ABBREVIATIONS:
- ACTH
- adrenocorticotropic hormone
- CD
- Cushing disease
- RFE
- recursive feature elimination
- ROC
- receiver operator characteristic curve.
Cushing disease (CD) is caused by pituitary adenomas that secrete adrenocorticotropic hormone (ACTH) inappropriately, leading to cortisol hypersecretion by the adrenal glands. Comprising 10% of clinically appreciated pituitary adenomas and 70% of all cases of hypercortisolemia, CD represents 0.7 to 2.4 cases per million people per year.1-3 CD is most often diagnosed in the third or fourth decade of life, and it is 8-fold more common in women. Symptoms of CD include obesity, diabetes, hypertension, osteoporotic fractures, psychiatric changes, and cognitive decline.4,5 Patients with CD suffer from up to a 5-fold increased mortality risk.2,4,6
Transsphenoidal adenomectomy, using a microscopic or endoscopic approach, remains the gold-standard treatment for CD. Whereas early postoperative remission rates range from 67% to 95%, series with extended follow-up periods report recurrence rates as high as 36% with a mean time to recurrence of 15 to 50 months.1
Accurate predictors for remission in CD remain inconsistent. Some studies suggest that large tumor size, invasion into adjacent structures, and aggressive tumor histology predict recurrence, whereas smaller tumor size and greater neurosurgical experience are associated with lower rates of recurrence.1,7 However, a meta-analysis suggested that age, sex, tumor size, and macroscopic invasion are not reliable predictors of recurrence.8
Postoperative biochemical markers (ie, cortisol and ACTH) have been shown to predict outcome, but routine postoperative steroid use may complicate interpretation.7,9-11 An intraoperative predictor for outcome in CD would be valuable for assessing resection of ACTH-secreting tissue intraoperatively and for use when steroids are replaced postoperatively. In this study, we sought to evaluate whether intraoperative measurement of peripheral blood ACTH levels could predict outcome after surgery for CD.
METHODS
A retrospective chart review was performed on all endoscopic endonasal surgeries for CD performed by 2 neurosurgeons (DP, 2010-2020, and DH, 2018-2020) at the Ohio State University Medical Center between July 2010 and May 2020. Cushing syndrome was screened for with 2 late-night salivary cortisol or 24-hour urine free cortisol or a low-dose dexamethasone suppression test.12 CD was verified by a normal or high-normal baseline ACTH, cortisol suppression after high-dose dexamethasone or inferior petrosal sinus sampling in equivocal cases, and the presence of a pituitary adenoma on MRI.12
Steroids were not administered preoperatively or intraoperatively. ACTH was measured preoperatively, thrice during surgery (during opening, during tumor resection, and during closure), and on postoperative day 1 (POD1). ACTH measurements were batched, processed using Siemens IMMULITE 2000 XPI assays, and resulted within 4 days. Procedure length (minutes) was measured from incision start to incision close. Cortisol was measured every 6 hours postoperatively until nadir. Steroid replacement was started if cortisol nadir dropped below 3 mcg/dL or patients became symptomatic. All patients were documented to have ACTH-immunostaining adenomas.
Biochemical remission was defined by either a 24-hour urinary cortisol of <45 μg/24 hours or 2 late-night salivary cortisol of <100 ng/dL at the most recent follow-up appointment.3,12 These values were in accordance with the standard reference ranges of the Ohio State University Wexner Medical Center laboratory. Biochemical remission was evaluated after patients were weaned off exogenous steroids.
Baseline characteristics are reported as means with standard deviations or percentages with counts. For categorical variables, significance was assessed using risk estimation and odds ratios to measure each variable's association with remission. For continuous variables, significance was assessed using 2-tailed paired samples for dependent groups and Student t tests for independent groups. The results of the paired samples t test were confirmed with nonparametric Wilcoxon signed-rank tests. Predictors of remission were identified through logistic regression, using remission as the dependent variable, and quality of fit evaluated through areas under the curve (AUC) of the receiver operator characteristic curve (ROC). The alpha was set at 0.05 to determine significance. Statistical analyses were performed using IBM SPSS Statistics.
A logistic regression machine learning model was created using both regression and nonregression algorithms to identify patterns in the data that may not be explicitly programmed.13 Our model incorporated final intraoperative ACTH, minimum intraoperative ACTH, early morning POD 1 cortisol, and postoperative cortisol nadir. Given the small study sample, only hormone variables that had no missing values were included for model creation. Training data were randomly divided into 5-folds. For each of the 5-folds, 1-fold was used as validation while the other 4-folds were used for training. AUCs were calculated for each validation fold, and the average of the 5 calculated AUCs was reported. A Mann-Whitney U test was used to calculate the P value for the 5 calculated AUCs compared with random AUC of 0.5. Recursive feature elimination was then used to generate models with fewer variables.
This study was approved by the Ohio State University Wexner Medical Center Institutional Review Board (protocol 2020H0221). This study was granted a waiver of consent given the retrospective nature of the study, minimal risk posed to patient (ie, risk of loss of confidentiality), and that the information used is normally accessed during clinical care. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline was implemented in this manuscript.
RESULTS
Patient Demographics and Tumor Characteristics
Intraoperative peripheral blood ACTH levels were measured during 55 of 58 eligible surgeries for biochemically confirmed CD. In 3 cases, the intraoperative ACTH levels were not available because of laboratory error and removed. The remaining 55 surgical encounters consisted of 49 patients, as identified by unique medical record numbers. Four patients had reoperations; 1 patient had 3 operations.
The mean age of this population was 44.6 (±14.1) years, and 80% were female (Table 1). CD was confirmed biochemically with a positive low-dose dexamethasone suppression (44.9%, n = 22), cortisol suppression after high-dose dexamethasone (42.9%, n = 21), or inferior petrosal sinus sampling (12.2%, n = 6). Of note, 25.5% of tumors were macroadenomas, 25.5% invaded the cavernous sinus, 10.9% exhibited suprasellar extension, and 21.8% had recurred after prior operations. The mean procedure length was 340.1 (±92.80) minutes, with a median of 325.5 minutes.
TABLE 1. -
Patient Demographics and Tumor Characteristics
| Characteristic |
Value |
| Patient demographics (n = 55) |
|
| Age at surgery in years (mean ± SD) |
44.6 (±14.1) |
| Sex (percent, number) |
|
| Male |
20.0% (11) |
| Female |
80.0% (44) |
| Tumor characteristics |
|
| Size (percent, number) |
|
| Microadenoma |
74.6% (41) |
| Macroadenoma |
25.5% (14) |
| Cavernous sinus invasion (percent, number) |
|
| Yes |
25.5% (14) |
| No |
74.6% (41) |
| Suprasellar extension (percent, number) |
|
| Yes |
10.9% (6) |
| No |
89.1% (49) |
| History of prior operation (percent, number) |
|
| Yes |
21.8% (12) |
| No |
78.2% (43) |
| Type of resection |
|
| Extracapsular |
34.5% (19) |
| Piecemeal |
65.5 (36) |
Descriptive statistics regarding the demographics and pituitary adenoma characteristics for our study sample including mean ± standard deviation, percent, and number.
Outcomes
The mean length of follow-up was 2.73 (±2.11 years), with a median follow-up of 2.07 years. Twelve patients had less than 1 year of follow-up on analysis. Remission was achieved in 69.1% (38/55) of operations. Of those achieving remission, 63.2% of patients (n = 24) were confirmed with late-night salivary cortisol. Excluding the 14 tumors with cavernous sinus invasion, the remission rate was 78.0% (32/41) with 68.5% of these patients (n = 22) confirmed with late-night salivary cortisol. In total, 82.4% (28/34) of cases without cavernous sinus invasion or a history of prior operation demonstrated remission.
Cavernous sinus invasion (odds ratio [OR]: 0.211; 95% CI: 0.058-0.767; P = .014), suprasellar extension (OR: 0.181; 95% CI: 0.029-1.105; P = .045), and history of prior operation (OR: 0.216; 95% CI: 0.056-0.833; P = .02) were all less likely to be associated with remission.
Intraoperative ACTH Levels Predict Outcome
Peripheral plasma ACTH levels were measured thrice intraoperatively: before, during, and after tumor dissection was complete. Mean plasma ACTH across the 3 time points were not significantly different (Figure 1A). Early morning preoperative and POD1 ACTH levels were also collected (Figure 1B). Intraoperative ACTH was higher than preoperative and postoperative ACTH levels (Table 2). Wilcoxon signed-rank tests confirmed these differences (Table 2).
;)
FIGURE 1.: Perioperative peripheral plasma ACTH levels. A, ACTH levels were drawn before dissection (labeled “First”), during dissection (labeled “Second”), and after dissection (labeled “Third”) intraoperatively (1 pg/mL = 0.2202 pmol/L). B, Mean early morning preoperative, mean intraoperative, and mean early morning postoperative day 1 ACTH levels (pg/mL; 1 pg/mL = 0.2202 pmol/L). Error bars represent standard error. ACTH, adrenocorticotropic hormone.
TABLE 2. -
Mean, Median Plasma Preoperative, Intraoperative, and Postoperative Adrenocorticotropic Hormone (ACTH) Levels (pg/mL)
| Perioperative time point |
n |
Mean (SD) |
Median |
95% CI |
| Preoperative |
48 |
69.25 (57.11) |
46.63 |
51.16-91.86 |
| Intraoperative |
55 |
158.0 (171.7) |
79.00 |
112.2-203.8 |
| Postoperative |
55 |
46.43 (106.3) |
15.80 |
18.07-74.79 |
| Comparative analysis |
P value |
95% CI |
| Preoperative—intraoperative |
.001 |
−153.6 to −38.62 |
| Intraoperative—postoperative |
<.001 |
56.09-170.8 |
| Preoperative—postoperative |
.154 |
−7.595 to 46.68 |
| Nonparametric Wilcoxon signed-rank tests |
P value |
Z |
| Preoperative—intraoperative |
<.001 |
1308.5 |
| Intraoperative—postoperative |
<.001 |
31.0 |
| Preoperative—postoperative |
<.001 |
172.5 |
Preoperative, intraoperative, and postoperative plasma ACTH (pg/mL; 1 pg/mL = 0.2202 pmol/L) descriptive statistics, comparative analysis, and Wilcoxon signed-rank test results. Seven patients did not have preoperative plasma ACTH laboratory results listed in their medical record. Bonferroni correction was applied for comparative analysis.
The final ACTH level measured intraoperatively predicted outcome (AUC = 0.766; 95% CI 0.626-0.907; P value = .002; Figure 2). The minimum intraoperative ACTH level (AUC = 0.747; 95% CI 0.603-0.891; P value = .004) and the maximum subtracted by minimum intraoperative ACTH level (AUC = 0.743; 95% CI 0.603-0.884; P value = .007) also predicted remission.
FIGURE 2.: ROC curves for A, final, B, minimum, and C, maximum minus minimum intraoperative ACTH levels. D, Area under the curve, 95% confidence interval, and P value (for the null hypothesis that AUC = 0.5) for the ROC curves. ACTH, adrenocorticotropic hormone; ROC, receiver operator characteristic.
A subset analysis of cases with 18 months or less of follow-up (n = 23) was conducted. A total of 78.3% were in remission, while 22.7% had recurred. The final intraoperative ACTH level predicted outcome (AUC = 0.978; 95% CI 0.924-1.000; P value <.001). The minimum intraoperative ACTH level (AUC = 0.983; 95% CI 0.939-1.000; P value < .001) and the maximum subtracted by minimum intraoperative ACTH level (AUC = 0.911; 95% CI 0.768-1.000; P value = .028) also predicted remission.
Final intraoperative ACTH levels in all 55 operations were rank ordered and categorized based on remission status (Figure 3A and Table 3). Notably, within the quartile of operations with the lowest final intraoperative ACTH, 92% (12/13) were found to be in remission. By contrast, in the quartile of operations with the highest intraoperative ACTH, 38% (5/13) were found to be in remission. The odds ratio of remission (Figure 3B) in patients with the lowest quartile final intraoperative ACTH compared with patients with the highest quartile final intraoperative ACTH was 23.4 (95% CI 2.325-235.536, P value = .002).
FIGURE 3.: Final intraoperative ACTH and long-term remission. A, Final intraoperative ACTH levels for all 55 patients rank-ordered and color-coded for remission (blue) vs nonremission (red). B, Odds ratio, confidence interval, and P value for the effect of cavernous invasion (no vs yes), suprasellar extension (no vs yes), history of prior operation (no vs yes), and final intraoperative ACTH (lowest quartile vs highest quartile) on long-term remission. ACTH, adrenocorticotropic hormone.
TABLE 3. -
Sensitivity vs Specificity of Final Intraoperative ACTH for Remission
| Remission if ACTH ≤ |
Sensitivity |
Specificity |
| In pg/mL |
In pmol/L |
| 4.6 |
1.0 |
0.0% |
100.0% |
| 5.8 |
1.3 |
2.6% |
100.0% |
| 7.5 |
1.7 |
2.6% |
94.1% |
| 10.3 |
2.3 |
5.3% |
94.1% |
| 13.8 |
3.0 |
7.9% |
94.1% |
| 18.3 |
4.0 |
10.5% |
94.1% |
| 20.9 |
4.6 |
13.2% |
94.1% |
| 21.8 |
4.8 |
15.8% |
94.1% |
| 23.2 |
5.1 |
18.4% |
94.1% |
| 24.4 |
5.4 |
21.1% |
94.1% |
| 25.5 |
5.6 |
23.7% |
94.1% |
| 26.8 |
5.9 |
26.3% |
94.1% |
| 30.5 |
6.7 |
28.9% |
94.1% |
| 34.5 |
7.6 |
31.6% |
94.1% |
| 36.6 |
8.1 |
34.2% |
94.1% |
| 43.3 |
9.5 |
36.8% |
94.1% |
| 48.9 |
10.8 |
39.5% |
94.1% |
| 49.6 |
10.9 |
42.1% |
94.1% |
| 52.5 |
11.5 |
44.7% |
94.1% |
| 57.0 |
12.5 |
47.4% |
94.1% |
| 62.8 |
13.8 |
47.4% |
88.2% |
| 67.4 |
14.8 |
47.4% |
82.4% |
| 68.9 |
15.2 |
50.0% |
82.4% |
| 69.3 |
15.2 |
52.6% |
82.4% |
| 69.8 |
15.4 |
55.3% |
82.4% |
| 70.8 |
15.6 |
57.9% |
82.4% |
| 71.9 |
15.8 |
60.5% |
82.4% |
| 72.8 |
16.0 |
60.5% |
76.5% |
| 73.6 |
16.2 |
63.2% |
76.5% |
| 76.3 |
16.8 |
65.8% |
76.5% |
| 81.7 |
18.0 |
65.8% |
70.6% |
| 87.2 |
19.2 |
68.4% |
70.6% |
| 96.0 |
21.1 |
71.1% |
70.6% |
| 109.0 |
24.0 |
73.7% |
70.6% |
| 116.5 |
25.7 |
76.3% |
70.6% |
| 124.5 |
27.4 |
76.3% |
64.7% |
| 148.5 |
32.7 |
78.9% |
64.7% |
| 168.5 |
37.1 |
81.6% |
64.7% |
| 183.5 |
40.4 |
81.6% |
58.8% |
| 196.0 |
43.2 |
84.2% |
58.8% |
| 218.0 |
48.0 |
86.8% |
58.8% |
| 240.0 |
52.8 |
86.8% |
52.9% |
| 245.5 |
54.1 |
86.8% |
47.1% |
| 254.5 |
56.0 |
86.8% |
41.2% |
| 262.0 |
57.7 |
86.8% |
35.3% |
| 290.5 |
64.0 |
86.8% |
29.4% |
| 331.5 |
73.0 |
89.5% |
29.4% |
| 351.0 |
77.3 |
92.1% |
29.4% |
| 401.5 |
88.4 |
94.7% |
29.4% |
| 470.5 |
103.6 |
94.7% |
23.5% |
| 589.0 |
129.7 |
97.4% |
23.5% |
| 707.5 |
155.8 |
97.4% |
17.6% |
| 789.0 |
173.7 |
97.4% |
11.8% |
| 860.0 |
189.4 |
97.4% |
5.9% |
| 1061.0 |
233.6 |
100.0% |
5.9% |
| 1251.0 |
275.5 |
100.0% |
0.0% |
ACTH, adrenocorticotropic hormone.
Sensitivity vs Specificity table of final intraoperative (after tumor dissection) plasma ACTH (pg/mL; 1 pg/mL = 0.2202 pmol/L) for remission (see “Methods” section for definition of biochemical remission).
Machine Learning Techniques Predict Outcome
Recursive feature elimination identified the variables most important for discriminating outcome, and a model was generated based on POD1 early morning cortisol and final intraoperative ACTH (Figure 4A). This model had an average AUC of 0.80 and a P value of .0027 (Figure 4B and 4C). Isolated models with only POD1 cortisol and only final intraoperative ACTH had an AUC = 0.72 (P value = .07) and AUC = 0.78 (P value = .003), respectively. Coefficients for the logistic regression are presented (Figure 4D).
FIGURE 4.: Machine learning analysis of ACTH and cortisol values. A, Schematic describing RFE. B, AUC for prediction of long-term outcome as variables are eliminated using RFE. C, P value for prediction of long-term outcome as variables are eliminated using RFE. D, Coefficients for logistic regression machine learning model using 2 variables: final intraoperative ACTH and POD1 cortisol. ACTH, adrenocorticotropic hormone; AUC, area under the curve; POD1, postoperative day 1; RFE, recursive feature elimination.
DISCUSSION
Functional adenomas, particularly CD, are uniquely challenging for pituitary surgeons. They may be difficult to identify on preoperative MRI, and complete resection is important for achieving biochemical remission. Postoperative serum cortisol and ACTH may indicate a need for reoperation because of lack of biochemical control.11,14,15 Studies have shown that lower ACTH and cortisol levels in the immediate postoperative period are predictive of remission.16-20 In addition, shorter time to postoperative ACTH nadir predicted remission.17 Reoperation has been recommended on patients with a postoperative cortisol nadir of >10 mcg/dL.19
Key Results
In this series of 55 operations for CD, intraoperative ACTH levels were measured and outcome was followed for a mean of 2.73 years and a median of 2.07 years. In agreement with previous studies, cavernous sinus invasion, suprasellar extension, and history of prior operation were associated with a lower chance of remission.
We show, for the first time, that intraoperative ACTH levels can predict outcome. We hypothesized that, given the short half-life between 5 and 23 minutes of ACTH in the circulation,21,22 complete resection of ACTH-secreting tumor tissue could be associated with a drop in serum ACTH intraoperatively. Indeed, the final intraoperative ACTH predicted outcome. Notably, being in the lowest or highest quartile of the final intraoperative ACTH was highly predictive, with an odds ratio of 23.4. The reduction in ACTH seen from the maximum value measured to the minimum value measured intraoperatively also predicted outcome, albeit with a lower discriminative ability. Machine learning techniques validated an algorithm incorporating final intraoperative ACTH and POD1 cortisol that is highly predictive of outcome. Removing final intraoperative ACTH from the model reduced its discriminative ability, validating the importance of intraoperative ACTH in predicting outcome.
Interpretation
These results warrant the consideration of both postoperative cortisol and intraoperative ACTH as predictors of remission. Previous studies have found postoperative cortisol nadir predicted remission in CD at a greater sensitivity than this study.23,24 However, long-term follow-up identified nearly one-third of patients who were in remission at 1 year had later recurred.1,24 Demonstration of hypocortisolemia remains the standard for remission, and degree of hypocortisolism remains valuable in predicting recurrence.1 Our study does not refute this, but the results suggest that postoperative cortisol and intraoperative ACTH have better predictive capacity together than either do alone.
Previous studies have examined intraoperative serum ACTH during surgery for CD. One study focusing on short-term remission concluded that intraoperative ACTH does not predict remission status.25 However, our results suggest that intraoperative ACTH can predict remission and that its discriminative ability increased in a short-term subset analysis. Manipulation of the pituitary gland can induce an increase in serum ACTH for 2 to 5 hours after resection of ACTH-secreting tumor.26,27 These findings suggest that, depending on surgical technique used, the induced short-term release of ACTH could mask the drop in serum ACTH associated with complete tumor resection. We used extracapsular resection technique28 whenever possible, minimizing disruption of the tumor and pseudocapsule and lessening the ACTH spike seen, which may explain the association of intraoperative ACTH and remission presented in this study. Analyses calculating the difference between maximum and minimum intraoperative ACTH as a proxy for ACTH spike were found to be significant but less accurate in its predictive capacity than minimum or final intraoperative ACTH values alone (Figure 2). Intraoperative ACTH predicts complete resection after surgery for ectopic ACTH-secreting tumors.29 Our work agrees with this concept and suggests that these findings may be extended to CD.
Generalizability
An intraoperative test assessing whether the tumor has been completely resected in real time would be valuable to surgeons. While it would be technically possible to obtain rapid results of ACTH testing, this is not available at most institutions. At our institution, serum ACTH samples are batched and processed 1 day per week. This study could serve as a proof of concept to encourage the adoption of rapid ACTH testing at pituitary surgery centers. For instance, if there is a significant drop in ACTH after resection, a surgeon may choose to refrain from further resection to prevent complications. This may circumvent the need for intraoperative MRI to assess resection cavity—an effect that could reduce time spent under anesthesia and operative costs. On the contrary, if no drop in ACTH is realized, the surgeon could continue resection, preventing the need for a reoperation. While the results of our study do not yet justify sweeping changes, it warrants further investigation in larger, prospective multicenter trials. Postoperative ACTH and cortisol remain the gold standard after surgery for CD and should continue to be analyzed when considering reoperation.
Limitations
Several limitations should be considered when evaluating the results presented here. First, as mentioned, many centers do not currently perform rapid ACTH testing. This study is limited by variable follow-up time given that patients with CD may experience recurrence years after remission.1,24 Studies have identified potential erroneous results because of analytical interference in Siemens IMMULITE 2000 XPI assays.30,31 Although concerning for patients with subclinical hypercortisolism whose treatment relies solely on laboratory values, patients with CD rely on a constellation of imaging, clinical, and laboratory findings. Interference was not identified in control patients, including 1 with CD.30 Our study is limited by using multiple biochemical tests for remission. One study suggests midnight salivary cortisol is superior, but debate remains around which should be the standard for determining remission.1,3,32
The machine learning approach presented has been applied only to this sample and needs to be validated using data sets collected at additional centers. Even in the author's data set, not every patient with a low final intraoperative ACTH benefited from remission. Finally, analyses were conducted under the assumption that all 55 cases are independent of each other; there may be interpatient correlation given inclusion of multiple operations of the same patient.
CONCLUSION
We present a series of 55 operations for Cushing disease in which peripheral blood ACTH levels were measured intraoperatively. Cavernous sinus invasion, suprasellar extension, and history of a prior operation were associated with nonremission. Intraoperative ACTH, especially levels of final intraoperative ACTH in the lowest or highest quartile, predicted outcome. Machine learning techniques identified an algorithm that combined intraoperative ACTH and POD1 cortisol that was highly discriminative for outcome. These data suggest that further investigation of intraoperative testing to predict outcome in Cushing disease is warranted.
Funding
This study did not receive any funding or financial support.
Disclosures
The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Dr Prevedello is a consultant for Stryker Corp., Medtronic Corp., Integra, BK Medical, and Vycor. He has received royalties from KLS-Martin, Mizuho, and ACE Medical. Dr Carrau receives royalties for instruments (through a university contract) from KLS. Dr Zachariah receives research support from the Mississippi Center for Translational Research. He owns stock in Exelixis and Zinnia Health, which he has also consulted for. He has also participated in separate research projects supported by Hitachi Healthcare and BK Ultrasound that are unrelated to this project. Dr Kirschner serves/has served on advisory boards for the following companies: Corcept Therapeutics and Recordati Rare Disease.
REFERENCES
1. Pendharkar AV, Sussman ES, Ho AL, Hayden Gephart MG, Katznelson L. Cushing's disease: predicting long-term remission after surgical treatment. Neurosurg Focus. 2015;38(2):E13.
2. Lindholm J, Juul S, Jørgensen JO, et al. Incidence and late prognosis of cushing's syndrome: a population-based study. J Clin Endocrinol Metab. 2001;86(1):117-123.
3. Lonser RR, Nieman L, Oldfield EH. Cushing's disease: pathobiology, diagnosis, and management. J Neurosurg. 2017;126(2):404-417.
4. Feelders RA, Pulgar SJ, Kempel A, Pereira AM. The burden of Cushing's disease: clinical and health-related quality of life aspects. Eur J Endocrinol. 2012;167(3):311-326.
5. Sharma ST, Nieman LK, Feelders RA. Comorbidities in Cushing's disease. Pituitary. 2015;18(2):188-194.
6. Clayton RN, Jones PW, Reulen RC, et al. Mortality in patients with Cushing's disease more than 10 years after remission: a multicentre, multinational, retrospective cohort study. Lancet Diabetes Endocrinol. 2016;4(7):569-576.
7. Lonser RR, Wind JJ, Nieman LK, Weil RJ, DeVroom HL, Oldfield EH. Outcome of surgical treatment of 200 children with Cushing's disease. J Clin Endocrinol Metab. 2013;98(3):892-901.
8. Roelfsema F, Biermasz NR, Pereira AM. Clinical factors involved in the recurrence of pituitary adenomas after surgical remission: a structured review and meta-analysis. Pituitary. 2012;15(1):71-83.
9. Patil CG, Prevedello DM, Lad SP, et al. Late recurrences of Cushing's disease after initial successful transsphenoidal surgery. J Clin Endocrinol Metab. 2008;93(2):358-362.
10. Starke RM, Reames DL, Chen CJ, Laws ER, Jane JA. Endoscopic transsphenoidal surgery for cushing disease: techniques, outcomes, and predictors of remission. Neurosurgery. 2013;72(2):240-247; discussion 247.
11. Abdelmannan D, Chaiban J, Selman WR, Arafah BM. Recurrences of ACTH-secreting adenomas after pituitary adenomectomy can be accurately predicted by perioperative measurements of plasma ACTH levels. J Clin Endocrinol Metab. 2013;98(4):1458-1465.
12. Nieman LK, Biller BMK, Findling JW, et al. The diagnosis of Cushing's syndrome: an Endocrine Society Clinical Practice guideline. J Clin Endocrinol Metab. 2008;93(5):1526-1540.
13. Muhlestein WE, Akagi DS, Davies JM, Chambless LB. Predicting Inpatient length of stay after brain tumor surgery: developing machine learning ensembles to improve predictive performance. Neurosurgery. 2019;85(3):384-393.
14. Biller BMK, Grossman AB, Stewart PM, et al. Treatment of adrenocorticotropin-dependent Cushing's syndrome: a consensus statement. J Clin Endocrinol Metab. 2008;93(7):2454-2462.
15. Lindsay JR, Oldfield EH, Stratakis CA, Nieman LK. The postoperative basal cortisol and CRH tests for prediction of long-term remission from Cushing's disease after transsphenoidal surgery. J Clin Endocrinol Metab. 2011;96(7):2057-2064.
16. Flitsch J, Knappe UJ, Lüdecke DK. The use of postoperative ACTH levels as a marker for successful transsphenoidal microsurgery in Cushing's disease. Zentralbl Neurochir. 2003;64(1):6-11.
17. Abellán-Galiana P, Fajardo-Montañana C, Riesgo-Suárez P, Pérez-Bermejo M, Ríos-Pérez C, Gómez-Vela J. Prognostic usefulness of ACTH in the postoperative period of Cushing's disease. Endocr Connect. 2019;8(9):1262-1272.
18. Asuzu D, Chatain GP, Hayes C, et al. Normalized Early postoperative cortisol and ACTH Values predict nonremission after surgery for Cushing disease. J Clin Endocrinol Metab. 2017;102(7):2179-2187.
19. Hameed N, Yedinak CG, Brzana J, et al. Remission rate after transsphenoidal surgery in patients with pathologically confirmed Cushing's disease, the role of cortisol, ACTH assessment and immediate reoperation: a large single center experience. Pituitary. 2013;16(4):452-458.
20. Srinivasan L, Laws ER, Dodd RL, et al. The dynamics of post-operative plasma ACTH values after transsphenoidal surgery for Cushing's disease. Pituitary. 2011;14(4):312-317.
21. López FJ, Negro-Vilar A. Estimation of endogenous adrenocorticotropin half-life using pulsatility patterns: a physiological approach to the evaluation of secretory episodes. Endocrinology. 1988;123(2):740-746.
22. Veldhuis JD, Iranmanesh A, Naftolowitz D, Tatham N, Cassidy F, Carroll BJ. Corticotropin secretory dynamics in humans under low glucocorticoid feedback. J Clin Endocrinol Metab. 2001;86(11):5554-5563.
23. Ironside N, Chatain G, Asuzu D, et al. Earlier post-operative hypocortisolemia may predict durable remission from Cushing's disease. Eur J Endocrinol. 2018;178(3):255-263.
24. Alexandraki KI, Kaltsas GA, Isidori AM, et al. Long-term remission and recurrence rates in Cushing's disease: predictive factors in a single-centre study. Eur J Endocrinol. 2013;168(4):639-648.
25. Graham KE, Samuels MH, Raff H, Barnwell SL, Cook DM. Intraoperative adrenocorticotropin levels during transsphenoidal surgery for Cushing's disease do not predict cure. J Clin Endocrinol Metab. 1997;82(6):1776-1779.
26. Pimentel-Filho FR, Cukiert A, Miyashita F, et al. Adrenocorticotropin levels do not change during early recovery of transsphenoidal surgery for ACTH-secreting pituitary tumors. J Endocrinol Invest. 2001;24(2):83-87.
27. Czirják S, Bezzegh A, Gál A, Rácz K. Intra- and postoperative plasma ACTH concentrations in patients with Cushing's disease cured by transsphenoidal pituitary surgery. Acta Neurochir (Wien). 2002;144(10):971-977; discussion 977.
28. Oldfield EH, Vortmeyer AO. Development of a histological pseudocapsule and its use as a surgical capsule in the excision of pituitary tumors. J Neurosurg. 2006;104(1):7-19.
29. Raff H, Shaker JL, Seifert PE, Werner PH, Hazelrigg SR, Findling JW. Intraoperative measurement of adrenocorticotropin (ACTH) during removal of ACTH-secreting bronchial carcinoid tumors. J Clin Endocrinol Metab. 1995;80(3):1036-1039.
30. Yener S, Demir L, Demirpence M, et al. Interference in ACTH immunoassay negatively impacts the management of subclinical hypercortisolism. Endocrine. 2017;56(2):308-316.
31. Korchia J, Freeman KP. Validation study of canine serum cortisol measurement with the Immulite 2000 Xpi cortisol immunoassay. J Vet Diagn Invest. 2021;33(5):844-863.
32. Carrasco CA, Coste J, Guignat L, et al. Midnight salivary cortisol determination for assessing the outcome of transsphenoidal surgery in Cushing’s disease. J Clin Endocrinol Metab. 2008;93(12):4728-4734.
COMMENT
This study presents 55 CD operations and evaluates intraoperative ACTH levels retrospectively to see if they are predictive of remission. This study shows progress in possible techniques for identifying patients for further exploration or early reoperation, but the direct application of the results is unclear. Concerns about the study which bear addressing include:
The diagnosis of Cushing's disease should include at least 1 test confiming a central source (such as low vs high dose dex suppression or IPSS). The authors have changed their listed testing to include ACTH levels after high dose dex testing, but this is not the appropriate lab and raises concern about the adequacy of diagnosis.
The lack of rapidity of ACTH results is a potentially fatal flaw in this study. Most labs are unable to produce these results real time. These typically take several hours and may take up to a day depending on the lab. Obviously, the ability to utilize this would greatly increase its value. If results were not available immediately intraoperatively, then the accuracy of these results should be compared with simple regular postoperative cortisol levels which have been shown to correlate with remission. Indeed, exactly how this technique or data would be applied in a patient is unclear since the ACTH levels are not significantly different from each other but only predictive based on a machine learning protocol or AUC compared to other cases.
Paul A. Gardner
Pittsburgh, Pennsylvania, USA
