462 Original article
Pelvic lymph node status prediction in melanoma patients with inguinal lymph node metastasis Sandro Pasqualia, Simone Mocellina, Francesco Bigolina, Antonella Vecchiatob, Maria C. Montescoc, Antonio Di Maggiod and Carlo R. Rossia,b The extent of the groin lymph node (LN) dissection for melanoma is still being debated, particularly in the case of micrometastasis (sentinel lymph node positive). We tested the predictive values of the criteria for pelvic dissection currently suggested by national guidelines (number of positive inguinal LN, Cloquet’s LN status, and preoperative computed tomographic scan) and the inguinal lymph node ratio (LNR, the ratio between metastatic and excised LNs) to identify patients with pelvic metastasis. We analyzed the predictive values of the above-mentioned criteria in 157 patients who underwent an ilioinguinal dissection, with a focus on their negative predictive values (NPV), which might help identify low-risk patients who might safely avoid pelvic dissection, pelvic dissection reduction, and error rate. Fortyfour (26.7%) patients had pelvic LN metastasis. In patients with micrometastasis (17.3% had pelvic LN metastasis), LNR less than 0.1 and Cloquet’s LN status achieved clinically relevant NPV (95.7 and 95.5%, respectively) and pelvic dissection reduction (38.4 and 84.6%, respectively), whereas the error rate was 1.7 and 3.0%, respectively. Lower NPVs were observed for number of positive inguinal LNs (88.6%) and computed tomographic scan (78.4%).
Introduction The management of patients with melanoma lymph node (LN) metastasis in the groin is still being debated, given the morbidity of the procedure and the limited available evidence on the survival benefit for lymphadenectomy [1–6]. Inguinal lymphadenectomy is routinely performed in patients with groin LN metastasis from melanoma. Pelvic dissection, however, is recommended by several national clinical practice guidelines and by several authors in patients with a clinically positive LN, radiological imaging showing pelvic LN metastasis [computed tomographic (CT) scan], at least three positive inguinal LNs, and metastasis in the Cloquet’s LN (which is considered the ‘sentinel node’ of the pelvic field) [7–9]. Other guidelines and authors’ suggestions include performing an elective ilioinguinal lymphadenectomy when subclinical [sentinel lymph node biopsy (SLNB) positive] or clinical LN metastasis is detected in the groin [10–12]. Given these different recommendations, it is not surprising that surgeons have different opinions on how to tailor the extension of lymphadenectomy to a single patient, particularly in cases with a positive SLNB [13], where the risk of harboring pelvic disease is 6–16% [14–19]. 0960-8931 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
Accuracy was enhanced when these criteria were considered in multivariable models. In patients with macrometastasis (36.8% had pelvic LN metastasis), LNR and current selection criteria achieved low NPVs and a high error rate. Avoiding pelvic dissection may be safe in sentinel lymph node-positive patients with LNR less than 0.1. The prediction of pelvic metastasis seems to be less accurate for patients with clinically positive LNs. Melanoma Res 24:462–467 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins. Melanoma Research 2014, 24:462–467 Keywords: lymph node dissection, melanoma, pelvic lymph node a Surgery Branch, Department of Surgery, Oncology and Gastroenterology, University of Padova, bMelanoma and Sarcomas Unit, cPathology Unit and d Radiology Unit, Veneto Institute of Oncology, IRCCS, Padova, Italy
Correspondence to Sandro Pasquali, MD, Surgery Branch, Department of Surgery, Oncology and Gastroenterology, University of Padova, via Giustiniani 2, 35128 Padova, Italy Tel: + 39 049 821 1851; fax: + 39 049 651891; e-mail: [email protected] Received 10 February 2014 Accepted 6 June 2014
In the absence of a randomized trial that can compare survival and morbidity after inguinal or ilioinguinal dissection for patients with groin LN metastasis, the identification of selection criteria for pelvic lymphadenectomy characterized by a high negative predictive value (NPV, that is, the pelvic LNs are really negative when predicted as negative) may be helpful in selecting patients who can avoid a pelvic lymphadenectomy with a low error rate (i.e. the possibility when a patient predicted as pelvic LN negative has melanoma metastasis in the pelvic LN) [20–22]. This is of particular importance for patients with a low risk of pelvic disease, such as in the case of a positive SLNB [14–19]. At our institution, ilioinguinal dissection is performed as the standard surgery for patients with inguinal LN metastasis, thus offering the opportunity to search for predictors that can potentially identify patients at a low risk of having pelvic disease. To identify these patients, we have studied the predictive value of the abovementioned selection criteria (the number of positive inguinal LNs, the pelvic LN status at preoperative CT scan, and the Cloquet’s LN status) as well as the inguinal lymph node ratio (LNR, which is the ratio of the metastatic LNs over the excised LNs [23–28]). DOI: 10.1097/CMR.0000000000000109
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node metastasis in melanoma Pasquali et al. 463
Methods Retrospective data for patients with primary melanomas treated between 1990 and 2010 were retrieved from a prospectively collected database according to the following selection criteria: (a) ilioinguinal lymphadenectomy performed for regional LN metastasis located in the groin and (b) detailed number of positive LNs for pelvic (i.e. iliac and obturator) anatomic regions. At our institution, SLNB has been performed since 1993, and between 1990 and 1992, lymphadenectomy was performed only for clinically positive LNs. The study was approved by the Melanoma Research Committee of the Veneto Institute of Oncology. Patients had provided their consent for research purposes. The description of ilioinguinal dissection has been reported in detail elsewhere [25,29]. Cloquet’s LN, which was defined as the LN node removed from under the inguinal ligament and medial to the external iliac vein, was marked by the surgeons who performed the dissection and was evaluated using the standard pathological protocol for LNs of lymphadenectomy. Since 2000, CT scans have been performed routinely in patients with clinically positive inguinal LNs or with high-risk primary tumors [i.e. American Joint Committee on Cancer (AJCC) T4 substages], as recommended by the Italian guidelines [12]. Pelvic LNs were considered at risk of metastasis when they presented an enhanced contrast or had a major diameter greater than 1 cm. Analyses were carried out to assess the predictive value of the inguinal LNR, the number of positive LNs, the CT scan, and the Cloquet’s LN status to identify patients with a low risk of pelvic LN metastasis. The following variables were considered for the analyses: patients’ age and sex, primary tumor site (trunk, lower extremity), Breslow tumor thickness (as a continuous variable recorded in mm), ulceration (absent, present), Clark level (II–IV, V), mitotic rate of the dermal invasive melanoma (as a continuous variable indicating the number of mitoses per mm2 [30,31]), performance of SLNB (yes, no), LN tumor burden (micrometastasis, macrometastasis), number of excised and positive sentinel lymph node(s) (SLN), number of excised and positive LN(s) at lymphadenectomy, inguinal LNR (LNR A, < 0.1, LNR B, 0.11–0.25, LNR C, > 0.25 [24,25]), and pathological status of pelvic LNs (negative, positive). Radiological status (negative, positive) of pelvic LNs at CT scan and information on the pathological status of Cloquet’s LN (negative, positive) were considered when available. Fisher’s exact test for categorical variables and the Mann–Whitney U-test for continuous variables were used to assess the association between covariates and pelvic LN status. Pathology assessment of pelvic LNs was compared with the number of positive inguinal LN
(s), preoperative CT scans, Cloquet’s LN status, and LNR. Each case was classified as true positive (TP), true negative (TN), false positive (FP), or false negative (FN) to assess diagnostic accuracy, sensitivity, specificity, and the positive and the negative value of each predictor. The NPV was calculated as TN/(FN + TN). Pelvic lymphadenectomy reduction and error rates were calculated as (TP + FP)/(TP + FP + TN + FN) and FN/(TP + FP + TN + FN), respectively. The receiver operating characteristic (ROC) curve and the area under the curve were calculated for each selection criterion and also for criteria considered together. The number of positive inguinal LNs and the LNR were considered in separate predictive models to avoid the issue of multicollinearity. Statistical analyses were carried out with a 95% confidence interval and an α-error of 0.05. Data were computed using STATA 10/SE (StataCorp, College Station, Texas, USA).
Results Among 157 melanoma patients who fulfilled the inclusion criteria, pelvic LN metastasis was detected at pathological examination in 42 cases (26.7%), of whom 14/81 (17.3%) presented with micrometastasis and 28/67 presented with macrometastasis (36.8%). Patients with pelvic LN metastasis were older than pelvic LN-negative patients (P = 0.013), had thicker primary tumors (P = 0.011), a greater incidence of macrometastasis (P = 0.007), and a higher LNR (P < 0.001) and number of positive LNs (P < 0.001, Table 1). Micrometastasis (positive SLN)
The highest NPV with the lower error rate was achieved by the inguinal LNR (Table 2). The inguinal LNR and the pathological status of the Cloquet’s LNs had the highest NPV (95.7 and 95.5%, respectively): if these criteria were used to screen patients preoperatively, these NPVs would allow for pelvic lymphadenectomy reductions of 38.4 and 84.6%, respectively. These clinical benefits would be associated with errors of 1.7 and 3.0%, respectively. The number of positive inguinal LNs and the preoperative CT scan had lower NPVs (88.6 and 78.4%): if these criteria were used to screen patients preoperatively, these NPVs would allow for a high pelvic reduction rate (86.4 and 86.0%, respectively), but would be associated with a high error rate (9.9 and 18.6%, respectively) (Table 2). Macrometastasis (clinically positive LN)
Overall, NPVs were lower than the desirable 90% in case of macrometastasis and error rates were higher (Table 2). Among patients with micrometastasis, inguinal LNR had the highest NPV (85.7%): if inguinal LNR was used to screen patients preoperatively, this NPV would allow for
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
464
Melanoma Research 2014, Vol 24 No 5
Table 1 Distribution of clinicopathologic features among patients with negative and positive pelvic inguinal lymph node(s) after ilioinguinal lymphadenectomy Variables Age (years) Sex Female Male Primary tumor site Lower limb Trunk Unknown primary Tumor thickness (mm) Ulceration Absent Present LN tumor burden Positive SLN Clinically positive LN Positive inguinal LNs Excised inguinal LNs Inguinal LNR A (< 0.1) B (0.11–0.25) C (>0.25) AJCC N stage N1 (1 positive LN) N2 (2–3 positive LNs) N3 (≥4 positive LNs)
Negative pelvic LNs (N = 115) [N (%)]
Positive pelvic LNs (N = 42) [N (%)]
P-value
61.57 (95% CI 51.28–69.50)
0.013
52.96 (IQR 40.74–66.49) 70 (60.9) 45 (39.1) 86 20 9 Median 2.60
21 (50.0) 21 (50.0)
(74.8) (17.4) (7.8) (IQR 1.47–4.0)
35 3 2 Median 3.53
46 (51.1) 44 (48.9)
0.274
(83.3) (11.9) (4.8) (IQR 2.26–6.50)
0.321 0.011
19 (52.8) 17 (47.2)
67 (58.3) 48 (41.7) Median 1 (IQR 1–2) Median 10 (IQR 7.5–14)
1.000
0.007
14 (33.3) 28 (66.7) Median 3 (IQR 1–5.75) Median 11 (IQR 8–14)
< 0.001 0.505
24 (40.0) 25 (41.6) 11 (18.4)
15 (22.7) 23 (34.8) 28 (42.5)
< 0.001
61 (53.0) 29 (25.2) 25 (21.8)
0 (0.0) 13 (31.0) 29 (69.0)
< 0.001
Statistically significant values have been underlined to highlight relevant results. AJCC, American Joint Committee on Cancer; CI, confidence interval; IQR, interquartile range; LN, lymph node; LNR, lymph node ratio; SLN, sentinel lymph node.
Accuracy, sensitivity, specificity, positive predictive value, negative predictive value, pelvic lymphadenectomy reduction, and error rate of the lymph node ratio, number of positive inguinal lymph nodes, Cloquet’s lymph node status, and pelvic lymph node(s) status at computed tomographic scan
Table 2
Scenarios LNR A (< 0.1) Positive SLN Clinically positive LN ≥ 3 positive inguinal LNs Positive SLN Clinically positive LN Cloquet’s LN status Positive SLN Clinically positive LN Pelvic LN status at CT scan Positive SLN Clinically positive LN
Accuracy (%) Sensitivity (%) Specificity (%) PPV (%) NPV (%) Pelvic lymphadenectomy reduction (%) Error (%) 48.3 53.0
87.5 92.0
42.3 29.3
18.9 44.2
95.7 85.7
38.4 21.2
1.7 3.0
84.0 60.3
42.9 59.3
92.5 60.9
54.5 47.1
88.6 71.8
86.4 53.5
9.9 15.1
96.2 65.7
80.0 25.0
100.0 87.0
100.0 50.0
95.5 69.0
84.6 82.8
3.8 25.7
74.4 47.7
27.3 57.7
90.6 41.0
50.0 39.5
78.4 59.3
86.0 41.5
18.6 16.9
Pelvic lymphadenectomy reduction: patients predicted as pelvic LN negative/all the patients. Error rate: false negative/all the patients. False-negative patients were predicted as pelvic LN negative even they had pelvic LN metastasis. CT, computed tomography; LN, lymph node; LNR, lymph node ratio; NPV, negative predictive value; PPV, positive predictive value; SLN, sentinel lymph node.
a pelvic lymphadenectomy reduction of 21.1%, with an error rate of 3%. A number of positive inguinal LNs, Cloquet’s LN status, and preoperative CT scans had lower NPVs (71.8, 69.0, and 59.3% respectively): if these criteria were used to screen patients preoperatively, these NPVs would allow for a high pelvic reduction rate (53.5, 82.8, and 41.5%), but would be associated with a high error rate (15.1, 25.7, and 16.9%, respectively).
exploratory evaluation of prediction models. The number of positive inguinal LNs and the LNR were included in two separate multivariate models to avoid the issue of covariate collinearity as these variables share some information (i.e. the number of metastatic LNs). The two multivariable models were more accurate than single predictors, as shown by the areas under the ROC curves (Table 3).
Discussion Prediction model for patients with micrometastasis
Considering that the predictive criteria achieved a clinically relevant NPV and a low error rate in patients with micrometastasis (but not for macrometastasis), we performed an
The series analyzed in this study comprises a group of melanoma patients with LN metastasis who underwent an ilioinguinal LN dissection which, at our institution, is performed as the standard surgery for inguinal LN
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node metastasis in melanoma Pasquali et al. 465
Area under the receiver operating characteristic curve for each predictive variable and for two multivariable models, which consider the number of positive inguinal lymph node(s) and the inguinal lymph node ratio each, in patients with micrometastasis
Table 3
Number of positive inguinal LNa Inguinal LNRb Cloquet’s LN CT scan Model 1 (including number of positive inguinal LNa) Model 2 (including inguinal LNRb)
ROC area
SE
95% CI
0.677 0.613 0.719 0.688 0.784
0.070 0.045 0.064 0.073 0.078
0.566–0.778 0.503–0.723 0.605–0.811 0.579–0.789 0.685–0.873
0.816
0.070
0.713–0.892
CI, confidence interval; CT, computed tomography; LN, lymph node; LNR, lymph node ratio; ROC, receiver operating characteristic. a Categorical variable, cutoff 3. b Categorical variable, cutoff LNR A versus B and C.
melanoma metastasis. We investigated predictors of pelvic LN metastasis (the inguinal LNR, the number of positive inguinal LNs, the Cloquet’s LN status, and the preoperative CT) in patients with micrometastasis (i.e. who underwent lymphadenectomy for a positive SLNB) and macrometastasis (i.e. who underwent lymphadenectomy for clinically positive LNs) to identify patients at a low risk of pelvic LN metastasis.
dissection [32], can identify whether patients need a pelvic dissection with a high level of evidence. The predictive value of the inguinal LNR may be explained considering that this parameter includes the number of excised LNs. Patients with lower values of LNR (e.g. LNR A) can have less positive LNs or more excised LNs than patients who present higher values of LNR. Different hypotheses can explain the predictive role of the LNR. Intuitively, a thorough dissection allows identification of all metastatic LNs and provides an optimal stratification of the risk of harboring pelvic LN metastasis. Moreover, the number of excised LNs might be a surrogate of the immunological response against the tumor as the immune system shows a greater activity against melanoma cells in LNs that are far from those having melanoma metastasis [33]. This suggests that a greater number of inguinal nodes may limit the spread of melanoma cells to the pelvic nodes.
Inguinal LNR A (≤ 0.1) had an NPV of 95.7 and 85.7% in patients with micrometastasis and macrometastasis, respectively, for predicting the status of pelvic LNs that was associated with a potential reduction rate of 38.4 and 21.2%, respectively. The overall error rate was lower in patients with micrometastasis (1.7 vs. 3.0%). These results suggest that the inguinal LNR can be of particular relevance for patients with micrometastasis, who are at a low risk of harboring pelvic disease (17.2% in this study and 6–16% is other series [14–19]), and may be considered for a less extensive dissection.
The number of positive inguinal LNs, although associated independently with the pelvic LN status, shows lower NPVs than the desirable 90%, both in the case of micrometastasis (88.6%) and in the case of macrometastasis (71.8%). Although this parameter has been widely considered an independent predictor of pelvic LN metastasis, this study questions its predictive effectiveness in light of the relatively low NPVs and high error rates. Similarly, Strobbe and colleagues studied patients who underwent ilioinguinal dissection for clinically positive LNs, and reported an NPV of 78% for the number of positive inguinal LNs, a value far less than the desirable 90%. Further data are needed to ascertain whether the number of positive inguinal LNs identifies patients at a low risk of pelvic LN metastasis who may avoid pelvic dissection.
The association between inguinal LNR and pelvic LN metastasis was showed for the first time by Chu et al. [14] in 40 SLNB-positive patients. Our analysis adds meaningful information to the existing literature because it suggests that the presence of pelvic LN metastasis after a positive inguinal SLNB can be predicted with a high NPV (95.7%). Accordingly, roughly 40% of patients with a positive inguinal SLNB may avoid pelvic dissection, with only 1.7% of patients having LN disease left in their groin. The prediction of a pelvic LN metastasis can be limited by the routine methodology for the pathologic assessment of the pelvic LNs, which is standard hematoxylin and eosin (H&E) staining. This process may miss melanoma deposits in some LNs and therefore leads to a greater number of patients with a FN result, thus limiting the value of predictive factors and tools. Although H&E staining of the LNs after a lymphadenectomy remains the standard method to detect melanoma LN metastasis, only a randomized trial, such as the EAGLE FM Study from the Australia and New Zealand Melanoma Trial Group comparing the outcomes of inguinal and ilioinguinal
In this series, the pathological status of the Cloquet’s LN had an NPV of 95.5% (error rate 3.8%) and 69.0% (error rate 25.7%) in patients with micrometastasis and macrometastasis, respectively. Interestingly, if SLNB-positive patients with negative Cloquet’s LN had avoided pelvic dissection, the reduction rate would have been up to 84%, with an error rate of 3.8%. Previous studies report NPVs of 94–97% in the case of micrometastasis [34,35] and of 80–95% in the case of macrometastasis [36–38]. Despite these findings, in a study on 53 patients with micrometastasis, Chu et al. [34] noted that the Cloquet’s LN was rarely involved with melanoma cells (3.8%). A high concordance between the presence of metastasis in the Cloquet’s and pelvic LNs was observed (NPV of 94%), but, considering the low incidence of metastasis in the Cloquet’s node, the authors suggest avoiding performance of Cloquet’s LN examination in patients with micrometastasis. Remarkably, our series has a higher positivity rate of the Cloquet’s LN in the case of micrometastasis (15%, 4/26). Although the positivity rates of NPVs and Cloquet’s LN lead to consideration of
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
466
Melanoma Research 2014, Vol 24 No 5
Cloquet’s LN as an effective predictive tool after a positive SLNB, the error rate that we identified (3.8%) limits its clinical value along with other drawbacks intrinsic to the biopsy of the Cloquet’s LN. In fact, the lymph flow does not always pass from the inguinal to the pelvic LNs through the femoral canal, where the Cloquet’s LN is located [38,39]. Furthermore, in this study, we have investigated the diagnostic value of the preoperative CT scan, which poorly characterized the pelvic LNs. This is shown by the lack of independent predictive value and by the low NPVs, which are 78.4% in patients with micrometastasis (error rate 18.6%) and 59.3% in patients with macrometastasis (error rate 16.9%). Considering that CT scans have a diagnostic accuracy of greater than 90% for staging the LNs in patients with melanoma irrespective of their location [40], this imaging technique is expected to have a high diagnostic relevance for the characterization of pelvic LNs, especially for macrometastasis. Van der Ploeg et al. [41] reported an NPV of 91% for preoperative CT scans in patients with macrometastasis and concluded that pelvic dissection should be performed only in patients with positive pelvic LNs. In contrast, Allan et al. [42] studied the diagnostic value of the CT scan in 72 patients with macrometastasis and considered an NPV of 86% insufficient to identify patients who might safely avoid pelvic lymphadenectomy. As the characterization of the pelvic LNs by the CT scan seems more controversial than expected, authors are considering PETCT for selection of patients for pelvic dissection [43]. Finally, this study explored the predictive accuracy of the considered selection criteria in multivariable predictive tools for patients with micrometastasis who are at a lower risk of pelvic LN metastasis. In fact, a reduction in the extension of groin dissection is less safe in patients with macrometastasis because of the greater risk of LN disease and the lower predictive values shown by the current selection criteria. The area and the ROC curve was greater for multivariable models – two models were constructed that considered the positive inguinal LNs and the inguinal LNR separately to avoid the issue of multicollinearity than for each predictor (Table 3), suggesting that a predictive tool encompassing several selection criteria can be valuable in the identification of patients who may safely avoid pelvic dissection. There are some limitations to the clinical applicability of the results of this study. Clearly, the availability of inguinal LNR after surgery limits its clinical utility. Despite this drawback, given that roughly one-third of surgeons perform only inguinal dissection for groin melanoma metastasis irrespective of other risk factors [13], the inguinal LNR may be considered for patient counseling and follow-up as low LNR values suggest a high likelihood of having negative pelvic LNs. The value of LNR is influenced by the LN count, suggesting that
the achievement of satisfactory nodal retrieval is of great importance to allow risk stratification for pelvic metastasis according to the LNR [10,17]. Furthermore, if a predictive tool, such as a nomogram, will include inguinal LNR as well as other predictive variables (including those considered in this study and others, such as lymph flow drainage) and achieve high predictive values, pelvic dissection may be tailored on patient and tumor characteristics and performed as a second operation in the relatively small group of patients with pelvic disease after a positive SLNB (6–17.2%). Nevertheless, given these limitations and the above-mentioned risk of pathological downstage of LN metastasis, only a randomized trial, such as the EAGLE FM Study [32], can define the optimal management of patients with groin LN melanoma metastasis. Conclusion
Pending a randomized trial, this study offers interesting insights for the identification of SLN-positive patients at a low risk of pelvic LN metastasis. In SLN-positive patients, avoiding pelvic dissection may be safe when LNR is lower than 0.1 and Cloquet’s LN is negative, although the error rate was slightly higher in the latter case. The use of the number of positive inguinal LNs and the preoperative CT scan may offer a less accurate prediction, on the basis of the lower NPVs and higher error rate. This study also shows that the prediction of pelvic metastasis seems less accurate for patients with clinically positive LNs, suggesting that pelvic dissection cannot safely be avoided in this scenario.
Acknowledgements The authors thank Ms Chiara Longhi and Mrs Marilyn Ward for editing the manuscript. Conflicts of interest
There are no conflicts of interest.
References 1 2
3
4
5 6
7
Mack LA, McKinnon JG. Controversies in the management of metastatic melanoma to regional lymphatic basins. J Surg Oncol 2004; 86:189–199. Morton DL, Thompson JF, Cochran AJ, Mozzillo N, Elashoff R, Essner R, et al. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med 2006; 355:1307–1317. Guggenheim MM, Hug U, Jung FJ, Rousson V, Aust MC, Calcagni M, et al. Morbidity and recurrence after completion lymph node dissection following sentinel lymph node biopsy in cutaneous malignant melanoma. Ann Surg 2008; 247:687–693. Hughes TM, A’Hern RP, Thomas JM. Prognosis and surgical management of patients with palpable inguinal lymph node metastases from melanoma. Br J Surg 2000; 87:892–901. Hughes TM, Thomas JM. Combined inguinal and pelvic lymph node dissection for stage III melanoma. Br J Surg 1999; 86:1493–1498. Kretschmer L, Neumann C, Preusser KP, Marsch WC. Superficial inguinal and radical ilioinguinal lymph node dissection in patients with palpable melanoma metastases to the groin – an analysis of survival and local recurrence. Acta Oncol 2001; 40:72–78. National Comprehensive Cancer Network. Clinical practice guidelines in oncology. Melanoma (2013). Available at: http://www.nncn.org. [Accessed 1 March 2014].
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node metastasis in melanoma Pasquali et al. 467
8 Australian Cancer Network Melanoma Guidelines Revision Working Party. Clinical practice guidelines for the management of melanoma in Australia and New Zealand. Cancer Council Australia and Australian Cancer Network, Sydney and New Zealand Guidelines Group, Wellington, 2008. 9 Shada AL, Walters DM, Tierney SN, Slingluff CL Jr. Surgical resection for bulky or recurrent axillary metastatic melanoma. J Surg Oncol 2012; 105:21–25. 10 Spillane AJ, Cheung BL, Stretch JR, Scolyer RA, Shannon KF, Quinn MJ, et al. Proposed quality standards for regional lymph node dissections in patients with melanoma. Ann Surg 2009; 249:473–480. 11 Spillane AJ, Tucker M, Pasquali S. A pilot study reporting outcomes for melanoma patients of a minimal access ilio-inguinal dissection technique based on two incisions. Ann Surg Oncol 2011; 18:970–976. 12 Italian National Research Council. Progetto oncologia. Available at: http:// progettooncologia.cnr.it/bridge/attivita-direzione.html. [Accessed 1 March 2014]. 13 Pasquali S, Spillane AJ, de Wilt JH, McCaffery K, Rossi CR, Quinn MJ, et al. Surgeons’ opinions on lymphadenectomy in melanoma patients with positive sentinel nodes: a worldwide web-based survey. Ann Surg Oncol 2012; 19:4322–4329. 14 Chu CK, Delman KA, Carlson GW, Hestley AC, Murray DR. Inguinopelvic lymphadenectomy following positive inguinal sentinel lymph node biopsy in melanoma: true frequency of synchronous pelvic metastases. Ann Surg Oncol 2011; 18:3309–3315. 15 Mozzillo N, Caracò C, Marone U, Di Monta G, Crispo A, Botti G, et al. Superficial and deep lymph node dissection for stage III cutaneous melanoma: clinical outcome and prognostic factors. World J Surg Oncol 2013; 11:36. 16 Santinami M, Carbone A, Crippa F, Maurichi A, Pellitteri C, Ruggeri R, et al. Radical dissection after positive groin sentinel biopsy in melanoma patients: rate of further positive nodes. Melanoma Res 2009; 19:112–118. 17 Spillane AJ, Haydu L, McMillan W, Stretch JR, Thompson JF. Quality assurance parameters and predictors of outcome for ilioinguinal and inguinal dissection in a contemporary melanoma patient population. Ann Surg Oncol 2011; 18:2521–2528. 18 Van der Ploeg IM, Kroon BB, Valdés Olmos RA, Nieweg OE. Evaluation of lymphatic drainage patterns to the groin and implications for the extent of groin dissection in melanoma patients. Ann Surg Oncol 2009; 16:2994–2999. 19 Zdzienicki M, Rutkowski P, Nowecki ZI, van Akkooi AC, Michej W, Dziewirski W, et al. The analysis of the outcomes and factors related to iliac–obturator involvement in cutaneous melanoma patients after lymph node dissection due to positive sentinel lymph node biopsy or clinically detected inguinal metastases. Eur J Surg Oncol 2013; 39:304–310. 20 Mocellin S, Thompson JF, Pasquali S, Montesco MC, Pilati P, Nitti D, et al. Sentinel node status prediction by four statistical models: results from a large bi-institutional series (n =1132). Ann Surg 2009; 250:964–969. 21 Pasquali S, Mocellin S, Campana LG, Vecchiato A, Bonandini E, Montesco MC, et al. Maximizing the clinical usefulness of a nomogram to select patients candidate to sentinel node biopsy for cutaneous melanoma. Eur J Surg Oncol 2011; 37:675–680. 22 Mocellin S, Ambrosi A, Montesco MC, Foletto M, Zavagno G, Nitti D, et al. Support vector machine learning model for the prediction of sentinel node status in patients with cutaneous melanoma. Ann Surg Oncol 2006; 13:1113–1122. 23 Mocellin S, Pasquali S, Rossi CR, Nitti D. Validation of the prognostic value of lymph node ratio in patients with cutaneous melanoma: a populationbased study of 8177 cases. Surgery 2011; 150:83–90. 24 Spillane AJ, Cheung BL, Winstanley J, Thompson JF. Lymph node ratio provides prognostic information in addition to american joint committee on cancer N stage in patients with melanoma, even if quality of surgery is standardized. Ann Surg 2011; 253:109–115. 25 Rossi CR, Mocellin S, Pasquali S, Pilati P, Nitti D. N-ratio: a novel independent prognostic factor for patients with stage-III cutaneous melanoma. Ann Surg Oncol 2008; 15:310–315.
26
27
28
29
30
31
32
33
34
35
36
37 38
39
40
41
42 43
Berger AC, Fierro M, Kairys JC, Berd D, Sato T, Andrel J, et al. Lymph node ratio is an important and independent prognostic factor for patients with stage III melanoma. J Surg Oncol 2012; 105:15–20. Grotz TE, Huebner M, Pockaj BA, Perkins S, Jakub JW. Limitations of lymph node ratio, evidence-based benchmarks, and the importance of a thorough lymph node dissection in melanoma. Ann Surg Oncol 2013; 20:4370–4377. Egger ME, Scoggins CR, Martin RC II, Stromberg AJ, Quillo AR, McMasters KM, Callender GG. The lymph node ratio has limited prognostic significance in melanoma. J Surg Res 2013; 179:10–17. Rossi CR, De Salvo GL, Bonandini E, Mocellin S, Foletto M, Pasquali S, et al. Factors predictive of nonsentinel lymph node involvement and clinical outcome in melanoma patients with metastatic sentinel lymph node. Ann Surg Oncol 2008; 15:1202–1210. Azzola MF, Shaw HM, Thompson JF, Soong SJ, Scolyer RA, Watson GF, et al. Tumor mitotic rate is a more powerful prognostic indicator than ulceration in patients with primary cutaneous melanoma: an analysis of 3661 patients from a single center. Cancer 2003; 97:1488–1498. Francken AB, Shaw HM, Thompson JF, Soong SJ, Accortt NA, Azzola MF, et al. The prognostic importance of tumor mitotic rate confirmed in 1317 patients with primary cutaneous melanoma and long follow-up. Ann Surg Oncol 2004; 11:426–433. ANZMTG 01.12. Inguinal or ilio-inguinal lymphadenectomy for patients with metastatic melanoma to groin lymph nodes and no evidence of pelvic disease on PET/CT scan – a randomised phase III trial (EAGLE FM). Available at: http://www.anzmtg.org/doc/16034%20EAGLE%20FM% 20Hamburg%20Poster%2027Jun13_al.pdf. [Accessed 28 September 2013]. Cochran AJ, Huang RR, Lee J, Itakura E, Leong SP, Essner R. Tumourinduced immune modulation of sentinel lymph nodes. Nat Rev Immunol 2006; 6:659–670. Chu CK, Zager JS, Marzban SS, Gimbel MI, Murray DR, Hestley AC, et al. Routine biopsy of Cloquet’s node is of limited value in sentinel node positive melanoma patients. J Surg Oncol 2010; 102:315–320. Essner R, Scheri R, Kavanagh M, Torisu-Itakura H, Wanek LA, Morton DL. Surgical management of the groin lymph nodes in melanoma in the era of sentinel lymph node dissection. Arch Surg 2006; 141:877–882, discussion 882–884. Illig L, Aigner KR, Biess B, Göhl J, Holle R, Hundeiker M, et al. Diagnostic excision of the Rosenmüller’s node. Screening for occult metastases before elective regional lymph node dissection in patients with lower limb melanoma? Cancer 1988; 61:1200–1206. Coit DG, Brennan MF. Extent of lymph node dissection in melanoma of the trunk or lower extremity. Arch Surg 1989; 124:162–166. Strobbe LJ, Jonk A, Hart AA, Peterse JL, Wobbes T, Nieweg OE, Kroon BB. The value of Cloquet’s node in predicting melanoma nodal metastases in the pelvic lymph node basin. Ann Surg Oncol 2001; 8:209–214. Van der Ploeg IM, Valdés Olmos RA, Kroon BB, Nieweg OE. Tumor-positive sentinel node biopsy of the groin in clinically node-negative melanoma patients: superficial or superficial and deep lymph node dissection? Ann Surg Oncol 2008; 15:1485–1491. Xing Y, Bronstein Y, Ross MI, Askew RL, Lee JE, Gershenwald JE, et al. Contemporary diagnostic imaging modalities for the staging and surveillance of melanoma patients: a meta-analysis. J Natl Cancer Inst 2011; 103:129–142. Van der Ploeg AP, van Akkooi AC, Schmitz PI, van Geel AN, de Wilt JH, Eggermont AM, Verhoef C. Therapeutic surgical management of palpable melanoma groin metastases: superficial or combined superficial and deep groin lymph node dissection. Ann Surg Oncol 2011; 18:3300–3308. Allan CP, Hayes AJ, Thomas JM. Ilioinguinal lymph node dissection for palpable metastatic melanoma to the groin. ANZ J Surg 2008; 78:982–986. Wevers KP, Bastiaannet E, Poos HP, van Ginkel RJ, Plukker JT, Hoekstra HJ. Therapeutic lymph node dissection in melanoma: different prognosis for different macrometastasis sites? Ann Surg Oncol 2012; 19:3913–3918.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node status prediction in melanoma patients with inguinal lymph node metastasis Sandro Pasqualia, Simone Mocellina, Francesco Bigolina, Antonella Vecchiatob, Maria C. Montescoc, Antonio Di Maggiod and Carlo R. Rossia,b The extent of the groin lymph node (LN) dissection for melanoma is still being debated, particularly in the case of micrometastasis (sentinel lymph node positive). We tested the predictive values of the criteria for pelvic dissection currently suggested by national guidelines (number of positive inguinal LN, Cloquet’s LN status, and preoperative computed tomographic scan) and the inguinal lymph node ratio (LNR, the ratio between metastatic and excised LNs) to identify patients with pelvic metastasis. We analyzed the predictive values of the above-mentioned criteria in 157 patients who underwent an ilioinguinal dissection, with a focus on their negative predictive values (NPV), which might help identify low-risk patients who might safely avoid pelvic dissection, pelvic dissection reduction, and error rate. Fortyfour (26.7%) patients had pelvic LN metastasis. In patients with micrometastasis (17.3% had pelvic LN metastasis), LNR less than 0.1 and Cloquet’s LN status achieved clinically relevant NPV (95.7 and 95.5%, respectively) and pelvic dissection reduction (38.4 and 84.6%, respectively), whereas the error rate was 1.7 and 3.0%, respectively. Lower NPVs were observed for number of positive inguinal LNs (88.6%) and computed tomographic scan (78.4%).
Introduction The management of patients with melanoma lymph node (LN) metastasis in the groin is still being debated, given the morbidity of the procedure and the limited available evidence on the survival benefit for lymphadenectomy [1–6]. Inguinal lymphadenectomy is routinely performed in patients with groin LN metastasis from melanoma. Pelvic dissection, however, is recommended by several national clinical practice guidelines and by several authors in patients with a clinically positive LN, radiological imaging showing pelvic LN metastasis [computed tomographic (CT) scan], at least three positive inguinal LNs, and metastasis in the Cloquet’s LN (which is considered the ‘sentinel node’ of the pelvic field) [7–9]. Other guidelines and authors’ suggestions include performing an elective ilioinguinal lymphadenectomy when subclinical [sentinel lymph node biopsy (SLNB) positive] or clinical LN metastasis is detected in the groin [10–12]. Given these different recommendations, it is not surprising that surgeons have different opinions on how to tailor the extension of lymphadenectomy to a single patient, particularly in cases with a positive SLNB [13], where the risk of harboring pelvic disease is 6–16% [14–19]. 0960-8931 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
Accuracy was enhanced when these criteria were considered in multivariable models. In patients with macrometastasis (36.8% had pelvic LN metastasis), LNR and current selection criteria achieved low NPVs and a high error rate. Avoiding pelvic dissection may be safe in sentinel lymph node-positive patients with LNR less than 0.1. The prediction of pelvic metastasis seems to be less accurate for patients with clinically positive LNs. Melanoma Res 24:462–467 © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins. Melanoma Research 2014, 24:462–467 Keywords: lymph node dissection, melanoma, pelvic lymph node a Surgery Branch, Department of Surgery, Oncology and Gastroenterology, University of Padova, bMelanoma and Sarcomas Unit, cPathology Unit and d Radiology Unit, Veneto Institute of Oncology, IRCCS, Padova, Italy
Correspondence to Sandro Pasquali, MD, Surgery Branch, Department of Surgery, Oncology and Gastroenterology, University of Padova, via Giustiniani 2, 35128 Padova, Italy Tel: + 39 049 821 1851; fax: + 39 049 651891; e-mail: [email protected] Received 10 February 2014 Accepted 6 June 2014
In the absence of a randomized trial that can compare survival and morbidity after inguinal or ilioinguinal dissection for patients with groin LN metastasis, the identification of selection criteria for pelvic lymphadenectomy characterized by a high negative predictive value (NPV, that is, the pelvic LNs are really negative when predicted as negative) may be helpful in selecting patients who can avoid a pelvic lymphadenectomy with a low error rate (i.e. the possibility when a patient predicted as pelvic LN negative has melanoma metastasis in the pelvic LN) [20–22]. This is of particular importance for patients with a low risk of pelvic disease, such as in the case of a positive SLNB [14–19]. At our institution, ilioinguinal dissection is performed as the standard surgery for patients with inguinal LN metastasis, thus offering the opportunity to search for predictors that can potentially identify patients at a low risk of having pelvic disease. To identify these patients, we have studied the predictive value of the abovementioned selection criteria (the number of positive inguinal LNs, the pelvic LN status at preoperative CT scan, and the Cloquet’s LN status) as well as the inguinal lymph node ratio (LNR, which is the ratio of the metastatic LNs over the excised LNs [23–28]). DOI: 10.1097/CMR.0000000000000109
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node metastasis in melanoma Pasquali et al. 463
Methods Retrospective data for patients with primary melanomas treated between 1990 and 2010 were retrieved from a prospectively collected database according to the following selection criteria: (a) ilioinguinal lymphadenectomy performed for regional LN metastasis located in the groin and (b) detailed number of positive LNs for pelvic (i.e. iliac and obturator) anatomic regions. At our institution, SLNB has been performed since 1993, and between 1990 and 1992, lymphadenectomy was performed only for clinically positive LNs. The study was approved by the Melanoma Research Committee of the Veneto Institute of Oncology. Patients had provided their consent for research purposes. The description of ilioinguinal dissection has been reported in detail elsewhere [25,29]. Cloquet’s LN, which was defined as the LN node removed from under the inguinal ligament and medial to the external iliac vein, was marked by the surgeons who performed the dissection and was evaluated using the standard pathological protocol for LNs of lymphadenectomy. Since 2000, CT scans have been performed routinely in patients with clinically positive inguinal LNs or with high-risk primary tumors [i.e. American Joint Committee on Cancer (AJCC) T4 substages], as recommended by the Italian guidelines [12]. Pelvic LNs were considered at risk of metastasis when they presented an enhanced contrast or had a major diameter greater than 1 cm. Analyses were carried out to assess the predictive value of the inguinal LNR, the number of positive LNs, the CT scan, and the Cloquet’s LN status to identify patients with a low risk of pelvic LN metastasis. The following variables were considered for the analyses: patients’ age and sex, primary tumor site (trunk, lower extremity), Breslow tumor thickness (as a continuous variable recorded in mm), ulceration (absent, present), Clark level (II–IV, V), mitotic rate of the dermal invasive melanoma (as a continuous variable indicating the number of mitoses per mm2 [30,31]), performance of SLNB (yes, no), LN tumor burden (micrometastasis, macrometastasis), number of excised and positive sentinel lymph node(s) (SLN), number of excised and positive LN(s) at lymphadenectomy, inguinal LNR (LNR A, < 0.1, LNR B, 0.11–0.25, LNR C, > 0.25 [24,25]), and pathological status of pelvic LNs (negative, positive). Radiological status (negative, positive) of pelvic LNs at CT scan and information on the pathological status of Cloquet’s LN (negative, positive) were considered when available. Fisher’s exact test for categorical variables and the Mann–Whitney U-test for continuous variables were used to assess the association between covariates and pelvic LN status. Pathology assessment of pelvic LNs was compared with the number of positive inguinal LN
(s), preoperative CT scans, Cloquet’s LN status, and LNR. Each case was classified as true positive (TP), true negative (TN), false positive (FP), or false negative (FN) to assess diagnostic accuracy, sensitivity, specificity, and the positive and the negative value of each predictor. The NPV was calculated as TN/(FN + TN). Pelvic lymphadenectomy reduction and error rates were calculated as (TP + FP)/(TP + FP + TN + FN) and FN/(TP + FP + TN + FN), respectively. The receiver operating characteristic (ROC) curve and the area under the curve were calculated for each selection criterion and also for criteria considered together. The number of positive inguinal LNs and the LNR were considered in separate predictive models to avoid the issue of multicollinearity. Statistical analyses were carried out with a 95% confidence interval and an α-error of 0.05. Data were computed using STATA 10/SE (StataCorp, College Station, Texas, USA).
Results Among 157 melanoma patients who fulfilled the inclusion criteria, pelvic LN metastasis was detected at pathological examination in 42 cases (26.7%), of whom 14/81 (17.3%) presented with micrometastasis and 28/67 presented with macrometastasis (36.8%). Patients with pelvic LN metastasis were older than pelvic LN-negative patients (P = 0.013), had thicker primary tumors (P = 0.011), a greater incidence of macrometastasis (P = 0.007), and a higher LNR (P < 0.001) and number of positive LNs (P < 0.001, Table 1). Micrometastasis (positive SLN)
The highest NPV with the lower error rate was achieved by the inguinal LNR (Table 2). The inguinal LNR and the pathological status of the Cloquet’s LNs had the highest NPV (95.7 and 95.5%, respectively): if these criteria were used to screen patients preoperatively, these NPVs would allow for pelvic lymphadenectomy reductions of 38.4 and 84.6%, respectively. These clinical benefits would be associated with errors of 1.7 and 3.0%, respectively. The number of positive inguinal LNs and the preoperative CT scan had lower NPVs (88.6 and 78.4%): if these criteria were used to screen patients preoperatively, these NPVs would allow for a high pelvic reduction rate (86.4 and 86.0%, respectively), but would be associated with a high error rate (9.9 and 18.6%, respectively) (Table 2). Macrometastasis (clinically positive LN)
Overall, NPVs were lower than the desirable 90% in case of macrometastasis and error rates were higher (Table 2). Among patients with micrometastasis, inguinal LNR had the highest NPV (85.7%): if inguinal LNR was used to screen patients preoperatively, this NPV would allow for
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
464
Melanoma Research 2014, Vol 24 No 5
Table 1 Distribution of clinicopathologic features among patients with negative and positive pelvic inguinal lymph node(s) after ilioinguinal lymphadenectomy Variables Age (years) Sex Female Male Primary tumor site Lower limb Trunk Unknown primary Tumor thickness (mm) Ulceration Absent Present LN tumor burden Positive SLN Clinically positive LN Positive inguinal LNs Excised inguinal LNs Inguinal LNR A (< 0.1) B (0.11–0.25) C (>0.25) AJCC N stage N1 (1 positive LN) N2 (2–3 positive LNs) N3 (≥4 positive LNs)
Negative pelvic LNs (N = 115) [N (%)]
Positive pelvic LNs (N = 42) [N (%)]
P-value
61.57 (95% CI 51.28–69.50)
0.013
52.96 (IQR 40.74–66.49) 70 (60.9) 45 (39.1) 86 20 9 Median 2.60
21 (50.0) 21 (50.0)
(74.8) (17.4) (7.8) (IQR 1.47–4.0)
35 3 2 Median 3.53
46 (51.1) 44 (48.9)
0.274
(83.3) (11.9) (4.8) (IQR 2.26–6.50)
0.321 0.011
19 (52.8) 17 (47.2)
67 (58.3) 48 (41.7) Median 1 (IQR 1–2) Median 10 (IQR 7.5–14)
1.000
0.007
14 (33.3) 28 (66.7) Median 3 (IQR 1–5.75) Median 11 (IQR 8–14)
< 0.001 0.505
24 (40.0) 25 (41.6) 11 (18.4)
15 (22.7) 23 (34.8) 28 (42.5)
< 0.001
61 (53.0) 29 (25.2) 25 (21.8)
0 (0.0) 13 (31.0) 29 (69.0)
< 0.001
Statistically significant values have been underlined to highlight relevant results. AJCC, American Joint Committee on Cancer; CI, confidence interval; IQR, interquartile range; LN, lymph node; LNR, lymph node ratio; SLN, sentinel lymph node.
Accuracy, sensitivity, specificity, positive predictive value, negative predictive value, pelvic lymphadenectomy reduction, and error rate of the lymph node ratio, number of positive inguinal lymph nodes, Cloquet’s lymph node status, and pelvic lymph node(s) status at computed tomographic scan
Table 2
Scenarios LNR A (< 0.1) Positive SLN Clinically positive LN ≥ 3 positive inguinal LNs Positive SLN Clinically positive LN Cloquet’s LN status Positive SLN Clinically positive LN Pelvic LN status at CT scan Positive SLN Clinically positive LN
Accuracy (%) Sensitivity (%) Specificity (%) PPV (%) NPV (%) Pelvic lymphadenectomy reduction (%) Error (%) 48.3 53.0
87.5 92.0
42.3 29.3
18.9 44.2
95.7 85.7
38.4 21.2
1.7 3.0
84.0 60.3
42.9 59.3
92.5 60.9
54.5 47.1
88.6 71.8
86.4 53.5
9.9 15.1
96.2 65.7
80.0 25.0
100.0 87.0
100.0 50.0
95.5 69.0
84.6 82.8
3.8 25.7
74.4 47.7
27.3 57.7
90.6 41.0
50.0 39.5
78.4 59.3
86.0 41.5
18.6 16.9
Pelvic lymphadenectomy reduction: patients predicted as pelvic LN negative/all the patients. Error rate: false negative/all the patients. False-negative patients were predicted as pelvic LN negative even they had pelvic LN metastasis. CT, computed tomography; LN, lymph node; LNR, lymph node ratio; NPV, negative predictive value; PPV, positive predictive value; SLN, sentinel lymph node.
a pelvic lymphadenectomy reduction of 21.1%, with an error rate of 3%. A number of positive inguinal LNs, Cloquet’s LN status, and preoperative CT scans had lower NPVs (71.8, 69.0, and 59.3% respectively): if these criteria were used to screen patients preoperatively, these NPVs would allow for a high pelvic reduction rate (53.5, 82.8, and 41.5%), but would be associated with a high error rate (15.1, 25.7, and 16.9%, respectively).
exploratory evaluation of prediction models. The number of positive inguinal LNs and the LNR were included in two separate multivariate models to avoid the issue of covariate collinearity as these variables share some information (i.e. the number of metastatic LNs). The two multivariable models were more accurate than single predictors, as shown by the areas under the ROC curves (Table 3).
Discussion Prediction model for patients with micrometastasis
Considering that the predictive criteria achieved a clinically relevant NPV and a low error rate in patients with micrometastasis (but not for macrometastasis), we performed an
The series analyzed in this study comprises a group of melanoma patients with LN metastasis who underwent an ilioinguinal LN dissection which, at our institution, is performed as the standard surgery for inguinal LN
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node metastasis in melanoma Pasquali et al. 465
Area under the receiver operating characteristic curve for each predictive variable and for two multivariable models, which consider the number of positive inguinal lymph node(s) and the inguinal lymph node ratio each, in patients with micrometastasis
Table 3
Number of positive inguinal LNa Inguinal LNRb Cloquet’s LN CT scan Model 1 (including number of positive inguinal LNa) Model 2 (including inguinal LNRb)
ROC area
SE
95% CI
0.677 0.613 0.719 0.688 0.784
0.070 0.045 0.064 0.073 0.078
0.566–0.778 0.503–0.723 0.605–0.811 0.579–0.789 0.685–0.873
0.816
0.070
0.713–0.892
CI, confidence interval; CT, computed tomography; LN, lymph node; LNR, lymph node ratio; ROC, receiver operating characteristic. a Categorical variable, cutoff 3. b Categorical variable, cutoff LNR A versus B and C.
melanoma metastasis. We investigated predictors of pelvic LN metastasis (the inguinal LNR, the number of positive inguinal LNs, the Cloquet’s LN status, and the preoperative CT) in patients with micrometastasis (i.e. who underwent lymphadenectomy for a positive SLNB) and macrometastasis (i.e. who underwent lymphadenectomy for clinically positive LNs) to identify patients at a low risk of pelvic LN metastasis.
dissection [32], can identify whether patients need a pelvic dissection with a high level of evidence. The predictive value of the inguinal LNR may be explained considering that this parameter includes the number of excised LNs. Patients with lower values of LNR (e.g. LNR A) can have less positive LNs or more excised LNs than patients who present higher values of LNR. Different hypotheses can explain the predictive role of the LNR. Intuitively, a thorough dissection allows identification of all metastatic LNs and provides an optimal stratification of the risk of harboring pelvic LN metastasis. Moreover, the number of excised LNs might be a surrogate of the immunological response against the tumor as the immune system shows a greater activity against melanoma cells in LNs that are far from those having melanoma metastasis [33]. This suggests that a greater number of inguinal nodes may limit the spread of melanoma cells to the pelvic nodes.
Inguinal LNR A (≤ 0.1) had an NPV of 95.7 and 85.7% in patients with micrometastasis and macrometastasis, respectively, for predicting the status of pelvic LNs that was associated with a potential reduction rate of 38.4 and 21.2%, respectively. The overall error rate was lower in patients with micrometastasis (1.7 vs. 3.0%). These results suggest that the inguinal LNR can be of particular relevance for patients with micrometastasis, who are at a low risk of harboring pelvic disease (17.2% in this study and 6–16% is other series [14–19]), and may be considered for a less extensive dissection.
The number of positive inguinal LNs, although associated independently with the pelvic LN status, shows lower NPVs than the desirable 90%, both in the case of micrometastasis (88.6%) and in the case of macrometastasis (71.8%). Although this parameter has been widely considered an independent predictor of pelvic LN metastasis, this study questions its predictive effectiveness in light of the relatively low NPVs and high error rates. Similarly, Strobbe and colleagues studied patients who underwent ilioinguinal dissection for clinically positive LNs, and reported an NPV of 78% for the number of positive inguinal LNs, a value far less than the desirable 90%. Further data are needed to ascertain whether the number of positive inguinal LNs identifies patients at a low risk of pelvic LN metastasis who may avoid pelvic dissection.
The association between inguinal LNR and pelvic LN metastasis was showed for the first time by Chu et al. [14] in 40 SLNB-positive patients. Our analysis adds meaningful information to the existing literature because it suggests that the presence of pelvic LN metastasis after a positive inguinal SLNB can be predicted with a high NPV (95.7%). Accordingly, roughly 40% of patients with a positive inguinal SLNB may avoid pelvic dissection, with only 1.7% of patients having LN disease left in their groin. The prediction of a pelvic LN metastasis can be limited by the routine methodology for the pathologic assessment of the pelvic LNs, which is standard hematoxylin and eosin (H&E) staining. This process may miss melanoma deposits in some LNs and therefore leads to a greater number of patients with a FN result, thus limiting the value of predictive factors and tools. Although H&E staining of the LNs after a lymphadenectomy remains the standard method to detect melanoma LN metastasis, only a randomized trial, such as the EAGLE FM Study from the Australia and New Zealand Melanoma Trial Group comparing the outcomes of inguinal and ilioinguinal
In this series, the pathological status of the Cloquet’s LN had an NPV of 95.5% (error rate 3.8%) and 69.0% (error rate 25.7%) in patients with micrometastasis and macrometastasis, respectively. Interestingly, if SLNB-positive patients with negative Cloquet’s LN had avoided pelvic dissection, the reduction rate would have been up to 84%, with an error rate of 3.8%. Previous studies report NPVs of 94–97% in the case of micrometastasis [34,35] and of 80–95% in the case of macrometastasis [36–38]. Despite these findings, in a study on 53 patients with micrometastasis, Chu et al. [34] noted that the Cloquet’s LN was rarely involved with melanoma cells (3.8%). A high concordance between the presence of metastasis in the Cloquet’s and pelvic LNs was observed (NPV of 94%), but, considering the low incidence of metastasis in the Cloquet’s node, the authors suggest avoiding performance of Cloquet’s LN examination in patients with micrometastasis. Remarkably, our series has a higher positivity rate of the Cloquet’s LN in the case of micrometastasis (15%, 4/26). Although the positivity rates of NPVs and Cloquet’s LN lead to consideration of
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
466
Melanoma Research 2014, Vol 24 No 5
Cloquet’s LN as an effective predictive tool after a positive SLNB, the error rate that we identified (3.8%) limits its clinical value along with other drawbacks intrinsic to the biopsy of the Cloquet’s LN. In fact, the lymph flow does not always pass from the inguinal to the pelvic LNs through the femoral canal, where the Cloquet’s LN is located [38,39]. Furthermore, in this study, we have investigated the diagnostic value of the preoperative CT scan, which poorly characterized the pelvic LNs. This is shown by the lack of independent predictive value and by the low NPVs, which are 78.4% in patients with micrometastasis (error rate 18.6%) and 59.3% in patients with macrometastasis (error rate 16.9%). Considering that CT scans have a diagnostic accuracy of greater than 90% for staging the LNs in patients with melanoma irrespective of their location [40], this imaging technique is expected to have a high diagnostic relevance for the characterization of pelvic LNs, especially for macrometastasis. Van der Ploeg et al. [41] reported an NPV of 91% for preoperative CT scans in patients with macrometastasis and concluded that pelvic dissection should be performed only in patients with positive pelvic LNs. In contrast, Allan et al. [42] studied the diagnostic value of the CT scan in 72 patients with macrometastasis and considered an NPV of 86% insufficient to identify patients who might safely avoid pelvic lymphadenectomy. As the characterization of the pelvic LNs by the CT scan seems more controversial than expected, authors are considering PETCT for selection of patients for pelvic dissection [43]. Finally, this study explored the predictive accuracy of the considered selection criteria in multivariable predictive tools for patients with micrometastasis who are at a lower risk of pelvic LN metastasis. In fact, a reduction in the extension of groin dissection is less safe in patients with macrometastasis because of the greater risk of LN disease and the lower predictive values shown by the current selection criteria. The area and the ROC curve was greater for multivariable models – two models were constructed that considered the positive inguinal LNs and the inguinal LNR separately to avoid the issue of multicollinearity than for each predictor (Table 3), suggesting that a predictive tool encompassing several selection criteria can be valuable in the identification of patients who may safely avoid pelvic dissection. There are some limitations to the clinical applicability of the results of this study. Clearly, the availability of inguinal LNR after surgery limits its clinical utility. Despite this drawback, given that roughly one-third of surgeons perform only inguinal dissection for groin melanoma metastasis irrespective of other risk factors [13], the inguinal LNR may be considered for patient counseling and follow-up as low LNR values suggest a high likelihood of having negative pelvic LNs. The value of LNR is influenced by the LN count, suggesting that
the achievement of satisfactory nodal retrieval is of great importance to allow risk stratification for pelvic metastasis according to the LNR [10,17]. Furthermore, if a predictive tool, such as a nomogram, will include inguinal LNR as well as other predictive variables (including those considered in this study and others, such as lymph flow drainage) and achieve high predictive values, pelvic dissection may be tailored on patient and tumor characteristics and performed as a second operation in the relatively small group of patients with pelvic disease after a positive SLNB (6–17.2%). Nevertheless, given these limitations and the above-mentioned risk of pathological downstage of LN metastasis, only a randomized trial, such as the EAGLE FM Study [32], can define the optimal management of patients with groin LN melanoma metastasis. Conclusion
Pending a randomized trial, this study offers interesting insights for the identification of SLN-positive patients at a low risk of pelvic LN metastasis. In SLN-positive patients, avoiding pelvic dissection may be safe when LNR is lower than 0.1 and Cloquet’s LN is negative, although the error rate was slightly higher in the latter case. The use of the number of positive inguinal LNs and the preoperative CT scan may offer a less accurate prediction, on the basis of the lower NPVs and higher error rate. This study also shows that the prediction of pelvic metastasis seems less accurate for patients with clinically positive LNs, suggesting that pelvic dissection cannot safely be avoided in this scenario.
Acknowledgements The authors thank Ms Chiara Longhi and Mrs Marilyn Ward for editing the manuscript. Conflicts of interest
There are no conflicts of interest.
References 1 2
3
4
5 6
7
Mack LA, McKinnon JG. Controversies in the management of metastatic melanoma to regional lymphatic basins. J Surg Oncol 2004; 86:189–199. Morton DL, Thompson JF, Cochran AJ, Mozzillo N, Elashoff R, Essner R, et al. Sentinel-node biopsy or nodal observation in melanoma. N Engl J Med 2006; 355:1307–1317. Guggenheim MM, Hug U, Jung FJ, Rousson V, Aust MC, Calcagni M, et al. Morbidity and recurrence after completion lymph node dissection following sentinel lymph node biopsy in cutaneous malignant melanoma. Ann Surg 2008; 247:687–693. Hughes TM, A’Hern RP, Thomas JM. Prognosis and surgical management of patients with palpable inguinal lymph node metastases from melanoma. Br J Surg 2000; 87:892–901. Hughes TM, Thomas JM. Combined inguinal and pelvic lymph node dissection for stage III melanoma. Br J Surg 1999; 86:1493–1498. Kretschmer L, Neumann C, Preusser KP, Marsch WC. Superficial inguinal and radical ilioinguinal lymph node dissection in patients with palpable melanoma metastases to the groin – an analysis of survival and local recurrence. Acta Oncol 2001; 40:72–78. National Comprehensive Cancer Network. Clinical practice guidelines in oncology. Melanoma (2013). Available at: http://www.nncn.org. [Accessed 1 March 2014].
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Pelvic lymph node metastasis in melanoma Pasquali et al. 467
8 Australian Cancer Network Melanoma Guidelines Revision Working Party. Clinical practice guidelines for the management of melanoma in Australia and New Zealand. Cancer Council Australia and Australian Cancer Network, Sydney and New Zealand Guidelines Group, Wellington, 2008. 9 Shada AL, Walters DM, Tierney SN, Slingluff CL Jr. Surgical resection for bulky or recurrent axillary metastatic melanoma. J Surg Oncol 2012; 105:21–25. 10 Spillane AJ, Cheung BL, Stretch JR, Scolyer RA, Shannon KF, Quinn MJ, et al. Proposed quality standards for regional lymph node dissections in patients with melanoma. Ann Surg 2009; 249:473–480. 11 Spillane AJ, Tucker M, Pasquali S. A pilot study reporting outcomes for melanoma patients of a minimal access ilio-inguinal dissection technique based on two incisions. Ann Surg Oncol 2011; 18:970–976. 12 Italian National Research Council. Progetto oncologia. Available at: http:// progettooncologia.cnr.it/bridge/attivita-direzione.html. [Accessed 1 March 2014]. 13 Pasquali S, Spillane AJ, de Wilt JH, McCaffery K, Rossi CR, Quinn MJ, et al. Surgeons’ opinions on lymphadenectomy in melanoma patients with positive sentinel nodes: a worldwide web-based survey. Ann Surg Oncol 2012; 19:4322–4329. 14 Chu CK, Delman KA, Carlson GW, Hestley AC, Murray DR. Inguinopelvic lymphadenectomy following positive inguinal sentinel lymph node biopsy in melanoma: true frequency of synchronous pelvic metastases. Ann Surg Oncol 2011; 18:3309–3315. 15 Mozzillo N, Caracò C, Marone U, Di Monta G, Crispo A, Botti G, et al. Superficial and deep lymph node dissection for stage III cutaneous melanoma: clinical outcome and prognostic factors. World J Surg Oncol 2013; 11:36. 16 Santinami M, Carbone A, Crippa F, Maurichi A, Pellitteri C, Ruggeri R, et al. Radical dissection after positive groin sentinel biopsy in melanoma patients: rate of further positive nodes. Melanoma Res 2009; 19:112–118. 17 Spillane AJ, Haydu L, McMillan W, Stretch JR, Thompson JF. Quality assurance parameters and predictors of outcome for ilioinguinal and inguinal dissection in a contemporary melanoma patient population. Ann Surg Oncol 2011; 18:2521–2528. 18 Van der Ploeg IM, Kroon BB, Valdés Olmos RA, Nieweg OE. Evaluation of lymphatic drainage patterns to the groin and implications for the extent of groin dissection in melanoma patients. Ann Surg Oncol 2009; 16:2994–2999. 19 Zdzienicki M, Rutkowski P, Nowecki ZI, van Akkooi AC, Michej W, Dziewirski W, et al. The analysis of the outcomes and factors related to iliac–obturator involvement in cutaneous melanoma patients after lymph node dissection due to positive sentinel lymph node biopsy or clinically detected inguinal metastases. Eur J Surg Oncol 2013; 39:304–310. 20 Mocellin S, Thompson JF, Pasquali S, Montesco MC, Pilati P, Nitti D, et al. Sentinel node status prediction by four statistical models: results from a large bi-institutional series (n =1132). Ann Surg 2009; 250:964–969. 21 Pasquali S, Mocellin S, Campana LG, Vecchiato A, Bonandini E, Montesco MC, et al. Maximizing the clinical usefulness of a nomogram to select patients candidate to sentinel node biopsy for cutaneous melanoma. Eur J Surg Oncol 2011; 37:675–680. 22 Mocellin S, Ambrosi A, Montesco MC, Foletto M, Zavagno G, Nitti D, et al. Support vector machine learning model for the prediction of sentinel node status in patients with cutaneous melanoma. Ann Surg Oncol 2006; 13:1113–1122. 23 Mocellin S, Pasquali S, Rossi CR, Nitti D. Validation of the prognostic value of lymph node ratio in patients with cutaneous melanoma: a populationbased study of 8177 cases. Surgery 2011; 150:83–90. 24 Spillane AJ, Cheung BL, Winstanley J, Thompson JF. Lymph node ratio provides prognostic information in addition to american joint committee on cancer N stage in patients with melanoma, even if quality of surgery is standardized. Ann Surg 2011; 253:109–115. 25 Rossi CR, Mocellin S, Pasquali S, Pilati P, Nitti D. N-ratio: a novel independent prognostic factor for patients with stage-III cutaneous melanoma. Ann Surg Oncol 2008; 15:310–315.
26
27
28
29
30
31
32
33
34
35
36
37 38
39
40
41
42 43
Berger AC, Fierro M, Kairys JC, Berd D, Sato T, Andrel J, et al. Lymph node ratio is an important and independent prognostic factor for patients with stage III melanoma. J Surg Oncol 2012; 105:15–20. Grotz TE, Huebner M, Pockaj BA, Perkins S, Jakub JW. Limitations of lymph node ratio, evidence-based benchmarks, and the importance of a thorough lymph node dissection in melanoma. Ann Surg Oncol 2013; 20:4370–4377. Egger ME, Scoggins CR, Martin RC II, Stromberg AJ, Quillo AR, McMasters KM, Callender GG. The lymph node ratio has limited prognostic significance in melanoma. J Surg Res 2013; 179:10–17. Rossi CR, De Salvo GL, Bonandini E, Mocellin S, Foletto M, Pasquali S, et al. Factors predictive of nonsentinel lymph node involvement and clinical outcome in melanoma patients with metastatic sentinel lymph node. Ann Surg Oncol 2008; 15:1202–1210. Azzola MF, Shaw HM, Thompson JF, Soong SJ, Scolyer RA, Watson GF, et al. Tumor mitotic rate is a more powerful prognostic indicator than ulceration in patients with primary cutaneous melanoma: an analysis of 3661 patients from a single center. Cancer 2003; 97:1488–1498. Francken AB, Shaw HM, Thompson JF, Soong SJ, Accortt NA, Azzola MF, et al. The prognostic importance of tumor mitotic rate confirmed in 1317 patients with primary cutaneous melanoma and long follow-up. Ann Surg Oncol 2004; 11:426–433. ANZMTG 01.12. Inguinal or ilio-inguinal lymphadenectomy for patients with metastatic melanoma to groin lymph nodes and no evidence of pelvic disease on PET/CT scan – a randomised phase III trial (EAGLE FM). Available at: http://www.anzmtg.org/doc/16034%20EAGLE%20FM% 20Hamburg%20Poster%2027Jun13_al.pdf. [Accessed 28 September 2013]. Cochran AJ, Huang RR, Lee J, Itakura E, Leong SP, Essner R. Tumourinduced immune modulation of sentinel lymph nodes. Nat Rev Immunol 2006; 6:659–670. Chu CK, Zager JS, Marzban SS, Gimbel MI, Murray DR, Hestley AC, et al. Routine biopsy of Cloquet’s node is of limited value in sentinel node positive melanoma patients. J Surg Oncol 2010; 102:315–320. Essner R, Scheri R, Kavanagh M, Torisu-Itakura H, Wanek LA, Morton DL. Surgical management of the groin lymph nodes in melanoma in the era of sentinel lymph node dissection. Arch Surg 2006; 141:877–882, discussion 882–884. Illig L, Aigner KR, Biess B, Göhl J, Holle R, Hundeiker M, et al. Diagnostic excision of the Rosenmüller’s node. Screening for occult metastases before elective regional lymph node dissection in patients with lower limb melanoma? Cancer 1988; 61:1200–1206. Coit DG, Brennan MF. Extent of lymph node dissection in melanoma of the trunk or lower extremity. Arch Surg 1989; 124:162–166. Strobbe LJ, Jonk A, Hart AA, Peterse JL, Wobbes T, Nieweg OE, Kroon BB. The value of Cloquet’s node in predicting melanoma nodal metastases in the pelvic lymph node basin. Ann Surg Oncol 2001; 8:209–214. Van der Ploeg IM, Valdés Olmos RA, Kroon BB, Nieweg OE. Tumor-positive sentinel node biopsy of the groin in clinically node-negative melanoma patients: superficial or superficial and deep lymph node dissection? Ann Surg Oncol 2008; 15:1485–1491. Xing Y, Bronstein Y, Ross MI, Askew RL, Lee JE, Gershenwald JE, et al. Contemporary diagnostic imaging modalities for the staging and surveillance of melanoma patients: a meta-analysis. J Natl Cancer Inst 2011; 103:129–142. Van der Ploeg AP, van Akkooi AC, Schmitz PI, van Geel AN, de Wilt JH, Eggermont AM, Verhoef C. Therapeutic surgical management of palpable melanoma groin metastases: superficial or combined superficial and deep groin lymph node dissection. Ann Surg Oncol 2011; 18:3300–3308. Allan CP, Hayes AJ, Thomas JM. Ilioinguinal lymph node dissection for palpable metastatic melanoma to the groin. ANZ J Surg 2008; 78:982–986. Wevers KP, Bastiaannet E, Poos HP, van Ginkel RJ, Plukker JT, Hoekstra HJ. Therapeutic lymph node dissection in melanoma: different prognosis for different macrometastasis sites? Ann Surg Oncol 2012; 19:3913–3918.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
