Introduction: A recent multiregional whole-exome sequencing of 48 tumour samples from 9 gastric adenocarcinomas discovered PCLO mutations in 23 (47.9%) tumour samples. Based on that unexpected high prevalence of PCLO mutations, we hypothesized a tumour biological significance of PCLO in gastric cancer (GC). Methods: Tumour samples (whole tissue sections) obtained from 466 patients resected for therapy-naive GC were stained with an anti-PCLO antibody. The histoscore for tumour cells and the presence of immunostaining of stromal cells and tumour vessels was documented for each case. An algorithm for PCLO immunopositivity was formed and correlated with clinicopathological patient characteristics. Results: 175 GCs were classified as PCLO positive within tumour cells, and 291 as negative. Stromal cells were positive for PCLO in 106 cases and tumour vessels in 84. PCLO-positive GCs more often showed an intestinal phenotype, a lower T category and were more commonly associated with Helicobacter pylori infection. A separate analysis of PCLO expression in intestinal and diffuse type GCs, respectively, showed no significant correlations. Patients with PCLO negative/low tumour cells showed a shortened overall (14.0 ± 1.4 vs. 16.0 ± 1.8 months) and tumour-specific survival (15.0 ± 1.6 months vs. 17.9 ± 3.6). Comparison of PCLOs genotype with its phenotype in 48 tumour samples obtained from nine cases showed no direct correlations with missense mutations. Conclusion: Our data provide evidence that PCLO is differentially expressed in GC and might delay tumour progression.

Gastric cancer (GC) is the eighth most common cancer in Germany, accounting for 3.8% of all deaths [1]. Although the incidence has steadily decreased in recent decades, the prognosis remains poor. The 5-year relative survival rate is only 33%, due to late diagnosis and diagnosis at an advanced stage.

Several risk factors for GC have been identified, including Helicobacter pylori (H. pylori) colonisation of the gastric mucosa, alcohol and tobacco consumption, high salt intake and obesity [2‒4]. Classification and identification of disease subgroups may help develop more specific and effective therapies to improve both life expectancy and quality of life [5, 6].

Piccolo presynaptic cytomatrix protein (PCLO) helps stabilise the presynaptic cytomatrix and is involved in vesicle trafficking. It is regularly expressed in the brain and neuronal structures. Previous studies have shown an association between PCLO and psychiatric disorders, major depression, and pontocerebellar hypoplasia type III [7, 8]. Interestingly, PCLO is also frequently mutated and amplified in oesophageal squamous cell carcinoma (ESCC), and Zhang et al. [9] provided evidence that the mutations of PCLO may lead to an unfavourable patient prognosis in ESCC.

In a recent study using multiregional whole-exome sequencing (WES) of 48 tumour samples from nine gastric adenocarcinomas, we found PCLO mutations in 23 (47.9%) tumour samples [10]. Based on this unexpectedly high prevalence of PCLO mutations, we hypothesised that PCLO may also have a tumour biological significance in GC. To test this hypothesis, we performed an immunohistochemical study of the expression and putative tumour biological significance of PCLO in a cohort of 466 well-characterised, therapy-naive GCs.

Ethic Statement

Our project was granted ethical clearance by the Local Ethics Committee of the University Hospital Schleswig-Holstein, Campus Kiel, Germany, in agreement with the Helsinki Declaration (D 453/10 and D 549/20).

Patients and Tumour Samples

Discovery Group

For the discovery group, we prospectively enrolled 9 patients with an adenocarcinoma of the stomach or oesophagogastric junction at the University Hospital Schleswig-Holstein, Campus Kiel, between 2016 and 2017. The mean age of the patients was 68 years, with a range of 50–85 years. Inclusion criteria were adequate size of the primary tumour (diameter >3 cm) to allow multiregional tissue sampling without compromising the surgical pathological evaluation of the resection specimen. After tumour resection, specimens were sent to the pathology department on ice. Depending on the size of the primary tumour, between 3 and 10 samples were taken from the primary tumour by core needle biopsy and frozen at −80°C until further use. Macroscopic images of the surgical resection were taken before and after tissue sampling to facilitate anatomical reconstruction of the sampling procedure and correlation with paraffin blocks. A total of 45 samples were obtained from the primary tumours. In one case, three additional samples were taken from three separate lymph node metastases. Finally, a total of 48 tumour samples were sent for WES. The detailed genomic data of this cohort have been described previously [4].

Validation Group

We retrospectively reviewed all patients who underwent total or partial gastrectomy for adenocarcinoma of the stomach or gastroesophageal junction between 1997 and 2009. Specimens were obtained from the archive of the Department of Pathology, University Hospital Schleswig-Holstein, Campus Kiel. The following patient characteristics were retrieved from the electronic database: age, sex, tumour location, tumour size, depth of tumour invasion, number of resected lymph nodes and number of lymph nodes with metastases, distant metastases, stage of disease, lymphatic invasion, venous invasion, grading, and residual tumour status. All four subtypes according to TCGA classification (see below) were represented [11].

Patients were included if histology confirmed a primary adenocarcinoma of the stomach or gastroesophageal junction. Patients were excluded if histology revealed a type of tumour other than adenocarcinoma or if they had received perioperative chemotherapy or radiotherapy. The date of patient death was obtained from the Epidemiological Cancer Registry of Schleswig-Holstein, Germany. Follow-up data for surviving patients were obtained from hospital records and general practitioners. Our study cohort included formalin-fixed and paraffin-embedded tissue samples. All patient-related data were pseudonymised after inclusion in the study. All patients of the discovery and validation cohort were white patients from northern Germany treated at a single center.

Assessment of Further Clinicopathological Patient Characteristics

The pTNM stage of all study patients was determined according to the 8th edition of the UICC guidelines [12]. All tumours were classified according to the Laurén classification [13]. H. pylori infection was assessed histologically using modified Giemsa staining and polymerase chain reaction. H. pylori-specific DNA was detected by a PCR-based assay targeting the 16S rRNA gene of H. pylori, as previously described [14]. Epstein-Barr virus-encoded RNA was detected using the EBER probe (Novocastra) and the BondMax detection system according to the manufacturer's instructions (Leica Microsystems GmbH) [15]. MSI status was assessed by IHC using antibodies against MLH1, PMS2, MSH2, and MSH6. For each case with reduced or absent nuclear staining, a subsequent molecular comparison of the allelic profiles of the mononucleotide repeat markers BAT-25, BAT-26, NR-21, NR-24, and NR-27 in the tumour and corresponding normal tissue was performed [16]. HER2 and MET status was also assessed as previously described [17, 18].

Immunohistochemistry

Immunohistochemical staining was performed with rabbit polyclonal antibodies against PCLO (dilution 1:200; ab110427; Abcam plc; Discovery Drive, Cambridge Biomedical Campus, Cambridge, CB2 0AX, UK) using the BondMax Autostainer (Leica, Germany). Immunoreaction was visualised using the Bond™ Polymer Refine Detection Kit (brown labelling; Novocastra; Leica Microsystems, Wetzlar, Germany). Immunostaining was evaluated using a Leica microscope (Leica DM 1000).

Immunostaining Assessment

Each tumour was scored using a semiquantitative approach combining the intensity of immunostaining and the percentage of positive cells in the tumour. The intensity of immunostaining of tumour cells showed no (0), weak (1+), moderate (2+), or strong (3+) staining, as shown in Figure 1. The percentage of positive tumour cells showing the defined staining intensities (0, 1+, 2+, 3+) was increased with respect to all tumour cells visible on each tissue sample and always added up to a total of 100% tumour cells.

Fig. 1.

Reference slides of the four different immunostaining intensities, i.e., no immunostaining (0, a), weak (1+; b), moderate (2+; c), and strong immunostaining (3+; d). Anti-PCLO antibody, haemalaun counterstain. Original magnification 400-fold (a, b, c, d).

Fig. 1.

Reference slides of the four different immunostaining intensities, i.e., no immunostaining (0, a), weak (1+; b), moderate (2+; c), and strong immunostaining (3+; d). Anti-PCLO antibody, haemalaun counterstain. Original magnification 400-fold (a, b, c, d).

Close modal

The HScore was calculated using the formula: HScore=0×percentageofimmunonegativetumourcells+1×percentageofweaklystainedtumourcells+2×percentageofmoderatelystainedtumourcells+3×percentageofstronglystainedtumourcells, resulting in a possible HScore between 0 and 300. Tumour cells with no detectable staining were scored as 0. The maximum possible HScore was 300 if all cells of a given tumour sample showed strong staining: [0 × 0%] + [1 × 0%] + [2 × 0%] + [3 × 100%] = 300. Since we did not know a priori, which “cut-off” value of PCLO expression might be biologically relevant, we used a stepwise exploratory approach. First, we divided the cohort at the median HScore, into “low/negative” (HScore = 0) and “high/positive” (HScore ≥10) groups. Second, we divided the cohort into four quartiles for a pairwise comparison and analysis of each quartile. Third, we documented the immunostaining for PCLO in vessels and stroma of each specimen.

Statistical Analysis

SPSS version 27.0 (IBM Corp., Armonk, NY, USA) was used for statistical analyses. Correlations between non-ordinal variables were determined using Fisher’s exact test, while those between ordinal variables were determined using Kendall’s tau test. A significance level of 0.05 was used. To correct for the false discovery rate within the correlations, we applied the Simes (Benjamini-Hochberg) procedure (false discovery rate correction) [19]. All p values are uncorrected. Median survival with 95% confidence intervals was calculated using the Kaplan-Meier method. Differences between median survival rates were tested using the log-rank test.

A total of 466 cases met all study criteria. The clinicopathological characteristics of the patients in our validation group are summarised in Table 1. Regarding tumour type, 239 (51.3%) cases were classified as intestinal according to Laurén, 146 (31.3%) as diffuse, 31 (6.7%) as mixed, and 50 (10.7%) cases were unclassifiable. EBV was detected in 20 cases (4.3%). 35 (7.5%) GCs were microsatellite unstable. H. pylori was positive in 62 cases (13.3%). Overall survival (OS) and tumour-specific survival (TSS) data were available for 454 and 425 cases, respectively, with a median follow-up of 12.8 months (range: 0–142.7 months). The median OS was 15.5 months, and the median TSS was 16.6 months.

Table 1.

Clinicopathological patient characteristics

PCLO expressionPCLO expression in stromaPCLO expression in vessels
Totallow (HScore = 0)high (HScore >0)absentpresentabsentpresent
N(%)n(%)n(%)n(%)n(%)n(%)n(%)
Total 
   466 (100) 291 (62.4) 175 (37.6) 360 (77.3) 106 (22.7) 382 (82.0) 84 (18.0) 
Gender n pa   466   0.844 466   0.734 466   0.173 
 Female   179 (38.4) 113 (63.1) 66 (36.9) 140 (78.2) 39 (21.8) 141 (78.8) 38 (21.2) 
 Male   287 (61.6) 178 (62.0) 109 (38.0) 220 (76.7) 67 (23.3) 241 (84.0) 46 (16.0) 
Age group n pa   466   0.104 466   0.378 466   0.149 
 <68 years   234 (50.2) 155 (66.2) 79 (33.8) 185 (79.1) 49 (20.9) 198 (84.6) 36 (15.4) 
 ≥68 years   232 (49.8) 136 (58.6) 96 (41.4) 175 (75.4) 57 (24.6) 184 (79.3) 48 (20.7) 
Localization n pa   463   0.836 463   0.905 463   1.000 
 Proximal stomach   145 (31.1) 92 (63.4) 53 (36.6) 112 (77.2) 33 (22.8) 119 (82.1) 26 (17.9) 
 Distal stomach   318 (68.2) 197 (61.9) 121 (38.1) 247 (77.7) 71 (22.3) 261 (82.1) 57 (17.9) 
Laurén phenotype n pa   466   0.015 466   0.315 466   0.788 
 Intestinal   239 (51.3) 135 (56.5) 104 (43.5) 179 (74.9) 60 (25.1) 192 (80.3) 47 (19.7) 
 Diffuse   146 (31.3) 106 (72.6) 40 (27.4) 111 (76.7) 34 (23.3) 122 (83.6) 24 (16.4) 
 Mixed   31 (6.7) 20 (64.5) 11 (35.5) 27 (87.1) (12.9) 27 (87.1) (12.9) 
 Unclassifiable   50 (10.7) 30 (60.0) 20 (40.0) 42 (84.0) (16.0) 41 (82.0) (18.0) 
Grading n pa   466   0.043 466   0.364 466   0.573 
 G1/G2   111 (23.8) 60 (54.1) 51 (45.9) 82 (73.9) 29 (26.1) 89 (80.2) 22 (19.8) 
 G3/G4   355 (76.2) 231 (65.1) 124 (34.9) 278 (78.3) 77 (21.7) 293 (82.5) 62 (17.5) 
pT category n pb   466   0.011 466   0.800 466   0.858 
 pT1a/T1b   59 (12.7) 30 (50.8) 29 (49.2) 48 (81.4) 11 (18.6) 49 (83.1) 10 (16.9) 
 pT2   53 (11.4) 32 (60.4) 21 (39.6) 42 (79.2) 11 (20.8) 46 (86.8) (13.2) 
 pT3   185 (39.7) 112 (60.5) 73 (39.5) 128 (74.6) 47 (25.4) 145 (78.4) 40 (21.6) 
 pT4a/T4b   169 (36.3) 117 (69.2) 52 (30.8) 132 (78.1) 37 (21.9) 142 (84.0) 27 (16.0) 
pN category n pb   465   0.263 465   0.871 465   0.849 
 pN0   136 (29.2) 82 (60.3) 54 (39.7) 108 (79.4) 28 (20.6) 115 (84.6) 21 (15.4) 
 pN1   62 (13.3) 35 (56.5) 27 (43.5) 44 (71.0) 18 (29.0) 44 (71.0) 18 (29.0) 
 pN2   82 (17.6) 52 (63.4) 30 (36.6) 61 (74.4) 21 (25.6) 68 (82.9) 14 (17.1) 
 pN3a/b   185 (39.7) 121 (65.4) 64 (34.6) 146 (78.9) 39 (21.1) 154 (83.2) 31 (16.8) 
pM category n pa   466   0.299 466   0.672 466   0.646 
 pM0   379 (81.3) 234 (61.7) 145 (38.3) 291 (76.8) 88 (23.2) 312 (82.3) 67 (17.7) 
 pM1   87 (18.7) 57 (65.5) 30 (34.5) 69 (79.3) 18 (20.7) 70 (80.5) 17 (19.5) 
UICC stage n pb   465   0.130 465   0.851 465   0.778 
 IA/IB   82 (17.6) 46 (56.1) 36 (43.9) 65 (79.3) 17 (20.7) 69 (84.1) 13 (15.9) 
 IIA/IIB   99 (21.2) 59 (59.6) 40 (40.4) 74 (74.7) 25 (25.3) 79 (79.8) 20 (20.2) 
 IIIA/IIIB/IIIC   197 (42.3) 128 (65.0) 69 (35.0) 151 (76.6) 46 (23.4) 163 (82.7) 34 (17.3) 
 IV   87 (18.7) 57 (65.5) 30 (34.5) 69 (79.3) 18 (20.7) 70 (80.5) 17 (19.5) 
LN ratio n pa   465   0.088 465   0.659 465   0.399 
 Low (<0.189)   227 (48.7) 134 (59.0) 93 (41.0) 173 (76.2) 54 (23.8) 182 (80.2) 45 (19.8) 
 High (≥0.189)   238 (51.1) 156 (65.5) 82 (34.5) 186 (78.2) 52 (21.8) 199 (83.6) 39 (16.4) 
pL category n pa   448   0.267 448   0.736 448   1.000 
 pL0   218 (46.8) 138 (63.3) 80 (36.7) 166 (76.1) 52 (23.9) 178 (81.7) 40 (18.3) 
 pL1   230 (49.4) 138 (60.0) 92 (40.0) 179 (77.8) 51 (22.2) 188 (81.7) 42 (18.3) 
pV category n pa   447   0.644 447   0.595 447   0.702 
 pV0   397 (85.2) 246 (62.0) 151 (38.0) 307 (77.3) 90 (22.7) 325 (81.9) 22 (18.1) 
 pV1   50 (10.7) 29 (58.0) 21 (42.0) 37 (74.0) 13 (26.0) 40 (80.0) 10 (20.0) 
pR status n pa   461   0.883 461   0.494 461   0.142 
 pR0   405 (86.9) 252 (62.2) 153 (37.8) 312 (77.0) 93 (23.0) 328 (81.0) 77 (19.0) 
 pR1/R2   56 (12.0) 36 (64.3) 20 (35.7) 46 (82.1) 10 (17.9) 50 (89.3) (10.7) 
HER2 status n pa   437   1.000 437   0.531 437   0.365 
 Negative   401 (86.1) 254 (63.3) 147 (36.7) 312 (77.8) 89 (22.2) 328 (81.8) 73 (18.2) 
 Positive   36 (7.7) 23 (63.9) 13 (36.1) 30 (83.3) (16.7) 32 (88.9) (11.1) 
MET status n pa   455   1.000 455   0.825 455   0.812 
 Negative   424 (91.0) 266 (62.7) 158 (37.3) 325 (76.7) 99 (23.3) 347 (81.8) 77 (18.2) 
 Positive   31 (6.7) 20 (64.5) 11 (35.5) 25 (80.6) (19.4) 25 (80.6) (19.4) 
H. pylori status n pa   395   0.015 395   0.148 395   0.009 
 Negative   333 (71.5) 213 (64.0) 120 (36.0) 256 (76.9) 77 (23.1) 276 (82.9) 57 (17.1) 
 Positive   62 (13.3) 29 (46.8) 33 (53.2) 42 (67.7) 20 (32.3) 42 (67.7) 20 (32.3) 
EBV status n pa   452   0.105 452   0.586 452   0.227 
 Negative   432 (92.7) 266 (61.6) 166 (38.4) 334 (77.3) 98 (22.7) 352 (81.5) 80 (18.5) 
 Positive   20 (4.3) 16 (80.0) (20.0) 17 (85.0) (15.0) 19 (95.0) (5.0) 
MSI status n pa   451   0.279 451   0.835 451   0.249 
 MSS   416 (89.3) 265 (63.7) 151 (36.3) 323 (77.6) 93 (22.4) 344 (82.7) 72 (17.3) 
 MSI   35 (7.5) 19 (54.3) 16 (45.7) 28 (80.0) (20.0) 26 (74.3) (25.7) 
OS, months pc   454 0.204 454 0.984 454 0.970 
 Total/events/censored   454 (22.0) 284/225/59 170/129/41 352/273/79 102/81/21 372/286/86 82/68/14 
 Median survival   15.0 1.1 14.0±1.4 16.0±1.8 15.6±1.2 14.1±1.8 14.1±1.3 16.7±1.8 
 95% CI   12.9–17.1 11.3–16.7 12.4–19.6 13.2–18.1 10.6–17.6 11.4–16.7 13.1–20.3 
TSS, months pc   425 0.086 425 0.367 425 0.449 
 Total/events/censored   425 (31.8) 270/192/78 155/98/57 329/229/100 96/61/35 348/238/110 77/52/25 
 Median survival   16.6±1.4 15.0±1.6 17.9±3.6 16.6±1.6 16.6±4.5 15.6±1.6 17.9±4.1 
 95% CI   13.8–19.4 11.8–18.2 13.8–19.4 13.5–19.7 7.8–25.4 12.4–18.9 13.8–25.9 
PCLO expressionPCLO expression in stromaPCLO expression in vessels
Totallow (HScore = 0)high (HScore >0)absentpresentabsentpresent
N(%)n(%)n(%)n(%)n(%)n(%)n(%)
Total 
   466 (100) 291 (62.4) 175 (37.6) 360 (77.3) 106 (22.7) 382 (82.0) 84 (18.0) 
Gender n pa   466   0.844 466   0.734 466   0.173 
 Female   179 (38.4) 113 (63.1) 66 (36.9) 140 (78.2) 39 (21.8) 141 (78.8) 38 (21.2) 
 Male   287 (61.6) 178 (62.0) 109 (38.0) 220 (76.7) 67 (23.3) 241 (84.0) 46 (16.0) 
Age group n pa   466   0.104 466   0.378 466   0.149 
 <68 years   234 (50.2) 155 (66.2) 79 (33.8) 185 (79.1) 49 (20.9) 198 (84.6) 36 (15.4) 
 ≥68 years   232 (49.8) 136 (58.6) 96 (41.4) 175 (75.4) 57 (24.6) 184 (79.3) 48 (20.7) 
Localization n pa   463   0.836 463   0.905 463   1.000 
 Proximal stomach   145 (31.1) 92 (63.4) 53 (36.6) 112 (77.2) 33 (22.8) 119 (82.1) 26 (17.9) 
 Distal stomach   318 (68.2) 197 (61.9) 121 (38.1) 247 (77.7) 71 (22.3) 261 (82.1) 57 (17.9) 
Laurén phenotype n pa   466   0.015 466   0.315 466   0.788 
 Intestinal   239 (51.3) 135 (56.5) 104 (43.5) 179 (74.9) 60 (25.1) 192 (80.3) 47 (19.7) 
 Diffuse   146 (31.3) 106 (72.6) 40 (27.4) 111 (76.7) 34 (23.3) 122 (83.6) 24 (16.4) 
 Mixed   31 (6.7) 20 (64.5) 11 (35.5) 27 (87.1) (12.9) 27 (87.1) (12.9) 
 Unclassifiable   50 (10.7) 30 (60.0) 20 (40.0) 42 (84.0) (16.0) 41 (82.0) (18.0) 
Grading n pa   466   0.043 466   0.364 466   0.573 
 G1/G2   111 (23.8) 60 (54.1) 51 (45.9) 82 (73.9) 29 (26.1) 89 (80.2) 22 (19.8) 
 G3/G4   355 (76.2) 231 (65.1) 124 (34.9) 278 (78.3) 77 (21.7) 293 (82.5) 62 (17.5) 
pT category n pb   466   0.011 466   0.800 466   0.858 
 pT1a/T1b   59 (12.7) 30 (50.8) 29 (49.2) 48 (81.4) 11 (18.6) 49 (83.1) 10 (16.9) 
 pT2   53 (11.4) 32 (60.4) 21 (39.6) 42 (79.2) 11 (20.8) 46 (86.8) (13.2) 
 pT3   185 (39.7) 112 (60.5) 73 (39.5) 128 (74.6) 47 (25.4) 145 (78.4) 40 (21.6) 
 pT4a/T4b   169 (36.3) 117 (69.2) 52 (30.8) 132 (78.1) 37 (21.9) 142 (84.0) 27 (16.0) 
pN category n pb   465   0.263 465   0.871 465   0.849 
 pN0   136 (29.2) 82 (60.3) 54 (39.7) 108 (79.4) 28 (20.6) 115 (84.6) 21 (15.4) 
 pN1   62 (13.3) 35 (56.5) 27 (43.5) 44 (71.0) 18 (29.0) 44 (71.0) 18 (29.0) 
 pN2   82 (17.6) 52 (63.4) 30 (36.6) 61 (74.4) 21 (25.6) 68 (82.9) 14 (17.1) 
 pN3a/b   185 (39.7) 121 (65.4) 64 (34.6) 146 (78.9) 39 (21.1) 154 (83.2) 31 (16.8) 
pM category n pa   466   0.299 466   0.672 466   0.646 
 pM0   379 (81.3) 234 (61.7) 145 (38.3) 291 (76.8) 88 (23.2) 312 (82.3) 67 (17.7) 
 pM1   87 (18.7) 57 (65.5) 30 (34.5) 69 (79.3) 18 (20.7) 70 (80.5) 17 (19.5) 
UICC stage n pb   465   0.130 465   0.851 465   0.778 
 IA/IB   82 (17.6) 46 (56.1) 36 (43.9) 65 (79.3) 17 (20.7) 69 (84.1) 13 (15.9) 
 IIA/IIB   99 (21.2) 59 (59.6) 40 (40.4) 74 (74.7) 25 (25.3) 79 (79.8) 20 (20.2) 
 IIIA/IIIB/IIIC   197 (42.3) 128 (65.0) 69 (35.0) 151 (76.6) 46 (23.4) 163 (82.7) 34 (17.3) 
 IV   87 (18.7) 57 (65.5) 30 (34.5) 69 (79.3) 18 (20.7) 70 (80.5) 17 (19.5) 
LN ratio n pa   465   0.088 465   0.659 465   0.399 
 Low (<0.189)   227 (48.7) 134 (59.0) 93 (41.0) 173 (76.2) 54 (23.8) 182 (80.2) 45 (19.8) 
 High (≥0.189)   238 (51.1) 156 (65.5) 82 (34.5) 186 (78.2) 52 (21.8) 199 (83.6) 39 (16.4) 
pL category n pa   448   0.267 448   0.736 448   1.000 
 pL0   218 (46.8) 138 (63.3) 80 (36.7) 166 (76.1) 52 (23.9) 178 (81.7) 40 (18.3) 
 pL1   230 (49.4) 138 (60.0) 92 (40.0) 179 (77.8) 51 (22.2) 188 (81.7) 42 (18.3) 
pV category n pa   447   0.644 447   0.595 447   0.702 
 pV0   397 (85.2) 246 (62.0) 151 (38.0) 307 (77.3) 90 (22.7) 325 (81.9) 22 (18.1) 
 pV1   50 (10.7) 29 (58.0) 21 (42.0) 37 (74.0) 13 (26.0) 40 (80.0) 10 (20.0) 
pR status n pa   461   0.883 461   0.494 461   0.142 
 pR0   405 (86.9) 252 (62.2) 153 (37.8) 312 (77.0) 93 (23.0) 328 (81.0) 77 (19.0) 
 pR1/R2   56 (12.0) 36 (64.3) 20 (35.7) 46 (82.1) 10 (17.9) 50 (89.3) (10.7) 
HER2 status n pa   437   1.000 437   0.531 437   0.365 
 Negative   401 (86.1) 254 (63.3) 147 (36.7) 312 (77.8) 89 (22.2) 328 (81.8) 73 (18.2) 
 Positive   36 (7.7) 23 (63.9) 13 (36.1) 30 (83.3) (16.7) 32 (88.9) (11.1) 
MET status n pa   455   1.000 455   0.825 455   0.812 
 Negative   424 (91.0) 266 (62.7) 158 (37.3) 325 (76.7) 99 (23.3) 347 (81.8) 77 (18.2) 
 Positive   31 (6.7) 20 (64.5) 11 (35.5) 25 (80.6) (19.4) 25 (80.6) (19.4) 
H. pylori status n pa   395   0.015 395   0.148 395   0.009 
 Negative   333 (71.5) 213 (64.0) 120 (36.0) 256 (76.9) 77 (23.1) 276 (82.9) 57 (17.1) 
 Positive   62 (13.3) 29 (46.8) 33 (53.2) 42 (67.7) 20 (32.3) 42 (67.7) 20 (32.3) 
EBV status n pa   452   0.105 452   0.586 452   0.227 
 Negative   432 (92.7) 266 (61.6) 166 (38.4) 334 (77.3) 98 (22.7) 352 (81.5) 80 (18.5) 
 Positive   20 (4.3) 16 (80.0) (20.0) 17 (85.0) (15.0) 19 (95.0) (5.0) 
MSI status n pa   451   0.279 451   0.835 451   0.249 
 MSS   416 (89.3) 265 (63.7) 151 (36.3) 323 (77.6) 93 (22.4) 344 (82.7) 72 (17.3) 
 MSI   35 (7.5) 19 (54.3) 16 (45.7) 28 (80.0) (20.0) 26 (74.3) (25.7) 
OS, months pc   454 0.204 454 0.984 454 0.970 
 Total/events/censored   454 (22.0) 284/225/59 170/129/41 352/273/79 102/81/21 372/286/86 82/68/14 
 Median survival   15.0 1.1 14.0±1.4 16.0±1.8 15.6±1.2 14.1±1.8 14.1±1.3 16.7±1.8 
 95% CI   12.9–17.1 11.3–16.7 12.4–19.6 13.2–18.1 10.6–17.6 11.4–16.7 13.1–20.3 
TSS, months pc   425 0.086 425 0.367 425 0.449 
 Total/events/censored   425 (31.8) 270/192/78 155/98/57 329/229/100 96/61/35 348/238/110 77/52/25 
 Median survival   16.6±1.4 15.0±1.6 17.9±3.6 16.6±1.6 16.6±4.5 15.6±1.6 17.9±4.1 
 95% CI   13.8–19.4 11.8–18.2 13.8–19.4 13.5–19.7 7.8–25.4 12.4–18.9 13.8–25.9 

aFisher’s exact test.

bKendall’s tau test.

cLog-rank test.

Immunohistochemistry

PCLO expression was examined in whole-mount tissue sections (Fig. 1). Cytoplasmic immunostaining was found in tumour, stromal, endothelial, muscle, and nerve cells.

Of the 466 cases, 175 (37.6%) showed PCLO immunostaining of tumour cells, while 291 (62.4%) cases were completely PCLO-immunonegative for tumour cells. Positive tumour cell staining ranged from weak (1+) to strong (3+) (Fig. 1). Immunostaining of at least 5% of tumour cells was found in 172 cases for weak (1+), 60 for moderate (2+), and 16 for strong immunostaining (3+). A combination of two staining intensities, i.e., 0, 1+, 2+, 3+, was found in 98 cases, of three in 49 cases and of all four in 14 cases. These data show that the expression of PCLO in GC is heterogeneous.

The HScore calculated from the percentages for each intensity in the tumour ranged from 0 to 190. When divided into quartiles and mostly negative cases, quartiles 1 and 2 were counted together. Quartile 3 counted 60 cases (12.9%) and quartile 4, 115 cases (24.7%). Within the stroma, 106 GCs were positive, and within the vessels, 84 GCs were positive.

Correlation with Clinicopathological Patient Characteristics

To correlate PCLO expression with different clinicopathological patient characteristics, we dichotomised the cohort at the median HScore into PCLO positive/high (HScore>0) and PCLO negative/low (HScore = 0). Interestingly, PCLO positive/high was more commonly found in intestinal and unclassifiable type GCs, locally less advanced GCs (T category), GCs with a lower lymph node ratio, and H. pylori-positive cases. However, after correction for multiple testing none of these associations proved to be significant (Table 1). No correlation was found after separation into quartiles and following pairwise comparison.

We then correlated the expression of PCLO with patient survival. Patients with PCLO negative/low showed a shortened OS and TSS (Fig. 2). The median OS was 14.0 ± 1.4 month for PCLO negative/low compared with 16.0 ± 1.8 months for PCLO positive/high (p = 0.204). Median TSS was 15.0 ± 1.6 months for PCLO negative/low versus 17.9 ± 3.6 months for PCLO positive/high (p = 0.086; Table 1; Fig. 2). Thus, low or loss of expression was associated with a worse prognosis. PCLO expression in stroma and vessels did not correlate with any clinicopathological patient characteristics (Table 1).

Fig. 2.

Kaplan-Meier plot showing OS (a) and TSS (b) in relation to the PCLO expression.

Fig. 2.

Kaplan-Meier plot showing OS (a) and TSS (b) in relation to the PCLO expression.

Close modal

Genotype-Phenotype Correlation

Finally, we correlated genotype and phenotype in nine cases from the discovery cohort that had previously undergone multiregional sequencing (n = 48 tumour samples) [10]. Table 2 lists all nine cases with between 3 and 10 tissue samples per case and mutations (missense mutations and nonsense mutations) detected by WES. Of the nine cases, five had PCLO mutations. In three cases, all tumour samples harboured a PCLO mutation (case #3, #6, and #9). Two cases (case #4 and #8) had a mutation in some but not all samples and four cases (case #1, #2, #5, and #7) had no PCLO mutations. Interestingly, two cases (case #6 and #9) had two different mutations. These data provide evidence of the intratumoral genetic heterogeneity of PCLO in GC.

Table 2.

Correlation of genotype and phenotype: shown are the Hscores of PCLO of 48 tumour samples from 9 patients of the discovery cohort with known mutation status

Case numberSampleGeneVariant classAmino acid changeAllele frequencyGeneVariant classAmino acid changeAllele frequencyPCLO_Hscore
#1 G13406         
G13407         20 
G13408         
G13409         
#2 G13401         10 
G13402         
G13404         80 
#3 G04240 PCLO Nonsense p.Q375Pfs*6 0.1695     30 
G04241 PCLO Nonsense p.Q375Pfs*6 0.210191083     10 
G04242 PCLO Nonsense p.Q375Pfs*6 0.242990654     20 
G04244 PCLO Nonsense p.Q375Pfs*6      70 
G04245 PCLO Nonsense p.Q375Pfs*6      10 
G04283 PCLO Nonsense p.Q375Pfs*6      30 
#4 G09209         
G09210 PCLO Missense p.A497T 0.0759     
G09211         
G09212         
G13370 PCLO Missense p.A497T 0.0289     
#5 G13389         40 
G13390         
G13391         100 
G13392         
G13393         
G13394         
G13395         10 
G13396         60 
G13397         
G13398         
#6 G13383 PCLO Missense p.P2517L 0.1429     
G13384 PCLO Missense p.P2517L 0.1421     
G13385 PCLO Missense p.P2517L 0.125 PCLO Missense p.A885T 0.125 
G13386 PCLO Missense p.P2517L 0.2222     130 
G13387 PCLO Missense p.P2517L 0.1088     50 
#7 G13378         
G13379         
G13380         
G13381         120 
#8 G09201 PCLO Missense p.S496P 0.034188034     40 
G09202 PCLO Missense p.S496P 0.034482759     140 
G09203 PCLO Missense p.S496P 0.03875969     20 
G09204 PCLO Missense p.S496P 0.063583815     100 
G09205 PCLO Missense p.S496P 0.017647059     40 
G09206         
#9 G13372 PCLO Missense p.S496P 0.0787 PCLO Missense p.A497T 
G13373 PCLO Missense p.S496P 0.04624277 PCLO Missense p.A497T 0.027472527 
G13374 PCLO Missense p.S496P 0.02953586     
G13375 PCLO Missense p.S496P 0.04166667 PCLO Missense p.A497T 
G13376 PCLO Missense p.S496P 0.07734807 PCLO Missense p.A497T 0.0833 
Case numberSampleGeneVariant classAmino acid changeAllele frequencyGeneVariant classAmino acid changeAllele frequencyPCLO_Hscore
#1 G13406         
G13407         20 
G13408         
G13409         
#2 G13401         10 
G13402         
G13404         80 
#3 G04240 PCLO Nonsense p.Q375Pfs*6 0.1695     30 
G04241 PCLO Nonsense p.Q375Pfs*6 0.210191083     10 
G04242 PCLO Nonsense p.Q375Pfs*6 0.242990654     20 
G04244 PCLO Nonsense p.Q375Pfs*6      70 
G04245 PCLO Nonsense p.Q375Pfs*6      10 
G04283 PCLO Nonsense p.Q375Pfs*6      30 
#4 G09209         
G09210 PCLO Missense p.A497T 0.0759     
G09211         
G09212         
G13370 PCLO Missense p.A497T 0.0289     
#5 G13389         40 
G13390         
G13391         100 
G13392         
G13393         
G13394         
G13395         10 
G13396         60 
G13397         
G13398         
#6 G13383 PCLO Missense p.P2517L 0.1429     
G13384 PCLO Missense p.P2517L 0.1421     
G13385 PCLO Missense p.P2517L 0.125 PCLO Missense p.A885T 0.125 
G13386 PCLO Missense p.P2517L 0.2222     130 
G13387 PCLO Missense p.P2517L 0.1088     50 
#7 G13378         
G13379         
G13380         
G13381         120 
#8 G09201 PCLO Missense p.S496P 0.034188034     40 
G09202 PCLO Missense p.S496P 0.034482759     140 
G09203 PCLO Missense p.S496P 0.03875969     20 
G09204 PCLO Missense p.S496P 0.063583815     100 
G09205 PCLO Missense p.S496P 0.017647059     40 
G09206         
#9 G13372 PCLO Missense p.S496P 0.0787 PCLO Missense p.A497T 
G13373 PCLO Missense p.S496P 0.04624277 PCLO Missense p.A497T 0.027472527 
G13374 PCLO Missense p.S496P 0.02953586     
G13375 PCLO Missense p.S496P 0.04166667 PCLO Missense p.A497T 
G13376 PCLO Missense p.S496P 0.07734807 PCLO Missense p.A497T 0.0833 

To correlate genotype with phenotype, we performed immunohistochemical staining of tissue sections obtained from 48 paraffin blocks (=samples) covering the anatomical regions of the tumours from which tissue samples were obtained for WES. The histoscore was determined for each sample (Table 2, PCLO_Hscore). For PCLO, the histoscore ranged from 0 to 140. With 27 (56.3%) negative samples, the median was 0. Ten samples from two cases were completely immunonegative (case #4 and #9). Tissue samples from seven cases showed a heterogeneous immunostaining pattern and no direct correlation was found between protein expression and missense or nonsense mutations (Table 2, PCLO_Hscore).

Comprehensive genetic analysis of malignant tumours leads to the discovery of mutations in genes whose tumour-biological significance for the respective tumour type is often incomplete or even unknown. In a recent study using multiregional WES of GC, we detected PCLO mutations in 23 of 48 tumour samples. Six mutations were nonsense mutations (p.Q375Pfs*6) and 22 were missense mutations (ten p.S496p, six p.A497T, five p.P2517L, and one p.A885T) (Table 2) [10]. In addition, two cases each harboured two different mutations of the PCLO [10]. Wang et al. [20] also pointed out that PCLO is one of the most frequently mutated genes in GC. Based on these genetic data, we set out to test the hypothesis that PCLO is expressed in GC and has a major biological significance.

At the beginning of our study, the role of PCLO in GC was almost unknown. Previous studies have shown that PCLO contributes to tumour aggressiveness in ESCC and that PCLO knockdown significantly attenuates ESCC malignancy [9]. Our study provides evidence that data from ESCC cannot be directly extrapolated to GC. Our data showed that among the different phenotypes of GC, intestinal-type GCs were significantly more likely to express PCLO compared to diffuse type GCs (43.5% vs. 27.4%). This finding supports the notion that the tumour biological significance of PCLO may depend on the tumour phenotype. In general, intestinal type GCs have a better prognosis. It was interesting to note that the number of PCLO-positive GCs decreased with increasing local tumour growth, suggesting that PCLO may act as a tumour suppressor. The putative tumour suppressive role of PCLO was underlined by the shortened OS and TSS of PCLO negative/low GCs.

Regarding H. pylori status, we found that PCLO was more frequently lost in H. pylori-negative GCs. This may be an epiphenomenon, as in our cohort H. pylori was less commonly found in diffuse type GCs [14]. Alternatively, H. pylori positive and negative GCs may differ in their aetiology, which may have an impact on the expression of PCLO. However, as we did not find an association between PCLO expression and molecular subtypes of GC, i.e., EBV, MSI, or HER2 status, phenotype may be more important. H. pylori-positive GCs were more likely to be PCLO positive/high. This finding supports the notion that loss of PCLO is a late event and not necessary for tumour initiation. This may also explain why multiple mutations were found in the same tumour [10]. Loss of PCLO could be a late, subclonal event, i.e., occurring at more advanced tumour stages.

Many previous studies in the same cohort have demonstrated substantial intra- and intertumoral heterogeneity in GC for various biomarkers, either due to intra- and intertumoral genetic heterogeneity or cancer cell plasticity [15, 17, 18, 21]. Our study demonstrated that this heterogeneity also applies to PCLO, as a combination of two staining intensities, i.e., 0, 1+, 2+, 3+, was found in 98 cases, of three in 49 cases and of all four in 14 cases.

There are several putative explanations for the intra- and intertumoral heterogeneity in GC. Multiregional sequencing has recently shown that PCLO mutations can be clonal and subclonal [4]. Subclonal mutations contribute to intratumoral genetic heterogeneity and hence to divergent histological phenotypes. Two cases harboured two different mutations of PCLO (case #6 and #9; Table 2) [10].

Focussing on mutations alone can be misleading and analysis of expression patterns and tumour context (i.e., intestinal vs. diffuse type) provides valuable additional information. According to the Knudson two-hit model, both alleles of a tumour suppressor must be inactivated, e.g., by mutation, loss of heterozygosity, or epigenetic silencing, to induce a phenotypic change. Our validation cohort provided no direct evidence that missense mutations are associated with loss of protein expression of PCLO. However, as we did not examine loss of heterozygosity and methylation status in our discovery cohort, we cannot exclude that the second allele was still functional. At least, the discovery cohort supports the presence of intratumoral heterogeneity at the genetic and phenotypic level.

Limitations

Our observational study design provides no functional data, and no comment can be made on the effect nonsense and missense mutations may have on GC biology. This might be rather complex and complicated to explore since the COSMIC database lists >400 different PCLO mutations in GC, the vast majority being point mutations. Using the search terms “cancer” and “PCLO” in PubMed provides only a limited number of studies. None addressed the tumour biological significance of PCLO in adenocarcinomas of the stomach.

Summing up, our small study provides evidence that PCLO is differentially expressed in GC and may have a tumour suppressive function in a context-dependent manner. Further studies on this frequently mutated gene in GC are warranted.

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1964 and later versions. Our project was granted ethical clearance by the Local Ethics Committee of the University Hospital Schleswig-Holstein, Campus Kiel, Germany, in agreement with the Helsinki Declaration (D 453/10 and D 549/20). Consent was not required for this retrospective study on archival tissue specimens in accordance with local or national guidelines (Ethics Committee of the University Hospital Schleswig-Holstein, Campus Kiel, Germany [D 453/10 and D 549/20]).

The authors have no conflicts of interest to declare.

There was no third-party funding.

Conceptualization, supervision, project administration, funding acquisition: C.R.; methodology: M.B., S.K., H.-M.B.; software: H.-M.B.; formal analysis: data curation, writing—review and editing, all authors; visualization: M.B., H.-M.B.; investigation: M.B., C.R.; resources: C.R., S.K. All authors have read and agreed to the published version of the manuscript.

All data are included in the manuscript. Further enquiries can be directed to the corresponding author.

1.
(2019) ZfKZiRK-I. Datensatz des ZfKD auf Basis der epidemiologischen Landeskrebsregisterdaten, verfügbare Diagnosejahre bis 2016. Version: epi2018_1. 2019.
2.
Crew KD, Neugut AI. Epidemiology of gastric cancer. World J Gastroenterol. 2006;12(3):354–62.
3.
Herrero R, Park JY, Forman D. The fight against gastric cancer: the IARC Working Group report. Best Pract Res Clin Gastroenterol. 2014;28(6):1107–14.
4.
Amieva M, Peek RMJr. Pathobiology of Helicobacter pylori-induced gastric cancer. Gastroenterol. 2016;150(1):64–78.
5.
Smyth EC, Nilsson M, Grabsch HI, van Grieken NC, Lordick F. Gastric cancer. Lancet. 2020;396(10251):635–48.
6.
Röcken C. Predictive biomarkers in gastric cancer. J Cancer Res Clin Oncol. 2022;149(1):467–81.
7.
Ahmed MY, Chioza BA, Rajab A, Schmitz-Abe K, Al-Khayat A, Al-Turki S, et al. Loss of PCLO function underlies pontocerebellar hypoplasia type III. Neurology. 2015;84(17):1745–50.
8.
Mbarek H, Milaneschi Y, Hottenga JJ, Ligthart L, de Geus EJC, Ehli EA, et al. Genome-wide significance for PCLO as a gene for major depressive disorder. Twin Res Hum Genet. 2017;20(4):267–70.
9.
Zhang W, Hong R, Xue L, Ou Y, Liu X, Zhao Z, et al. Piccolo mediates EGFR signaling and acts as a prognostic biomarker in esophageal squamous cell carcinoma. Oncogene. 2017;36(27):3890–902.
10.
Röcken C, Amallraja A, Halske C, Opasic L, Traulsen A, Behrens HM, et al. Multiscale heterogeneity in gastric adenocarcinoma evolution is an obstacle to precision medicine. Genome Med. 2021;13(1):177.
11.
The Cancer Genome Atlas Research Network. Comprehensive molecular characterization of gastric adenocarcinoma. Nature. 2014;513(7517):202–9.
12.
Brierley J, Gospodarowicz MK, Wittekind C. TNM classification of malignant tumours. 8th ed. Wiley Blackwell; 2017.
13.
Lauren P. The two histologic main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. Acta Pathol Microbiol Scand. 1965 1965;64:31–49.
14.
Warneke V, Behrens HM, Haag J, Balschun K, Böger C, Becker T, et al. Prognostic and putative predictive biomarkers of gastric cancer for personalized medicine. Diagn Mol Pathol. 2013;22(3):127–37.
15.
Böger C, Krüger S, Behrens HM, Bock S, Haag J, Kalthoff H, et al. Epstein-Barr virus-associated gastric cancer reveals intratumoral heterogeneity of PIK3CA mutations. Ann Oncol. 2017;28(5):1005–14.
16.
Mathiak M, Warneke VS, Behrens HM, Haag J, Böger C, Krüger S, et al. Clinicopathologic characteristics of microsatellite instable gastric carcinomas revisited: urgent need for standardization. Appl Immunohistochem Mol Morphol. 2017;25(1):12–24.
17.
Warneke VS, Behrens HM, Böger C, Becker T, Lordick F, Ebert MP, et al. Her2/neu testing in gastric cancer: evaluating the risk of sampling errors. Ann Oncol. 2013 3;24(3):725–33.
18.
Metzger ML, Behrens HM, Böger C, Haag J, Kruger S, Röcken C. MET in gastric cancer: discarding a 10% cutoff rule. Histopathology. 2016;68(2):241–53.
19.
Simes RJ. An improved Bonferroni procedure for multiple tests of significance. Biometrika. 1986;73(3):751–4.
20.
Wang H, Shen L, Li Y, Lv J. Integrated characterisation of cancer genes identifies key molecular biomarkers in stomach adenocarcinoma. J Clin Pathol. 2020;73(9):579–86.
21.
Böger C, Behrens HM, Röcken C. Ki67--An unsuitable marker of gastric cancer prognosis unmasks intratumoral heterogeneity. J Surg Oncol. 2016;113(1):46–54.