, but this group of 448 represents a real world sample, chosen before any results were known to ensure it would be representative.

The numbers are now high enough that not a lot of change will be seen. The little grey bars at the top of the graph indicate the 95% confidence intervals. As you can see the more results there are the narrower this range (shorter bar) and it is 20:1 that the "real" result falls inside this range.





So we can see that SVR rates in GT1 are in the mid nineties, and for GT3 in the mid eighties. 323/448 are now at or past the SVR12 point with 50% (224/448) now having SVR4 (or more) results to hand.



What we can also see is that SVR4 is 96.4% durable through to SVR12 with the overall average SVR4 slipping from 91.1% (204/224) to 88.6% (163/184) so if you get SVR4 there is only a 3.6% chance you won't get SVR12 (the odds are now 27:1 in your favour).

Finally the rates of treatment experience and cirrhosis are high compared to manufacturer trials, so there is a degree of negative selection bias. In GT3 we have 38% cirrhosis and 45% past treatment experience (in the easy to treat PEG/Riba GT) so it's not entirely unexpected that we are not seeing the 90% suggested by the originator trials (the easy to cure patients got cured by PEG/Riba). A long term observation in medicine is that real world results are always a bit worse than what the trials suggest.

One of the really good reasons not to be involved in HCV treatment was the 100% certainty that SVR rates of < 100% predict some patients will fail. Individual relapses are heartbreaking, and to be frank I admire the stamina and fortitude of those doctors who had to treat using PEG/Riba given the much lower SVR rates, but to my eyes the rebels have been well served by generics.

Much appreciated, grateful for your time on this.
In good health to all in treatment.

Fantastic wealth of real info.

Morning all,

I don't understand the bit about the little grey bars. What is a 'confidence rate'? Can someone who understands statistics please explain?

Thanks, RSF

Hello Dr. Freeman,

thank you for these explanations. Although 100% cure rate would be a wonderful success rate, a success rate of more than 90% is also an impressive one.
I believe that any HCV patient who is currently on treatment with generics (or will be on treatment) has 2 key questions in mind:
- Am I among the ~95% who will get cured?
- Are the generics really effective?

For the moment (no mass treatment being deployed worldwide) the level of anxiety is somehow high among the patients, yet in the following months, with more and more data available about SVR12/24, the level of trust will increase and taking the treatment will be (I hope) something normal in tems of accessibility and confidence.

Cheers,
RHF

I don't understand the bit about the little grey bars. What is a 'confidence rate'? Can someone who understands statistics please explain?

I could point you at this en.wikipedia.org/wiki/Binomial_proportion_confidence_interval but the formulas might make your eyes bleed so let's work through it. It's actually pretty easy to understand.

Say we have a coin and we toss it once. It comes up heads. We now have 100% (1/1) heads, but it could just as easily have been 100% tails. It was impossible to get the expected 50:50 result with a single toss, but even with 2 tosses it is 1:4 of being 2 heads, 1:2 of being a head and a tail, and 1:4 of being two tails. With 3 tosses the chances of 3 heads are 1:8, 4 tosses 1:16, 5 tosses 1:32 so you can see that the more we toss the coin the less likely we are to get an "all heads" rubbish result.

If you go to this page: statpages.info/confint.html you can put in 1 for x and 1 for N and get the confidence interval on our single coin toss which is 0.0250 - 1.000 (2.5% to 100%). You can change the values for x (say heads) and N (say tosses) and see what happens.

Now say we toss it 3 times and get 2/3 - 66% heads. The confidence interval is 0.094 - 0.9916 (9.4% to 99.16%)

Now say we toss it 30 times and get 20/30 - still 66% heads. The confidence interval is 0.4719 - 0.8271 (47.19% - 82.71%)

See how the confidence interval gets narrower the more tests we do? I kind of makes sense that the more we test this, the closer we are going to get to the right answer.

Now say we toss it 300 times and get 200/300. Now we have tossed it so many times that random chance starts to evaporate because we have had lots of tries. The confidence interval is now 0.6102 - 0.7198 - narrower again, but....

This interval does not include 0.5 aka 50% aka 50:50 so we can conclude that the coin is loaded.

So in short the more results we have the more confident we are that the value is correct and the confidence interval tells us how much higher or lower the real result could reasonably be if we do an infinite number of trials.

What's "reasonable"?

Reasonable in this case is a mathematical definition because confidence intervals exist at different levels. We have been looking at the 95% confidence interval level that says - with mathematical precision - that 95% of the time, the results will fall within this range.

Thanks very much for the explanation Dr Freeman

RSF

Good timing as usual Dr. F
I was trying to explain to my wife what all the "hullabaloo" about relapsing was about on recent postings and here you come to the rescue, right on time and right to the point. Very helpful. Thank you.

Matt

"Individual relapses are heartbreaking, and to be frank I admire the stamina and fortitude of those doctors who had to treat using PEG/Riba given the much lower SVR rates, but to my eyes the rebels have been well served by generics."

I did imagine an ever so slight look resignation and despair with medicos that had treated me with this stuff - the look of someone chasing their losses on a poker machine and not catching up.

That said though, relapses on generics have proven to be a bit of a lousy time for me, but there was not the 'thank God that's [interferon/riba] over; I can get back to actually doing my work again" aspect about it.

Appears from Dr Freeman's post I am a 14 percenter?



Jeff

Sabrecat, you have the new Sofosbuvir Velpatasvir generics (Sofvel from Beacon being the first) coming on line that now make Gen3 as curable as Gen 1 We're all getting cured here! Dr. Freeman and Dr. Devsam to the rescue!

Thinking of you Sabrecat
Yup totes get what you feel about the old relapse having been there myself like many others too, and I am sure you are totally focused again on getting the bugger out for once and for all. I won't ever forget that feeling of sheer determination at relapse and the machine I turned into on a mission that I was never gonna quit. New meds coming through thank Gawd, and things are moving so much faster now too. Another better thing than those poopy old days.
All relapsers have my heart and everyone here I can see are getting behind anyone who has not cleared easily on this easy DAA tx we have nowadays.
Sabrecat, it is in reach I am certain.
Love from Ariel

Come on in Kathleen, the water's fine!

Hello Kathleen and Welcome......

Hep C generics are probably one of the only exceptions to the rule "If it seems to good to be true, it must be".

In the case of generics, the 100,000% markup on the originator medications is what allows the prices to be 1-2%

Even if you don't treat, it's nice to know you could, if you wanted to.

As a patient, it's got to be a cruel and unusual punishment to know that a cure exists, but not for me.