said
Mohammadhadi Khorrami, M.S., a Ph.D. candidate from the Department of
Biomedical Engineering, Case Western Reserve University School of Engineering
in Cleveland, Ohio, who, along with Monica Khunger, M.D., from the Department
of Internal Medicine at Cleveland Clinic, led the study.
The findings of the study appear in the journal
Radiology:
Artificial Intelligence.
Predicting
Response to Treatment in NSCLC Using Radiomics
The study team set out to identify the role of radiomic
patterns both within the
lung tumor as well as a peri-tumor
area to predict response to chemotherapy, overall survival as well as the risk
of progression in patients with
non-small cell lung cancer (NSCLC).- The study
included 125 patients treated in Cleveland Clinic with pemetrexed-based platinum doublet cancer treatment
- In this
randomized study, the participants were grouped randomly into a training
set and validation set. The training
set had 53 patients with NSCLC, while the validation set had 72 patients
- The training set
included equal numbers of patients who had responded to NSCLC chemotherapy
(responders) and an equal number of patients who had not (non-responders)
- A computer
analyzed the CT scans of lung cancer to pinpoint unique patterns of heterogeneity both inside as well as the
area around the tumor
- These image
patterns were then matched against the CT scans of both responders and
non-responders and these unique feature patterns present were then used to
train a machine learning system to identify those patients who were likely
to respond to chemotherapy
- The results
demonstrated that the radiomic patterns derived from both within the tumor
and the area around the tumor were able
to differentiate between patients who responded to chemotherapy from those
who showed no response
- Also, the
radiomic features were able to
predict the rate of
progression and overall patient survival
- When patterns
within the tumor alone were analyzed, the accuracy of prediction was 0.68
but when patterns both within the tumor and peri-tumor area were analyzed,
the accuracy went up to 0.77
The findings of the study suggest that analysis of radiomic
patterns both within the tumor and the area surrounding the tumor
can accurately predict patients who are likely
to respond to chemotherapy in NSCLC as well as the rate of progression and overall survival.
"This is the first study to demonstrate that computer-extracted
patterns of heterogeneity, or diversity, from outside the tumor were predictive
of response to chemotherapy," Dr. Khunger said.
"This
is very critical because it could allow for predicting in advance of therapy
which patients with lung cancer are likely to respond or not. This, in turn, could
help identify patients who are
likely to not respond to chemotherapy for alternative therapies such as
radiation or immunotherapy." Although, the possible reason for the difference in
response to chemotherapy is not clear, the study team believes that increased
fibrous tissue content of tumor may make it more responsive to chemotherapy.
Summary
Analyzing unique radiomic patterns of CT images can help
predict patient response to chemotherapy in lung cancer as well as rate of
progression and risk of tumor recurrence. This information will help treating
doctors plan management accordingly.
Reference: - Combination of Peri- and Intratumoral Radiomic Features on Baseline CT Scans Predicts Response to Chemotherapy in Lung Adenocarcinoma - (https://pubs.rsna.org/doi/10.1148/ryai.2019180012)
Source: Medindia