Introduction: In MRI-negative Cushing’s disease, inferior petrosal sinus sampling (IPSS) can help
confirm the pituitary as the source of excess ACTH. Prior studies have demonstrated
that IPSS does not correlate with pituitary adenoma laterality; however, IPSS-guided
resection continues to be practiced. This study evaluated different patterns of angiographic
cross filling of IPSS venograms to categorize the risk of pituitary adenoma mislateralization.
Methods: A single center, retrospective review of cases between 1998 and 2017 was performed
in which patients underwent IPSS for a presumed ACTH-secreting pituitary adenoma with
negative/equivocal imaging. IPSS venograms were evaluated for angiographic cross filling
into the contralateral IPS and the presence of basilar venous plexus filling along
the clivus. Angiographic filling patterns were subdivided into three groups ([Fig. 1]): Group 1: no angiographic cross filling when either IPS was injected; Group 2:
angiographic cross filling that occurred with the injection of one IPS but not the
other (directional cross filling); Group 3: angiographic cross filling that occurred
when either IPS was injected (bilateral cross filling). The contralateral and ipsilateral
ACTH response ratios (ratio of change in ACTH levels measured at 10 minutes compared
with 2 minutes after CRH administration) were compared between each group. The rate
of agreement between IPSS lateralization and intraoperative lateralization was calculated.
A multivariate logistic regression was performed with multiple venogram characteristics
to create a model of factors that predicts an increased risk of mislateralization
by IPSS.
Fig. 1
Results: Forty-two patients were included in the study, with 9 patients classified in group
1 (21.4%), 17 in group 2 (40.5%), and 16 in group 3 (38.1%). The mean fold increase
of ACTH levels in the ipsilateral or contralateral IPS was not significantly different
between groups; however, there was a trend of increasing contralateral and decreasing
ipsilateral ACTH response ratio from group 1 to group 3 ([Fig. 2]). Patients in group 1 and group 2 had a 77.8 and 94.1% rate of agreement between
IPSS and intraoperative lateralization, respectively, while patients in group 3 only
had a 68.8% rate of agreement. Multivariate analysis revealed that bilateral cross
filling had an increased risk of mislateralization by IPSS by 10 times (OR: 10.4,
95% CI: 1.3–232.1, p = 0.057), while directional cross filling had a decreased risk of mislateralization
(OR: 0.096, 95% CI: 0.0043–0.80, p = 0.057), though this was outside of statistical significance. A multivariate logistic
regression model designed to predict IPSS mislateralization using angiographic IPS
cross filling, basilar venous plexus filling, and the bilateral cross filling pattern
achieved an AUC of 0.78 (95% CI: 0.61–0.95, p = 0.015; [Fig. 3]).
Fig. 2
Fig. 3
Conclusion: Cases in which the IPS showed bilateral cross filling following contrast injection
from either side demonstrated an increased risk of mislateralization during IPSS,
while directional cross filling was associated with a lower risk. The presence of
basilar venous plexus filling may be an effect modifier of this association, as models
including this variable demonstrated a significant improvement in predicting mislateralization
by IPSS.
