Melanoma is the deadliest form of skin cancer and its incidence has doubled every 20 years in France, where this cancer is responsible of more than 1600 deaths each year.

Patients with early diagnosis have good prognosis and can be generally cured by surgery only. Advanced melanoma however has a very bad prognosis.

Loco-regional lymph nodes are usually the first distant localization in metastatic melanoma. Lymph node dissection is then the recommended treatment, although it's impact on survival has never been proven.

In the same way, the benefit risk profile of interferon as adjuvant treatment after lymph node dissection is still much debated.

Recently, new treatments either with immunotherapy (ipilimumab, nivolumab) or by the targeted therapy dabrafenib/trametinib in patients with BRAF mutation have shown an impact on survival in the adjuvant setting after lymph node dissection.

But, it has not yet been established if this strategy has a benefit gain compared to starting those treatments only in the metastatic setting after watchful follow-up.

Moreover, if these novel therapies (targeted therapies: TT, immunotherapies : IT) demonstrated for the first time a real benefit in terms of survival or of responses rates in melanoma, physicians and patients had to address new problems, such as the management of unusual adverse events.

Partial and dissociated responses can also be seen with those new treatments. Some patients will have complete response in some lesions, stabilization in others and progression in a few. It is to be expected that one of the real key points of this therapy is to be found here, as this situation is commonly seen, and it would probably be a poor choice to stop a treatment that is globally effective for progression of only 1 or 2 lesions, in a patient otherwise stabilized.

That is the context in which interventional radiology (IR) should be considered as an extremely efficient option. IR is a real medical revolution in the last 2 decades.

It provides not only the opportunity to determine the characteristics of residual lesion (fibrosis, necrosis, metastasis, or sarcoidosis,…) by biopsy, but allows also their targeted destruction through different technics (cryotherapy, radiofrequency, laser,…).

The investigators are fortunate to have in their institution one of the best IR department of the world (headed by Prof. Afshin GANGI), with a technical platform unique in Europe that allows IR through ultrasound, scan, petscan and MRI.

To the best of their knowledge, Immunotherapy associated with IR has not been performed so far.

This association could in theory:

1. Combine immunotherapy with tumoral necrosis, which inherently increases the effects of immunotherapy by massive tumoral leakage of danger signals and tumoral antigens;
2. Allow direct injection in targeted zones, where the beneficial effect is desired, and thus increase the expected immune response;
3. Reduce side effects related to immunotherapy, by reducing quantities injected; which seems particularly important in the (neo)-adjuvant setting.

That's why the investigators are willing to conduct this pilot project, the objectives of which are:

1. Providing a proof of the feasibility of this association,
2. Obtaining preliminary insights on the effects on non-targeted lesions,
3. Adding a translational research to establish the effect on tumor antigenic expression and the immune response.

Inclusion Criteria:

1. Subject with previously treated (or unknown) primary stage IIIB/IIIC melanoma with lymph node metastases, with at least one ≥1 cm3 that is accessible to cryotherapy under interventional radiology (IR), assessed at the inclusion visit (V1).
2. ECOG performance status (0 to 2) at the selection visit (V0)
3. If woman or man are of reproductive age, they should have no desire to procreate for the duration of their participation in the study, agreeing to use a highly effective contraception method*, with at least one barrier method (condom or diaphragm), and for 5 months after the infusion of Nivolumab for women and 7 months for men
4. Leukocytes >2000/mm3, neutrophils >1500/mm3, platelets >100,000/mm3, haemoglobin >9g/dl at the selection visit (V0)
5. Total bilirubin <1.5 mg/dl except for patients with Gilbert's syndrome who may have total bilirubin <3.0 mg/dl at the selection visit (V0)
6. Liver function: SGOT, SGPT, ALP <2.5 N, if liver metastases SGOT and SGPT <5 N at the selection visit (V0)
7. Serum creatinine <1.5 N or creatinine clearance >40 ml/min (using the Cockcroft-Gault equation) at the selection visit (V0)

Exclusion Criteria:

1. Pregnancy (women of childbearing potential : positive blood pregnancy test at the selection visit (V0) or breastfeeding

2. History of hypersensitivity to ipilimumab, nivolumab or one of the excipients

3. History of severe hypersensitivity to a monoclonal antibody

4. History of positive tests for HIV or Acquired Immunodeficiency Syndrome (HIV assessed at the selection visit (V0))

5. Positive tests for HCV or HBV indicating an acute or chronic infection at the selection visit (V0)

6. Patient presenting with an active, known or suspected autoimmune disease. The following, however, may participate:

   • patient with type 1 diabetes or hypothyroidism requiring substitute hormone therapy only;
   • patient with psoriasis, vitiligo or alopecia;
   • patient with conditions that are not known to reoccur without an exogenous triggering agent.

7. History of active neoplasia during the last 3 years with the exception of localised curable cancers considered to be cured, such as basal or squamous cell carcinomas, superficial bladder cancer and prostate, colon or breast carcinoma in situ.

8. Active systemic infection

9. Patient with a condition requiring systemic steroid treatment (at a dose >10 mg/prednisone equivalent or at unstable dose) or another immunosuppressant within/following 14 days of study drugs administration. Inhaled steroids and treatment for adrenal insufficiency, however, are permitted.

10. Contraindication for the cryotherapy procedure as assessed by the radiologist (due to tumour size or proximity to a vascular or nerve structure giving the procedure an unacceptable level of risk) at the inclusion visit (V1)

11. Clotting disorder that may interfere with the cryoablation, assessed at the selection visit (V0)

12. Contraindication for MRI with gadolinium-based contrast media

13. History of uveal melanoma

14. Patient having received prior treatment for their melanoma, in particular, patient previously treated with interferon fewer than 3 months ago or with an anti-PD-1, anti-PD-L1, anti-PD-L2, anti-CD-137 or anti-CTLA-4 antibody. Patients who have undergone primary melanoma surgery and the excision of in-transit metastases, as well as adjuvant therapy with interferon if completed over 3 months ago are eligible.

15. Contraindication for general and/or local anaesthesia

16. Patient operated on under general anaesthesia, within the 4 weeks prior to the planned Nivolumab infusion.

17. Patient operated on epidural anaesthesia, within the 72 h prior to the planned Nivolumab infusion.

18. History of anterior organ transplantation, including allograft stem cell transplantation.

19. History of interstitial lung disease