Prognostic and Clinicopathological Significance of CCND1/Cyclin D1 Upregulation in Melanomas: A Systematic Review and Comprehensive Meta-Analysis

Simple Summary The incidence of cutaneous melanoma is increasing worldwide, currently responsible for 287,723 new cases and 60,712 deaths per year (GLOBOCAN, IARC, WHO). It should be also highlighted that some less frequent subtypes of melanomas—i.e., acral, uveal, and mucous melanoma—are responsible for significant morbidity associated with metastasis, responding typically worse to newer therapies. Therefore, new biomarkers are needed to improve the prognosis in individual patients. In this sense, the present study showed that CCND1/cyclin D1 upregulation is a common molecular oncogenic alteration in melanomas that probably favors the growth and expansion on cutaneous primary melanomas. Furthermore, immunohistochemical cyclin D1 overexpression strongly predicted a higher Breslow thickness, currently considered the most relevant prognostic factor in individual patients with melanomas. Finally, special attention should be paid to the CCND1/cyclin D1 complex in mucosal melanomas, whose upregulation was strikingly altered. Abstract Our objective was to evaluate the prognostic and clinicopathological significance of cyclin D1 (CD1) overexpression/CCND1 amplification in melanomas. We searched studies published before September 2019 (PubMed, Embase, Web of Science, Scopus). We evaluated the quality of the studies included (QUIPS tool). The impact of CD1 overexpression/CCND1 amplification on overall survival and relevant clinicopathological characteristic were meta-analyzed. We performed heterogeneity, sensitivity, small-study effects, and subgroup analyses. Forty-one studies and 3451 patients met inclusion criteria. Qualitative evaluation demonstrated that not all studies were performed with the same rigor, finding the greatest risk of bias in the study confounding domain. Quantitative evaluation showed that immunohistochemical CD1 overexpression had a statistical association with Breslow thickness (p = 0.007; OR = 2.09,95% CI = 1.23–3.57), significantly higher frequency of CCND1/cyclin D1 abnormalities has been observed in the primary tumor compared to distant metastases (p = 0.004), revealed also by immunohistochemical overexpression of the protein (p < 0.001; OR = 0.53,95% CI = 0.40–0.71), while the CCND1 gene amplification does not show association (p = 0.43); while gene amplification, on the contrary, appeared more frequently in distant metastases (p = 0.04; OR = 1.70,95% CI = 1.01–2.85) and not in the primary tumor. In conclusion, CCND1/cyclin D1 upregulation is a common molecular oncogenic alteration in melanomas that probably favors the growth and expansion of the primary tumor. This upregulation is mainly consequence to the overexpression of the cyclin D1 protein, and not to gene amplification.


CCND1/cyclin D1 and overall survival by alteration
HR, hazard ratio; CI, confidence intervals. Random-effects model (based on the DerSimonian and Laird method) pooling hazard ratios.   Figure S7. Forest plot graphically representing the association between CCND1/cyclin D1 alterations and Overall Survival in Cutaneous Melanomas by immunohistochemical pattern.

CCND1/cyclin D1 and overall survival by immunohistochemical pattern
HR, hazard ratio; CI, confidence intervals. Random-effects model (based on the DerSimonian and Laird method) pooling hazard ratios.       Figure S18. Forest plot graphically representing the association between CCND1/cyclin D1 alterations and types of melanoma (nodular vs. SSM/LMM/AM) in Cutaneous Melanomas by geographic area. OR, odds ratio; CI, confidence intervals. Fixed-effect model (Mantel-Haenszel method) pooling odds ratios.   Figure S19. Forest plot graphically representing the association between CCND1/cyclin D1 alterations and types of melanoma (nodular vs. SSM/LMM/AM) in Cutaneous Melanomas by immunohistochemical pattern. OR, odds ratio; CI, confidence intervals. Fixed-effect model (Mantel-Haenszel method) pooling odds ratios.   Figure S21. Forest plot graphically representing the association between CCND1/cyclin D1 alterations and tisular expression (distance metastasis vs. primary tissue) in Cutaneous Melanomas.

CCND1/cyclin D1 and Distance metastasis vs. Primary tissue by alteration
OR, odds ratio; CI, confidence intervals. Fixed-effect model (Mantel-Haenszel method) pooling odds ratios.       Figure S24. Forest plot graphically representing the frequency of CCND1/cyclin D1 alterations in Uveal Melanomas.

Meta-analysis on the frequency of CCND1/cyclin D1 alterations in uveal Melanoma
ES, estimation; CI, confidence intervals. Random-effects model (based on the DerSimonian and Laird method) using pooled proportions.

Meta-analysis on the association between CCND1/cyclin D1 alterations and Largest Basal Dimension in Uveal Melanoma
OR, odds ratio; CI, confidence intervals. Fixed-effect model (Mantel-Haenszel method) pooling odds ratios.  Figure S26. Forest plot graphically representing the association between CCND1/cyclin D1 alterations and pathology (epithelioid vs. spindle/mixed cell shape) in Uveal Melanomas.

Meta-analysis on the association between CCND1/cyclin D1 alterations and Pathology in Uveal Melanoma
OR, odds ratio; CI, confidence intervals. Fixed-effect model (Mantel-Haenszel method) pooling odds ratios.   Cyclin D1/CCND1 alterations in mucosal melanoma(%)   Figure 35. A funnel plot of estimated logOR against its standard error, graphically representing the analysis of small-study effects on Type of Cutaneous Melanomas. SE, standard error; OR, odds ratio. The black vertical line corresponds to the pooled estimated prevalence. The two diagonal intermittent lines represent the pseudo-95% confidence interval. The circles represent the published studies.  Figure 36. A funnel plot of estimated logOR against its standard error, graphically representing the analysis of small-study effects on Ulceration in Cutaneous Melanomas. SE, standard error; OR, odds ratio. The black vertical line corresponds to the pooled estimated prevalence. The two diagonal intermittent lines represent the pseudo-95% confidence interval. The circles represent the published studies.  Figure 37. A funnel plot of estimated logOR against its standard error, graphically representing the analysis of small-study effects on Ulceration in Cutaneous Melanomas. SE, standard error; OR, odds ratio. The black vertical line corresponds to the pooled estimated prevalence. The two diagonal intermittent lines represent the pseudo-95% confidence interval. The circles represent the published studies.  Table S3. Sensitivity analysis of the studies pooled in the meta-analysis on the association between CCND1/cyclin D1 alterations and overall survival in cutaneous melanoma Table S4. Sensitivity analysis of the studies pooled in the meta-analysis on the association between CCND1/cyclin D1 alterations and Breslow Thickness in cutaneous melanoma Sensitivity analysis ("leave-one-out" method) of the meta-analysis results, sequentially omitting one study at a time. Sensitivity analysis ("leave-one-out" method) of the meta-analysis results, sequentially omitting one study at a time.

Review question.
What is the clinicopathological and prognostic significance of CCND1/cyclin D1 alterations in patients with melanomas?

Searches.
PubMed Scopus -TITLE-ABS-KEY(("cyclin d1" OR "cyclind1" OR "ccnd1" OR "ccnd 1") AND ("melanoma")) In adittion, a manual screening of the references lists of the included articles will also be performed to find relevant articles that might have been missed by the original search strategy.

Condition or domain being studied.
Malignant melanoma is the most aggressive form of skin cancer with unpredictable behavior. Although multiple approaches have been used in its treatment, the mortality of melanoma patients has barely improved in recent decades. Therefore, in addition to conventional clinicopathological prognostic parameters (e.g. Breslow thickness), the early prediction of the evolution of these patients through the use of molecular biomarkers could also be a promising tool in daily clinical practice.

Participants/population.
Patients diagnosed with melanoma.

Intervention(s), exposure(s).
We will evaluate studies in which CCND1/cyclin D1 alterations were analyzed in tumor tissue from patients diagnosed with melanoma.
Differences in the amplification levels of CCND1 will be categorized as positive for the exposition group, and negative for the control group, based on the cut-off value chosen by the authors. Differences in the expression of cyclin D1 will be categorized as high expression/overexpression for the exposition group and low expression for the control group, based on the cut-off value chosen by authors.

Comparator(s)/control.
Control group will be represented by the group of patients with melanomas with negative CCND1 amplification or low cyclin D1 expression.

Types of study to be included.
Criteria for eligibility of a study included in the qualitative and quantitative analysis will be: -Original research articles published in English language.
-To assess CCND1/cyclin D1 alterations (gene amplification or protein expression) in biopsies from patients diagnosed with melanomas.
-To analyze the associations between CCND1/cyclin D1 alterations with survival endpoints (overall and/or disease-free survival) and/or clinicopathological features (type of melanoma, subtype, Breslow thickness, ulceration, N and M status, clinical stage, Clark level, growth phases, regression, satellitosis, mitotic index, anatomic location, sex, age and/or ethnicity).
Criteria for eligibility of a study excluded in the qualitative and quantitative analysis will be: -Reviews and meta-analyses, case reports, editorials, letters, meeting abstracts, personal opinions or commentaries, books chapters and non-English language articles.
-No melanoma (Differents skin cancers will be excluded).
-Experimental analysis conducted in vitro or in animal models.
-Lack of the precedent essential survival endpoints or clinicopathological features.

Risk of bias (quality) assessment.
The risk of bias in individual studies will be assessed using the Quality in Prognostic Studies (QUIPS) tool, developped by Cochrane prognosis methods group. Specifically, it contains 6 domains: study participation, study attrition, prognostic factor measurement, outcome measurement, study confounding and statistical analysis and reporting. Each domain will be rated as low, moderate or high risk of bias for each study.
Strategy for data synthesis.
Differences in the alterations of CCND1/cyclin D1 will be categorized as positive/high and negative/low, according to the cut-off value chosen by the authors. Odds ratios (OR) and 95% confidence intervals (CI) will be used as the measure of association to determine the correlations between CCND1/cyclin D1 alterations and clinicopathological features. Hazard ratios (HR) and 95%CI will be used as the measure of association to estimate the impact of CCND1/cyclin D1 alterations on time-to-event parameters (overall survival and disease-free survival). In meta-analyses, OR and HR will be pooled when appropriate, using both fixed-effect and/or random-effects models. Forest plots will be used to examine the overall effect. Heterogeneity between studies will be checked using the χ² based Cochran's Q test (p<0.10). We will quantify the proportion of heterogeneity between studies with the I² statistic. In addition, sensitivity analyses will be performed to explore the influence of each individual study on the estimation of the overall effect and to test the reliability of the overall pooled results. Funnel plots will be constructed when appropriate, to assess small-study effects such as publication bias. Egger's test (p<0.10) will also be used to statistically assess funnel plots asymmetry. All statistical analyses will be performed with Stata version 14 (Stata Corp, College Station, TX, USA) using user-written commands. In meta-analyses, a two-tailed p-value <0.05 will be considered statistically significant.

Analysis of subgroups or subsets.
We will perform different subgroup analyses (by continent, type/subtype of melanoma, cut off value and immunostaining pattern) to asses potential confounding factors as a possible heterogeneity source, and to explore the relations between CCND1/cyclin D1 alterations and the precedent outcomes in these subgroups.