Background
Where were we with the Lopinavir/Ritonavir story? Let’s backtrack for a moment.
Lopinavir/Ritonavir showed some efficacy against SARS, but as triple therapy in combination with ribavirin. A while ago, some studies tested lopinavir/ritonavir alone against COVID-19 and this didn’t work. Consequently, that was set aside for a while. However, there were lingering questions about whether lopinavir/ritonavir might work when combined with ribavirin given synergy between these agents.1
Well, now lopinavir/ritonavir are back, along with two of their friends: ribavirin and interferon-1 beta (interferon-1b). The addition of ribavirin comes as no real surprise (honestly, ribavirin is a little late to this party). Adding interferon-1b is an interesting idea, as this drug may augment early innate immune responses to help control viral load. Overall, combination therapy with lopinavir/ritonavir, ribavirin, and interferon-1b represents a kitchen-sink strategy to early-stage COVID-19.
Hung et al: Triple combination of interferon beta-1b, lopinavir-ritonavir, and ribavirin in the treatment of patients admitted to hospital with COVID-19: an open-label, randomized, phase 2 trial.
This was a multi-center, open-label RCT involving patients hospitalized with mild COVID-19 who presented within <14 days of symptoms.2 Patients were randomized in a 2:1 ratio into two groups:
- Combination therapy group: Patients with symptoms <7 days received lopinavir/ritonavir 400 mg/100 mg BID for 14 days, ribavirin 400 mg BID for 14 days, and interferon-1b. To avoid exacerbation of inflammation, interferon was given only every other day up until seven days after the initiation of symptoms. For example, patients presenting 1 day after symptom initiation would receive three doses of interferon. Alternatively, patients presenting >7 days after symptom initiation received lopinavir/ritonavir and ribavirin without any interferon-1b.
- Control group: Treatment with lopinavir/ritonavir 400 mg/100 mg BID for 14 days.
Key inclusion criteria were:
- Symptom duration of 14 days or less.
- National Early Warning Score (NEWS) of two or greater (NEWS is an early warning system designed to quantify physiological derangement, as shown below).
- No history of prolonged QT, arrhythmia, or heart block.
- No history of depression.
- No pregnancy or lactation.
127 patients were recruited (86 in the combination therapy group and 41 receiving only lopinavir/ritonavir). Retention and follow-up were excellent, as shown below. Among 86 patients in the combination therapy group, 52 were admitted within <7 days of symptoms and received some interferon therapy, whereas the remaining 34 patients were admitted >7 days of symptoms and received no interferon.
Baseline characteristics are shown below. The median duration of symptoms prior to start of study treatment was 5 days (inter-quartile range of 3-7 days). The groups are well matched. Patients generally were only mildly ill (for example, only 7% of patients had dyspnea and the absolute lymphocyte counts were only mildly reduced).
The primary endpoint was the time from treatment initiation to a negative nasopharyngeal PCR, which was shorter in the combination therapy group (7 days vs. 12 days, p=0.001). Combination therapy also hastened clinical recovery and reduced the hospital duration of stay. Essentially, every important outcome was positive:
Combination therapy appears to reduce viral load within 1-2 days (figure below). This eventually translated into a delayed reduction in interleukin-6 levels as well:
Post-hoc subgroup analysis divided out patients treated within one week of symptoms compared to patients treated after one week of symptoms. Only patients treated within one week of symptom onset experienced a difference in viral clearance and symptom improvement (table below). This is difficult to interpret. It could reflect that early treatment is required. It could alternatively reflect that interferon-1b treatment is required (because only patients presenting within <1 week of symptom onset received interferon-1b). Or it could reflect under-powering of these subgroups.
Adverse events included nausea and diarrhea, with no difference between the two groups. This may be a bit misleading, as both groups were exposed to lopinavir/ritonavir. Thus, it’s certain that combination therapy does increase the rate of side-effects compared to no therapy at all. One patient in the control group discontinued lopinavir/ritonavir due to elevated transaminases.
deep thoughts
#1) so, this is what a positive trial looks like
Over the last several weeks, we’ve seen neutral trial after neutral trial. In desperation, this may lead us to lower our standards, to the point where we will grasp at any tiny benefit. Within this increasingly desolate context, it is nice to see something which actually looks like a truly positive trial:
- Numerous metrics of viral load show biological improvement following therapy.
- Numerous clinical metrics show clinically meaningful improvement with therapy.
The natural comparison here is to the Wang et al. trial on remdesivir.3 Both trials are very similar (for example, both involved 2:1 randomization and a sensible array of clinical and virologic endpoints). Despite being a larger trial, the Wang et al. study found no meaningful differences in clinical or virologic endpoints from remdesivir. The contrast is striking. One limitation in this comparison is that patients in the current trial presented earlier than in the Wang et al. trial (after a median illness duration of 5 days vs. 10 days, respectively).
#2) additional RCTs are urgently needed to confirm and extend these findings
This trial essentially tested a kitchen-sink strategy with three different agents. This opens up numerous possibilities, for example:
- All of these agents are required for maximal efficacy.
- Interferon-1b is really the active agent.
- Lopinavir/ritonavir plus ribavirin is the active agent.
Further trials are urgently needed to differentiate between these possibilities (and, in so doing, validate that one or more of these agents are indeed beneficial). These drugs are fairly benign, but they all have costs and side-effects – so it is of paramount importance to determine which ones truly work (if any).
#3) important pre-requisites for using an early anti-viral therapy (it can’t function in a vacuum)
Let’s imagine, for a moment, that we do find an anti-viral therapy (which, if given early in the disease course, guarantees a cure). That sounds terrific. However, any drug must function within the context of a healthcare system.
To obtain substantial benefit from an early-acting antiviral drug, the following pre-requisites are required:
- Basic society-wide interventions to reduce spread (e.g. social distancing, effective isolation of infected people, and excellent hygiene). Without these interventions, the number of cases is so high that early treatment becomes logistically impossible.
- Broad testing and contact testing. These allow identification of infected people early in the disease process, when anti-viral therapies are effective.
Hung et al. was performed on Hong Kong, an excellent example of leveraging public health strategies to prevent the spread of COVID-19 (nice discussion of this here). Thus, a critical factor silently undergirding the success of this study was the success of Hong Kong’s overall healthcare system.
Alternatively, current conditions in the United States would make it difficult for an early anti-viral strategy to work:
- Currently the number of infected people is growing at a rate of ~25,000 per day (even during lockdown). As the country opens up, this number will likely grow. This enormous volume of new cases would make it challenging to treat all of these patients with anti-viral therapy. For example, this would require ~700,000 tablets of lopinavir/ritonavir per day using the above treatment regimen.
- Most COVID-19 patients present to medical attention relatively late, outside a window where anti-viral therapy would be maximally effective.
Spending millions on anti-viral therapies in the context of a dysfunctional public health system is like giving a Ferrari to a 13-year-old boy. Yes, the Ferrari works. Yeah, it’s pretty cool. But no, that’s not a winning strategy.
#4) could this therapy actually improve mortality or avoid intubation?
This study involved patients who were mildly ill and did well – regardless of which treatment they received. In the current United States context, most of these patients may not have been admitted to the hospital at all. The study demonstrates biological and clinical efficacy, but the question remains as to whether these medications could improve “harder” endpoints (e.g. avoidance of intubation).
My guess might be yes – if you could treat enough patients early in the disease course, you would probably prevent some patients from deteriorating. Assuming this is true, the next question is – what is the number needed to treat? It might be extremely large (e.g. 50-200 patients with mild disease to prevent one intubation). Treating these many patients would involve a large cost and also an increased risk of incurring uncommon side-effects. Thus, a strategy of early testing and aggressive treatment might conceivably cause more iatrogenic harm than it’s worth.
This uncovers the fundamental paradox of anti-viral therapies:
- Early treatment is effective – but it’s unclear early on who will get sick (so this requires treating a larger group of patients).
- Restricting treatment to sicker patients decreases the number of patients who would require treatment. However, by the time patients become more severely ill, they have already progressed onwards in their disease course and are less likely to respond to anti-viral therapy.
Studies involving a sicker cohort of patients would be needed to clarify whether anti-viral therapy could improve outcomes among such patients (which may be a more important clinical issue).
- Hung et al. demonstrates biological and clinical efficacy using combination therapy with lopinavir/ritonavir, ribavirin, and interferon-1b early in the course of COVID-19. This study is consistent with some prior evidence using combinations of lopinavir/ritonavir plus ribavirin with SARS and MERS.1,4
- This study illustrates that moderate-sized RCTs can reveal differences in viral load and clinical course, potentially serving as a benchmark for study design. In comparison, the recent trial on remdesivir included more patients – without showing persuasive signs of efficacy.3
- Based on the combination design of this trial, further studies are urgently needed to validate this medication combination and determine which components are necessary for efficacy (e.g. is interferon-1b truly required for efficacy?).
- It’s unclear whether early anti-viral therapy could pragmatically be implemented within the United States. For example, if this therapy were currently made available, it might be largely utilized among patients presenting with late-stage disease, yielding minimal benefit.
prior studies on lopinavir/ritonavir
- Young et al. – earliest study on lopinavir/ritonavir (PulmCrit)
- Cao et al. – more definitive study that lopinavir/ritonavir monotherapy doesn't work (PulmCrit)
- Repository of notes on lopinavir/ritonavir.
references
- 1.Chu CM. Role of lopinavir/ritonavir in the treatment of SARS: initial virological and clinical findings. Thorax. Published online March 1, 2004:252-256. doi:10.1136/thorax.2003.012658
- 2.Hung IF-N, Lung K-C, Tso EY-K, et al. Triple combination of interferon beta-1b, lopinavir–ritonavir, and ribavirin in the treatment of patients admitted to hospital with COVID-19: an open-label, randomised, phase 2 trial. The Lancet. Published online May 2020. doi:10.1016/s0140-6736(20)31042-4
- 3.Wang Y, Zhang D, Du G, et al. Remdesivir in adults with severe COVID-19: a randomised, double-blind, placebo-controlled, multicentre trial. The Lancet. Published online April 2020. doi:10.1016/s0140-6736(20)31022-9
- 4.Park S, Lee J, Son J, et al. Post-exposure prophylaxis for Middle East respiratory syndrome in healthcare workers. J Hosp Infect. 2019;101(1):42-46. doi:10.1016/j.jhin.2018.09.005
- PulmCrit Wee: Rational selection of infusion rate based on loading dose - June 25, 2024
- PulmCrit: PPIs are safe and effective for GI prophylaxis… the end. - June 18, 2024
- PulmCrit: Bilevel Sequence Intubation (BSI) – The new standard - June 17, 2024
Josh, there should be enough data available to identify patients needing antiviral therapy only by looking at their preexisting comorbidities. If we selected a population with a risk of e.g. 20% of needing critical care, treating them with antivirals very early should show great results. I’m thinking of a CHADSVASC-like score for COVID-19
agree, if we do develop early antiviral therapy that works it would probably need to be restricted to higher risk patients, based on epidemiology data and possibly lab data (e.g. https://emcrit.org/pulmcrit/csi/). But even so, you would still need to treat lots of at-risk patients in order to catch a few patients who would get severe disease – because none of these prognostic models or epidemiologic risk factors are 100% perfect.
Great summary and thoughts. Determining if interferon is truly needed is critical because it requires a needle whereas the other two are pills.
keep it up Josh
you are one of the several brightest shining lights in this pandemic. keep it up. and thank you.
my of my favorite lines is: “Spending millions on anti-viral therapies in the context of a dysfunctional public health system is like giving a Ferrari to a 13-year-old boy. Yes, the Ferrari works. Yeah, it’s pretty cool. But no, that’s not a winning strategy.”
Josh, at some point I’d like you to consider discussing anybody testing. I wonder if that same logic applies there as well.
tom, from Merced
This is a great addition to the Hung journal found in the New England Journal of Medicine.
LOVE your graphics!