Introduction
Diffuse large B-cell lymphoma (DLBCL), the most common form of non-Hodgkin’s lymphoma (NHL), causes 30–40% of NHL in adults. DLBCL patients usually present with tumors in single or multiple lymph nodes or extranodal sites, and the tumors grow rapidly. In all DLBCL cases, the common type accounts for 80–85%, and the rare type, consisting of large B cells, accounts for approximately 15–20%. The WHO classification system classified DLBCL into activated B-cell-like (ABC) and germinal center B cell-like (GCB) but was unable to classify DLBCL according to gene expression characteristics, indicating that DLBCL is a heterogeneous disease. Based on this finding, pathologists, in addition to conducting a preliminary diagnosis, need to conduct a more detailed study of the existing DLBCL subtypes to explore potential markers associated with differences in prognosis for further classification studies.
P63 expression is tissue specific, and it is restricted to epithelial cells and certain subpopulations of basal cells in normal tissues. However, some P63 positive cells can occasionally be seen in the germinal centers of lymph nodes and seem to be related to the development of lymphoma [
1‐
3]. Because of an alternate intronic promoter and alternative splicing, P63 encodes two major different isoforms [
4], a full length transactivating (TA) domain (TAP63/P63) and a truncated form without the N-terminus (∆NP63/P40). The function of TAP63/P63 is similar to P53, which can regulate the expression of P53 downstream target genes, block the cell cycle, induce apoptosis, and may be a candidate tumor suppressor. However, ∆NP63/P40 can antagonize P53 or P63 by competitively binding DNA sites, thereby promoting cell proliferation and inhibiting apoptosis, and it functions similar to an oncogene. In a range of human cancers, the expression of P63 was proven to be associated with tumor development. Kenji Hibi et al. [
5] found that overexpression of ∆NP63/P40 may lead to increased epithelial stem cell renewal and promote tumor growth in squamous cell carcinoma. However, there were controversial conclusions on the role of P63 and its prognostic significance in lymphoma. A study from Japan showed the expression of the P63 protein in 34% of cases with poor overall survival (OS) [
6]. However, a study from America found that a subset of DLBCL (32% of cases) expressed P63, but it did not correlate with overall survival [
7]. China has a large number of DLBCL patients, but no study thus far has reported P63 expression in Chinese DLBCL patients. We evaluated P63 expression by immunohistochemistry to determine the expression of P63 and its prognostic significance in a Chinese DLBCL cohort.
Discussion
P63, an important transcription factor, was discovered in 1998 and is located on chromosome 3q27–28. The P63 gene has structural and functional homology with the P53 gene family, regulating downstream target genes, activating various signaling pathways, and participating in the regulation of a variety of biological functions. P63 is at the key node of the regulation network, involved in mechanisms of tumorigenesis and development, such as cell cycle regulation, apoptosis, differentiation and cell adhesion and migration. It is well known that P53 is usually a tumor suppressor gene, but many studies have found that P63 may promote tumor development in human primary tumors and cell lines.
DLBCL (diffuse large B-cell lymphoma) exhibits clinical heterogeneity and responds differently to treatment and prognosis. Although survival rates can be estimated based on clinical parameters, recent literature reports that a group of tumor suppressor proteins and oncogenic proteins are associated with prognosis [
10]. However, at present, there are contradictory results about the prognostic significance of P63 in lymphoma, especially in DLBCL. In addition, in our daily pathology work, especially in some needle biopsy cases, DLBCL may mimic carcinoma cells, presenting a round, oval, or polygonal shape and clear nuclei that are positive for P63, and we found that P63 is expressed in a considerable proportion of DLBCL. Under the circumstances, it may easily be misguided. In our cohort, we found there was no P40 (a specific marker of squamous cell carcinoma) expression in DLBCL, which may be extremely useful for the differential diagnosis of poorly differentiated squamous cell carcinoma versus DLBCL, especially in small sample needle biopsies.
P63 is a particularly useful marker in the differential diagnosis of lymphoma as well, with a high positive predictive value of 96% for primary mediastinal large B cell lymphomas, but very rare in CHL (classical Hodgkin’s lymphoma) [
11]. Zamò A et al. [
12] also found that P63 was a useful diagnostic marker of primary mediastinal large B-cell lymphoma at both the protein and mRNA levels. Shi QY et al. [
13] used immunohistochemical methods to show that tumor cells of mediastinal large B-cell lymphoma were highly positive for P63 (84%, 16/19), and their results were consistent with ours (81%, 17/21). However, few studies have investigated the expression and prognosis of P63 in DLBCL. In 2002, Di Como et al. [
2] found a P63-positive population in non-Hodgkin’s B-cell lymphoma and normal lymph nodes and detected the expression of P63 in peripheral blood lymphocytes by Western blot. In addition, P63 expression levels are higher in tumor lymphocytes compared to normal lymphoid tissue. However, they did not explore the prognostic significance of P63 positive cases. In a study regarding 64 primary cutaneous large B cell lymphomas, Alistair Robson et al. [
14] demonstrated that the expression of P63 protein was significantly increased in pc DLBCLL (70%) compared to pc FCCL (12%) (
P < 0.001). Evaluation by Ki67 immunostaining revealed that high expression of P63 was associated with a higher proliferation index (
P = 0.015). However, when analyzing the current data, they did not find a significant association between P63 expression and patient outcome. Hedvat et al. [
7] demonstrated that 32% of cases expressed P63 in the nuclei of DLBCL cells, and P63 expression is associated with the proliferation index; they also concluded that P63 did not correlate with overall survival. Interestingly, after that, conflicting results regarding P63 expression and its prognostic role in DLBCL were published. In 2005, Park CK et al. [
15] found that P63 was expressed in 32/61 cases (52.5%) of DLBCL, including 15 strong expression cases in their cohorts, and it was a significant poor survival factor in DLBCL (
p = 0.0228). Fukushima et al. [
6] investigated and evaluated the expression pattern of P63 in B-cell lymphomas and found that P63 was expressed in the nuclei of tumor cells in 22 of 65 (34%) DLBCL cases, and they also reported that P63 is a disadvantageous factor for prognosis in DLBCL. However, although Hallack et al. [
16] observed P63 expression in 15.1% of DLBCL cases, no correlation was found between P63 expression and OS (
p = 0.09). Furthermore, they proposed that being P63(+) provided a protective effect on high-intermediate and high risk DLBCL, and DLBCL patients with P63(+) have a better DFS than negative cases [
16]. A recent study concerning high-risk diffuse large B-cell lymphoma also reported that P63 expression confers significantly better survival outcomes [
17].
In the present study, we characterized the expression of P63 in a series of 159 Chinese DLBCL cases and found that 76 (47.8%) cases were P63 positive regardless of the intensity and percentage, and the results were completely consistent with studies from Japan and Korea. In addition, we found that 4/8 (50%) EBV+ DLBCL cases expressed P63 in the nucleus, while in 2/8 (20%) cases, it was highly expressed. Although the sample size was small, the results were almost the same as DLBCL in our cohort, indicating that EBV status had no effect on P63 expression. At the same time, we also discovered that P63 positive patients had a worse prognostic value, which was inconsistent with previous studies based on non-Asian populations. We think a possible cause of the inconsistent findings is ethnic differences because all of the reports of poor prognosis were Asians. This conclusion needs further research and confirmation with large samples. Other possible reasons might be technical factors related to staining quality, cut-off value, data interpretation, treatment regimens and follow-up time. For instance, Hallack et al. [
16] used a 50% cut-off to indicate immunohistochemical positivity. In our study, a cut-off value of “≥6” was used based on the adjusted Allred scoring system and ROC curve, and this difference should be tested in other studies to confirm its repeatability.
Furthermore, we did not observe a significant correlation between P63 expression patterns and subgroups of GCB-like DLBCL and non-GCB DLBCL, which is consistent with the research of Hallack [
16]. We discovered that P63 expression was correlated with the Ki67 proliferative index in DLBCL, and P63 gene amplification has been reported to correlate with the Ki67 proliferative index in lung cancer [
18]. Miller et al. [
19] reported that a tumor proliferation index of > 80% was associated with poorer survival in previously untreated patients with aggressive NHL, and we also confirmed this correlation in our DLBCL cohort. Several reports [
20‐
24] have revealed that P53 mutations and P53 overexpression are common events in DLBCL. Their findings were very consistent, as P53 abnormalities were associated with a poor prognostic indicator. In addition, P53 expression in ≥50% of lymphoma cells was used as a cut-off value, which is consistent with our study. In our combined analysis, it was found that the prognosis for P63 and P53 co-expression cases is extremely poor. Whether these DLBCL patients need radical treatment still requires further research in the future.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.