S‐100B as an extra selection tool for FDG PET/CT scanning in follow‐up of AJCC stage III melanoma patients

Abstract Background and Objectives This current study assessed the value of S‐100B measurement to guide fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) scanning for detecting recurrent disease in stage III melanoma patients. Methods This study included 100 stage III melanoma patients in follow‐up after curative lymph node dissection. Follow‐up visits included physical examination and S‐100B monitoring. FDG PET/CT scanning was indicated by clinical symptoms and/or elevated S‐100B. Results Of 100 patients, 13 (13%) had elevated S‐100B without clinical symptoms, of whom 7 (54%) showed disease evidence upon FDG PET/CT scanning. Twenty‐six patients (26%) had clinical symptoms with normal S‐100B and FDG PET/CT revealed metastasis in 20 (77%). Three patients had clinical symptoms and elevated S‐100B, and FDG PET/CT revealed metastasis in all three (100%). Overall, FDG PET/CT scanning revealed metastasis in 30 of the 42 patients (71.4%). For seven recurrences, elevated S‐100B prompted early detection of asymptomatic disease; 10% of all asymptomatic patients in follow‐up, 23% of all patients with recurrent disease. Conclusion S‐100B cannot exclude recurrent disease during follow‐up of stage III melanoma. However, adding S‐100B measurement to standard clinical assessment can guide FDG PET/CT scanning for detecting recurrent melanoma.


| INTRODUCTION
The incidence of cutaneous melanoma has increased worldwide over recent decades. 1 In the Netherlands, 1563 new cases were diagnosed in 1990, and this number grew to 6743 in 2017. 2 Mortality has increased at a lower rate, with 348 melanoma-related deaths in 1990 in the Netherlands, and 767 in 2016. The lower rise in mortality is because the increased incidence largely involves more cases of thin melanoma, likely due to improved awareness and earlier melanoma detection. 1,3 In melanoma patients, the goal of follow-up surveillance is the cost-effective detection of recurrence at an early stage, based on the assumption that early surgical and/or systemic treatment will improve disease-free survival (DFS), melanoma-specific survival (MSS), and overall survival (OS). There are no clinical data to support this assumption. Until now, data on the effectiveness of routine imaging for recurrence detection in follow-up is limited. Data with respect to an impact on the quality of life in melanoma patients with intensive follow-up schedules are lacking. 4 The melanoma biomarker S-100B reportedly shows strong correlations with distant metastasis-free survival and OS in stage IIB-III melanoma patients. 5 The serum concentration of S-100B is correlated with disease stage, and S-100B is an independent predictor of melanoma prognosis in patients undergoing therapeutic lymph node dissection (TLND) for nodal macro-metastases. 6,7 German melanoma follow-up guidelines added the melanoma biomarker S-100B and Italian guidelines added both S-100B and fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) scanning, in addition to regular patient history and physical examination. 8,9 Specifically, S-100B measurement has been recommended for use in some follow-up guidelines in the selection of stage III patients to undergo FDG PET/CT scanning. However, the added value of this screening is unknown. 10,11 Assessment of the melanoma marker could potentially contribute to the detection of asymptomatic disease recurrence in stage III melanoma, and therewith reduce the number of routine FDG PET/CT scans. As long as scientific data on the effect of standard scanning regimens are lacking, a strategy using a biomarker as a trigger for scanning in asymptomatic patients could be an interesting alternative.
In the present study, we primarily aimed to assess the added value of the biomarker S-100B as a selection tool before FDG PET/ CT scanning for the detection of recurrent disease in stage III melanoma patients. Our secondary objective was to evaluate the associated costs of this follow-up strategy.

| Patients
This investigation included all patients with stage III melanoma who underwent curative treatment with complete lymph node dissection (CLND) for a positive sentinel node, or with TLND for macrometastases, and were treated at the Division of Surgical Oncology of the University Medical Center Groningen (UMCG), the Netherlands.
The study protocol was applied to all stage III melanoma patients who were in follow-up in 2015, and to all newly diagnosed patients since 2015. Study data were collected during the period 2015-2018.
Patients who underwent off-protocol FDG PET/CT imaging during this time period were excluded from the present analysis. Data collection was conducted according to the declaration of Helsinki ethical principles for medical research involving human subjects. 12

| Follow-up
Outpatient follow-up visits included patient medical history, physical examination, and serum S-100B and LDH laboratory testing following the UMCG protocol (Table 1).
Serum S-100B level laboratory calculations were performed as previously described. 7 The S-100B cut-off value was ≥0.

| Costs
For all patients participating in the UMCG follow-up protocol, we calculated the follow-up costs of the detection of asymptomatic and symptomatic recurrences, including S-100B measurement, as well as the total costs of FDG PET/CT scanning. Data were acquired from the Patient Financial Department of the UMCG.

| Costs
The clinical assessment, to guide FDG PET/CT scanning for the detection of recurrent disease, without the financial, logistical, and radiation burdens of a standard scanning follow-up scheme.
In cases of cutaneous melanoma, S-100B serum concentrations are a prognostic marker of metastatic disease. 5,7 Serum concentrations of S-100B correlate with disease stage, although large variation is observed with or without S-100B elevation. 6 Previous findings suggest that S-100B levels may be influenced by the melanoma metastasis location and by variations in the ability of melanoma cells to produce S-100B. [14][15][16] Together with the limited S-100B elevation in patients with low tumor load, it is difficult to designate S-100B as a solid indicator of recurrence. 17 In the current study, disease recurrence was detected on with clinical symptoms and normal S-100B. These data correspond with previous findings that elevated S-100B was the only sign in 20% of patients with disease progression. 16 In the present series, 33% of patients that recurred IV disease had increased S-100B, which is in line with prior reports of increased S-100B levels in 4% to 100% of patients with stage IV disease. 6 In stage II and III melanoma patients, the reported sensitivity and specificity of S-100B for recurrent disease varies from 29% to 43% and 93% to 94%, respectively. 7,11,18 To compare with other tumor markers, the widely accepted colorectal cancer biomarker carcino-embryonic antigen (CEA) has a 41% to 97% sensitivity, which is somewhat higher, and a 52% to 100% specificity, which is comparable to that of S-100B. 19 A recent study revealed that 1.5% of all CEA measurements from curatively treated patients with stage I-III colorectal cancer ultimately led to recurrence detection. 20,21 As with S-100B, a normal CEA level does not exclude recurrent disease. 22 Tumor markers can be used in cancer detection and diagnosis, but are mainly used in follow-up to detect recurrent disease in an early phase. 23  diagnosis. Adding S-100B measurement to follow-up could be a way to support this demand, when patients are still asymptomatic.
Future research is needed to optimize its use, to assess the absolute survival gain, and compare to the efficacy and costs of this follow-up method with those of standard scanning protocols.
Research should also focus in the future on patient and tumor characteristics that may predict the sensitivity of S-100B during follow-up, with the aim of identifying patient subgroups in which S-100B shows higher sensitivity, to maximize the effectiveness of this tool.

| CONCLUSIONS
S-100B cannot exclude recurrent disease during follow-up of stage III melanoma. However, adding S-100B measurement to standard clinical assessment can effectively guide FDG PET/CT scanning for detecting recurrent melanoma. Future studies are needed to determine whether this protocol is a good alternative to follow-up regimens that include standard scheduled FDG PET/CT scans.

EAD received a research grant from the Groningen Melanoma
Sarcoma Foundation. The authors wish to express their gratitude to those who took care of the melanoma patients in this S-100B study.