NN are growth-arrested melanocytic clusters that are usually diagnosed in SNs draining melanoma, breast, vulvar, or penile cancer. Unlike most melanomas, NN usually have bland cytomorphology and no mitotic activity. With the exception of intranodal blue nevi, NN do not express HMB-45 or express it only weakly.(Gonzàlez-Farré, Ronen et al. 2020) Histological images of NN are shown in Fig. 2. Recently, p16 (mostly positive for NN, Fig. 2)(Piana, Tagliavini et al. 2015) and PRAME (mostly negative for NN)(See, Finkelman et al. 2020) have been introduced in the differential diagnosis to SN metastases of melanoma.
The NN prevalence reported in the literature vary considerably, from 0 to 22 percent.(McCarthy, Palmer et al. 1974, Fisher, Hill et al. 1994, Biddle, Evans et al. 2003, Holt, Sangueza et al. 2004, Gambichler, Scholl et al. 2013, Piana, Tagliavini et al. 2015, Smith, Coelho et al. 2016, Kim, Seo et al. 2018, de Beer, van Diest et al. 2019, Yamashita, Pockaj et al. 2020) The likelihood of an NN diagnosis is certainly influenced by many factors, including the type of cancer, the pathologic protocol used, the lymphatic drainage conditions, the size of the lymph nodes, the number of lymph nodes removed, and the frequency and extent of malignant metastases. Using an extensive pathologic protocol, we observed an overall prevalence of 15.7% in a large population with SNB for cutaneous melanomas. The specific prevalences for cervical, axillary, and groin SNs were 16.5%, 19.4%, and 9.8%, respectively.
Regarding etiology, it has been speculated that during embryogenesis NN develop from melanocytic precursor cells that migrate from the neural crest to the skin, taking the dorsolateral developmental pathway. (Kos, Reedy et al. 2001) Melanocytic precursors migrate through the mesenchyme, where the lymphatic system is formed almost simultaneously. It has been argued that the location of NN predominantly in fibrous structures of lymph nodes, the association of NN with congenital nevi (Bowen, Duffy et al. 2015), and the very presence of neurocutaneous melanosis would support the theory of cell arrest during embryogenesis.(Carson, Wen et al. 1996)
On the other hand, there are a number of good arguments that NN stem from melanocytic cells of the skin that have migrated via afferent lymphatic vessels: (1) In agreement with Carson et al.(Carson, Wen et al. 1996), at least in univariate analysis, we documented a significantly increased probability of NN in patients whose primary melanomas were histologically associated with a cutaneous nevus. (2) Invasion of lymphatic vessels by nevus cells has been observed within cutaneous nevi.(McCarthy, Palmer et al. 1974, Bell, Hill et al. 1979, Subramony and Lewin 1985, Hara 1993) (3) Cutaneous nevi in the drainage area of lymph nodes with NN have been observed in several case series.(McCarthy, Palmer et al. 1974, Hara 1993, Fontaine, Parkhill et al. 2002, Holt, Sangueza et al. 2004, Hu, Ren et al. 2020) (4) NN have been much more frequent in SNs than in non-SNs (Carson, Wen et al. 1996, Holt, Sangueza et al. 2004, Gambichler, Scholl et al. 2013) or lymphadenectomy specimens.(Ridolfi, Rosen et al. 1977, Fisher, Hill et al. 1994) (5) NN have not been described in deeply located nodes of the thorax or abdomen, which do not drain the skin. (McCarthy, Palmer et al. 1974) (6) Lymph nodes from melanoma surgery are more likely to contain NN than superficial lymph nodes excised for other cancers.(Biddle, Evans et al. 2003, Piana, Tagliavini et al. 2015) (7) Electron microscopic studies of NN have revealed ultrastructural features identical to those of intradermal nevus cells.(Erlandson and Rosen 1982)
We demonstrated that NN can be found in SNs related to all growth patterns of primary melanoma. This includes rare growth patterns such as desmoplastic and nevoid melanomas. The SNs of three out of four patients with malignant blue nevus displayed an NN.
Using multivariate analyses, we discovered a very robust association between multiple nevi of the skin and the prevalence of NN, both in SN-negative and SN-positive patients. To the best of our knowledge, only one small case-control study including 22 patients with NN has confirmed this observation.(Ribero, Longo et al. 2017) We found that the midline location of the primary melanoma as well as bidirectional lymphatic drainage resulted in a higher number of excised SNs, which logically increases the chance of detecting NN. As many as 31% of patients with upper trunk melanoma had bidirectional lymphatic drainage.(Kretschmer, Hellriegel et al. 2019) A lower NN prevalence in inguinal lymph nodes has already been reported. (McCarthy, Palmer et al. 1974, Gambichler, Scholl et al. 2013) By reporting prevalence for cervical SNs for the first time, we can state a significantly higher NN prevalence for SNs draining the upper part of the body, which hold true in multivariate logistic regression. This coincides with the distribution of cutaneous nevi. Also their density seems to be higher on the upper half of the body.(Echeverría, Bulliard et al. 2014) Many studies have shown an association between sun exposure and the number of skin nevi. Cervical lymph nodes drain a relatively small area of skin that is chronically exposed to sunlight. Axillary lymph nodes drain large areas of intermittently sun-exposed skin, which is characterized by high skin nevus count.(Newton-Bishop, Chang et al. 2010) In a longitudinal study, an increase in skin nevi after 15 years of observation was registered only on the upper parts of the body.(Ribero, Zugna et al. 2020) We confirmed that primary melanomas located at the arm were associated with higher NN rates than leg-located melanomas.(Yamashita, Pockaj et al. 2020) It has been shown that also the density of cutaneous nevi of the arms is greater than that of nevi of the legs.(Harrison, Buettner et al. 1999) Ultraviolet light seems to induce the expression of growth factors involved in the early migration process of malignant melanocytes.(Wäster, Orfanidis et al. 2017)
Like others, we found no sex-specific difference. Fitzpatrick skin phototype was not associated with NN, an observation not previously reported. The influence of multiple solar lentigines was not convincing because it had different signs depending on SN status (Table 3).
We can only speculate about the possible implications of our surprising observation that patients with melanoma in whom at least one parent had undergone another life-threatening cancer were significantly more likely to be diagnosed with NN. In parallel, multiple nevi of the skin were also highly significantly associated with a history of cancer in the parents. Genome-wide association studies have described a larger number of susceptibility loci for melanoma that are related not only to nevus count, pigment type, and tanning response but also to telomere maintenance and DNA repair mechanisms. From this perspective, a link between multiple melanocytic nevi in skin and lymph nodes and a susceptibility to develop other tumor entities seems potentially explainable.(Landi, Bishop et al. 2020)
Analyzing 56 SN-negative patients with NN, Yamashita et al observed a non-significantly lower recurrence-rate, compared with purely SN-negatives and suggested a metastasis-protective effect of NN.(Yamashita, Pockaj et al. 2020) Most previous studies (Gambichler, Scholl et al. 2013, Smith, Coelho et al. 2016, Kim, Seo et al. 2018, de Beer, van Diest et al. 2019, El Sharouni, Laeijendecker et al. 2021) have concluded that the presence of NN in SN-negative cases does not affect survival. We can generally confirm this statement but found a trend toward slightly higher mortality in SN-negatives with NN (Fig. 1A). A large Dutch registry study of 11,274 patients confirmed a similar trend. (de Beer, van Diest et al. 2019) Possibly, in daily practice SN metastases are misinterpreted as NN in rare cases. For example, metastases of nevoid melanomas can strongly resemble NN.(Biddle, Evans et al. 2003, Davis, Patil et al. 2016) Very small metastases, metastases in intracapsular lymphatic vessels, NN in the presence of concurrent SN metastases, and lack of HMB-45 reactivity of SN metastases represent further diagnostic pitfalls. (Biddle, Evans et al. 2003, Gonzàlez-Farré, Ronen et al. 2020, Lezcano, Pulitzer et al. 2020) Very rarely, metastases of other kinds of cancer, e. g., lobular breast cancer may mimic NN.(Fisher, Hill et al. 1994)
Prevalence and prognostic impact of NN diagnosis in SN-positive patients have not been reported. In our study, patients who were both SN and NN positive had a slightly higher survival rate than SN-positive patients without NN. We demonstrated that the likelihood of an NN diagnosis decreases with increasing Breslow thickness and, relatedly, with increasing SN metastasis size. This alone explains why the SN-positive patients with concurrent NN diagnosis fared somewhat better (Fig. 1B).
Our study has several limitations, including retrospective data analysis, the failure to record the number of SNs affected with NN per subject, the microanatomical location of NN within SNs, and the histologic specificities of NN according to the growth patterns of primary melanoma. Multiple testing is another problem; we cannot exclude the possibility that some error probabilities are due to chance.
In summary, using multivariate analyses, we found a highly significantly increased prevalence of NN in SNs in patients with upper body melanomas, midline primary melanomas, and those with more than 50 nevi of the skin. A history of non-cutaneous malignancies in parents was associated with increased NN prevalence, as were multiple nevi of the skin. When the primary melanoma was histologically associated with a cutaneous nevi, NNs were also more frequently found in SNs. In SN-positive patients, the likelihood of an NN diagnosis decreased with increasing metastatic diameter. The slightly worse melanoma-specific survival of SN-negative patients with NN that we observed suggests that careful differentiation of NN and SN metastases is sometimes difficult and should be done very carefully.