Outcome of multidisciplinary management of adrenocortical carcinoma: a retrospective study

Background: Adrenocortical Carcinoma (ACC) is a rare and highly malignant tumor with a high recurrence rate. Reoperation to attain negative resection margin is recommended for selected patients. However, there is limited data on the outcome of unresectable recurrent tumor treated with palliative radiotherapy (RT) and transhepatic arterial embolization (TAE). Method: A retrospective analysis of the Peking Union Medical College Hospital was conducted. Patients with pathologically con�rmed ACC were included. Result: Ninety-two patients ful�lled the criteria. The 5-year, 3-year, 1-year overall survival (OS) rates were 27.2%, 44.6% and 71.7% respectively, and median OS was 32.1 months. Age < 50 years, ENSAT stage I/II, surgery and negative surgical margin were associated with improved OS (all P <0.05). A Ki67 index < 20% was a predictor of better disease-free interval (P =0.011). Of 51 patients with recurrent ACC, local treatment (reoperation, palliative RT and TAE) improved OS after recurrence from 14.5 months to 41.6 months. Local therapy and mitotane were both prognostic factors for improved OS after recurrence. Conclusion: multidisciplinary treatment is the major therapeutic approach for ACC. Resectable cases should receive complete resection to attain negative margin. Local treatment such as palliative RT and TAE may extend OS for unresectable recurrence.


Introduction
Adrenocortical carcinoma (ACC) is a rare and highly aggressive malignancy with a median overall survival of 35.2 months [1].Women develop ACCs more often than men with female/male ratio 1.5 to 2.5 :1 [2,3].In spite of low annual incidence of 1-2 case per million population [4], most tumors are diagnosed at an advanced stage with invasion of adjacent organs or metastatic disease [5].Based on the tumor staging classi cation suggested by the European Network for Study of Adrenal Tumors (ENSAT), the estimated 5-year overall survival of stage I, II, III and IV are 82%, 61%, 50% and 13%, respectively [6].
The mainstay of treatment for primary ACC is complete radical resection to attain negative surgical margins [7].However, local recurrence and metastasis rates remain high, with up to 85% patients in previously reported cohort experiencing local recurrence [8].Adjuvant therapy such as mitotane and radiotherapy were suggested for patients with a high risk of recurrence, despite inconsistent results in published studies [9,10].
There is no consistent agreement in the management of recurrent ACC and few effective chemotherapy regimens can be offered for these recurrent ACC patients.Mitotane monotherapy is recommended for unresectable patients, but not all of these patients can achieve longer survival time [11].As for isolated recurrent ACC, if complete resection is possible, aggressive surgical resection seems to be the best method to prolong survival.Some studies have proved that the median survival after recurrence could be more than 60 months in patients receiving reoperation [5].For patients with resectable liver metastases, surgical resection should be considered with a 5-year overall survival of 39% [12].There is limited data about palliative radiotherapy (RT), transhepatic arterial embolization (TAE) and radiofrequency ablation (RFA) in the management of recurrent ACC patients.These available researches supported the bene t of palliative RT for unresectable locally or distant metastatic disease, such as in bone or lung [13].
In the present study, based on ACC patients treated in Peking Union Medical College Hospital (PUMCH) for 30 years, we analyzed the predictive clinicopathologic and therapeutic factors associated with the prognosis and survival of ACC patients.Furthermore, we evaluated the bene t of reoperation, palliative radiotherapy and TAE in local recurrence or distant metastases.

Methods
The clinical data of all patients with ACC treated at PUMCH over a 30-year period from February 1, 1990 to January 1, 2019 were retrospectively collected.Medical records at PUMCH were retrospectively reviewed for demographic data, clinical and histo-pathologic information, and treatment parameters.
Dates of death, cancer recurrence and metastasis were con rmed by either querying the medical records or making telephone interviews.Data of 98 patients having pathological diagnosis could be included.
One patient with early postoperative mortality (<30 days) and 5 patients with inadequate clinical records were excluded.
The last time of follow-up for this study was October 1, 2019.Overall survival (OS) was de ned as the interval from the date of diagnosis to the date of death by ACC or the nal follow-up, with patients alive at last follow up censored on that date.Disease-free interval (DFI) was de ned as the interval from the date of the initial treatment to the date of rst relapse or metastasis by ACC or the nal follow-up, with patients censored on the date of last follow up if alive without disease progression on that date.For tumor stage, we used the European Network for the Study of Adrenal Tumors (ENSAT) [14] staging system to evaluate the ACC stage.Stage and stage II were combined and classi ed as the reference.
Continuous variables were described with medians and interquartile range (IQR), and Mann-Whitney U tests were used to compare the differences between the groups.Survival analysis was conducted using Kaplan-Meier methods and log-rank tests.Multivariate analysis was carried out using Cox proportional hazards regression methods.P values <0.05 were considered statistically signi cant.All analyses were performed using SPSS 23.0.

Overall survival in whole group
Of 92 patients, 46 were female, with a female/male ratio of 1:1, which was contradicted with previous reported studies.The median age at rst diagnosis of ACC diagnosis was 43.5 years old (IQR: 37.0-57.0),and all the patients were adults except for a 2-year child.66.3% patients were <50 years and 33.7% were ≥50 years.According to the ENSAT staging system, 10.9% patients presented with stage I, 47.8% with stage II, 26.1% with stage III and 15.2% with stage IV.In terms of the location of primary tumor occurrence, 56.5% occurred in the right adrenal gland and 43.5% in the left.In terms of clinical symptoms, 35.9% patients had endocrine symptoms (20 with Cushing's syndrome, 9 with sex hormone abnormality and 4 with primary hyperaldosteronism), 27.2% had tumor-related symptoms, and 36.9% didn't show any clinical symptoms.Demographic characteristics of the studied population are summarized in Table 1.
In patients who received chemotherapy, 86.1% included mitotane as monotherapy or a part of their regimen.In patients who received radiotherapy, 50% received adjuvant radiation after resection, with a median dose of 56 Gy (range, 40-60) in a median of 28 fractions (range, 20-30 fractions).All adjuvant RT were external radiation including Tomotherapy, IMRT and VMAT, using a linear accelerator of 6MV to 18MV.In general, adjuvant RT was well tolerated, with reports of mild to moderate nausea (n=6), fatigue (n=4), abdominal pain (n=1).The total course of RT was not interrupted due to side effects.
We evaluated the signi cance of various clinical and treatment factors that may in uence the prognosis and survival.In the entire group, age < 50 years old, ENSAT stage I/II, Ki-67<20%, surgery and negative surgical margin were all associated with superior OS by univariate analysis.However, those who received any form of chemotherapy only demonstrated an insigni cant survival bene t compared with those who did not receive chemotherapy.The local recurrence rate of adjuvant RT patients was signi cantly lower than that of non-RT patients (16.7% vs. 51.7%,p = 0.027), but the latter had a tendency to prolong OS (23.13 vs. 44.40,p= 0.61).Multivariate analysis of above variate showed that age < 50 years, ENSAT stage I/II and negative surgical margin were independent predictors of prolonged overall survival (P < 0.05).
All the characteristics mentioned above were re-evaluated on DFI (Table 2).Negative surgical margin and Ki67 index<20% were associated with improved DFI (P < 0.05).ACC in left side, ENSAT stage I/II and postoperative adjuvant radiotherapy tended to improve DFI.In multivariate analysis, negative surgical margins (P = 0.001) and Ki67 index<20% (P = 0.008) were still independent predictors of improved DFI.

Overall survival after recurrence
We further analyzed the 51 patents who recurred after the initial operative resection.Among them, 37 had negative margins, 9 had positive margins and 5 were unknown.According to the initial ENSAT stage, 10 patients presented with stage I, 27 with stage II, and 14 with stage III.The median interval from primary operation to rst recurrence was 14.2 months (IQR: 8.8-28.5).Thirty-two patients had local recurrence and 29 patients had distant metastases, 10 patients had both local recurrence and distant metastases.
According to the primary treatment after recurrence, these patients were divided into local treatment group (n=29) and expectant treatment group (n=22).
Therapy modality of the local treatment group included reoperation (n=17), palliative RT (n=8) and TAE (n=4) (Fig. 3&4).All the 17 operated patients had locoregional resection: 7 had resection of only the tumor bed, and 10 had resection of adjacent organs (6 radical nephrectomies, 2 partial nephrectomies, 2 partial hepatectomy, 1 cholecystectomy and 1 segmental colectomy).No death occurred within 30 days after operation.Of the 8 patients received palliative RT, the median dose was 45 Gy (range 24-60Gy) in a median of 12 fractions (range 6-30 times).The sites of the treated regions were the adrenal tumor bed (n=3), bone metastasis (n=2), lung metastasis (n=1), abdominal mass (n=1) and psoas major metastasis (n=1).All patients obtained either symptom relief or tumor shrinkage through imaging assessment after RT.In terms of TAE, the 4 patients underwent 2, 3, 11 and 13 courses respectively.The median diameter of hepatic metastases was 6.3cm (range, 3.9-17.8).Through imaging assessment, their tumor sizes did not enlarge signi cantly for at least half a year after receiving treatment.Besides, all these patients could tolerate with TAE.Demographic characteristics of the two groups are summarized in Table 3.There were no statistically signi cant differences in age, gender, clinical manifestation, tumor distribution, primary surgical approach, surgical margin and mitotane use between the two groups.In the local treatment group, the proportion of patients with initial ENSAT stage I/II (stage I: 21.4% vs. 17.4%, stage II: 64.3% vs. 39.1%,p = 0.033), local recurrence (75% vs. 47.8%,p=0.048) and DFI>12 (75% vs.47.8%, p=0.048) is signi cantly higher than that of the expectant treatment group.Median OS after recurrence of the local treatment group was 41.6 months (IQR: 27.4-66.1),while it was 14.5 months (IQR: 5.7-24.9) of the expectant group (P=0.001) (Fig. 2).There was no difference in OS after recurrence among these patients who received reoperation, palliative RT and TAE (41.9 months [27.4-69.1],41.5 months [21.9-63.7],40.1 months [26.5-

97.8], respectively).
In univariate analysis, independent prognostic factors for higher OS were female, original ENSAT stage I/II, local treatment for recurrence, chemotherapy (All P<0.05).Age <50 years old (P=0.073),liver metastases (P=0.093),DFI >12 months between the initial resection and the diagnosis of recurrence (P=0.079)tended to improve OS after recurrence (Table 4).The original tumor size, ENSAT stage, Ki67 index, and the primary surgical technique were not predictive factors for OS after recurrence.In multivariate analysis, local therapy for recurrence, mitotane use and female were independently factors for improved OS after recurrence (P <0.05).

Discussion
ACC constitutes very rare and aggressive malignancies that leads to highly recurrence and metastasis.It is really challengeable to assess the management about this disease.Consequently, relevant studies are generally of reports from Europe and North America.In this retrospective study, with the largest sample size in Asia population, we analyzed the clinicopathologic and therapeutic factors associated with the prognosis and survival of ACC.And we further compared two groups with recurrent disease: the rst group received local therapy for recurrence, and the second group had expectant therapy.We demonstrated a signi cant survival advantage of local treatment (reoperation, palliative RT, TAE) in recurrent ACC.
Previous studies have proved that tumor stage, surgical resection margin, mitotic index, Ki67 index, age and overproduction of cortisol were associated with prognosis [7,[15][16][17].The results of this study indicated that age <50 years old, ENSAT stage I/II, negative resection margin were signi cant factors with better prognosis of ACC.The tumor size and clinical symptoms had no obvious impact on the prognosis, which could be explained partly by our small sample size.Although some previous studies [18,19] and our study found that laparotomy and laparoscopy had similar impact on OS, most researchers believed that laparoscopy for ACC resection would lead to more frequent and earlier recurrence [9,19,20].Therefore, most guidelines recommend laparotomy as a standard surgical treatment for ACC method [19].
In addition to surgical resection margin, Ki67 index is also an important factor to evaluate the risk of recurrence.It has been used to select patients for adjuvant chemotherapy [16].The cutoff value of Ki67 index for high and low recurrence risk of ACC has not yet been determined [21].However, a high-grade ACC de ned partly by a high mitotic rate (and/or Ki67 score> 20%) has a higher risk of recurrence [22] when compared with a low-grade ACC.A recent study showed that patients with a Ki67 index <10% had an OS after initial surgery of 181 months, while patients with a Ki67 index ≥20% had on OS of only 42 months [16].In our study, patients with Ki67 index ≥20% also had a shorter OS (32.2 vs. 44.3months, P =0.023) and recurred more quickly after initial surgery (DFI: 10.8 vs. 16.4 months; P = 0.016) than patients with Ki67 index <20%.Therefore, ACC patients with Ki67 index ≥20% may need to take more active measures to prevent recurrence.And it was interesting to note that patient with right ACC had a shorter DFI and a higher distant metastasis rate at initial recurrence (62.5% vs. 26.7%,p=0.003).There are several reasons for this observation.First, right adrenocortical tumors are in a narrower space among liver, inferior vena cava and kidney with a higher potential of invasion compared with left tumors.Second, resection of left organs (kidney, spleen, pancreas) is easier than resection of right organs (liver, inferior vena cava).Third, there is only one venous return of the right adrenal gland, which is more likely to rupture under the compression of the tumor, increasing the risk of tumor cells entering the blood vessels.
Locally recurrent or distant metastatic ACC is currently uncurable.If complete surgical resection is possible, surgical treatment should be performed to obtain a negative resection margin [23].Reoperation can signi cantly improve the prognosis when the recurrence can be completely removed, especially in patients with a disease-free survival (DFI) >12 months.Simon et al included 59 patients who recurred after initial operative resection, and the OS of 29 reoperated patients was signi cantly better than that of the 30 non-operated patients (91m vs. 15m, p <0.001) [5].Baur et al reported 43 patients who underwent liver metastases resection with a 5-year survival rate of 51.3% and a median OS of 76.1 months [12].The prognosis of 17 reoperated patients in our study was also better than that of patients in expectant treatment group.Furthermore, 6 patients underwent ³2 reoperations achieving with an OS of 80.5 months.
This indicated that repeated operative resection for recurrent ACC could achieve long-term survival in selected patients.
Mitotane has been used in the treatment of unresectable and recurrent ACC, but the results were not exactly the same [10,24].Mitotane acts on adrenocortical cell mitochondria to inhibit CYP11B1 and CYP11A1 causing necrosis of adrenocortical cells [25].The toxicity pro le of mitotane limits tolerability and requires frequent drug monitoring.It should be noted that the results of surgery cannot be achieved with mitotane monotherapy and combined chemotherapy [26].Most patients with unresectable metastases die within 1 year making it urgent to nd a way to extend their lives.Therefore, more and more attention has been put on local palliative treatment such as RT and RFA with deeper understanding.
ACC was once thought to be radiation-resistant in previous small studies [27,28].However, with the application of modern radiation techniques, more and more retrospective studies showed that postoperative adjuvant RT can reduce the local recurrence rate.Sabolch et al [29] performed a casecontrol analysis on 20 patients who received adjuvant RT with a median irradiation dose was 55Gy.The local recurrence rate of adjuvant RT group was signi cantly reduced to 5% vs. 60% of non-RT group, although there was no difference in OS.The largest study [30] on this subject included 39 patients with a median irradiation does of 55Gy, and found that adjuvant RT signi cantly improved the local recurrence (33.3% vs. 71.8%)and OS (77.7% vs. 48.6% at 3 years).In the present study, only 16.7% patients in adjuvant RT group experienced local recurrences, which was signi cantly lower than 51.7% in non-RT group.Large-scale prospective randomized studies are needed to better understand the role of adjuvant radiotherapy in the management of ACC, especially to gure out whether the reduction in the local recurrence rate would accompany with better OS.
There is limited evidence to support the use of palliative RT for unresectable ACC.Polat et al [13] reported 51 of 91 patients gained bene t (ie, pain relief, reduction in paresthesia or paralysis) from palliative RT for advanced ACC.Ho et al [31] reported 10 of 12 patients had either a clinical or radiographic response after palliative RT, although the long-term effect is unknown.Our study also showed that 8 patients gained bene t after palliative RT with an improved median OS of 73.8 months.This could be partly explained by the active local treatment of these 8 patients, of which 2 patients received ³3 times of palliative RT and 1 patient received RFA after second recurrence.By the way, the patients in our study received more precise and advanced RT technology, including IMRT, VAMT and Tomotherapy.
In the case of metastatic ACC, liver is one of the most commonly involved organs beside the lung [12].Liver resection in the case of liver metastases can achieve long-term survival, while the OS of patients with unresected hepatic metastases was only 10.1 months.RFA may provide short-term local control for small liver metastases with a 5-year OS rate of 29% [32,33].TAE is frequently used in the treatment of unresectable liver cancer and metastases.However, the number of patients receiving TAE for hepatic metastases of ACC is small due to its rarity.In our study, 4 patients received TAE for unresectable liver metastases after initial recurrence, two of whom had more than 10 courses of embolization.The median OS of these 4 patients was 76.5m, indicating repeated TAE can achieve long-term control of unresectable liver metastases of ACC, although there may be selection bias, which is a common problem in the study of this disease.
The main limitations of this study are its retrospective, non-random design and inconsistency in treatment management due to the long-time span.The number of patients was small and not all patients were thoroughly followed because some of them were treated and managed in several different hospitals.Most of the current researches are retrospective studies, and the heterogeneity of treatment preference in different institutions is relatively large.Prospective studies will help understand the potential role of local treatment such as palliative RT and TAE in the treatment of recurrent ACC.
In summary, multidisciplinary treatment is the major therapeutic approach for ACC.Physicians need to choose the most appropriate local and systemic treatment for each individual patient according to the physical condition and tumor characteristics.Resectable cases should receive complete resection to attain negative margin for both primary and recurrent ACC.In terms of unresectable recurrence, whether locally recurrent or distant metastatic, local treatment such as palliative RT and TAE may not only relieve symptoms, but also extend OS through repeated treatment.

Figures
Figures

Figure 3 Treatment
Figure 3

Table 1 .
Demographic and treatment features associated OS

Table 3 .
Demographic treatment features in relapsed patients