Simeprevir, Sofosbuvir, and the Limitations of the COSMOS

July 31st, 2014

Simeprevir, Sofosbuvir, and the Limitations of the COSMOS

These are exciting times for hepatitis C treatment, as the approval of simeprevir and sofosbuvir in late 2013 have made curing this disease a whole lot easier.

Since that sentence barely conveys the transformative nature of this medical advance, allow me this tortured analogy: Before simeprevir and sofosbuvir, curing hepatitis C was like making a transatlantic journey by ship, and you had to stay in steerage — always a long and painful trip, but you’d get there if you could stand it. Now the cure is like a business-class flight — much shorter, safer, and more comfortable, but you just have to have the resources to pay for it.

But our treatments are not perfect. I was reminded of this fact when someone treated with simeprevir and sofosbuvir — or “SIM-SOF” as it’s commonly abbreviated — just relapsed a month after finishing 12 glorious, side-effect–free weeks of treatment.

In hindsight you could perhaps have predicted that this patient — or this type of patient — would be a treatment failure on these new drugs. There was prior treatment failure on interferon/ribavirin; he has genotype 1a (and hence probably Q80K, didn’t test for it); he’s overweight; he has several other medical problems, including diabetes.

And, probably of greatest importance, he has cirrhosis, diagnosed by liver biopsy several years ago.

The treatment failure sent me scurrying over to the reason we use this combination anyway — the excellent HCV treatment guidelines and the remarkable COSMOS study (just published in the Lancet) of simeprevir and sofosbuvir, with and without ribavirin, in patients with HCV genotype 1.

Here’s what recommended in the guidelines for patients who have failed prior treatment:

Daily sofosbuvir (400 mg) plus simeprevir (150 mg), with or without weight-based RBV (1000 mg [75 kg]) for 12 weeks is recommended for retreatment of HCV genotype 1 infection, regardless of subtype or IFN eligibility.

And here are the relevant data cited:

Among those null responders with a Metavir fibrosis stage of 3 or 4 (n=47) who received 12 weeks of sofosbuvir and simeprevir, SVR4 was observed in 14 (93%) of 15 patients in the ribavirin-containing arm and 100% (all 7 participants) in the RBV-free arm.

A few things stand out from reviewing the study and the guidelines, aside from the high response rate:

The sample size in the COSMOS study was small.
The number with actual cirrhosis was even smaller, as they combined stages Metavir 3 and 4.
The number with actual cirrhosis who did not receive ribavirin and received only 12 weeks of treatment was smaller still.
And did I mention — the sample size in this study was small?

The reason I’m harping on the sample size is a lesson taught early in every Stats 101 course — namely, that a small sample size means that the outcome will not be very precise, with lots of potential wobble around the results.

The most recent data presentation of the COSMOS study allows us to drill down a bit more on the data. There were 7 patients — just 7 — with cirrhosis who received only simeprevir and sofosbuvir for 12 weeks. 6 out 7 were cured (SVR 12), for a response rate of 86%.

And the lower bound of the 95% confidence interval for this proportion? 60%, at least according to this web gizmo that does the math for us, if I’ve chosen the right “equation”. So the occurrence of relapses should not be a huge surprise, despite the > 90% response rate for the study overall.

Some take-home lessons:

Unlike the real thing, the COSMOS study is very small. (There’s a larger study of this combination ongoing.)
This small sample size means we can’t predict response rates with much accuracy, especially in the hardest to treat patients who were not heavily represented in the study.
SIM-SOF (note it’s rarely “SOF-SIM”) is still a great advance in HCV treatment, but treatment failures will happen.

So what to do next? The good news is that resistance to sofosbuvir is rare even in treatment failures, and that the HCV treatment cosmos is about to get bigger — the next class of drugs (NS5A inhibitors) is on the way soon, and the data on sofosbuvir plus ledipasvir or daclatasvir look very promising.

A reason for optimism even for those for whom SIM-SOF didn’t do the trick.

HIV and ID Observations

An ongoing dialogue on HIV/AIDS, infectious diseases,
all matters medical, and some not so medical.

July 31st, 2014