ELSEVIER
Check for updates
Prognostic significance of lymph node count in surgically treated patients with T2-4 stage nonmetastatic adrenocortical carcinoma
Anis Assada,*, Francesco Barlettaa,b,c, Reha-Baris Incesua,d, Lukas Scheipnera,e, Simone Morraa,f , Andrea Baudoa,g,h, Cristina Cano Garciaa,1, Zhe Tiana, Sascha Ahyaid, Nicola Longof, Felix K.H. Chung, Shahrokh F. Shariat),k,l,m, Derya Tilkid,n,º, Alberto Brigantib,C, Fred Saada, Pierre I. Karakiewicza
a Cancer Prognostics and Health Outcomes Unit, Division of Urology, University of Montréal Health Center, Montréal, Québec, Canada b Unit of Urology/Division of Oncology, Gianfranco Soldera Prostate Cancer Lab, IRCCS San Raffaele Scientific Institute, Milan, Italy · Vita-Salute San Raffaele University, Milan, Italy
d Martini-Klinik Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany e Department of Urology, Medical University of Graz, Graz, Austria
Department of Neurosciences, Reproductive Sciences and Odontostomatology, University of Naples Federico II, Naples, Italy
§ Department of Urology, IRCCS Ospedale Galeazzi - Sant’Ambrogio, Milan, Italy ” Department of Urology, IRCCS Policlinico San Donato, Milan, Italy
Department of Urology, University Hospital Frankfurt, Goethe University Frankfurt am Main, Frankfurt am Main, Germany
¿Department of Urology, Comprehensive Cancer Center, Medical University of Vienna, Vienna, Austria
k Department of Urology, Weill Cornell Medical College, New York, NY
l Department of Urology, University of Texas Southwestern Medical Center, Dallas, TX
” Hourani Center of Applied Scientific Research, Al-Ahliyya Amman University, Amman, Jordan
” Department of Urology, University Hospital Hamburg-Eppendorf, Hamburg, Germany
° Department of Urology, Koc University Hospital, Istanbul, Turkey
Received 15 October 2023; received in revised form 11 February 2024; accepted 1 April 2024
ABSTRACT
Purpose: The role of lymphadenectomy and the optimal lymph node count (LNC) cut-off in nonmetastatic adrenocortical carcinoma (nmACC) are unclear.
Methods: Within the Surveillance, Epidemiology, and End Results (SEER) database, surgically treated nmACC patients with T2-4 stages were identified between 2004 and 2020. We tested for cancer-specific mortality (CSM) differences according to pathological N-stage (pN0 vs. pN1) and two previously recommended LNC cut-offs (≥4 vs. ≥5) were tested in pN0 and subsequently in pN1 subgroups in Kaplan-Meier plots and multivariable Cox regression models.
Results: Of 710 surgically treated nmACC patients, 185 (26%) underwent lymphadenectomy and were assessable for further analyses based on available LNC data. Of 185 assessable patients, 152 (82%) were pN0 and 33 (18%) were pN1. In Kaplan-Meier analyses, CSM- free survival was 74 vs. 14 months (A 60 months, P ≤ 0.001) in pNO vs. pN1 patients, respectively. In multivariable analyses, pN1 was an independent predictor of higher CSM (HR:3.13, P < 0.001). In sensitivity analyses addressing pN0, LNC cut-off of ≥4 was associated with lower CSM (multivariable hazard ratio [HR]: 0.52; P = 0.002). In sensitivity analyses addressing pN0, no difference was recorded when a LNC cut-off of ≥5 was used (HR:0.60, P=0.09). In pN1 patients, neither of the cut-offs (≥4 and ≥5) resulted in a statistically significant stratification of CSM rate, and neither reached independent predictor status (all P > 0.05).
The research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. *Corresponding author. Tel .: +514-890-8000. E-mail address: anis.assad10@gmail.com (A. Assad).
Conclusions: Lymphadenectomy provides a prognostic benefit in nmACC patients and identifies pN1 patients with dismal prognosis. Conversely, in pNO patients, a LNC cut-off ≥4 identifies those with particularly favorable prognosis. @ 2024 Elsevier Inc. All rights reserved.
Keywords: Adrenalectomy; Adrenocortical carcinoma, Lymph node dissection, Lymphadenectomy, Lymph node examination
1. Introduction
Current guideline recommendations are in favor of rou- tine loco-regional lymphadenectomy in nonmetastatic adre- nocortical carcinoma (nmACC) [1-3]. However, historical data suggest that it is not universally implemented. Specifi- cally, less than 30% of patients underwent lymphadenec- tomy in most series [2,4-7]. Additionally, there is a lack of consensus on the optimal lymph node count (LNC) cut-off points for lymphadenectomy, leading to the utilization of different LNC cut-offs in previous studies, including LNC ≥4, LNC ≥5, and even relying on the “surgeon’s effort to perform a lymphadenectomy” [6-10]. Two large scale pop- ulation studies that relied on the National Cancer Database (NCDB; Panjwani et al. : 2004-2013, n = 827; and Deschner et al. : 2004-2015, n=897) addressed this topic [9,11]. However, the prognostic benefit of LNC ≥4 was only iden- tified by Panjwani et al. The limitation of the NCDB data- base relies in lack of cancer-specific mortality (CSM) information which may undermine the ability of the analy- sis to test for the effect of lymphadenectomy on more spe- cific cancer control related endpoints such as CSM or recurrence. To address the uncertainty about the benefit of lymphadenectomy on nmACC, we relied on the Surveil- lance, Epidemiology and End Results (SEER) database where CSM is available and can be interpreted as the most ultimate cancer control endpoint. In addition to testing the effect of lymph node involvement (pN1) vs. no-lymph node involvement (pN0) on CSM, we also tested the effect of various LNC that have been previously recommended at lymphadenectomy for nmACC [6]. Specifically, LNC ≥4 as well as LNC≥5 have been previously recommended. We postulated that pN1 will have a profound effect on CSM. Moreover, we postulated that various previously recom- mended LNC may help in better discriminating between pN0 patients and possibly between pN1 patients.
2. Materials and methods
2.1. Data source and study population
The SEER database, which covers approximately 48% of the United States (US) population, is representative of the US population with respect to demographic composition and cancer incidence rates. Within the SEER database (2004-2020) [12], we identified patients aged 18 years or older, with histologically confirmed ACC (International Classification of Disease for Oncology [ICD-O-3] site code
“C74.0:Cortex of adrenal gland”/“C74.9:Adrenal gland, NOS”; histologic code: “8370/3: Adrenal cortical carcinoma”, “8010/3: Carcinoma, NOS”, and “8140/3: Adenocarcinoma, NOS”). Histological status categorized as “8010/3: Carcinoma, NOS” and “8140/3: Adenocarcinoma, NOS,” along with site code “C74.9: Adrenal gland, NOS” were excluded due to their uncertain nature. Patients diag- nosed with metastasis at the time of ACC diagnosis were excluded (NAACCR items: #2850 [2004-2015], #960 [2016-2017], #776 (2018-2020). Surgical status was con- firmed using the “Surgery of Primary Site” code (NAACCR item #1290). Adrenalectomy was assigned for patients who underwent local tumor resection, partial resection, or radi- cal resection of the primary site (“Surgery of Primary site:20-90”). Patients who did not undergo surgery of the primary site (“Surgery of Primary site:00”), and those with treatment type coded as local tumor ablation (“Surgery of Primary site:10”), or not otherwise specified (“Surgery of Primary site:99”) were excluded.
According to current recommendations for lymphadenec- tomy only stages T2-4 were included [2,3,6]. The two groups of interest consisted of pN0 vs. pN1 patients. Autopsy or death certificate only cases were also excluded. LNC status was determined using the “Regional nodes examined code (RNE) [NAACCR item 830]“. This included categories such as: no lymph node examination performed (“RNE: 00”), varying numbers of lymph node examinations performed (“RNE: 01-90”), and uncertain or unknown lymph node examination status (“RNE: 95-99”). Patients with uncertain or unknown LNC status (“RNE:95-99”) were subsequently excluded from the analysis (n = 15, Fig. 1).
2.1.1. Variables of interest
Demographic covariates consisted of age at diagnosis (years, continuously coded), sex, race/ethnicity (Caucasian, non-Caucasian). Tumor characteristics consisted of primary tumor size (millimeters, continuously coded), tumor side, pathological N stage (pN0, pN1), and LNC. Patients who did not undergo pathological lymph node examination were labelled pNx. Systemic and radiation therapies were coded as received, not received or unknown. CSM represented the endpoint of interest.
2.2. Statistical analyses
Six analytical steps were performed. First, descriptive sta- tistics compared pN0 vs. pN1 vs. pNx patients. Second,
Surveillance, Epidemiology, and End Results (SEER) database 2004-2020 on adrenocortical carcinoma (ACC) patients n=1619
Excluded:
· Confirmed ACC histology: No/Unknown (n=86)
· Patients <18 yo (n=65)
· Autopsy or death certificate cases (n=2)
· Metastatic status: Yes/Unknown (n=588)
Non-metastatic ACC (nmACC) n=860
Excluded:
· Tx (n=33)
· T1 (n=74)
T 2-4 stage nmACC n=753
Excluded:
· Surgical treatment: No/Unknown (n=28)
Surgically treated patients with T 2-4 stage nmACC n=725
Excluded:
· Lymph node examined status: Unknown (n=15)
Surgically treated patients with T 2-4 stage nmACC and known lymph node examined status n=710
estimated annual percentage changes for LNC ≥4 and LNC≥5 in patients who underwent lymphadenectomy were quantified. Third, Kaplan-Meier plots displayed rates of CSM according to pN0 vs. pN1 in the overall cohort. Fourth, to fur- ther examine the correlation between pN-stage and CSM, a multivariable Cox regression models were fitted with adjust- ment variables consisting of: age at diagnosis, T stage, N stage, modality of treatment (surgery only vs. surgery plus systemic therapy and/or radiation [ST/RT]). Fifth, we tested the prognostic effect of LNC on CSM using two LNC cut- offs (≥4 and ≥5) in pN0 and pN1 subgroups. Finally, we ana- lyzed CSM Hazard ratios (HR) of pN0 and pN1 nmACC patients according to LNC status ( ≥4 and ≥5).
All statistical analyses were performed in the R software environment for statistical computing and graphics (version 4.2.2 for macOS; www.r-project.org) [13]. All tests were two sided, with a level of significance set at P < 0.05.
3. Results
3.1. Descriptive characteristics
Of 710 nmACC patients, 185 (26%) underwent lympha- denectomy with pathological lymph node examination. Of
185 patients with pathologically assessed lymph nodes, 152 (82%) patients were pN0 and 33 patients (18%) were pN1. In 185 pathological lymph nodes examined patients, 77 (41%) patients had four or more pathological lymph nodes examined and 64 (35%) patients had five or more pathologi- cal lymph nodes examined (Figure 2). The rates of both LNC ≥4 and ≥5 remained stable over the study period (P=0.9) (Figure 3).
Differences between pN0 vs. pN1 patients consisted of younger age for pN0 patients (pN0: 54 vs. pN1: 58 years, P=0.025) and larger primary tumor size for pN0 patients (pN0: 140 vs. pN1:120, P ≤ 0.001). Addi- tionally, pN1 patients exhibited higher rates of T4 stage (pN0: 23% vs. pN1:42% vs. pNx:13% , P ≤ 0.001), as well as higher rates of surgery plus ST/RT (pN0: 40% vs. pN1:64% vs. pNx:44% mm, P=0.047). In 185 patients who underwent pathological lymph nodes exam- ination, pN1 patients had a higher median LNC (pN0: 2 vs. pN1: 4 , P ≤ 0.017), as well as higher rates of LNC ≥4 (pN0: 39% vs. pN1: 55%, P ≤ 0.001) and LNC≥5 (pN0: 32% vs. pN1: 45%, P ≤ 0.001). Last but not least, no differences were recorded regarding ethnic- ity (P=0.9), gender (P=0.6), year of surgery (P=0.2), and tumor laterality (P=0.08) (Table 1).
40
34%
30
Percentage
20
17%
10
8%
7%
5%
3%
3%
2%
2%
3%
1%
1%
1%
1%
2%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
1%
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
24
26
27
28
34
35
36
37
38
40
Number of examined lymph nodes
100
06
Estimated Annual Percent Change (EAPC) for lymph nodes counts ≥4: - 0.12% (CI :- 4.5% to +4.3%, p=0.9)
80
Estimated Annual Percent Change (EAPC) for lymph nodes counts 25 :- 0.20% (CI :- 4.8% to +4.6%, p=0.9)
70
Rates of LNC ≥4 (%)
60
50
40
30
20
10
0
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Year of surgery
3.2. Survival analyses
The median CSM-free survival for the entire cohort regardless of the N-stage status was 77 months (95% confi- dence interval [CI]: 66-164 months). Median CSM-free sur- vival according to pN1 vs. pN0 was 14 vs. 74 months
(4 =60 months; P < 0.0001). Five-year-CSM-free survival was 18.5% vs. 55.4% for pN1 and pN0 patients, respec- tively (Figure 4).
In multivariable analyses, pN1 was an independent pre- dictor of higher CSM (HR:3.13, CI: 1.95-5.03, P < 0.001). Additionally, T3 (HR: 2.08 , P <0.001), T4 (HR: 1.90,
| Characteristic | N | Overall (N=710 )ª | pN0 N = 152 (21%)ª | pN1 N=33 (4.6%) | pNx N= 525 (74%)ª | P valueb |
|---|---|---|---|---|---|---|
| Age, median (IQR) | 710 | 56 (44, 66) | 54 (42, 63) | 58 (49, 61) | 57 (45, 68) | 0.025 |
| Ethnicity, n (%) | 710 | 0.9 | ||||
| Caucasian | 500 (70%) | 108 (71%) | 23 (70%) | 369 (70%) | ||
| Non-Caucasian | 210 (30%) | 44 (29%) | 10 (30%) | 156 (30%) | ||
| Gender, n (%) | 710 | 0.6 | ||||
| Female | 447 (63%) | 90 (59%) | 21 (64%) | 336 (64%) | ||
| Primary tumor size, median (IQR) | 676 | 108 (80, 145) | 140 (101, 188) | 120 (92, 148) | 100 (75, 135) | <0.001 |
| Tumor laterality, median (IQR) | 706 | 0.2 | ||||
| Left | 392 (56%) | 93 (62%) | 18 (55%) | 281 (54%) | ||
| T stage, n (%) | 710 | <0.001 | ||||
| T2 | 395 (56%) | 69 (45%) | 9 (27%) | 317 (60%) | ||
| T3 | 200 (28%) | 48 (32%) | 10 (30%) | 142 (27%) | ||
| T4 | 115 (16%) | 35 (23%) | 14 (42%) | 66 (13%) | ||
| Year of diagnosis, n (%) | 710 | 0.2 | ||||
| 2004-2011 | 303 (43%) | 61 (40%) | 19 (58%) | 223 (42%) | ||
| 2012-2020 | 407 (57%) | 91 (60%) | 14 (42%) | 302 (58%) | ||
| Modality of treatment, n (%) | 710 | 0.047 | ||||
| Surgery only | 399 (56%) | 91 (60%) | 12 (36%) | 296 (56%) | ||
| Surgery plus ST/RTC | 311 (44%) | 61 (40%) | 21 (64%) | 229 (44%) | ||
| Number of examined nodes, median (IQR) | 185 | 2 (1,7) | 2 (1,6) | 4 (2,9) | 0 (0%) | 0.017 |
| Lymph node count, n (%) | 710 | <0.001 | ||||
| 1-3 nodes | 108 (15%) | 93 (61%) | 15 (45%) | 0 (0%) | ||
| ≥4 nodes | 77 (11%) | 59 (39%) | 18 (55%) | 0 (0%) | ||
| ≥5 nodes | 64 (9%) | 49 (32%) | 15 (45%) | 0 (0%) |
ª Median (IQR); n (%).
b Kruskal-Wallis rank sum test; Pearson’s chi-square test.
” ST/RT: Systemic therapy and/or radiation therapy.
P < 0.001) , age (HR: 1.02, P <0.001) and contemporary year of surgery (HR: 0.75 , P=0.03) were also identified as independent predictors of CSM (Table 2).
3.3. Sensitivity analyses
3.3.1. Prognostic benefit of LNC cut-off ≥4 and ≥5 in pN0 patients
Of 152 pN0 patients, 59 (39%) patients had a LNC≥4. Median CSM according to LNC≥4 vs. LNC 1-3 was 125 vs. 58 months (4 =67 months; P < 0.015; Fig. 3). In multivariable analyses, LNC≥4 was associated with lower CSM (HR: 0.52, CI: 0.30-0.91, P=0.02) (Table 3).
Of 152 pN0 patients, 49 (31%) patients had a LNC≥5. Median CSM according to LNC ≥5 vs. LNC 1-4 was 125 vs. 64 months (4 = 61 months; P < 0.08; Fig. 3). No sta- tistically significant difference was recorded in univariable analyses (HR:0.60, CI: 0.34-1.06, P=0.08). Similarly, in multivariable analyses, LNC≥5 did not reach independent predictor status (HR:0.60, CI: 0.33-1.08, P=0.09 (Table 3).
3.3.2. Prognostic benefit of LNC cut-off ≥4 and ≥5 in pN1 patients
Of 33 pN1 patients, 18 (55%) patients had a LNC≥4. Median CSM according to LNC≥4 vs. LNC 1-3 was
14 vs. 29 months (4 =15 months; P < 0.1; Fig. 3). No statistically significant difference was recorded in uni- variable analyses (HR:1.90, CI:83-4.37, P=0.1). Simi- larly, in multivariable analyses, LNC≥4 did not reach independent predictor status (HR:1.69, CI:0.70-4.06, P=0.2).
Of 33 pN1 patients, 15 (45%) patients had a LNC ≥5. Median CSM according to LNC ≥5 vs. LNC 1-4 was 9.5 vs. 19 months (A=10.5 months; P < 0.2; Fig. 3). No statistically significant difference was recorded in univariable analyses (HR:1.77, CI:80-3.92, P=0.2). Similarly, in multivariable analyses, LNC≥4 did not reach independent predictor status (HR:1.51, CI:0.64-3.58, P=0.3).
3.3.3. Comparison of survival between pNx patients and pN0 patients with LNC 1-3 vs. LNC≥4
The median CSM-free survival for the pNx patients was 103 months (95% confidence interval [CI]: NA-69 months). Median CSM-free survival according to pNx vs. pN0 with LNC 1-3 vs. pN0 with LNC≥4 was 103 vs. 58 vs. 125 months ( P < 0.04). Five-year-CSM- free survival was 67.9% vs. 57.8% vs. 47.1% for pN0 with LNC≥4 , pNx and pN0 with LNC 1-3 patients, respectively.
a)
b)
Entire cohort
Entire cohort
CSM
CSM
- Entire cohort
- pNo - pN1
1.00
1.00
0.75
survival probability
0.75
survival probability
0.50
0.50
Log-rank
0.25
0.25
p < 0.0001
60 mo
0.00
0.00
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
0
4
8
12
16
20
24
28
32
36 40
44
48
52
56
60
64
68
72
76
80
84
Months
Months
Number at risk
Number at risk
pNo
152 141
133
123
111
99
86
84
78
73
65
58
55
53
51
46
43
42
38
35
34
32
Entre cohort
710 664 608 541 494 440 396 359 333 306 279 260 241 233 217 205 197 186 172 160 151 141
PN1
33
30
21
17
15
12
9
7
6
5
4
3
3
3
3
2
2
1
1
1
1
1
0
.
12
16
20
24
32
-
4
44
M
-
-
7
-
N
0
4
8
12
16
20
24
32
36
40
Months
14
48
52
56
60
68
72
5
80
54
Months
c)
d)
LNC 1-3 vs LNC 24 in pNO patients
LNC 1-4 vs LNC ≥5 in pNO patients
CSM
- LNC1-3 - LNC24
CSM
- LNC1-4 - LNC25
1.00
1.00
survival probability
0.75
survival probability
0.75
0.50
0.50
0.25
Log-rank
Log-rank
p = 0.015
0.25
67 mo
p = 0.077
61 mo
0.00
0.00
0
6
12
18
24
30
36
42
48
54
60
6 72 78
90
102 1 108
114
12
126
Months
0
6
12
18
24
30
36
42
48
54
60
66 72 78
VA
102 2 108
Months
90 96
8 1
12
126
Number at risk
Number at risk
LNCS-3
93
81
72
55
43
41
38
33
31
28
26
23
22
20
18
18
16
14
14
13
12
11
LNG1-4
103
91 81
64
52
49
44
37
35
32
29
25
24
22
20
20
18
15
15
14
13
12
INON
59
53
51
48
43
40
35
27
2
24
20
19
16
15
14
12
1
12
9
8
8
8
7
LNCAS
49
43
42
39
34
32
29
23
20
20
17
17
4
13
1
12
10
10
8
7
7
7
6
0 8 2 11 24 30 36 42 88 54 60 6 72 78 84 90 16 12 100 154 130 150
Vores
Months
| Median CSM-FS (CI 95%) | CSM-FS at 60 mo | |
|---|---|---|
| Entire cohort | 77 mo (66-164) | 55.3% |
| Median CSM-FS (CI 95%) | CSM-FS at 60 mo | |
|---|---|---|
| pNO | 74 mo (58-NA) | 55.4% |
| PN1 | 14mo (8-29) | 18.5% |
| Median CSM-FS (CI 95%) | CSM-FS at 60 mo | ||
|---|---|---|---|
| LNC 1-3 | 58 mo (23-NA) | 47% | |
| LNC ≥4 | 125 mo (74-NA) | 68% | |
| Median CSM-FS (CI 95%) | CSM-FS at 60 mo | |
|---|---|---|
| LNC 1-4 | 64 mo (30-NA) | 50% |
| LNC ≥5 | 125 mo (74-NA) | 66% |
Figure 4. Kaplan-Meier plots displaying the cancer-specific mortality-free survival (CSMFS) rates of patients with stage T2-4 non-metastaticc adrenocorti- cal>cal carcinoma (mACC) in the a) en>re cohort; b) according to pathological N-stage status in the en>re cohort; c) according to lymph node cut-off ≥4 sta- tus in pN0 patients; d) according to lymph node cut-off ≥5 status in pN0 patients.
4. Discussion
The optimal LNC cut-off for lymphadenectomy in nmACC is unclear. We addressed this knowledge gap in a very large contemporary population of nmACC patients. We recorded several noteworthy observations.
First, Of 710 surgically treated nmACC patients, 185 (26%) underwent lymphadenectomy and were assessable for further analyses according to LNC criteria, as described in the methodology of previous NCDB studies [9,11]. The rate is comparable to previous historical series. For exam- ple, Tran et al. based on a previous SEER study (1988-
2009, n = 320), Reibetanz et al. based on the German ACC registry, (GACCR, 1981-2009, n =283, as well as Gerry et al. based on U.S. Adrenocortical Carcinoma Study Group database (ACSG,1993-2014, n = 120) respectively reported pathological lymph node examination rates of 29%, 30%, and 34% [4,6,8,14]. However, the rate observed in the cur- rent study was higher than those reported in previous histor- ical studies that relied on NCDB. Specifically, Tella et al. (NCDB, 2004-2015, n = 842), Panjwani et al. (2004-2013, n = 827), and Deschner et al. (2004-2015, n = 897) respec- tively reported pathological lymph node examination rates of 18%, 19%, and 16% [9-11]. This discrepancy can be
| Variable | Multivariable analysis | ||
|---|---|---|---|
| Hazard ratio | 95% CI | P value | |
| Pathological N stage status | |||
| pN0 | - | - | - |
| pN1 | 3.13 | 1.95-5.03 | <0.001 |
| pNx | 1.02 | 0.76-1.37 | 0.9 |
| T stage | |||
| T2 | - | - | - |
| T3 | 2.08 | 1.59-2.73 | <0.001 |
| T4 | 1.90 | 1.36-2.65 | <0.001 |
| Age | 1.02 | 1.01-1.03 | <0.001 |
| Modality of treatment | |||
| Surgery only | - | - | - |
| Surgery plus ST/RT | 1.19 | 0.92-1.53 | 0.2 |
| Year of diagnosis | |||
| Historical (2004-2011) | - | - | - |
| Contemporary (2012-2020) | 0.75 | 0.59-0.97 | 0.03 |
Adjusted for : Age, gender, year of diagnosis, T stage, and modality of treatment status.
attributed to a large proportion of patients being excluded due to unknown ENSAT stage information within the NCDB cohort. An additional explanation for the discrep- ancy in pathological lymph node examination rates may be related to differences in T stage consideration for inclusion and/or exclusion. For example, Panjwani et al. excluded patients with T4 stage in whom lymphadenectomy is more frequently performed [11]. Similarly, previous studies that relied on NCDB included T1 stage patients in whom lym- phadenectomy is infrequently performed [9-11]. Taken together, the above findings emphasize the low rate of path- ological lymph node examination in surgically treated nmACC. This observation is worrisome as it reflects a
deviation from the recommended standard of care for nmACC patients, and is considered a negative indicator of quality of care [1-3].
Second, of 185 patients who underwent pathological lymph node examination, 33 (18%) patients were classified as pN1. This rate was comparable to those reported in pre- vious large-scale population-based studies. For example, in 147 patients who underwent pathological lymph node examination, Dreschner et al. reported a 16.3% pN1 rate in nmACC patients [9]. Similarly, in 156 patients who under- went lymph node examination, Panjwani et al. reported a 22.4% pN1 rate in nmACC patients [11]. Interestingly, in the current study, a higher proportion of pN1 patients had ≥4 lymph nodes examined than in pN0 patients (39% vs. 55%, P ≤ 0.001). In summary, the similarity of pN1 rates in the current study and other SEER and NCDB studies vali- dates the recorded and reported pN1 rates.
Third, we tested for CSM differences in 185 patients who underwent pathological lymph node examination according to presence or absence of pN1 stage. Median CSM-free survival according to pN1 vs. pN0 was 14 vs. 74 months (4 =60 months; P < 0.0001). Resulting five-year- CSM-free survival of 18.5% vs. 55.4% for pN1 and pN0 patients, respectively. Additionally, in multivariable analy- ses, pN1 was an independent predictor of higher CSM with a HR of 3.13 (P ≤ 0.001). The higher CSM rates in pN1 patients recorded in the current study is in agreement with previous studies that relied on NCDB (P < 0.02), SEER (P=0.02), and GACCR( P=0.058) databases [4,9,11,14]. Moreover, the independent predictor status of pN1 recorded in the current study is in agreement with previous studies that relied on NCDB, (Panjwani et al. [HR: 3.02 , P = 0.02] and Deschner et al. [HR: 2.00, P = 0.02]) [9,11]. Compared to other studies, the current study is based on the largest and most contemporary population of nmACC patient who underwent pathological lymph node examination.
| Univariable analyses | Multivariable analyses | |||||
|---|---|---|---|---|---|---|
| Variable | Hazard ratio | 95% CI | P value | Hazard ratio | 95% CI | P value |
| Analyses with LNC threshold: ≥4 nodes In pN0 population | ||||||
| LNC | ||||||
| LNC 1-3 | - | - | - | |||
| LNC≥4 | 0.51 | 0.30-0.88 | 0.02 | 0.52 | 0.30-0.91 | 0.02 |
| Analyses with LNC threshold: ≥5 nodes In pN0 population | ||||||
| LNC | ||||||
| LNC 1-4 | - | - | - | |||
| LNC≥5 | 0.60 | 0.34-1.06 | 0.08 | 0.60 | 0.33-1.08 | 0.09 |
Adjusted for : Age, gender, year of diagnosis, T stage, modality of treatment status and lymph node count cut-off.
Fourth, we also tested whether previously recommended LNC cut-off (≥4 and ≥5) may add prognostic discrimina- tion in either pN0 or pN1 patients. We found that LNC≥4 improved prognostic ability in stratifying patients based on their pathological N-stage status. Specifically, in pN0 patients, LNC cut-off ≥4 was an independent predictor sta- tus of lower CSM (HR: 0.52, P=0.02). To the best of our knowledge the current study provides first evidence of inde- pendent predictor status where LNC≥4 pN0 patients exhibit significantly lower CSM. Previous studies addressing LNC≥4 where limited in their endpoints since only overall mortality (OM) was available but not CSM. In those OM endpoints-based analyses, LNC≥4 pN0 patients exhibited more favorably OM [11]. However, the statistical signifi- cant of more favorable OM, was only based on univariable analyses, no multivariable based results were reported. In consequence, the current study provides the most robust proof that lymphadenectomy with LNC≥4 adds the ability to discriminate between patients with very favorably low CSM vs. others. LNC≥5 failed to exhibit statistically signif- icant CSM differences, as well as failed to achieve indepen- dent predictor status. In consequence, LNC≥4 should be encouraged in clinical practice and when it is achieved, it may also be used in prognostic consideration in pN0 patients.
Fifth, testing of LNC cut-offs in pN1 neither reached sta- tistical significance, nor achieved independent predictor sta- tus when either LNC≥4 or LNC≥5 represented candidate predictors. In consequence, CSM-free survival of pN1 patients is invariably poor and additional LNC considera- tions appear of limited value and interest.
The therapeutic benefit of lymphadenectomy in nmACC remains uncertain, with conflicting results observed across several studies [4,5,8,9]. These discrepancies may be attrib- uted to variations in definitions and inclusion criteria for lymphadenectomy. In a recent meta-analysis conducted by Hendricks et al., studies including stage I-III ACC patients demonstrated a survival advantage from lymphadenectomy, while those involving stage I-IV ACC patients did not show a survival benefit [6]. Another contributing factor is that most studies are retrospective in nature, which exposes them to potential selection bias. While lymphadenectomy is recommended for suspected ACC, studies indicate that patients with smaller tumors, T1-2 stage, and those under- going laparoscopic or robotic surgery resection are less likely to undergo lymphadenectomy [4,8,9]. Indeed, we observed the same pattern in our study as patients with pNx status displayed less aggressive tumor features compared to those with pN0 and pN1, including smaller tumor size (100 vs. 140 vs. 120 mm, P < 0.001) and a higher propor- tion of T2 stage (60% vs. 45% vs. 27%, P < 0.001). How- ever, our findings emphasize the prognostic advantages of lymph node examination, offering essential insights for patient counseling. When compared to pNx patients and pN0 with LNC 1-3, those with pN0 with LNC≥4 exhibited
higher overall survival (125 vs. 103 vs. 58 months; P < 0.04). In this context, pNx patients represent a heteroge- neous group, which presents challenges in interpretation. Therefore, accurate surgical lymph node staging is pivotal for evaluating patient prognosis and may impact decisions regarding systemic and/or radiation therapy.
The current study is not devoid of limitations. First, despite strict statistical analyses potential biases may still be operational due to its retrospective design. Indeed, inher- ent selection biases can be challenging to control, and other unknown variables may have influenced the decision to per- form lymphadenectomy, such as the patient’s performance status, technical considerations, tumor burden, level of pre- operative suspicion of malignancy and surgeon expertise. Second, SEER database lacks information on potential prognostic factors, including mitotic index, venous inva- sion, surgical margin status, and endocrine tumor activity. These details are crucial in a clinical setting to guide adju- vant treatment decisions. For instance, adjuvant systemic therapy and radiation therapy are often recommended in nmACC cases with a ki-67 >20%, vascular invasion, or intraoperative tumor spillage, as these features are associ- ated with a higher risk of recurrence and progression [3,15]. However, these limitations apply to all population-based analyses that were based on NCDB, SEER, GACCR, or other population-based data repositories. Third, limited details regarding the type of treatment were available. Spe- cifically, SEER database does not provide information regarding specific systemic therapies used as well as the timing of adrenalectomy (adjuvant vs. neoadjuvant). While previous studies using other population, databases have demonstrated that open surgery was associated with a higher rate of lymphadenectomy and that lymphadenec- tomy does not increase morbidity, these variables were not assessable in the SEER database [4,6,9,10]. Therefore, the observational nature of studies available to assess the effi- cacy of LNC may be subject to selection bias, limiting the ability to establish causality between LNC status and improved survival. Unfortunately, no randomized clinical trial has been conducted or is likely to be conducted due to the rarity of the disease. Therefore, these results must be interpreted within the boundaries of such limitations.
5. Conclusions
Lymphadenectomy provides a prognostic benefit in nmACC patients and identifies pN1 patients with dismal prognosis. Conversely, in pN0 patients, a LNC cut-off ≥4 identifies those with particularly favorable prognosis.
CRediT authorship contribution statement
Anis Assad: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Software, Supervision, Validation,
Visualization, Writing - original draft, Writing - review & editing. Francesco Barletta: Conceptualization, Data cura- tion, Formal analysis, Writing - original draft. Reha-Baris Incesu: Conceptualization, Data curation, Formal analysis, Writing - original draft, Writing - review & editing. Lukas Scheipner: Conceptualization, Data curation, For- mal analysis, Writing - original draft, Writing - review & editing. Simone Morra: Conceptualization, Data curation, Formal analysis, Writing - original draft. Andrea Baudo: Conceptualization, Data curation, Formal analysis, Writing - original draft. Cristina Cano Garcia: Conceptualization, Data curation, Formal analysis, Writing - original draft. Zhe Tian: Data curation, Formal analysis, Methodology, Software. Sascha Ahyai: Methodology, Supervision, Vali- dation, Writing - original draft. Nicola Longo: Methodol- ogy, Supervision, Writing - original draft. Felix K.H. Chun: Methodology, Supervision, Writing - original draft. Shahrokh F. Shariat: Methodology, Supervision, Writing - original draft. Derya Tilki: Methodology, Supervision, Writing - original draft. Alberto Briganti: Methodology, Supervision, Writing - original draft. Fred Saad: Method- ology, Supervision, Writing - original draft. Pierre I. Kar- akiewicz: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing - original draft, Writing - review & editing.
Declaration of generative AI and AI-assisted technologies in the writing process
None.
Data availability statement
All data generated for analyses were from the Surveil- lance, Epidemiology, and End Results Research Plus (SEER) database. The code for the analyses will be made available upon request.
Declaration of competing interest
The authors declare that there is no conflict of interests.
Ethics consent statement
All analyses and their reporting followed the SEER reporting guidelines. Due to the anonymously coded design of the SEER database, study-specific Institutional Review Board ethics approval was not required.
Patient consent statement, permission to reproduce material from other sources and clinical trial registration
Not applicable.
Acknowledgments
The authors have no acknowledgments.
References
[1] Else T, Kim AC, Sabolch A, et al. Adrenocortical carcinoma. Endocr Rev 2014;35(2):282-326.
[2] Fassnacht M, Dekkers O, Else T, et al. European Society of Endocri- nology Clinical Practice Guidelines on the management of adrenocor- tical carcinoma in adults, in collaboration with the European Network for the Study of Adrenal Tumors. Eur J Endocrinol 2018;179(4):G1- G46.
[3] Shah MH, Goldner WS, Benson AB, et al. Neuroendocrine and adre- nal tumors, version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2021;19(7):839-68.
[4] Reibetanz J, Jurowich C, Erdogan I, et al. Impact of lymphadenec- tomy on the oncologic outcome of patients with adrenocortical carci- noma. Ann Surg 2012;255(2):363-9.
[5] Icard P, Goudet P, Charpenay C, et al. Adrenocortical carcinomas: surgical trends and results of a 253-patient series from the French Association of Endocrine Surgeons study group. World J Surg 2001;25(7):891-7.
[6] Hendricks A, Müller S, Fassnacht M, et al. Impact of lymphadenec- tomy on the oncologic outcome of patients with adrenocortical carci- noma-a systematic review and meta-analysis. Cancers 2022;14 (2):291.
[7] Nilubol N, Patel D, Kebebew E. Does Lymphadenectomy Improve Survival in Patients with Adrenocortical Carcinoma? A Population- Based Study. World J Surg 2016;40(3):697-705.
[8] Gerry JM, Tran TB, Postlewait LM, et al. Lymphadenectomy for adrenocortical carcinoma: is there a therapeutic benefit? Ann Surg Oncol 2016;23(Suppl 5):708-13.
[9] Deschner BW, Stiles ZE, DeLozier OM, et al. Critical analysis of lymph node examination in patients undergoing curative-intent resec- tion for adrenocortical carcinoma. J Surg Oncol 2020;122(6):1152- 62.
[10] Tella SH, Kommalapati A, Yaturu S, Kebebew E. Predictors of sur- vival in adrenocortical carcinoma: an analysis from the national can- cer database. J Clin Endocrinol Metab 2018;103(9):3566-73.
[11] Panjwani S, Moore MD, Gray KD, et al. The impact of nodal dissec- tion on staging in adrenocortical carcinoma. Ann Surg Oncol 2017;24(12):3617-23.
[12] Overview of the SEER program. Accessed September 10, 2022. https://seer.cancer.gov/about/overview.html
[13] The R project for statistical computing. Accessed March 10, 2023. https://www.r-project.org/
[14] Tran TB, Liou D, Menon VG, Nissen NN. Surgical management of advanced adrenocortical carcinoma: a 21-year population-based anal- ysis. Am Surg 2013;79(10):1115-8.
[15] Fassnacht M, Assie G, Baudin E, et al. Adrenocortical carcinomas and malignant phaeochromocytomas: ESMO-EURACAN Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2020;31(11):1476-90.