1. Introduction

Benign tumours of the salivary glands — predominantly pleomorphic adenoma and papillary sialadenoma — represent diagnostically and biologically distinct entities whose quantitative characterisation has important implications for surgical planning, prognosis, and recurrence risk stratification. The application of Pearson product-moment correlation analysis to multi-parameter histomorphometric and clinical datasets provides a rigorous, reproducible framework for identifying biologically meaningful associations [1,2].

The present integrated analysis addresses two complementary objectives: (i) Chapter 5 — to determine the strength of pairwise correlations between 14 morphometric and IHC parameters measured in individual papillary sialadenoma specimens (n = 75, df = 73); and (ii) Chapter 3 — to quantify the agreement between retrospective (n = 200) and prospective (n = 200) cohorts across 26 matched clinical and diagnostic parameters (df = 398). Correlation strength is classified as: |r| < 0.30 = negligible; 0.30–0.49 = weak; 0.50–0.69 = moderate; 0.70–0.89 = strong; |r| ≥ 0.90 = very strong [3].

Morphometric and IHC Correlation Analysis

Statistical comparison between the control group (n = 15) and papillary sialadenoma (n = 60) using the independent-samples t-test revealed significant differences in 13 of 14 measured parameters (p ≤ 0.01). Serous glandular parameters alone did not reach statistical significance (p = 0.077 and p = 0.589). The results are summarised in Table 1.

Table 1. Group comparison of all 14 morphometric and IHC parameters. Values: Mean ± SD. t-test p-values. Fold change = Sialadenoma mean / Control mean. Red = p < 0.001; ×value colour: red > 3-fold, blue < 1-fold

Epithelial and Ductal Compartments

The total epithelial area was ×4.99 greater in papillary sialadenoma (80,500.70 ± 8,489.35 µm²) than in controls (16,122.59 ± 1,690.68 µm²; p < 0.0001). The total cell count increased ×6.87 (1,017.35 ± 123.14 vs. 148.07 ± 21.92; p < 0.0001). The striated duct system showed the most pronounced changes: duct area ×5.82 and cell count ×7.33 (both p < 0.0001), consistent with the papillary ductal proliferation that defines this tumour type histologically. Mucinous glandular elements expanded significantly (area ×2.33; cells ×2.32; p < 0.0001 and p < 0.01 respectively), while serous acinar parameters were statistically unchanged (p = 0.077 and p = 0.589), suggesting selective ductal and mucinous expansion with preserved serous architecture.

Nuclear-Cytoplasmic Ratio and Nuclear Morphology

The nuclear-cytoplasmic (N:C) ratio was significantly elevated (0.209 ± 0.077 vs. 0.083 ± 0.012; ×2.52; p < 0.0001), reflecting nuclear enlargement relative to cytoplasmic volume — a consistent feature of active epithelial proliferation. Nuclear area increased ×1.65 (24.875 ± 12.14 vs. 15.109 ± 4.127 µm²; p < 0.0001). These changes are within the range expected for benign neoplastic tissue and do not indicate malignant transformation.

Vascular Parameters

Venule area showed the most pronounced vascular change: ×3.67 increase (2,200.5 ± 500.3 vs. 600.37 ± 120.5 µm²; p < 0.0001). Arteriole area increased ×1.63 (1,300.4 ± 250.7 vs. 800.3 ± 150.2 µm²; p < 0.001). This asymmetric vascular remodelling — venular dilation exceeding arteriolar expansion — is consistent with tumour-driven venous congestion and passive vascular adaptation to the expanding epithelial mass, as described in other benign salivary gland neoplasms [4].

Immunohistochemical Expression Profiles

All four IHC markers were significantly upregulated (all p < 0.0001). p-63 labelling index increased ×3.94 (57.96 ± 2.70% vs. 14.72 ± 2.10%), confirming robust myoepithelial differentiation. p-40 increased ×3.13 (41.95 ± 1.80% vs. 13.40 ± 2.37%), reflecting basal cell phenotype. S-100 increased ×3.32 (55.71 ± 11.30% vs. 16.80 ± 7.30%), indicating dual luminal/myoepithelial expression. The Ki-67 proliferation index increased ×3.08 (30.47 ± 5.70% vs. 9.89 ± 2.34%), confirming active but non-malignant proliferation — Ki-67 indices > 50% are typical of high-grade salivary carcinomas [5], whereas 30% is consistent with benign dynamics.

Correlation Analysis — Chapter 5 (n = 75, df = 73)

Figure 1 shows the IHC marker expression comparison. All four markers are significantly elevated (p < 0.0001), with p-63 showing the highest fold change (×3.94).

Figure 1. IHC marker expression (% positive cells): Control vs. Papillary Sialadenoma. All differences p < 0.0001. Chart is editable: double-click to modify data in Microsoft Word

Figure 2 illustrates the fold change for all 14 parameters. Striated duct cell count shows the greatest relative increase (×7.33), followed by total cell count (×6.87). Serous parameters show fold changes below 1.0, confirming their statistical non-significance.

Figure 2. Fold change (Sialadenoma mean / Control mean) for all 14 parameters. Dashed reference at 1.0 = no change. Chart is editable in Microsoft Word

Pearson Correlation Matrix

The full 14×14 Pearson correlation matrix (Table 2) was computed on individual specimen data (n = 75, df = 73). The matrix reveals three major correlation clusters: (i) an IHC–Morphometric cluster (p-63, p-40, S-100, Ki-67 with epithelial area and cell count; r = 0.851–0.987); (ii) a Vascular–Morphometric cluster (venule area with epithelial area; r = 0.892); and (iii) a Nuclear parameter cluster (N:C ratio, nuclear area) with lower but significant correlations.

Table 2. Pearson correlation matrix — Chapter 5 (n = 75, df = 73). Bold: r ≥ 0.85. Colour: dark blue r ≥ 0.90, medium blue r ≥ 0.75, light blue r ≥ 0.60

Top-15 Correlation Pairs

Table 3. Top 15 Pearson correlations — Chapter 5 (n = 75, df = 73, all p < 0.001). Sorted by r descending. r² = coefficient of determination

Figure 3. Top 15 Pearson r values (Chapter 5, n=75, df=73). All pairs p < 0.001. Chart is editable in Microsoft Word

Scientific Interpretation of Chapter 5 Correlations

1. p-63 / p-40 Co-expression (r = 0.987, r² = 0.974)

The strongest correlation in the dataset was between p-63 and p-40 (r = 0.987, r² = 0.974, p < 0.001). The coefficient of determination of 97.4% indicates that virtually all variance in p-40 expression is explained by p-63 — a near-perfect linear relationship. Both markers are transcription factors expressed in nuclei of myoepithelial and basal cells of salivary glands. This co-regulation confirms that papillary sialadenoma cells maintain a coherent myoepithelial/basal identity, and supports the use of p-63 and p-40 as a complementary diagnostic panel rather than redundant markers.

2. Epithelial Area and IHC Markers (r = 0.855–0.969)

Epithelial area correlated very strongly with p-63 (r = 0.969), p-40 (r = 0.969), Ki-67 (r = 0.855), and moderately with S-100 (r = 0.829). Cell count showed equivalent associations (p-63: r = 0.963; p-40: r = 0.967; Ki-67: r = 0.855; S-100: r = 0.864). These high correlations indicate that morphological expansion and IHC marker expression are biologically co-regulated: as the tumour epithelial mass increases, the expanding myoepithelial cell population proportionally drives marker positivity. This is not simply a dilution effect — it reflects the maintenance of cellular phenotype across the proliferating tumour bulk.

3. Venule Area and Epithelial Expansion (r = 0.892)

The very strong correlation between venule area and epithelial area (r = 0.892, r² = 0.795, p < 0.001) provides quantitative evidence for proportional tumour-induced angiogenesis. The venular dilation (×3.67-fold) substantially exceeds arteriolar expansion (×1.63-fold), consistent with preferential venous adaptation to sustain perfusion of the expanding epithelial compartment. This vascular pattern has been described in VEGF-mediated angiogenesis in benign salivary gland neoplasms [4,6].

4. Ki-67 and Benign Proliferative Activity

Ki-67 correlated strongly with all other IHC markers (r = 0.855–0.876) and with morphometric parameters (r = 0.855). The mean Ki-67 index of 30.47 ± 5.70% is approximately three-fold above control tissue (9.89 ± 2.34%), yet substantially below the >50% threshold associated with high-grade salivary carcinomas [5]. This confirms active but non-malignant proliferative dynamics in papillary sialadenoma. The coupling of Ki-67 with p-63, p-40, and S-100 (r = 0.868–0.876) indicates that proliferation and myoepithelial differentiation are co-regulated processes in this tumour — a unified proliferative-differentiative programme rather than independent events.

Clinical and Diagnostic Correlation Analysis

Overall Cohort Agreement (r = 0.993)

The overall Pearson correlation between retrospective and prospective cohort profiles across all 26 matched clinical and diagnostic parameters was r = 0.993 (r² = 0.986, p < 0.001) — a near-perfect agreement. This extraordinary concordance has two major implications: (i) the retrospective archival data are methodologically equivalent to the standardised prospective protocol, validating their combined use in the analysis; and (ii) the underlying patient population is clinically homogeneous across the two observational periods, supporting the generalisability of findings.

Category-Level Correlations

Category-level Pearson r values ranged from 0.972 (presenting complaints) to 0.999 (symptom dynamics and FNAB cytology). All categories achieved statistical significance (p < 0.05). Table 4 and Figure 4 summarise these results.

Table 4. Category-level Pearson correlations between retrospective (n=200) and prospective (n=200) group profiles. All p < 0.05

Figure 4. Category-level Pearson correlation coefficients between retrospective and prospective cohort profiles (Chapter 3). All r ≥ 0.972. Chart is editable in Microsoft Word

Recurrence Analysis

Among 200 retrospective patients, 19.0% (n = 38) experienced tumour recurrence following primary surgery, with a mean interval to recurrence of 2.0 ± 0.5 years (range: 9 months – 4.2 years). Clinical features at recurrence differed markedly from primary presentation: rapid growth (63.2%), pain syndrome (31.6%), limited tumour mobility (28.9%), and facial deformity (21.1%). These features indicate more aggressive biological behaviour at recurrence, consistent with published data on pleomorphic adenoma [7]. The higher recurrence rate underscores the importance of complete primary excision with adequate margins and systematic postoperative surveillance.

Figure 5. Clinical features at recurrence among 38 patients with recurrent benign salivary gland tumours (retrospective group). Chart is editable in Microsoft Word

Cross-Parameter Correlations in Chapter 3 (n = 400, df = 398)

Cross-domain parameter correlations in the pooled cohort (n = 400) are predominantly weak (|r| = 0.10–0.17). This is clinically meaningful rather than a limitation: it confirms that ultrasound, sialography, MSCT, and FNAB each provide non-redundant, independent diagnostic information. The clinically significant associations are concentrated in three clusters.

Duration–Capsule Cluster. Disease duration > 2 years showed a significant negative correlation with clear US capsule visibility (r = −0.157, p = 0.0016), indicating that longer-standing tumours are more likely to exhibit capsular disruption — a finding directly relevant to surgical planning and risk of incomplete excision.

Ductal–Osseous Cluster. Sialographic moderate stenosis correlated significantly with MSCT jaw pressure (r = 0.173, p < 0.001), reflecting co-occurrence of periductal and periossal tumour expansion in advanced or large lesions. This cluster suggests a pathway from ductal involvement to mechanical effects on adjacent bone.

Structural–Cytological Cluster. Anechoic US inclusions correlated with lobular MSCT contours (r = 0.107) and non-informative FNAB (r = 0.116). Lobular MSCT contours also correlated with atypical FNAB cytology (r = 0.114). This confirms that structurally complex tumours — with cystic degeneration and irregular shape — yield diagnostically challenging cytological specimens, reinforcing the need for complementary imaging when FNAB results are atypical or non-informative.

Cohort Comparison: Presenting Complaints

Figure 6 shows the side-by-side comparison of presenting complaints between retrospective and prospective cohorts. The high Pearson r = 0.972 (p = 0.0012) confirms near-identical complaint profiles. A painless tumour mass was the dominant presentation in both groups (82.5% and 86.5%). The slightly lower rate of incidental detection in the prospective cohort (5.5% vs. 9.5%) likely reflects improved diagnostic awareness over time.

Figure 6. Presenting complaints at first consultation: Retrospective vs. Prospective groups. Pearson r = 0.972 (p = 0.0012). Chart is editable in Microsoft Word

2. Discussion

The present integrated correlation analysis of Chapters 3 and 5 contributes quantitative evidence at two levels: (i) the biological architecture of papillary sialadenoma at the tissue level; and (ii) the epidemiological and diagnostic reproducibility of the study design.

At the tissue level, the dominant finding — near-perfect co-expression of p-63 and p-40 (r = 0.987, r² = 0.974) — provides the strongest quantitative evidence yet that papillary sialadenoma maintains a coherent myoepithelial/basal differentiation programme throughout its proliferating mass. The co-expression of these markers with S-100 (r = 0.851–0.858) further confirms the dual luminal/myoepithelial nature of the tumour cell population. Importantly, the very strong correlations between morphometric parameters and IHC expression (r = 0.855–0.969) demonstrate that tumour growth and marker expression are not independent biological events — they are facets of the same proliferative-differentiative process.

The Ki-67 findings deserve particular attention. The three-fold elevation in proliferative index (30.47% vs. 9.89%) and its strong correlations with morphometric and IHC parameters (r = 0.855–0.876) confirm that papillary sialadenoma is biologically active, yet its Ki-67 level remains well within the benign range. This quantitative boundary is clinically important: it distinguishes papillary sialadenoma from high-grade carcinomas while explaining its tendency for local recurrence when incompletely excised.

At the study design level, the near-perfect cohort agreement (r = 0.993) validates the methodological consistency of data collection across two independent observational periods. The uniformly high category-level correlations (r = 0.972–0.999) confirm that the retrospective and prospective cohorts represent the same clinical entity with equivalent diagnostic profiles — a prerequisite for valid pooled analysis and comparative conclusions.

The 19.0% recurrence rate in the retrospective cohort is consistent with literature-reported rates for benign salivary gland tumours treated by simple enucleation without adequate cuff resection [7,8]. The clinical profile of recurrences — characterised by faster growth and more prominent symptoms — supports the hypothesis of clonal selection and altered biological behaviour in recurrent tumours.

3. Conclusions

The following principal conclusions are drawn from the integrated Pearson correlation analysis of Chapters 3 and 5:

p-63 and p-40 show the strongest correlation in the dataset (r = 0.987, r² = 0.974, p < 0.001), confirming near-perfect synchronous expression and a unified myoepithelial/basal differentiation programme in papillary sialadenoma.

Morphometric expansion (epithelial area, cell count) correlates very strongly with IHC marker expression (r = 0.855–0.969), demonstrating that structural growth and molecular differentiation are co-regulated biological events, not independent processes.

Venule area correlates very strongly with epithelial area (r = 0.892), providing quantitative evidence for proportional tumour-induced angiogenesis — venular dilation (×3.67) substantially exceeds arteriolar expansion (×1.63).

Ki-67 index (×3.08-fold, 30.47%) and its strong correlations with IHC and morphometric parameters confirm active but non-malignant proliferative dynamics, consistent with benign tumour behaviour.

Serous acinar parameters are statistically independent from the main correlation cluster, reflecting preserved serous glandular architecture in papillary sialadenoma.

Chapter 3: overall matched-pair Pearson r = 0.993 (r² = 0.986, p < 0.001) confirms near-perfect methodological equivalence between retrospective and prospective cohorts across all 26 clinical and diagnostic parameters.

Category-level correlations (r = 0.972–0.999) confirm consistent data quality across all seven diagnostic domains: complaints, duration, dynamics, US, sialography, MSCT, and FNAB.

Cross-domain clinical correlations are appropriately weak (|r| = 0.10–0.17), confirming diagnostic independence and validating the complementary value of multimodal diagnostic workup.

The duration–capsule–duct–osseous correlation cluster (r = 0.109–0.173) provides quantitative grounds for integrating temporal, ultrasound, and tomographic data in surgical planning.

A recurrence rate of 19.0% with an accelerated clinical profile at recurrence underscores the importance of complete primary excision and systematic long-term follow-up.