 It's a pleasure to present the results of this trial to you, which was done last year during a meningococcal meningitis outbreak in Niger. Some background, meningococcal meningitis outbreaks happen in what's called the meningitis belt of Africa, basically as the Sahel. There's seasonal, cyclical outbreaks that come during the dry season. Standard epidemic response is based on three pillars, reinforced surveillance, case management, and reactive vaccination. But we know that reactive vaccination often happens late. Supplies to vaccines are managed by the international coordinating group. It's tough. Vaccines are not stored in country. It takes a while to detect an outbreak, to make a request to get the vaccines in country, then to vaccinate, and then another 10 days before the vaccine actually becomes protective. As a result, its late impact of reactive vaccination is mitigated, and we know that it's better to vaccinate faster. It's just tough to overcome some of these logistic barriers. This is compounded by the fact that there's a limited vaccine supply. There are only 2.4 million doses of vaccine available for reactive campaigns this past epidemic season. Thankfully, it was calm. That would not even cover the entire city of Kano in Nigeria, let alone the entire meningitis belt. Unlike in the global north, antibiotic profile access for close contact of cases is not recommended in the African meningitis belt. That's for a few reasons. The first is because there's no evidence. There's like one pseudo trial from Sudan in the 1950s. But there's otherwise no real evidence. But then there's also concerns about logistics and resources. I talked about what we do during epidemics. That's already a lot of stuff. Without any evidence, is it worth investing other resources in going to people's houses or doing antibiotic reflexes? But then the calculus sort of changed with the emergence of serogroup C meningococcal disease, which started causing large-scale outbreaks in 2015 in the meningitis belt. And basically a WHO expert panel said, well, given that we don't have enough vaccines, someone should start looking into this idea of antibiotic prophylaxis, which is what we did. We designed a three-arm open-label cluster randomized trial to assess the impact of prophylaxis of single-dose ciprofloxacin, either to household contacts or to entire villages on overall attack rates during a meningitis epidemic. We're not talking about the individual level protective effectiveness of the ciprofloxacin. We're talking about, is this an effective outbreak strategy? And just so you know, this was something that was planned years in advance. And we had protocols approved and stops of ciprofloxacin preposition. The entire protocol was actually published. And you see the results of the reference there. And then we actually got to do it, which was even better. There were three arms. The first arm was a standard control arm, the standard care arm, what was the control arm. Case management happened as usual, but then otherwise nothing else happened. In the second arm, we gave ciprofloxacin to household contacts. And we were large with our definition of household. It was all the co-spouses. It was all the kids in these large compounds in a row and a share. And we gave it directly observed by a nurse at home within 24 hours of the notification of the case. And in the third arm, we organized mass distributions of ciprofloxacin, that sounds dramatic, but village-wide distributions of ciprofloxacin within 72 hours of the notification of the first case from a village during an epidemic. Now, all doses of ciprofloxacin that we gave were directly observed. We didn't give little baggies worth of cipro to take home. We gave every single dose was directly observed. And it was age-based dosing using an age-based scale. And we gave it to everyone from one day to 100 years. Children and pregnant women included. There were predefined study launch criteria, which in the setting of Niger was that two health districts across the epidemic threshold, two health areas of the same health district, across the epidemic threshold during the same week. That's because we wanted to make sure it was going to be large enough epidemic to make it worthwhile to actually start the trial. And villages were randomized after the first case was notified from that village. In a household prophylaxis arm, your household only got prophylaxis ones. You only got cipro ones. If three days later, your wife got meningitis, we didn't go back and give a second round of prophylaxis to that household. Similarly, in the village-wide prophylaxis arm, each village only got prophylaxis distributions ones. Because we were looking at attack rates, surveillance was very important. We had a dedicated surveillance nurse in each health area of the study and in each of the facilities. We used standard WHO definitions for meningitis. And then we also used the standard MOH, Ministry of Health, procedures for sample collection and flow to the reference lab, where PCR testing only was done. There were no trans-isolate tubes taken during the epidemic. And then also because we were interested in attack rates, we performed a door-to-door census in each of the villages that were included so we could have accurate denominators. Again, our primary outcome was the attack rate in each of the three arms. And then we also looked at potential confounders and adjusted using Poisson regression. Some of you are concerned about antibiotic resistance. I think probably everyone in the room is. It's the major question that was brought up to us that we don't want to create lots and lots of drug-resistant salmonella infections so that if we have to look at the risks and the benefits of any such strategy, even if we had a great result, we wouldn't want to do it if we create other problems. So to do this, we performed a sub-study that took 10 villages in the control arm and 10 villages in the village-wide prophylaxis arm. We recruited 20 individuals in each of those villages and we took stool samples from them on days 0, 7, and 28. Now, we're concerned about invasive salmonella. We can't really set up this surveillance over the long term that would be necessary. So we used prevalence, carriage prevalence, of resistant gut flora as our surrogate marker. So we plated up the stool samples from days 0, 7, and 28 on selective media that were selective for both sephatacem and ciprofloxacin to look for both ESBLs and ciproresistance and did quality control at a reference lab in Paris. In terms of the ethical considerations here, the cluster randomized trials are a little bit interesting. We followed the principles set out by the Ottawa statement on the conduct of cluster randomized trials. Village chiefs provided permission for their villages to be randomized, but we didn't have individual consent from each of the people that received ciprofloxacin. You'll see why I'm in this 72,000 people over the course of a month. It's a bit difficult. The timeline, it all happened very fast. On the 20th of April, the trials start criteria we were met. And within 36 hours, we took the decision at that point, and within 36 hours, we'd already started the trial. About three weeks later, it started raining. And then if you'll remember when I was saying that reactive vaccination always comes late. After that, the vaccine campaign started. And then the last village was included. And the last case was notified just over a month after the trial started, so it was very quick. In terms of the results, the baseline characteristics of the villages, to make it really simple, they were all sort of the same. This is what you want in a trial. It's rural Niger, maybe slightly larger than we had expected in terms of the size of the villages, but largely the same in terms of population structure. Also importantly, in terms of when they were randomized over the course of the epidemic, 47 of the 49 villages were eventually vaccinated and also in terms of when they were included versus when the rains came. Because obviously, rains are very important in terms of stopping meningococcal meningitis epidemics. So, primary results. In the standard care arm, there were 115 cases and the household prophylaxis arm, there were 91 cases and in the village-wide prophylaxis arm, there were 42 cases. That translates into attack rates of 451 per 100,000, 386 and 190. That's a significant difference in terms of reduction of attack rates in the village-wide prophylaxis arm. We looked at potential confounders. The only one that actually remained in the model was whether the village was included before or after the first rains. And it translates in the adjusted model to about a 60% reduction in attack rate in the village-wide prophylaxis, given within 72 hours of the notification of the first case in a village. In terms of laboratory results and laboratory confirmations, so these were suspected cases of meningitis. We had about 20% of the samples were sent for confirmation, which I admit is not as high a percentage as we would have preferred. All of those that were sent that were positive were positive for Sierra Group C meningococcus. And again, the numbers are small, but in the standard care arm, there were 28 samples sent, 16 were positive. In the household prophylaxis arm, there were 16 sent and five were positive. And again, small numbers, but there were no confirmed cases of Sierra Group C meningitis in the village-wide prophylaxis arm. To this point, I was very optimistic about the results, and then putting this graph sort of brings it home for me. You see the difference here. This is the histogram for each of the three different arms, the standard care on the top, the household prophylaxis in the middle, and the village-wide prophylaxis on the bottom. And what you see is that the difference really comes here in the first few days after randomization, which is really consistent with what we know about meningitis epidemics at a village-wide level. So maybe at a district level it might last six weeks, but at a village level it's a very rapid thing. It'll blue through a village in seven days, ten days. So this is very consistent with what we know about how the transmission of meningitis epidemics works. Interestingly, coverage, we didn't know how we were going to do. One of the questions was, if we give household prophylaxis, are we going to end up giving to half of the village, because these are big villages. We actually only need up to four percent of the population in these villages. And impressively, given the time constraints we were in, there was a huge increase in the distribution. And this is of everyone in the village, adults included. And then, because this has all been great so far, I'll show you the results that are less great. It's maybe not less great for the strategy, but it's certainly concerning. At baseline, the prevalence of carriage of a super-resistant bug in the fecal floor was 95 percent in both arms and didn't change significantly. For those of you who are really into it, the baseline carriage prevalence of ESBL bugs was over 90 percent in both arms. Now, again, this is not invasive disease. These were all E. coli. This is, let's be clear about what we're talking about. We're talking about carriage of resistant bugs. This was highly surprising and much higher than anything that's ever been described in the literature. All these results were completely concordant with the results we found in the reference lab in Paris. So, in conclusion, village-wide prophylaxis works. We significantly reduced the attack rates and it could be a really attractive epidemic response because it's faster, it's cheaper, you can stock drugs in the country. We would have preferred to have more confirmed cases, but the trends are the same. And then the question is, what do we do about the resistance? Hopefully, this is one place, hopefully it's not as bad in other places, but we also don't know what this means to carry these bugs. One thing that's sure is that people are well exposed to antibiotics in this part of the world. So, the next steps, we presented back to the WHO expert panel who thought the results were great, but would need more evidence before changing recommendations, particularly on antibiotic resistance patterns, but also what it might look like in an urban area where transmission patterns might be different. So, for this past season, we had superpre-presitions in both Niger and Nigeria with protocols that were pre-approved. We also made contingency plans for what would happen in the big epidemic in Borno where we couldn't do a trial. What would need to be documented. And thankfully, there wasn't a big epidemic this year. So many thanks to everyone that participated. This was a massive effort in the field to do a trial during an emergency. It's not easy. Thanks to everyone. Also, especially thanks to Bashir, who just presented who was the field investigator for both of these last two studies and whose birthday is today.