Abstract
Introduction:
Adrenocortical carcinoma (ACC) is a notoriously aggressive cancer with a dismal prognosis, especially for patients with metastatic disease. Metastatic ACC is classically a contraindication to operative management. Here, we evaluate the impact of primary tumor resection and metastasectomy on survival in metastatic ACC.
Methods:
We performed a retrospective cohort study of patients with metastatic ACC (2010–2019) utilizing the National Cancer Database. The primary outcome was overall survival (OS). Cox proportional hazards models were developed to evaluate the associations between surgical management and survival. Propensity score matching (PSM) was utilized to account for selection bias in receipt of surgery.
Results:
Of 976 subjects with metastatic ACC, 38% underwent surgical management. Median OS across all patients was 7.6 months. On multivariable Cox proportional hazards regression, primary tumor resection alone (HR: 0.523; p<0.001) and primary resection with metastasectomy (HR: 0.372; p<0.001) were significantly associated with improved OS. Metastasectomy alone had no association with OS (HR: 0.909; p=0.740). Primary resection with metastasectomy was associated with improved OS over resection of the primary tumor alone (HR: 0.636; p=0.018). After PSM, resection of the primary tumor alone remained associated with improved OS (HR 0.593; p<0.001), and metastasectomy alone had no survival benefit (HR 0.709; p=0.196) compared with non-operative management; combined resection was associated with improved OS over primary tumor resection alone (HR 0.575, p=0.008).
Conclusion:
In metastatic ACC, patients may benefit from primary tumor resection alone or in combination with metastasectomy, however further research is required to facilitate appropriate patient selection.
Keywords: adrenocortical carcinoma, ACC, metastasectomy, adrenalectomy
Introduction
Adrenocortical carcinoma (ACC) is a notoriously aggressive cancer. Patients often present with large, high-grade tumors and distant metastases at diagnosis. While disease-specific five-year survival in early-stage ACC may be as high as 60%, patients with metastatic disease have exceptionally poor outcomes, with a five-year survival as low as 10.2% [1–11]. With an estimated annual incidence of two or fewer cases per million, however, ACC remains quite rare [1, 2, 4–8].
While surgical resection is the mainstay of therapy in non-metastatic ACC, management of metastatic ACC relies primarily on systemic chemotherapy. Mitotane is the only drug approved for the treatment of metastatic ACC [12] and is often combined with systemic chemotherapy (etoposide, doxorubicin, and cisplatin (EDP)) following the FIRM-ACT trial [13, 14]. However, response rates to EDP-mitotane remain relatively poor [13], and no other broadly effective therapeutic options have been identified [15]. Preliminary research on immunotherapeutic agents and immune checkpoint inhibitors have demonstrated modest efficacy in specific subsets of patients, but overall results have been disappointing [12, 14, 15].
Classically, metastatic disease is a relative contraindication to surgical resection in ACC. Extensive study in other metastatic malignancies, however, has demonstrated benefit from both primary tumor resection and metastasectomy [16–18]. In ACC, retrospective studies of the California Cancer Registry (CCR) and the Surveillance, Epidemiology, and End Results (SEER) database have likewise demonstrated improved survival with primary tumor resection in metastatic ACC [19, 20]. Pulmonary and hepatic metastasectomy have also demonstrated benefit in both recurrence-free and overall survival [21–24]. However, the evidence in support of these interventions remains limited, and the relative impact of primary tumor resection versus metastasectomy is poorly defined.
Therefore, in this study, we aim to characterize the association between survival and surgical resection of 1) the primary tumor; and 2) metastatic disease in a large national cohort with metastatic ACC.
Methods
Data Source
We utilized the National Cancer Database (NCDB) to perform a retrospective cohort study of all patients presenting with metastatic ACC. The NCDB is a nationwide, facility-based dataset of cancer diagnoses from over 1,500 accredited programs, capturing 72% of incident malignancies in the United States per year [25, 26]. This study was deemed exempt from review by the Institutional Review Board of the University of Pennsylvania.
Study Cohort
All adult patients diagnosed with ACC from 2010–2019 were identified for inclusion using the ICD-O-3 histology codes for ACC (8370 and 8373) and ICD-10 diagnosis codes for malignant neoplasm of the adrenal cortex (C74.0). Of 4,788 patients with ACC, 1,198 (25%) had documented metastatic disease and were included. Patients who did not receive any therapy (n=161), who underwent debulking surgery (n=15), or who had missing surgical data (n=46) were excluded, leaving a final cohort of 976 patients (Figure 1).
Figure 1.
CONSORT flow diagram demonstrating final composition of the included cohort.
Definitions and Outcomes
Patients were categorized by surgical intervention received: 1) primary tumor resection, 2) metastasectomy, 3) primary tumor resection with metastasectomy, or 4) no surgical intervention. The primary outcome was overall survival (OS). Metastatic disease was determined by the presence of Stage IV disease (in patients staged according to the American Joint Committee on Cancer (AJCC), 8th edition) or by documentation of a metastatic site (in patients staged according to the AJCC, 7th edition). Pathologic stage of disease was used if available; otherwise, clinical stage was used. “Number of metastatic sites” was defined using NCDB indicators, available only for metastases specifically to the liver, lung, brain, and/or bone. While the NCDB categorizes location of surgery as primary site or distant site (metastasectomy), data on the anatomic location of metastasectomy was unavailable. The NCDB does not capture data regarding the temporal relationship of surgeries a subject received and so it was not possible to determine if metastasectomy was performed for synchronous or metachronous disease. It was not feasible to identify and exclude patients undergoing palliative surgery, as it is not possible to determine the site, type, or date of such procedure within NCDB. Patients treated with mitotane were classified as receiving chemotherapy in the NCDB.
Statistical Analysis
Demographic characteristics, tumor size, number of metastatic sites, and treatment modalities were examined, and descriptive statistics were calculated. Data were reported as means with standard deviations (SD) or medians with interquartile ranges (IQR) for continuous variables and frequencies with percentages for categorical variables. Group comparisons were performed using χ2 tests, Student’s t-tests, and Wilcoxon Rank-Sum tests, as appropriate. Multivariable logistic regression was utilized to determine associations of demographic and clinical characteristics with receipt of surgical intervention.
Kaplan-Meier survival analysis was performed based on clinical and treatment characteristics. Median survival and one-, three-, and five-year survival rates were calculated. Mantel-Haenszel tests were used to compare differences in survival. Univariable and multivariable Cox proportional-hazards models were created to determine associations between demographic and clinical characteristics, treatment receipt, and OS. Variables with significance below p=0.020 on univariable analysis as well as important demographic factors such as sex and race were incorporated into the multivariable model.
Propensity Score Matching
To account for the selection bias regarding which patients underwent surgical management in this cohort, propensity score matching (PSM) was used. Three matched groups were created. Patients who did not receive any surgical intervention were matched to those who received: 1) primary resection only, and 2) metastasectomy. Then, patients who received primary resection only were matched with those who received primary resection with metastasectomy. For each sub-cohort, a 1:1 match using the optimal algorithm without replacement and Mahalanobis distance was used. Age, sex, tumor size, Charlson-Deyo comorbidity index, and number of metastatic sites were selected as covariates.
Statistical analysis was performed utilizing Stata, version 17.0 [27] and R, version 4.2.2 using the ‘MatchIt’ package [28, 29].
Results
Cohort Characteristics
In total, 976 patients met inclusion criteria. The mean age was 55.2 years (SD: 15.4). Most patients were female (60%) and White race (84%). The median tumor size at presentation was 11.0 centimeters (IQR: 7.7–15.2). The mean number of metastatic sites was 1.4 (SD: 0.75). Of all patients, 59% had metastases to the lung, 57% to the liver, 20% to the bone, and 3.6% to the brain (Table 1).
Table 1.
Demographic and clinical characteristics of patients with metastatic ACC, stratified by surgical management. Bold indicates p<0.05.
| Total Cohort (n = 976) |
Operative Management (n = 379) |
Non-Operative Management (n = 597) |
p-value | |
|---|---|---|---|---|
| Mean Age, Years (SD) | 55.2 (15.4) | 53.9 (15.5) | 56.0 (15.3) | 0.035 |
| Sex, n (%) | ||||
| Male | 394 (40.4) | 142 (37.5) | 252 (42.2) | 0.141 |
| Female | 582 (59.6) | 237 (62.5) | 345 (57.8) | |
| Race, n (%) | ||||
| White | 820 (84.0) | 322 (85.0) | 498 (83.4) | 0.475 |
| Black | 96 (9.8) | 32 (8.4) | 64 (10.7) | |
| Other | 60 (6.2) | 25 (6.6) | 35 (5.8) | |
| Ethnicity, n (%) | ||||
| Hispanic | 111 (11.4) | 33 (8.7) | 78 (13.1) | 0.037 |
| Non-Hispanic | 865 (88.6) | 346 (91.3) | 519 (86.9) | |
| Insurance status, n (%) | ||||
| Private | 491 (51.4) | 202 (53.3) | 289 (48.4) | 0.421 |
| Public | 419 (43.8) | 155 (40.9) | 264 (44.2) | |
| Uninsured | 46 (4.8) | 29 (7.7) | 29 (4.9) | |
| Charlson-Deyo Score, n (%) | ||||
| 0 | 699 (71.6) | 272 (71.8) | 427 (71.5) | 0.8416 |
| 1 | 193 (19.8) | 76 (20.1) | 117 (19.6) | |
| 2 | 52 (5.3) | 24 (6.3) | 28 (4.7) | |
| ≥3 | 32 (3.3) | 7 (1.8) | 25 (4.2) | |
| Laterality, n (%) | ||||
| Left | 500 (51.2) | 199 (52.5) | 301 (50.4) | 0.001 |
| Right | 401 (41.1) | 167 (44.1) | 234 (39.2) | |
| Bilateral | 12 (1.2) | 2 (0.5) | 10 (1.7) | |
| Unspecified | 63 (6.5) | 11 (2.9) | 52 (8.7) | |
| Median Tumor Size, cm (IQR) | 11.0 (7.7–15.2) | 10.9 (7.5–14.6) | 12 (8–17) | 0.004 |
| Mean Number of Metastatic Sites, (SD) | 1.37 (0.75) | 1.15 (0.67) | 1.52 (0.76) | <0.001 |
| Location of Metastatic Sites | ||||
| Liver, n (%) | 550 (56.3) | 166 (43.8) | 384 (64.3) | <0.001 |
| Lung, n (%) | 568 (58.2) | 200 (52.8) | 368 (61.6) | 0.003 |
| Brain, n (%) | 35 (3.6) | 5 (1.3) | 30 (5.0) | 0.002 |
| Bone, n (%) | 189 (19.4) | 61 (16.1) | 128 (21.4) | 0.037 |
| Medical Therapy, n (%) | ||||
| Radiation | 606 (62.1) | 209 (55.1) | 397 (66.5) | <0.001 |
| Chemotherapy | 677 (69.4) | 237 (62.5) | 440 (73.7) | <0.001 |
Therapeutic Approach
In this cohort, 39% of patients (n=379) underwent operative management. Among patients who underwent surgery, 75% underwent primary resection only, 8.2% underwent metastasectomy only, and 17% underwent both primary resection and metastasectomy. Female sex (OR: 1.476; 95% CI 1.089–1.999; p=0.012) was associated with higher odds of primary tumor resection. Older age (OR: 0.988; 95% CI 0.977–0.998; p=0.022) and a greater number of metastatic sites (OR: 0.426; 95% CI 0.340–0.533; p<0.001) were associated with lower likelihood of undergoing primary tumor resection. Patients with larger primary tumors trended toward a higher likelihood of surgery (OR: 1.016; 95% CI 1.000–0.1032; p=0.050). Most patients underwent radiation therapy (62%) or systemic chemotherapy (69%); few were treated with immunotherapy (4%) or hormonal therapy (4%).
Unadjusted Survival
The median follow-up time was 7.2 months (IQR: 3.1–20.7). The median OS was 7.6 months (95% CI 6.8–8.7). One-year survival was 40.2% (95% CI 36.8–43.6%), three-year survival was 17.3% (95% CI 14.7%–20.1%), and five-year survival was 10.3% (95% CI 7.9%–13.0%). The median survival was higher in patients who underwent any surgery (16 vs. 5.2 months, p<0.001) compared with those managed non-operatively.
Survival was longest in patients who underwent primary tumor resection with metastasectomy (26.3 months; 95% CI 13.4–39.6), followed by patients who underwent primary resection only (15.3 months; 95% CI 10.9–18.2) and metastasectomy only (6.6 months; 95% CI 3.5–22.0). Survival was shortest in patients managed non-operatively (5.2 months; 95% CI 4.6–6.0) (Figure 2). All surgical interventions had significantly improved survival over non-operative management (primary resection with metastasectomy: p<0.0001; primary resection only: p<0.0001, metastasectomy only: p=0.0471). There were no significant differences in survival between surgical interventions on pairwise comparisons.
Figure 2.
Kaplan-Meier survival curves demonstrating unadjusted survival for patients by therapeutic approach. All operative interventions were associated with significantly improved survival over non-operative management (primary tumor resection with metastasectomy: p<0.0001; primary tumor resection alone: p<0.0001, metastasectomy alone: p=0.0471); however, there was no significant difference between any of the operative interventions on pairwise comparisons.
Cox Proportional Hazards Modeling
On univariable analysis, several demographic and clinical characteristics were associated with survival, as shown in Table 2. Resection of the primary tumor alone (HR: 0.549; 95% CI 0.464–0.650; p<0.001), metastasectomy alone (HR: 0.591; 95% CI 0.377–0.926; p=0.022), resection of the primary with metastasectomy (HR: 0.405; 95% CI 0.292–0.563; p<0.001), and chemotherapy (HR: 0.787; 95% CI 0.668–0.927; p=0.004) were associated with improved survival.
Table 2.
Cox proportional hazards models evaluating the associations between clinical and treatment characteristics and survival. Bold indicates p<0.05.
| Univariable | Multivariable | |||||
|---|---|---|---|---|---|---|
| HR | 95% CI | p-value | HR | 95% CI | p-value | |
| Age | 1.008 | 1.003–1.013 | 0.002 | 1.003 | 0.997–1.010 | 0.314 |
| Sex | ||||||
| Male | Ref | -- | -- | -- | -- | -- |
| Female | 0.965 | 0.829–1.123 | 0.644 | 1.104 | 0.933–1.306 | 0.248 |
| Race | ||||||
| White | Ref | -- | -- | -- | -- | -- |
| Black | 0.891 | 0.695–1.142 | 0.360 | 0.718 | 0.542–0.951 | 0.021 |
| Other | 0.935 | 0.681–1.283 | 0.675 | 0.859 | 0.604–1.221 | 0.397 |
| Insurance | ||||||
| Uninsured | Ref | -- | -- | -- | -- | -- |
| Private | 0.741 | 0.534–1.028 | 0.072 | 0.586 | 0.406–0.848 | 0.005 |
| Public | 0.908 | 0.653–1.264 | 0.569 | 0.705 | 0.479–1.037 | 0.076 |
| Charlson-Deyo (per point) | 1.220 | 1.110–1.355 | <0.001 | 1.218 | 1.082–1.372 | 0.001 |
| Tumor Size (per cm) | 0.994 | 0.987–1.003 | 0.161 | 0.999 | 0.991–1.006 | 0.745 |
| Number of Metastatic Sites | 1.357 | 1.231–1.496 | <0.001 | 1.301 | 1.167–1.450 | <0.001 |
| Operative Management | ||||||
| Non-Operative | Ref | -- | -- | -- | -- | -- |
| Resection of Primary | 0.549 | 0.464–0.650 | <0.001 | 0.523 | 0.430–0.635 | <0.001 |
| Metastasectomy | 0.591 | 0.377–0.926 | 0.022 | 0.909 | 0.518–1.595 | 0.740 |
| Primary + Metastasectomy | 0.405 | 0.292–0.563 | <0.001 | 0.372 | 0.260–0.532 | <0.001 |
| Medical Therapy | ||||||
| Chemotherapy | 0.787 | 0.668–0.927 | 0.004 | 0.645 | 0.532–0.782 | <0.001 |
| Radiation Therapy | 0.873 | 0.751–1.012 | 0.072 | 0.716 | 0.605–0.846 | <0.001 |
On multivariable modeling, Black race (HR: 0.718; 95% CI 0.542–0.951; p=0.021), private insurance (HR: 0.586; 95% CI 0.406–0.848; p=0.005), resection of the primary tumor (HR: 0.523; 95% CI 0.430–0.635; p<0.001) and resection of the primary with metastasectomy (HR: 0.372; 95% CI 0.260–0.532; p<0.001) were significantly associated with improved OS. Chemotherapy (HR: 0.645; 95% CI 0.532–0.782; p<0.001) and radiotherapy (HR: 0.716; 95% CI 0.605–0.846; p<0.001) were also associated with improved OS. In contrast, Charlson-Deyo score (HR: 1.2182; 95% CI 1.082–1.372; p=0.001) and greater number of metastatic sites (HR: 1.301; 95% CI 1.167–1.450; p<0.001) were associated with worse survival. Metastasectomy alone had no association with survival (HR: 0.909; 95% CI 0.518–1.595; p=0.740) (Table 2). When this model was restricted to only patients receiving primary tumor resection, the addition of metastasectomy was associated with improved OS over primary tumor resection alone (HR: 0.636; 95% CI 0.437–0.926; p=0.018).
Propensity Score Matching
PSM was performed to develop three matched cohorts: 1) subjects who underwent primary tumor resection alone matched with subjects who did not receive surgery, 2) subjects who underwent metastasectomy alone matched with subjects who did not receive surgery, and 3) subjects who underwent primary tumor resection alone with subjects who underwent primary tumor resection with metastasectomy. All matched cohorts were well-balanced with respect to age, sex, race, ethnicity, insurance status, comorbidities, tumor size and number of metastatic sites (Table 3).
Table 3.
Covariate balance in matched cohorts. Bold indicates p<0.05.
| Non-Operative vs. Primary Resection Match | Non-Operative vs. Metastasectomy Match | Surgery with vs. without Metastasectomy Match | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Non-Operative Management (n = 278) |
Primary Resection Only (n = 278) |
p-value | Non-Operative Management (n = 30) |
Metastasectomy Only (n = 30) |
p-value | Primary Resection Only (n = 63) |
Primary Resection with Metastasectomy (n = 63) |
p-value | |
| Mean Age, Years (SD) | 54.1 (14.7) | 53.8 (15.8) | 0.826 | 56.0 (13.1) | 56.4 (13.9) | 0.909 | 53.0 (13.5) | 51.9 (14.8) | 0.656 |
| Sex, n (%) | |||||||||
| Male | 111 (39.9) | 107 (38.5) | 0.728 | 18 (60.0) | 17 (56.7) | 0.793 | 15 (23.8) | 15 (23.8) | 1.00 |
| Female | 167 (60.1) | 171 (61.5) | 12 (40.0) | 13 (43.3) | 48 (76.2) | 48 (76.2) | |||
| Race, n (%) | |||||||||
| White | 231 (83.1) | 237 (85.3) | 0.153 | 25 (83.3) | 26 (86.7) | 0.237 | 50 (79.4) | 54 (85.7) | 0.642 |
| Black | 35 (12.6) | 23 (8.3) | 3 (10.0) | 0 (0.0) | 8 (12.7) | 6 (9.5) | |||
| Other | 12 (4.3) | 18 (6.5) | 2 (6.7) | 4 (13.3) | 5 (7.9) | 3 (4.8) | |||
| Hispanic Origin, n (%) | |||||||||
| Yes | 36 (13.0) | 25 (9.0) | 0.136 | 2 (6.7) | 5 (16.7) | 0.424 | 8 (12.7) | 2 (3.2) | 0.095 |
| No | 242 (87.1) | 253 (91.0) | 28 (93.3) | 25 (83.3) | 55 (87.3) | 61 (96.8) | |||
| Insurance Status, n (%) | |||||||||
| Private | 148 (55.2) | 146 (53.5) | 0.919 | 15 (51.7) | 19 (63.3) | 0.220 | 33 (52.4) | 36 (57.1) | 0.254 |
| Public | 107 (39.9) | 113 (41.4) | 11 (37.9) | 11 (36.7) | 23 (36.5) | 25 (39.7) | |||
| Uninsured | 13 (4.9) | 14 (5.1) | 3 (10.3) | 0 (0.0) | 7 (11.1) | 2 (3.2) | |||
| Charlson-Deyo Score, n (%) | |||||||||
| 0 | 201 (72.3) | 197 (70.9) | 0.981 | 24 (80.0) | 24 (80.0) | 1.000 | 46 (73.0) | 46 (73.0) | 1.000 |
| 1 | 57 (20.5) | 61 (21.9) | 4 (13.3) | 4 (13.3) | 9 (14.3) | 9 (14.3) | |||
| 2 | 15 (5.4) | 15 (5.4) | 2 (6.7) | 2 (6.7) | 6 (9.5) | 6 (9.5) | |||
| ≥3 | 5 (1.8) | 5 (1.8) | 0 (0.0) | 0 (0.0) | 2 (3.2) | 2 (3.2) | |||
| Laterality, n (%) | |||||||||
| Left | 137 (49.3) | 141 (50.7) | 0.001 | 17 (56.7) | 16 (53.3) | 1.000 | 26 (41.3) | 38 (60.3) | 0.032 |
| Right | 112 (40.3) | 130 (46.8) | 9 (30.0) | 10 (33.3) | 37 (58.7) | 24 (38.1) | |||
| Bilateral | 5 (1.8) | 1 (0.4) | 1 (3.3) | 0 (0.0) | 0 (0.0) | 0 (0.0) | |||
| Unspecified | 24 (8.6) | 6 (2.2) | 3 (10.0) | 4 (13.3) | 0 (0.0) | 1 (1.6) | |||
| Tumor Size, cm (IQR) | |||||||||
| Less than 5cm | 33 (11.9) | 32 (11.5) | 0.983 | 3 (10.0) | 3 (10.0) | 1.000 | 5 (7.9) | 6 (9.5) | 0.950 |
| 5 – 10 cm | 66 (23.7) | 65 (23.4) | 10 (33.3) | 10 (33.3) | 16 (25.4) | 16 (25.4) | |||
| Greater than 10 cm | 179 (64.4) | 181 (65.1) | 17 (56.7) | 17 (56.7) | 42 (66.7) | 41 (65.1) | |||
| Mean Number of Metastatic Sites, (SD) | 1.14 (0.63) | 1.11 (0.65) | 0.552 | 1.50 (0.86) | 1.50 (0.86) | 1.000 | 1.13 (0.63) | 1.13 (0.63) | 1.000 |
| Location of Metastatic Sites | |||||||||
| Liver, n (%) | 149 (53.6) | 110 (39.6) | 0.001 | 19 (63.3) | 9 (30.0) | 0.010 | 29 (46.0) | 43 (68.3) | 0.012 |
| Lung, n (%) | 128 (46.0) | 154 (55.4) | 0.027 | 15 (50.0) | 21 (70.0) | 0.114 | 36 (57.1) | 23 (36.5) | 0.020 |
| Brain, n (%) | 5 (1.8) | 0 (0.0) | 0.061 | 3 (10.0) | 5 (16.7) | 0.706 | 0 (0.0) | 0 (0.0) | 1.00 |
| Bone, n (%) | 36 (13.0) | 45 (16.2) | 0.279 | 8 (26.7) | 10 (33.3) | 0.573 | 6 (9.5) | 5 (7.9) | 0.752 |
| Medical Therapy, n (%) | |||||||||
| Radiation | 178 (64.0) | 153 (55.0) | 0.031 | 19 (63.3) | 21 (70.0) | 0.584 | 36 (57.1) | 33 (52.4) | 0.591 |
| Chemotherapy | 213 (76.6) | 174 (62.6) | <0.001 | 21 (70.0) | 15 (50.0) | 0.114 | 33 (52.4) | 43 (68.3) | 0.069 |
In the first matched cohort, subjects undergoing primary tumor resection had significantly lower rates of liver and lung metastases when compared with matched patients managed non-operatively (Table 3). On survival analysis, patients who underwent primary tumor resection had a significantly longer median survival (15.9 months, 95% CI 11.0–18.9) than matched patients managed non-operatively (6.0 months; 95% CI 4.8–7.5; p<0.001) (Figure 3A). Primary tumor resection was associated with improved survival compared to non-operative management on Cox proportional hazards regression (HR: 0.593; 95% CI 0.486–0.723; p<0.001).
Figure 3.
Kaplan-Meier survival curves for matched cohorts, comparing: A) patients who did not receive surgery versus those who underwent resection of the primary only; B) patients who did not receive surgery versus those who underwent metastasectomy only; and C) patients who underwent resection of the primary tumor only versus those who underwent resection of the primary tumor with metastasectomy.
In the second matched cohort, both cohorts were well-matched, except subjects who underwent metastasectomy had a significantly lower rate of liver metastases compared with matched patients managed non-operatively (Table 3). There was no difference in OS between patients who underwent metastasectomy alone (5.6 months; 95% CI 3.5–22.0) and matched patients treated non-operatively (4.6 months; 95% CI 1.9–11.5; p=0.271) (Figure 3B). Metastasectomy alone was not associated with a survival advantage compared to non-operative management on Cox proportional hazards regression (HR: 0.709; 95% CI 0.421–1.194; p=0.196).
In the third matched cohort, both cohorts were well-matched except subjects who underwent primary resection alone had a significantly lower rate of liver metastases and a significantly higher rate of lung metastases compared with matched patients who underwent primary resection with metastasectomy (Table 3). On survival analysis, patients who underwent primary tumor resection with metastasectomy had a significantly longer median survival (26.3 months, 95% CI 13.4–37.3) than matched patients managed non-operatively (10.1 months; 95% CI 6.3–20.0; p=0.011) (Figure 3C). Primary tumor resection with metastasectomy was associated with improved survival compared to primary tumor resection alone on Cox proportional hazards regression (HR: 0.575; 95% CI 0.383–0.864; p=0.008).
Discussion
In metastatic ACC, effective therapeutic options remain limited. Advancements over the past decade in multimodal therapy, chemotherapeutics, immunotherapy, and targeted therapy in other malignancies have not yet translated to improved survival within this patient population. This has, in part, been hindered by the rarity of ACC, which inherently limits the ability to perform rigorous prospective randomized controlled trials. Therefore, the best available data comes from large, multi-center retrospective studies.
To our knowledge, this is the largest study to date to specifically evaluate the role of metastasectomy either alone or in combination with primary tumor resection in metastatic ACC. We found that surgical resection of the primary tumor is associated with marked improvements in OS compared with non-operative management. Metastasectomy, when added to primary tumor resection, was associated with improved survival over non-operative management or primary tumor resection without metastasectomy. However, metastasectomy alone without primary tumor resection did not improve survival. These findings suggest that the status of the primary tumor may significantly impact disease trajectory regardless of metastatic disease burden, and that metastasectomy is likely beneficial when performed in pursuit of potentially curative resection.
Management of the primary tumor is often one of the first decision points in advanced ACC. In this investigation, we found that resection of the primary tumor alone was associated with improved survival after controlling for extent of metastatic disease. By stratifying patients on receipt of metastasectomy, we isolated the effect of primary resection independent of metastasectomy. Our finding that resection of the primary tumor confers a survival advantage is consistent with two recent studies that utilized the CCR and SEER databases to examine the role of surgery in metastatic ACC. While both of these investigations were notably limited by selection bias [19, 20], the consistency of this finding across multiple studies and patient populations suggests that there is a physiologic advantage associated with the reduction of primary tumor burden, translating to longer survival. Alternatively, this finding may represent excellent patient selection, with patients with less aggressive tumors or a better prognosis being more likely to undergo surgical resection. After matching patients based on tumor size and number of metastatic sites, however, resection of the primary tumor remained associated with improved survival.
The role of metastasectomy is not well-described in ACC. Studies generally demonstrate improved OS, especially in patients with metachronous rather than synchronous metastases [22, 30, 31]. Resection of pulmonary [21, 32] and liver metastases [23, 33] specifically appears safe and effective, however recurrence is common. In both of our analyses, we found improved OS in patients undergoing primary tumor resection with metastasectomy compared to primary tumor resection alone. Patients who underwent metastasectomy alone did not demonstrate a survival benefit over non-operative management. These findings likely indicate that, in the face of limited metastatic disease, attempts at complete, curative resection are reasonable and can improve survival.
Whereas combined adrenalectomy and metastasectomy may improve survival through complete margin-negative resection, the etiology of improved outcomes through resection of the primary tumor alone is unclear. We postulate this may be due to improved control of corticosteroid hormonal secretion, mass effect, or tumor thrombus through debulking or that perhaps the primary tumor serves as the principal reservoir for shedding metastatic disease. It may also be that patients are being selected for primary resection are fundamentally different than those who do not undergo surgery; for instance, they may have excess hormonal secretion or other locoregional symptoms that allow them to benefit more from resection. Regardless, we would suggest that in carefully selected patients, resection of the primary tumor should be considered, and that in patients where complete R0 resection is achievable, metastasectomy should be utilized in conjunction with primary tumor resection.
Palliative resection of the primary tumor or of metastases has not been extensively studied [34]. Our cohort did not have an adequate sample of patients undergoing debulking or palliative resection. However, patients undergoing metastasectomy alone are likely being treated palliatively for symptomatic control rather than for curative intent. While we cannot discern the role of palliative resection for the primary tumor, subjects undergoing metastasectomy alone did not demonstrate improved survival with surgery.
Our study also identified two unexpected significant relationships between private insurance and Black race and improved survival. Improved insurance coverage, decreased time to surgery, and/or higher socioeconomic status associated with private insurance may explain improved survival in this cohort. There may be dissimilarities in tumor phenotype by race or ethnic background explaining differential outcomes by this variable; for instance, altered tumor oncogenetic pathways may make one group more prone to more aggressive or perhaps more hormonally active tumor phenotypes. Further research is necessary to elucidate the mechanisms behind these relationships.
This retrospective study is primarily limited by selection bias. Patients with resectable primary tumors, lower volume metastatic disease, or metachronous metastatic disease may have been selected as good candidates for surgical therapy. Alternatively, there may be fundamental differences in tumor biology between groups selected for surgery and those selected for medical management, such as differences in excess corticosteroid secretion. In this study, we attempted to account for these disease-level factors through multivariable Cox regression, as well propensity score matching. Nonetheless, it is not possible to fully account for selection bias. Our findings may therefore reflect excellent surgical decision-making and appropriate patient selection rather than a universal surgical benefit in metastatic ACC. Our study is additionally limited by the data source. The NCDB does not capture data regarding the temporal relationship of surgeries a subject received, and it is therefore not possible to determine if metastasectomy was performed for synchronous or metachronous disease. This limitation is somewhat mitigated by the fact that we determined stage based on initial clinical and/or pathologic staging, which should limit the impact of metachronous disease versus synchronous disease on outcomes. The NCDB also does not provide information regarding which specific distant site was resected, nor the number of metastatic sites resected. While we were able to match for the number of metastatic sites, limited patient numbers made it infeasible to match for the specific sites of metastatic disease present and there were some differences regarding specific metastatic sites within our matched cohorts. Future studies should analyze comparative outcomes based on metastatic site to be resected. Finally, data regarding hormonal secretion for each patient are not captured in the NCDB. In patients with hormonally active tumors, debulking may improve quality of life or survival through reduction of hormone secretion, but this was unable to be assessed nor controlled for.
In conclusion, we found that resection of the primary tumor alone or in combination with metastasectomy improves OS in metastatic ACC, further supporting the importance of surgical management within this population.
Funding:
HW received funding from the National Institutes of Health, NCI grant #K08 CA270385. JMSB received funding from the NIH T32 Training Program in Surgical Oncology Research at Penn, grant #5T32CA251063-02.
Footnotes
Conflicts of interest: The authors declare no conflicts of interest.
References
- 1.Sharma E, Dahal S, Sharma P, et al. (2018) The characteristics and trends in adrenocortical carcinoma: a United States population based study. Journal of clinical medicine research 10:636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kebebew E, Reiff E, Duh Q-Y, et al. (2006) Extent of disease at presentation and outcome for adrenocortical carcinoma: have we made progress? World journal of surgery 30:872–878 [DOI] [PubMed] [Google Scholar]
- 3.Bilimoria KY, Shen WT, Elaraj D, et al. (2008) Adrenocortical carcinoma in the United States: treatment utilization and prognostic factors. Cancer: Interdisciplinary International Journal of the American Cancer Society 113:3130–3136 [DOI] [PubMed] [Google Scholar]
- 4.Golden SH, Robinson KA, Saldanha I, et al. (2009) Prevalence and incidence of endocrine and metabolic disorders in the United States: a comprehensive review. The Journal of Clinical Endocrinology & Metabolism 94:1853–1878 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wooten MD, King DK (1993) Adrenal cortical carcinoma. Epidemiology and treatment with mitotane and a review of the literature. Cancer 72:3145–3155 [DOI] [PubMed] [Google Scholar]
- 6.Luton J-P, Cerdas S, Billaud L, et al. (1990) Clinical features of adrenocortical carcinoma, prognostic factors, and the effect of mitotane therapy. New England Journal of Medicine 322:1195–1201 [DOI] [PubMed] [Google Scholar]
- 7.Fassnacht M, Allolio B (2009) Clinical management of adrenocortical carcinoma. Best practice & research Clinical endocrinology & metabolism 23:273–289 [DOI] [PubMed] [Google Scholar]
- 8.Else T, Kim AC, Sabolch A, et al. (2014) Adrenocortical carcinoma. Endocrine reviews 35:282–326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Vassilopoulou-Sellin R, Schultz PN (2001) Adrenocortical carcinoma: clinical outcome at the end of the 20th century. Cancer: Interdisciplinary International Journal of the American Cancer Society 92:1113–1121 [DOI] [PubMed] [Google Scholar]
- 10.Kostiainen I, Hakaste L, Kejo P, et al. (2019) Adrenocortical carcinoma: presentation and outcome of a contemporary patient series. Endocrine 65:166–174 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.David SO, Krieg S, Esposito I, et al. (2023) A Revised Version of the TNM Classification Leads to Optimized Predictive Performance in Patients with Adrenocortical Carcinoma. Hormone and Metabolic Research 55:227–235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kiseljak-Vassiliades K, Bancos I, Hamrahian A, et al. (2020) American Association of Clinical Endocrinology Disease state clinical review on the evaluation and management of adrenocortical carcinoma in an adult: a practical approach. Endocrine Practice 26:1366–1383 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fassnacht M, Terzolo M, Allolio B, et al. (2012) Combination chemotherapy in advanced adrenocortical carcinoma. New England Journal of Medicine 366:2189–2197 [DOI] [PubMed] [Google Scholar]
- 14.Jasim S, Habra MA (2019) Management of adrenocortical carcinoma. Current oncology reports 21:1–11 [DOI] [PubMed] [Google Scholar]
- 15.Altieri B, Ronchi CL, Kroiss M, et al. (2020) Next-generation therapies for adrenocortical carcinoma. Best Practice & Research Clinical Endocrinology & Metabolism 34:101434. [DOI] [PubMed] [Google Scholar]
- 16.Vogel JD, Felder SI, Bhama AR, et al. (2022) The American Society of Colon and Rectal Surgeons clinical practice guidelines for the management of colon cancer. Diseases of the Colon & Rectum 65:148–177 [DOI] [PubMed] [Google Scholar]
- 17.Hernandez Dominguez O, Yilmaz S, Steele SR (2023) Stage IV Colorectal Cancer Management and Treatment. Journal of Clinical Medicine 12:2072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Straker III RJ, Witmer HD, Deschner B, et al. , The Role of Surgery in Managing Primary and Metastatic Colorectal Cancer (2022) In: Cancer Metastasis Through the Lymphovascular System, Springer, p. 407–419 [Google Scholar]
- 19.Livhits M, Li N, Yeh MW, et al. (2014) Surgery is associated with improved survival for adrenocortical cancer, even in metastatic disease. Surgery 156:1531–1541 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wang S, Gao W-C, Chen S-S, et al. (2017) Primary site surgery for metastatic adrenocortical carcinoma improves survival outcomes: an analysis of a population-based database. OncoTargets and therapy 10:5311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kemp CD, Ripley RT, Mathur A, et al. (2011) Pulmonary resection for metastatic adrenocortical carcinoma: the National Cancer Institute experience. The Annals of thoracic surgery 92:1195–1200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Prendergast KM, Smith PM, Tran TB, et al. (2020) Features of synchronous versus metachronous metastasectomy in adrenal cortical carcinoma: analysis from the US adrenocortical carcinoma database. Surgery 167:352–357 [DOI] [PubMed] [Google Scholar]
- 23.Lo WM, Kariya CM, Hernandez JM (2019) Operative management of recurrent and metastatic adrenocortical carcinoma: A Systematic review. The American Surgeon 85:23–28 [PMC free article] [PubMed] [Google Scholar]
- 24.Gaujoux S, Al-Ahmadie H, Allen PJ, et al. (2012) Resection of adrenocortical carcinoma liver metastasis: is it justified? Annals of surgical oncology 19:2643–2651 [DOI] [PubMed] [Google Scholar]
- 25.Boffa DJ, Rosen JE, Mallin K, et al. (2017) Using the National Cancer Database for outcomes research: a review. JAMA oncology 3:1722–1728 [DOI] [PubMed] [Google Scholar]
- 26.Mallin K, Browner A, Palis B, et al. (2019) Incident cases captured in the National Cancer Database compared with those in US population based central cancer registries in 2012–2014. Annals of surgical oncology 26:1604–1612 [DOI] [PubMed] [Google Scholar]
- 27.StataCorp Stata Statistical Software: Release 17, College Station, TX, StataCorp LLC, 2021. [Google Scholar]
- 28.Stuart EA, King G, Imai K, et al. (2011) MatchIt: nonparametric preprocessing for parametric causal inference. Journal of statistical software [Google Scholar]
- 29.Team RC R: A language and environment for statistical computing (Version 4.2. 2)[Computer software]. R Foundation for Statistical Computing, 2021. [Google Scholar]
- 30.Datrice NM, Langan RC, Ripley RT, et al. (2012) Operative management for recurrent and metastatic adrenocortical carcinoma. Journal of surgical oncology 105:709–713 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Dy BM, Strajina V, Cayo AK, et al. (2015) Surgical resection of synchronously metastatic adrenocortical cancer. Annals of surgical oncology 22:146–151 [DOI] [PubMed] [Google Scholar]
- 32.op den Winkel J, Pfannschmidt J, Muley T, et al. (2011) Metastatic adrenocortical carcinoma: results of 56 pulmonary metastasectomies in 24 patients. The Annals of thoracic surgery 92:1965–1970 [DOI] [PubMed] [Google Scholar]
- 33.Baur J, Büntemeyer T-O, Megerle F, et al. (2017) Outcome after resection of adrenocortical carcinoma liver metastases: a retrospective study. BMC cancer 17:1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Sinclair TJ, Gillis A, Alobuia WM, et al. (2020) Surgery for adrenocortical carcinoma: When and how? Best Practice & Research Clinical Endocrinology & Metabolism 34:101408. [DOI] [PubMed] [Google Scholar]