NEW YORK (GenomeWeb) – While a great deal of emphasis has rightly been placed on the need to increase diversity in large genomic datasets, several researchers have also shown that the lack of access to cancer screening and treatment resources is hindering efforts to lower the rates of cancer-related deaths.
In Nigeria, for example, Obafemi Awolowo University researcher and surgeon Olusegun Alatise was seeing an increasing number of patients with colorectal cancer (CRC). The major problem, as he recently showed in his presentation at this year's American Association for Cancer Research conference, was that many of these patients died within a year of diagnosis because they were presenting with later-stage disease — their diagnoses came too late to give them a chance at an effective treatment.
So Alatise set about to change this dynamic. In partnership with colleagues at Memorial Sloan Kettering Cancer Center and the University of Alberta, he's working to develop a handheld point-of-care diagnostics device that would use a metabolomic signature in urine to screen patients and determine their risk for developing CRC. The device consists of a handheld biosensor that reads a disposable chip. The chip would contain the reagents that would react with the metabolites in a patient's urine — if a urine sample contains the correct metabolic signature, the biosensor will give a signal, indicating that the patient should be referred for further testing.
"This is not going to replace these normal screening tests like colonoscopy and fecal occult blood testing, but because the facilities [which would do]the screening are limited in resource-poor environments like mine, we need to have something that can do the initial screening that will be very low-cost," Alatise told GenomeWeb. "[With this diagnostic,] anybody can donate a sample, and then if something comes up, you can go for more expensive screening."
The challenge
Overall, he reported in his presentation at AACR, while the incidence of CRC has slowed or stabilized in many high-income countries, new diagnoses of the disease are slated to increase 77 percent between 2015 and 2030, and 62 percent of those new diagnoses will be in low-to middle-income countries (LMICs) like Nigeria.
Despite this, the International Agency for Research on Cancer's GLOBOCAN database — which includes estimates of the incidence of, mortality, and prevalence from 36 types of cancer and for all cancers combined in 185 countries of the world — only contained data representing 2 percent of the African continent as of 2017, and didn't have any registries in West Africa, he noted. Because of this, Alatise and his collaborators at the Memorial Sloan Kettering Cancer Center established the African Research Group for Oncology (ARGO) to develop a CRC database and conduct research on CRC incidence in Nigeria, the molecular features of the disease, and possibly conduct prospective clinical trials based on the data.
In a retrospective review of the available literature, Alatise found distinct differences between Nigerian CRC patients and American CRC patients. Along with presenting with later-stage disease, for example, the Nigerian patients also presented with CRC at a much younger age.
He then conducted a pilot study of CRC molecular profiles in 159 Nigerian patients, looking at microsatellite instability status, conducting methylation analysis and testing with the MSK-IMPACT tumor profiling assay, and performing germline mutation analysis in some samples.
He found that 27.7 percent of the patients in his cohort had MSI-high status compared to an average of about 10 percent of CRC patients with a status of MSI-high in the literature, and an average of 5.8 percent of African-American CRC patients in the MSK-IMPACT data. Also, 44 percent of the patients who had cancers with MSI-high status had tumors in their rectums, which is unusual as MSI-high tumors are usually present on the right side of the digestive tract, he added.
But this unusual finding also has implications for treatment. Because MSI-high is associated with an increase in immune infiltration and expression of immune checkpoint regulators, checkpoint inhibitors could be a particularly effective treatment for Nigerian CRC patients, Alatise said.
He also found a significantly low number of APC mutations in the Nigerian cohort compared to the African-American cohort and other groups in the MSK-IMPACT data, but a similar number of KRAS and TP53 mutations.
Further, he noted, 25 percent of the patients in the Nigerian cohort had hypermethylation. In the literature, hypermethylation is associated with BRAF mutations in CRC, but in the Nigerian cohort, BRAF mutations were largely absent.
In summary, Alatise said, the biology of Nigerian CRC patients is "intriguing," and knowledge of these biological differences may allow for better treatment, and for identifying high-risk patients in low-resource environments.
That knowledge, however, is hard to come by in an LMIC like Nigeria that doesn't have the resources to conduct genetic testing on every person who may be at risk for CRC. In fact, Alatise said in the interview, identifying people in Nigeria who may be at risk for CRC is challenging in a number of ways — access to colonoscopy is very limited, and about 30 percent to 40 percent of the population may suffer from some form of bleeding from an infective condition, making it challenging to use fecal occult blood testing.
"We can't recommend colonoscopy up front. We don't have enough resources to recommend colonoscopy for everybody. The system would be overwhelmed," Alatise said. "So, we looked for a surveillance device that we could use to stratify people at the level of the community before getting them to do [testing]."
Importantly, while fecal occult blood testing is thought to be cheap, the logistics to get it done can make it relatively expensive, especially in sub-Saharan Africa, he added. So, having a test that can at least stratify the people in the community that need to go for further screening could reduce the number of people going to larger academic centers for colonoscopies and other tests.
"It will help the system in so many ways, and we'll be able to pick a number of cases early," Alatise said.
A solution?
The researchers in Alberta had begun to look at metabolites of CRC, paring about 400 possible metabolites down to three. Their target was initially the detection of polyps, however. When he joined the collaboration, Alatise used the data he had gathered for the ARGO database on the molecular features of CRC in the Nigerian population and identified metabolites in urine that were more specific to the detection of early cancers. He was able to add two more metabolites to the signature, bringing the current total to five metabolites on the device.
It has a sensitivity of about 90 percent, he said, and its specificity is a little above 50 percent. As the test is only meant to stratify patients and refer high-risk patients for further testing, rather than as a replacement for colonoscopy or fecal occult blood testing, these levels of sensitivity and specificity are about what the team was aiming for, Alatise added.
Significantly, this same metabolic signature could also be used to test for CRC in other LMICs. "Currently we've verified the signature in our population and in Canadian and American populations — African-American and Caucasian groups. We've also tested it on Chinese urine samples that were available for study at the University of Alberta. And [the signature] seems to be ubiquitous," Alatise said. "Of course, like all devices there will be a first version, a second version, a third version. We will keep on learning. But the signal just seemed to be ubiquitous, so it is not peculiar to Nigerian samples."
In fact, Alatise would like as many LMICs as possible to undertake a pilot study using the diagnostic device. A grant for the device through the US National Institutes of Health calls for the manufacturing to take place as a partnership between a high-income country and an LMIC such as Nigeria. In order to make the manufacturing as cost effective as possible, at least some of the production has to take place in the LMICs. Further, Alatise said, the parts and resources for the handheld devices and the disposable sample chips also have to come from the LMICs. That's what will keep the cost down to about $1 per chip, he added.
After that, the team would recruit non-governmental organizations, universities, research organizations, and anyone else willing to help to distribute the tests to towns throughout these countries.
"We will rely on everybody. We want everybody to be on board. We are talking about a million people dying of a condition, that I know even if I carried these patients out of Nigeria to the best hospital in the United States, there's little we would be able to do — these are the kind of patients we see in low-middle-income countries. People that we cannot help," Alatise said. "So, it is so exciting that we are trying to get a device that can reduce that."
By the time AACR's 2020 meeting comes around, he added, the team will hopefully have conducted a pilot study and should be ready "to show the device to the whole world."