Introduction

Adrenocortical cancer (ACC) is a rare tumour with an incidence of 1–2/million population/year. [1] Complete surgical resection is the backbone of treatment for localised disease, with reported 5-year survival rates after curative intent resection ranging from 40 to 51%. [2,3,4] Factors currently commonly used for survival estimates are clinical parameters, such as hormonal functionality of the tumour [5, 6] and age of the patient [7], or pathological variables, such as stage of the tumour [8], proliferation index of the tumour (e.g. the Weiss score) [9] and Ki67 immunostaining. [10] However, the cohort of patients with ACC as a whole is very heterogeneous and long-term outcomes may therefore differ even among patients within the same tumour stage, due to the largely still unknown impact of different clinical, pathological and molecular factors [11].

Traditional risk factors are currently unable to stratify patients satisfactorily to a certain prognostic group; thus, molecular and genetic biomarkers have been proposed as potential new predictive markers. [12,13,14] However, these have not yet been proved to be feasible in routine clinical practice as many of these techniques remain restricted to research laboratories.

In recent years, neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) have been reported to predict oncologic outcomes for various other types of carcinomas such as gastric, colorectal, lung, pancreas, ovarian, hepatocellular and oesophageal. [15,16,17,18,19,20] Conversely to molecular testing, these assays are inexpensive and readily available for all patients to be used as inflammatory- and immunologic-based investigations.

Inflammation is a known major driver for the development and progression among various types of cancer. The immune cells most commonly associated with progression of disease include neutrophils and platelets. [21] The hypothesis is that a higher NLR and PLR are associated with a worse oncological outcome. There are presently limited data on this association between NLR and PLR and oncological outcomes for patients who underwent surgery for ACC. [22, 23] The aim of the current analysis was to assess the prognostic impact of baseline NLR and PLR for patients who underwent resection of their ACC.

Methods

All patients (n = 57) who underwent a resection of their ACC between 2000 and 2020 at Oxford University Hospital Trust (Oxford, UK) and for whom a complete blood cell count with differential count within a week prior to the operation was available were included. All data were prospectively collected in a database and retrospectively analysed as part of an ongoing audit on outcomes after surgery for endocrine tumours.

Management decisions

At time of diagnosis, all patients were formally discussed in our weekly multidisciplinary meeting. This meeting was attended by endocrine surgeons, oncologists, radiologist, pathologists, nuclear physicians and endocrinologists. All patients underwent a full endocrine assessment, and all patients also had preoperative imaging studies done, in the form of a thoraco-abdominal Computed Tomography (CT) scan, while more recent patients also underwent a preoperative 18F-FDG positron emission tomography (PET) scan. All tumour staging was decided in concordance with the international guidelines published by the European Network for Study of Adrenal Tumours (ENSAT) [8].

Open radical adrenalectomy with en bloc resection of the tumour and ipsilateral kidney was the standard surgical approach in patients who had a preoperative diagnosis of ACC, with further decisions about multi-organ resection made on a case-by-case basis. Laparoscopic surgery was performed only when the presenting diagnosis of incidentaloma was made in patients with a tumour of < 6 cm and the diagnosis of ACC was proved only postoperatively.

To make decisions on postoperative oncological treatment and follow-up, all patients were again discussed in our multidisciplinary meeting. While adjuvant therapy with mitotane was considered for all patients with ENSAT-stage 1 or 2, patients with a tumour staged as ENSAT 3 or 4 were offered to start mitotane treatment within 4–8 weeks after their operation. These patients had their plasma levels regularly monitored (target level 14–20 mg/dL). Our follow-up protocol consisted of cross-sectional imaging by CT scan (with or without a 18F-FDG-PET scan), according to clinical protocols. Recurrence of disease was defined as either radiological evidence of local or distant disease on cross-sectional imaging or on functional studies.

Data collected

Standard demographic (age, sex), clinical (site and functionality of the tumour), biochemical (neutrophil count in 109/L, lymphocyte count in 109/L and platelet count in 109/L within a week prior to surgery) and operative data (type and extent of surgery) were collected. Moreover, histological variables (size of the tumour, ki-67 index, [10] resection margin and TNM status [24]) were noted. Short-term outcomes (length of stay and development of post-operative morbidity [classified according to minor (Dindo–Clavien grade < 3) vs. major (Dindo–Clavien grade ≥ 3) [25] as well as oncological outcomes (recurrence of disease and overall survival) were recorded].

Statistical analyses

Summary statistics were obtained and presented as percentages or median values. Upon comparing categorical data, the Chi-square test, or if deemed appropriate Fisher's exact test, was used, while the Mann–Whitney U-test was used to compare continuous data.

NLR was calculated by division of the absolute neutrophil and lymphocyte counts measured in peripheral blood before surgery. PLR was calculated by division of thrombocytes and lymphocytes accordingly. For descriptive statistics, patients were subdivided into two groups after performance of a median split, the low group below the median and the high group above.

Factors associated with overall and recurrence free survival were examined using cross-tabs and the nonparametric product limit method. Cox proportional hazards models were developed using relevant clinicopathologic variables to determine the association of each with survival. Relative risks were expressed as hazard ratios (HRs) with 95% CIs. Overall, a p-value of less than 0.05 was considered significant. All statistical analyses were performed using IBM SPSS Statistics for Macintosh, Version 23.0 (IBM Corp. IMB SPSS statistics, Armonk, NY).

Results

Patient characteristics

The demographic, clinical and peri-operative data of the 57 patients who were included in the current analysis are detailed in Table 1. The median age at time of diagnosis was 53 years [range: 18–86]. In almost half of patients, the tumour was hormonally active (n = 26; 46%), with 15 patients (58%) having a cortisol-producing tumour.

Table 1 Demographic, clinical and operative characteristics of all patients included
Full size table

Neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR)

The median NLR for the whole cohort was 4.63, and the median PLR was 186.21. Table 2 details the demographic, clinical and operative characteristics of all patients stratified by patients NLR and PLR, grouped according to medial split. There are no differences with regard to age, sex or tumour characteristics and peri-operative data between the two groups (all p > 0.05).

Table 2 Demographic, clinical and operative characteristics of all patients included stratified by patients’ neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR)
Full size table

Details of surgery and histopathology

The majority of patients underwent an open resection of their ACC (n = 48; 84%). More details on the surgical procedure are summarised in Table 1.

Unfortunately, the N-stage of most patients was unknown [i.e. Nx; n = 37 (65%)], while 12 patients had no signs of lymph node metastases [i.e. N0; n = 12 (21%)], and in eight patients, metastases in lymph nodes were confirmed [n = 8 (14%)].

Short-term outcomes

The median length of stay was 7 days [range: 1–67]. Two patients (4%) died during their hospital admission after treatment for severe malignant Cushing syndrome. One patient died on post-operative day 36 after undergoing an open bilateral adrenalectomy and splenectomy with lymph node dissection. This patient developed intra-abdominal collections and subsequent sepsis and succumbed to multi-organ failure. Another patient died 45 days after undergoing an open left-sided adrenalectomy plus nephrectomy plus splenectomy and extraction of the tumour from the inferior vena cava through bypass surgery. They suffered a neurological deficit attributed to prolonged bypass use and died of airway complications after tracheostomy. Further details on post-operative morbidity are detailed in Table 1.

When looking at the NLR, there was no association found between a high ratio and the development of post-operative morbidity (p = 0.29), while among those with a high PLR the incidence of complications bordered on being significantly higher (p = 0.05). There was no increased morbidity when looking at sex or age, the hormonal status of the tumour or the location of the tumour (all p > 0.05).

Long-term outcomes and predictors of survival

At time of last follow-up, 28 (49%) patients had developed recurrent disease after a median recurrence-free survival (RFS) of 26 months, with a 1-year RFS of 70% and a 3-year RFS of 45%. While six patients (21%) developed only local recurrence of their disease, 19 patients (68%) developed recurrent disease at one or more distant locations (e.g. lung or liver) and two patients (7%) developed both local and distant recurrence of disease. A total of 12 patients (43%) underwent repeat resection of their recurrent ACC.

Upon exploring factors associated with an increased risk of recurrence of disease (Table 3), a high NLR did not seem to influence the risk of recurrent disease (HR = 1.57 [95%-CI: 0.73–2.38]; p = 0.25; Fig. 1a). Conversely, patients with a high PLR before surgery had an increased risk of developing recurrence of their disease, with patients with a higher than median PLR being at a more than two-fold increased risk [HR = 2.39 (95%-CI: 1.08–5.31); p = 0.03; Fig. 1b]. Other factors found to be associated with an increased risk of developing recurrent disease on univariate analyses were T-stage of 3 or 4 [HR = 2.72 (95%-CI: 1.23–5.99); p = 0.01] and the presence of capsular invasion on histological examination [HR = 2.39 (95%-CI: 1.09–5.22); p = 0.03]. Due to the small number of included patients, controlling for competing factors by performance of a multivariate analysis was not deemed feasible. Examining the subgroup of patients without metastatic disease, there was no difference in RFS when stratifying patients by NLR (p = 0.28) or PLR (p = 0.06).

Table 3 Unadjusted analyses of factors proposed to be associated with recurrence-free and overall survival after resection of ACC
Full size table
Fig. 1
figure 1

Kaplan–Meier curves comparing disease-free survival stratified by NLR (a.) and PLR (b.)

Full size image

The median overall survival (OS) for all patients was 33 months, with a 1-year OS of 79% and a 3-year OS of 49%. Table 3 displays the factors associated with OS for the whole cohort of patients. Specifically, patients within the high NLR group had a more than twofold increased risk of an early death [HR = 2.24 (95%-CI: 1.07–4.70); p = 0.03; Fig. 2a]. Moreover, when looking at the potency of the PLR as a prognostic factor, for patients with a higher than median PLR, the risk of an early death was more than four times higher [HR = 4.02 (95%-CI: 1.80–8.98); p = 0.001; Fig. 2b]. Other factors deemed to be associated with a shortened overall survival on univariate analyses were age at time of diagnosis [HR = 1.03 (95%-CI: 1.01–1.05); p = 0.006], the presence of local invasion or invasion into adjacent organs (i.e. T-stage of 3 or 4) [HR = 3.35 (95%-CI: 1.50–7.51); p = 0.003] or of positive lymph nodes (i.e. N1-status) [HR = 4.10 (95%-CI: 1.05–15.95); p = 0.04], signs of capsular invasion on histological examination [HR = 2.46 (95%-CI: 1.18–5.12); p = 0.02] or a positive resection margin (i.e. R1 or R2) [HR = 2.47 (95%-CI: 1.20–5.09); p = 0.02]. Multivariate analysis to exclude confounding was not deemed possible due to the limited sample size. For patients who did not have metastatic disease present at the time of surgery, their OS was shorter for patients with higher NLR (1-year OS of 74% and a 3-year OS of 33%) compared with low NLR (1-year OS of 90% and a 3-year OS of 74%; p = 0.024). Similarly, the OS was found to be shorter for those patients with a high PLR (1-year OS of 67% and 3-year OS of 25%) compared with low PLR (1-year OS of 96% and a 3-year OS of 72%; p < 0.001).

Fig. 2
figure 2

Kaplan–Meier curves comparing overall survival stratified by NLR (a.) and PLR (b.)

Full size image

Discussion

Adrenocortical carcinoma (ACC) is one of the most aggressive endocrine tumours. Due to its low incidence and possibly to the concurrent lack of centralised treatment for this patient group, robust risk factors to predict a worse long-term outcome have not yet been identified. [26] While the incidence of ACC seems to be rising [27, 28] and traditionally used variables have proved to be unable to stratify patients towards reliable survival groups, there is a need for more reliable prognostic markers. With molecular testing currently not readily available to be used in clinical settings, we explored the prognostic role of pre-operative NLR and PLR as potential cheap and widely available markers. Only two studies have addressed this question before in ACC. The first reported on 84 patients who underwent surgery for ACC at several institutions; 5-year RFS was lower when NLR and PLR were elevated and NLR was associated with decreased diseases-specific survival (HR 2.21) as well as and RFS (HR 1.99). [22] Others confirmed that the higher NLR group (NLR > 5) showed a significantly poorer overall survival than the lower NLR group. [23] We aimed to validate the findings from these previous two reports in a relatively large, single-centre cohort of patients undergoing curative intent surgery for ACC in order to increase the overall interest in the significance of using NLR and PLR as prognostic markers for this group of patients and to perhaps promote future prospective studies into its clinical implementation.

We report on a fairly typical cohort of patients undergoing curative intent surgery for their ACC, [26] including 57 patients with a median age of 53 years and a male-to-female ratio of almost 1:1.

About half of patients developed post-operative morbidity, while two patients died during their admission (4%). Peri-operative complications following surgery for ACC are relatively common, probably largely due to the extent of the surgery undertaken. [29, 30] While in our cohort, a higher NLR or PLR were not significantly increasing the incidence of morbidity, other groups have reported on this association in other types of diseases and noted that higher ratios of either of both increased the risk of developing systemic inflammatory response syndrome (SIRS) or sepsis. [31, 32] Although we could not demonstrate this in the current study, this might be an important finding, as Margonis et al. [30] reported that occurrence of postoperative complications after resection of ACC is an independent risk factor for a decreased long-term outcome. Therefore, the prevention of post-operative complications might be an important factor from an oncologic perspective and being able to stratify patients into a high-risk group before their surgery could therefore be beneficial in the longer term.

Overall median RFS was 26 months, with a 1-year RFS of 70% and a 3-year RFS of 45%, and a total of 12 patients (43%) underwent repeat resection of their recurrent disease. These statistics are similar to those reported by other groups who also described the feasibility of repeat surgery in select patients. [33,34,35] The median OS for all patients was 33 months, with a 1-year OS of 79% and a 3-year OS of 49% as corroborated by survival analyses in other cohorts. [8, 36,37,38] While a higher NLR did not seem to significantly influence the risk of recurrent disease, patients with a high PLR before surgery had a more than two-fold increased risk of developing recurrence of their disease. For OS, this association seemed to be even more profound, as the risk of an early death was increased more than twofold for patients within the high NLR and more than four times higher for those in the high PLR group. Similarly, an increasing number of studies in the literature have revealed that NLR and PLR indeed act as significant prognostic factors for both disease recurrence and overall survival in a variety of cancer types [15,16,17,18,19,20, 39, 40].

Although the exact molecular pathways through which the NLR and PLR are associated with poorer outcomes remain unelucidated, several hypotheses have been, mainly based on imbalances in the immune response to the tumour cells. [39, 40] Circulating cytokines and chemokines released by the malignant cells of solid tumours trigger a systemic inflammatory response, resulting in systematic alterations, such as an increase in neutrophils and platelets and a decrease in lymphocytes. [41] The subsequent occurrence of neutrophilia inhibits the cytolytic activity of certain immune cells, such as lymphocytes, activated T-cells and natural killer cells. [42] Moreover, neutrophils have shown to enhance angiogenesis and to promote primary tumour progression. [43] Additionally, platelets not only aid in the growth of primary tumour cells via angiogenesis, but also assist the tumour cells in evading the host immune system by protecting them from tumour lysis mediated by natural killer cells. [44,45,46] On the other hand, for patients with a higher NLR or PLR, their lymphocyte count is relatively low. Lymphocytes play an important role in the cell-mediated anti-tumour immune response. Several studies on different types of cancer state that lymphopenia is a predictive factor for a shorter RFS and OS, [47, 48] with others suggesting an improved outcome for those with an increased lymphocyte infiltration in tumours [49].

Conversely, some authors recently suggested that its impact on oncological outcomes is merely due to the relationship between the NLR or PLR and adverse certain tumour characteristics. [50, 51] While we could not fully exclude confounding by other variables as we were limited by our sample size to perform any adjusted analyses, we did not find any differences in patient and tumour variables between the low and high NLR or PLR groups, indicating NLR and PLR to act as surrogates for other characteristics.

The proliferation marker ki-67 index has been shown to be an independent prognostic factor for survival among patients undergoing resection for ACC. [10] This finding could be reproduced in our current analysis. The ki-67 index was, however, not consistently reported for patients included and was therefore only available in 24 patients (42%). Moreover, only the subgroups of an index of < 10% and ≥ 10% could thus be formed. This limited availability of the ki-67 results is most probably due to the ongoing reluctance of several pathologists to provide an exact value for this variable. This is presumably based on the known variability among the current practices for ki-67 scoring assessments, as interobserver variations set particular limitations around clinically relevant cut-off values. [52] The statistical power of our analysis involving this parameter could therefore be interpreted as non-robust and hence thus unable to demonstrate a significant correlation of the ki-67 index with survival in our cohort.

The present study has several limitations associated with its retrospective nature such as that we had to rely on the accuracy and completeness of the recordkeeping. Moreover, although our study has one of the largest numbers of patients from a single institution, the included number of patients is still relatively low due to the rarity of the disease. While this, on one hand, only stresses the highly selected nature of the cohort of patients with ACC who are considered for resection, the small sample size has statistical disadvantages. Specifically, we were unable to properly control for competing factor and exclude confounding, as multivariate analyses were not deemed appropriate due to the relatively small number of patients in each group or unavailability of necessary data.

While both NLR and PLR are variables available at the beginning of the decision-making process, they could potentially be used to stratify patients into certain risk groups and influence treatment and follow-up protocols. However, this potential feasibility of incorporating NLR and PLR in the decision-making process has not been assessed in the current study, but this could be the scope of a further prospective study. It is, nonetheless, likely that other parameters which are currently not widely available or in use yet (e.g. molecular testing and gene profiling) could prove to be much stronger predictors of long-term outcomes for patients undergoing curative intent surgical resection of their ACC and that these parameters should therefore be used to guide decision-making on the intensity of treatment and/or follow-up.

In summary, for patients undergoing curative intent surgical resection of their ACC both a higher pre-operative NLR and PLR strongly correlated with a diminished overall survival, while a higher PLR also predicted the recurrence of disease. This proposes NLR and PLR as prognostic markers which are inexpensive and readily available and could easily add additional prognostic information for patients undergoing surgery for ACC. This hypothesis should be validated in a multicentre study.