Neoadjuvant Therapy in Resectable Advanced Melanoma: Swiss Real-World Data

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript "Neoadjuvant therapy in resectable advanced melanoma: Swiss real-world data" reports real world data obtained with neoadjuvant therapy of malignant melanoma at one center. The study includes 31 patients with significant heterogeneity with regards to patient's characteristics as well treatment choice.

The manuscript contains a large amount of information, some of it distributed in different chapters and tables. The authors could try to shorten the manuscript without loss of information.

In the abstract, line 40, it should read "including 5 of 18 (28%)" instead of "6 of 18 (28%). 

An interesting aspect concerns the assessment of pathological response: in this study, all specimens from a given patient were taken into account, whereas in the NADINA trial only the lymph nodes were evaluated. This difference is important for the individual assessment of a patient’s response. It is also known that tumor heterogeneity is highest in the primary lesion, but response in lymph nodes may be more relevant for early relapse. The authors should consider discussing this point.

Author Response

We thank the reviewer for their most valuable comments, suggestions, and criticisms. We have carefully considered them and offer the following comments and corrections. Changes in the manuscript are underlined, manuscript text is written in italic and the reviewer’s comments are marked bold in the point-to-point letter.

Point-to-point reply to the reviewers’ comments

Comments to the authors

The manuscript "Neoadjuvant therapy in resectable advanced melanoma: Swiss real-world data" reports real world data obtained with neoadjuvant therapy of malignant melanoma at one center. The study includes 31 patients with significant heterogeneity with regards to patient's characteristics as well treatment choice.

The manuscript contains a large amount of information, some of it distributed in different chapters and tables. The authors could try to shorten the manuscript without loss of information.

Thank you for this helpful suggestion. In the current revised version, we have performed an additional reduction of the manuscript length.

In the abstract, line 40, it should read "including 5 of 18 (28%)" instead of "6 of 18 (28%). 

Thank you for this important comment, we corrected this mistake.

A major pathologic response (MPR) was achieved in 12 of 31 patients (38%) overall, in-cluding 5 of 18 (28%) in the NADINA cohort, 6 of 10 (60%) in the SWOG S1801 cohort and 1 of 3 (33%) in the mucosal cohort.

An interesting aspect concerns the assessment of pathological response: in this study, all specimens from a given patient were taken into account, whereas in the NADINA trial only the lymph nodes were evaluated. This difference is important for the individual assessment of a patient’s response.

Thank you for this helpful comment. We also see a big importance in the differences in pathologic evaluation and therefore accentuated this aspect in the revised version.

In the NADINA trial pathological response was evaluated only in the lymph node specimens, even if in-transit metastases were excised[26]. In the SWOG S1801 trial and in our study pathological response was evaluated across all specimens of a patient, which may have led to important differences in the individual assessment of patients’ responses.    

It is also known that tumor heterogeneity is highest in the primary lesion, but response in lymph nodes may be more relevant for early relapse. The authors should consider discussing this point.

Thank you for this attentive remark. We also thought about this important aspect but didn’t address it.

Another important factor may be the presence of lymph node involvement, which constituted a mandatory inclusion criterion in the NADINA trial, but not in the SWOG S1801 trial. Among the 5 patients in our cohort without lymph node involvement, we observed an MPR in 1 patient (20%, 1/5), two pPR (40%, 2/5), and two pNR (40%, 2/5). In the NEO-MEL study, MPR rates also differed by this factor, with 16% (3/19) for isolated in-transit metastases and 49% (39/80) for isolated lymph node disease [27]. An important topic to discuss in this context is the tumor heterogeneity [28]. The immune phenotype distribution is significantly different in metastatic samples compared to primary melanomas [29,30]. Primary tumors often show more lymphoid aggregates and histologic regression than metastases, suggesting a stronger initial immune presence, but response in lymph nodes may be more relevant for early relapse.

Out of these discrepancies rise two questions. Firstly, is the immunogenicity of primary tumors and subcutaneous metastases sufficient to elicit a meaningful preoperative response to immunotherapy, thereby justifying a neoadjuvant approach over direct excision followed by adjuvant therapy? Secondly, do the pathological changes observed in primary tumors or subcutaneous metastases reflect the patient’s anti-tumor response and thus help guide treatment decisions, such as the need for adjuvant therapy? Thirdly, is the pathologic response in a lymph node a better surrogate marker for response and relapse compared to subcutaneous metastases or primary tumors? Further translational studies are needed to analyze pathologic alterations and immunologic response patterns in the different tissues to clarify the efficiency of neoadjuvant therapy for patients with solely in-transit metastasis and locally advanced primary tumors.

We emphasized the aspect in the limitations as well.

Moreover, our cohort included patients without lymph node involvement, a group excluded from the NADINA trial, thereby broadening the clinical spectrum of neoadjuvant therapy. Consequently, pathological assessment differed from that in NADINA and should be considered when interpreting response rates. Along with the single-center design, limited sample size and relatively short follow-up, these factors limit the generalizability of our findings.

We hope that the corrections and modifications performed make our manuscript now acceptable for publication in Cancers.

 

We look forward to hearing from you about this matter at your earliest convenience.

 

Yours sincerely,

Lara Valeska Maul (on behalf of all co-authors)     

Reviewer 2 Report

Comments and Suggestions for Authors

I have read with interest the article by Blumenrother and coll. reporting the data from the Swiss real-world study on neoadiuvant therapy in advanced melanoma (stages III and IV). The article is interesting and the topic is of great interest.

The Authors report their results with the 2 available immunotherapies (IPI+NIVO and PEMBRO) in different clinical settings/features (stage III, stage Iv, cutaneous and mucosal melanoma). They evaluated a cohort of 31 patients in total and their different scores, either in term of pathological, radiological and metabolic results. This is of great importance and allow us to properly obtain various interpretations of data, surely useful and with great repercussion in clinical practice.

Notwithstanding, I believe that some changes are due, and the suggestions/criticisms are listed below:

• First, cutaneous and mucosal melanoma cases have been included in the paper. Since mucosal melanomas account for a minimal part of the cases (3), and also given that they present a peculiar behavior, and that such patients received different drug schedule and dosage, they should not be included in the results.
• The groups are composed by different number of patients (as expected) but any effort is required to obtain homogeneous groups (if any). For instance, adjuvant radiotherapy has been administered in such a small number of subjects, I would suggest to delete them.
• Figure 1; please report the data in bar diagram (instead of the cake)
• Moreover, in my opinion, a part of Table 1 (patients characteristics) should be moved to the Supplementary materials. Indeed, just the principal data needs to be reported in the Table along the text (i.e., ECOG, LDH could be reported in the Supplementary..)
• Accordingly, Table 2 could be shortened as only the principal data might be reported.
• In summary, I suggest to re-evaluate data once the cohort has been made by more homogeneous groups of patients. I am strongly convinced that having heterogenous groups of patients with different treatments could represent an important bias, thus leading to incorrect interpretation of data.

Author Response

We thank the reviewer for their most valuable comments, suggestions, and criticisms. We have carefully considered them and offer the following comments and corrections. Changes in the manuscript are underlined, manuscript text is written in italic and the reviewer’s comments are marked bold in the point-to-point letter.

 

Point-to-point reply to the reviewers’ comments

Comments to the authors

I have read with interest the article by Blumenrother and coll. reporting the data from the Swiss real-world study on neoadiuvant therapy in advanced melanoma (stages III and IV). The article is interesting and the topic is of great interest.

The Authors report their results with the 2 available immunotherapies (IPI+NIVO and PEMBRO) in different clinical settings/features (stage III, stage IV, cutaneous and mucosal melanoma). They evaluated a cohort of 31 patients in total and their different scores, either in term of pathological, radiological and metabolic results. This is of great importance and allow us to properly obtain various interpretations of data, surely useful and with great repercussion in clinical practice.

Notwithstanding, I believe that some changes are due, and the suggestions/criticisms are listed below:

First, cutaneous and mucosal melanoma cases have been included in the paper. Since mucosal melanomas account for a minimal part of the cases (3), and also given that they present a peculiar behavior, and that such patients received different drug schedule and dosage, they should not be included in the results.

Thank you for this important suggestion. As you know, mucosal melanoma is associated with a more aggressive disease course and generally poorer responses to current therapeutic approaches. Owing to the rarity of this melanoma subtype, data on neoadjuvant treatment remain scarce. In the SWOG S1801 trial, patients with mucosal melanoma were analyzed together with the overall study population. Given the distinct biological behavior you highlighted, we deliberately chose to summarize mucosal melanoma cases separately in our manuscript.

In our mucosal cohort, we reported a low 9-month EFS rate of 33%. These findings should be interpreted with caution because of the small sample size (n = 3).

This subgroup is therefore presented as a descriptive case series rather than as part of the main efficacy analysis, allowing us to report these rare real-world cases without conflating them with the cutaneous melanoma results.

4.8. Strengths and limitations

This is the first real-world analysis showing patients treated in line with both the NADINA and the SWOG S1801 protocol according to individual patient characteristics. While clinical trials have established the foundation for the neoadjuvant treatment concept, our real-world study provides valuable insights into everyday clinical practice, encompassing an unselected and pretreated patient population as well as rare melanoma subtypes. Because of the limited sample size in the mucosal cohort, these results should be interpreted as a distinct case series.

We hope that this approach adequately addresses your concern and that you will agree with this handling of the mucosal melanoma cases.

The groups are composed by different number of patients (as expected) but any effort is required to obtain homogeneous groups (if any). For instance, adjuvant radiotherapy has been administered in such a small number of subjects, I would suggest to delete them.

Thank you for this important remark. As adjuvant radiotherapy was an eligible treatment in both the NADINA and SWOG S1801 trials, excluding these patients would lead to a selection bias and reduce the real-world representativeness of our cohort. Given the small number of patients receiving radiotherapy, we did not perform separate subgroup analyses, but we have clarified this point and acknowledged it as a limitation in the revised manuscript.

In addition, the inhomogeneity of the patient cohort may have confounded the inter-pretation of treatment effects by introducing variability unrelated to the therapeutic intervention. Along with the single-center design, limited sample size, and relatively short follow-up, these factors limit the generalizability of our findings.

Figure 1; please report the data in bar diagram (instead of the cake)

Thank you very much for this valuable comment. In accordance with your suggestion, we have prepared a bar chart, which is provided in the word document below. However, this figure shows treatment response across different cohorts and compares between pathological and radiological assessments. After comparing the bar and the pie chart, we believe that the pie chart presentation offers a clearer and more intuitive overview of these multiple levels.

 

Moreover, in my opinion, a part of Table 1 (patients characteristics) should be moved to the Supplementary materials. Indeed, just the principal data needs to be reported in the Table along the text (i.e., ECOG, LDH could be reported in the Supplementary..)

Accordingly, Table 2 could be shortened as only the principal data might be reported.

Thank you for this important suggestion. We have now consolidated the key patient and treatment characteristics into a single table in the manuscript to improve clarity and overview (see manuscript). Given the inhomogeneity of the cohort you highlighted, we aimed to provide the reader with full transparency regarding the diversity of the patient groups as well as treatment characteristics and therefore added additional information to the appendix.

In summary, I suggest to reevaluate data once the cohort has been made by more homogeneous groups of patients. I am strongly convinced that having heterogenous groups of patients with different treatments could represent an important bias, thus leading to incorrect interpretation of data.

Thank you for this relevant remark, which we strongly take into consideration. As we highlight in the text our study does not aim to directly compare the NADINA and the SWOG protocol.

It should be emphasized that a direct comparison between the SWOG S1801 and the NADINA real-world cohort lacks scientific validity due to differing patient populations. Therefore, no conclusions regarding a favorable regimen can be drawn from our results. The choice between the NADINA and the SWOG S1801 protocol, if both are reimburs-able, should be individualized, taking all patient-specific factors into account, and de-cided by consensus within an interdisciplinary tumor board.

As in Switzerland both neoadjuvant regiments are reimbursable, there is a selection bias between the NADINA and the SWOG cohort.

Factors influencing the decision in favor of a combined ICI in our cohort included younger age, few comorbidities, high tumor load, low TMB, elevation of LDH and s100, pretreatments and high-risk subtypes as mucosal or acral lentiginous melanoma among others

As careful treatment decisions consider the individual patient characteristics, this bias will also be present with a bigger patient cohort and data pool in the future. Therefore, we consider it important to explicitly emphasize to the reader that these data reflect a real-world clinical setting. We fully agree that, given the limited sample size, this heterogeneity carries greater weight and represents an important limitation. Nevertheless, we strongly believe that reporting these data at the present time is highly relevant. As you are aware, in many countries patients still lack access to the substantial advances in neoadjuvant treatment strategies due to regulatory approval and reimbursement constraints. In this context, early real-world analyses are important to inform clinical practice and to help facilitate broader implementation of these approaches.

Finally, we emphasize that the data presented in this manuscript are intended to serve as a foundation for larger, multicenter analyses, which are currently being initiated by our group and will allow for more homogeneous patient stratification and more robust conclusions.

We hope that the corrections and modifications performed make our manuscript now acceptable for publication in Cancers.

 

We look forward to hearing from you about this matter at your earliest convenience.

 

Yours sincerely,

Lara Valeska Maul (on behalf of all co-authors)