Imaging Studies and the Impact of Intercurrent Illnesses

Joseph Pierro and David Raunig |

The COVID-19 pandemic has caused significant disruptions in our daily routines as well as clinical research. Many research institutions have postponed non-critical research for 3-12 months and ongoing studies face challenges due to the direct and indirect disruption caused by the pandemic. Additionally, a myriad of related factors has led to lost patient visits, protocol deviations and missing data (missing time points), including disturbances in drug supply logistics and patient site visits, investigator staff transitioned to remote work or deployed to clinical care duties, and COVID-19 infections. This data critically serves as the basis for sponsors and regulatory agencies to better understand and navigate the impact of COVID-19 and intercurrent events (ICE).

The FDA and ICH have issued guidance on ICE and define them as events that are not included in the considerations for randomization, and which can undermine a clinical trial if they become too numerous.1,2 Within the COVID-19 pandemic, ICE examples include trial subjects who miss visits because of the risk of going into the clinic, or subjects who pass away due to COVID-19 or are unable to participate because they are considered as high-risk for COVID-19 (e.g., elderly). A non-subject related example includes disruption to the drug distribution process due to lockdowns delaying enrollment and access to treatment.

None of the ICE consideration strategies laid out in the FDA and ICH provided guidelines are particularly effective, and our best advice is to anticipate and plan for changes and study deviations so you don’t have to use a corrective strategy at all. As such, we would like to share observations that we believe can be instructive in this moment, based on a number of different approaches that sponsors have employed to address ICE challenges. Central to our observations is the reality that imaging core labs (ICLs) are well positioned to innovate their offerings in partnership with sponsors to better meet the needs of both the COVID-19 pandemic, but also broader long-term changes to how we run clinical trials overall.

First, one study team mentioned switching from central to site-based reads. In our opinion, this is the exact opposite of what should be done given the impact on data quality and reader variability due to the inclusion of using numerous site readers instead of a limited pool of independent readers. As we’ve discussed in a past blog post, central readers are better able to modify their availability to read more easily than the local readers who will also have to deal with the demands of overcrowded ICUs. Additionally, imaging studies might find that an imaging site has to shut down or limit services, and a different scanner or imaging site must be used for a subsequent visit. Without anticipating this scenario, the eCRFs typically don’t have any way to capture the calibration phantom data that could be used to measure any scanner differences or provide a mechanism to allow the ICL to account or correct for them.

Another problem is dropping a reader for any one of a number of reasons, which would likely be at a site as explained above, but each ICL has procedures in place to accommodate this scenario. A bigger problem may be missing data due to subjects missing scans because they believe going into the clinic will place them at higher risk of COVID exposure. Solutions may include the use of mobile scanners or home-based assessments. Furthermore, in situations where time points may be missing or delayed there are ways to determine the likelihood of disease progression like measuring the rate of increase in target lesions and possibly measuring more than 5 lesions so that rate of progression can be used to interpolate or extrapolate progression to augment assessment of Progressive Disease (PD) or Not PD.

The pandemic has had an immediate effect on the conduct of clinical trials and it’s important to emphasize, as has been recognized by the FDA and other bodies, that the “old way” of doing things is probably gone forever. We now face a sense of urgency to change the way we run trials. The use of ICE strategies is not a cure-all when it comes to correcting for COVID-19 biases and these strategies are sometimes problematic, as we’ve noted above. The available options that ICLs can provide, since ICLs have done hundreds of trials with many sponsors, highlights the innovations that core labs must use as a part of their business model.

In closing, regulatory guidelines recommend clinical sponsors recognize that these types of occurrences may occur in ongoing trials or when planning new trials during the pandemic. As such, trial planning should include meetings or discussions with the relevant review bodies regarding the impact of ICEs on data that may not be collected and limitations to planned data collection/analyses and subject recruitment. In our experience, anticipating and preplanning how the changes and study deviations (e.g., missed RECIST study visits) impact the plan analysis and statistics is the optimal approach.

For additional perspective on impacts of COVID-19 on clinical trials, and solutions for navigating the associated disruptions, including ECG/EKG advancements and using virtual technology in oncology trials, visit our website and search for “COVID”.