UCLA Journal of Radiation Oncology APRIL 2023 - Flipbook - Page 16
UCLA RADIATION ONCOLOGY JOURNAL
For similar reasons, the treating physician could
not be blinded, as the treating physician would
need to order, manage, and coordinate different
procedures and different scans for patients on
the different arms. In theory, an independent
physician reviewer could have been used to
adjudicate toxicity. However, this would also
have been impractical for two major reasons.
also significantly lower with MRI guidance vs CT
guidance (0.0% vs 10.5%, P = .003).
After SBRT, urinary incontinence subdomain
scores decreased significantly more with CT
guidance vs MRI guidance at 1 month (11.3-point
vs 6.2-point decrement; P = .01), but were no
longer significantly different at 3 months). The
percentage of patients with a clinically relevant
increase in IPSS (>15 points) was significantly
greater with CT guidance at 1 month (19.4% vs
6.8%, P = .01), but not at 3 months (1.4% vs 4.1%,
P = .30).
First, the treatments are significantly different
enough that an off-hand comment by the
patient would immediately lead to unintentional
unblinding (e.g., a comment about “being treated
inside a tube” [only done on the MRI-guided
SBRT arm] or about “having seeds put in my
prostate” [referring to fiducial markers, only
placed in the CT-guided SBRT arm]).
Magnetic resonance imaging guidance vs CT
guidance also resulted in a significantly smaller
decrement in EPIC-26 bowel domain subscores
at 1 month (4.1-point vs 18.2-point decrement,
P < .001) but not at 3 months. A significantly
larger percentage of patients treated with CT
guidance experienced a clinically relevant
(≥12-point) decrement in EPIC-26 bowel domain
scores (50.0% vs 25.0%, P = .001).
Second, the trial was designed and launched at
the height of the COVID-19 pandemic. In fact,
two patients (one on each arm) died of COVID19-related complications before the 90-day
window to assess toxicity was complete, leading
to a final effective sample size of 154 patients.
Thus, axillary patient-physician encounters for
purely independent-adjudication in the context
of clinical research would have been highly
impractical, if even ethically permissible.
Finally, a numerically greater decrement in
EPIC-26 sexual domain scores was seen with
CT guidance vs MRI guidance at 3 months
(13.2-point vs 3-point decrement, P = .04; analysis
restricted to men who did not receive androgenTo attempt to mitigate this limitation, as well
deprivation therapy). Though the arms of the
as the limitation of whether physician-scored
trial were well-balanced with respect to relevant
toxicity is a relevant endpoint for patients,
variables like hydrogel use, nodal radiation,
we also collected patient-reported outcomes.
simultaneous integrated boost, and gland size,
Specifically, all patients received IPSS and
we did perform a post-hoc exploratory analysis
Expanded Prostate Cancer Index Composite-26
to see whether the benefit of MRI-guidance still
(EPIC-26) questionnaires at 1- and 3-months post- persisted after adjusting for these points. The
treatment. These were used to evaluate changes
results of this analysis suggested that trial arm
in urinary and bowel quality of life.
remained associated with a 60% reduction in
odds of grade ≥2 GU toxicity.
The primary endpoint results of the trial,
published in JAMA Oncology on Jan. 12, 2023,6
therefore included not only physician-scored
GU (and gastrointestinal, or GI) toxicity, but also
IPSS and EPIC-26 changes. Overall, the primary
endpoint of the trial was met and rates of grade
≥2 GU toxic effects were significantly lower with
MRI vs CT guidance (24.4% vs. 43.4%, P = .01).
Additionally, rates of grade ≥2 toxic effects were
An open question, and one for which we
encourage study, is whether alternative
imaging and treatment delivery platforms could
facilitate a similar reduction in PTV margins.
A major challenge in designing and executing
the MIRAGE trial was the need to randomize
between two significantly different treatment
platforms – a simple margin reduction trial
16