Anaplastic large cell lymphoma (ALCL), is a type of lymphoma that can be classified into two subtypes: ALK⁺ and ALK^-. Both subtypes are characterized by the presence of large anaplastic cells with diffuse and strong expression of CD30 on the cell membrane and in the Golgi region (1, 2). The diagnosis of ALK⁺ ALCL is usually straightforward. However, for most cases of ALK^- ALCL, no unifying genetic abnormality has been identified, making the diagnosis more reliant on morphology features and CD30 expression (3).

Despite its strong expression on anaplastic cells in ALK⁺ and ALK^- ALCL, CD30 is not a specific marker for these malignancies. In fact, a variety of T- or B-cell lymphomas, as well as reactive lymphoproliferative disorders, and non-hematopoietic lesions, may also express CD30 to varying degrees. Among these entities, peripheral T-cell lymphoma, not otherwise specified (PTCL, NOS) with CD30 expression can be particularly difficulty to distinguish from ALK^- ALCL. Although most cases of PTCL, NOS are negative or only weakly positive for CD30 in only a minority of tumor cells (4–6), a small subset exhibit diffuse and strong CD30 expression(CD30^high PTCL, NOS). In such cases, the distinction between ALK^- ALCL and PTCL, NOS becomes ambiguous, and the diagnosis may be subjective (5, 7).

Differentiating between these two conditions is clinically important since patients with ALK^- ALCL typically have better overall survival and respond more favorably to Brentuximab vedotin compared to those with PTCL, NOS (8). Unfortunately, due to the lack of widely accepted diagnostic criteria or reliable biomarkers to differentiate between ALK^- ALCL and CD30^high PTCL, NOS, accurately diagnosing these two conditions remains a challenge and requires careful consideration of clinical, morphological, and immunophenotypic features by an experienced pathologist.

Aberrant activation of the signal transducer and activator of transcription 3 (STAT3) has been identified as a key oncogenic event in both ALK ⁺ and ALK ^- ALCL (9–15). In ALK⁺ ALCL, the NPM::ALK fusion protein drives STAT3 activation by promoting the phosphorylation of STAT3 (15). In ALK^- ALCL, the STAT3 pathway can be activated by alternative upstream signals, such as JAK1/STAT3 mutations (5, 16), NFkB2::ROS1 and NFkB::TYK2 fusions (9). However, JAK1/STAT3 mutations are not commonly observed in PTCL, NOS (9). These findings raised the question of whether the activation status of STAT3 could be used to distinguish ALK^- ALCL from CD30^high PTCL, NOS. Notably, phosphorylation of STAT3 at tyrosine 705 (Y705) was found to be positive in 84-93% of ALK⁺ ALCL and 47-57% of ALK^- ALCL (17, 18). The phosphorylation status of STAT3 in PTCL, NOS remains largely unknown, and the phosphorylation of STAT3 at another activation site, serine 727 (S727) has not been well studied, in contrast to the frequent observation of positive phosphorylation at Y705 in both ALK ⁺ and ALK ^- ALCL.

This study aimed to develop an immunohistochemistry (IHC)-based classifier using two commercially available antibodies targeting different phosphorylation sites of STAT3, namely Y705 and S727. The goal was to determine whether the phosphorylation status of these residues could serve as a biomarker for distinguishing between ALK^- ALCL and CD30^high PTCL, NOS. Furthermore, the study investigated whether the presence of phosphorylated STAT3 (pSTAT3) could predict clinical outcomes in patients with ALCL or PTCL, NOS.

The diagnostic dilemma of ALK^- ALCL vs. CD30^high PTCL, NOS is one of the most challenging scenarios in the daily pathology practice. Although CD30 expression in PTCL, NOS is mostly weak and heterogeneous, some cases do express CD30 diffusely (80%) and strongly. Given the morphological and immunophenotypic overlap between ALK^- ALCL and PTCL, NOS, and no clinically available biomarker to differentiate them, the diagnosis between ALK^- ALCL and CD30^high PTCL, NOS could be challenging. We investigated whether phosphorylated STAT3 could help with the differential diagnosis and serve as a predictive biomarker by studying the expression of pSTAT3-Y705 and -S727 in 89 cases of ALK^+/- ALCL and PTCL, NOS, including 10 cases of CD30^high PTCL, NOS. Our results showed that pSTAT3-Y705 was more specific while pSTAT3-S727 was more sensitive in the differential diagnosis between ALK^- ALCL vs. CD30^high PTCL, NOS, with a relatively balanced sensitivity and specificity. Furthermore, the higher expression of pSTAT3-S727 in the TILs of PTCL, NOS was associated with a better prognosis.

Given the significance of STAT3 activation in tumorigenesis, researchers have undertaken different scoring strategies to evaluate pSTAT3 expression. However no cutoff value or scoring system has been widely accepted for the IHC staining of pSTAT3. In the present study, we used H-score, a well-established scoring system taking both staining extent and intensity into account, to assess pSTAT3-Y705 and -S727 expression. When studying the prognostic significance of pSTAT3-S727, considering the proportion of TILs in tumor areas might also affect patient prognosis, we scored S727_TILs by multiplying the H-score of S727_TILs by the percentage of TILs in tumor areas. However, a more concise and applicable semiquantitative scale may be needed in future clinical practice.

Despite differences in gene alterations between ALK⁺ ALCL and ALK^- ALCL, they share some deregulated transcription factors programs like STAT3 and NOTCH1 activation, IRF4, and c-MYC overexpression (9, 26–29). STAT3 activation has been reported in ALK⁺ ALCL and ALK^- ALCL as a well-defined oncogenic driver through different machanisms (9, 12–14, 29–31). In the present study, we found that both pSTAT3-Y705 and -S727 were diffusely expressed by tumor cells in ALCL, irrespective of ALK status; though pSTAT3-Y705 was more specific and pSTAT3-S727 is more sensitive. However, they were less frequently and variably expressed in PTCL, NOS, implying they could be reliable differential biomarkers between ALCL vs. PTCL, NOS. When narrowing down to differential diagnosis between ALK^- ALCL and CD30^high PTCL, NOS, considering a minority (13%) of ALK^- ALCL expressing pSTAT3-Y705 in a very low level (<30%), pSTAT3-S727 appears to be a better choice, with a sensitivity of 0.86 and specificity of 0.9. Our results suggest that pSTAT3-S727 IHC staining could help distinguish ALK^- ALCL from PTCL, NOS, even when CD30 is diffusely positive: if the H-score of pSTAT3-S727 expression surpasses 145, a diagnosis of ALK^- ALCL is favorable. This IHC-based classifier guides the differential diagnosis between ALK^- ALCL and CD30^high PTCL, NOS with a sensitivity of 0.86 and specificity of 0.9, suggesting pSTAT3-S727 is a promising differential biomarker for ALK^- ALCL vs. CD30^high PTCL, NOS.

For what reason the STAT3 phosphorylation status in PTCL, NOS differs from ALK^- ALCL remains unclear. In ALK⁺ ALCL, ALK kinase constitutively drives STAT3 phosphorylation (30). However, in ALK^- ALCL, only ~20% of ALK- ALCLs have activate mutations of JAK1 and/or STAT3 genes (9), which could not explain the high rate of pSTAT3 positivity both in our study and other researchers’ work (17, 18, 32). Additionally, compared with ALK^+/- ALCL, a lower proportion of PTCL, NOS expresses pSTAT3, reported also by other researchers (14, 18). In line with this, the mutations JAK1 and/or STAT3 was a rare event in PTCL, NOS: M. Rodríguez et al. detected JAK1/STAT3 mutations in 0/14 PTCL, NOS patients (33). However, E. Andersson et al. reported a higher rate of STAT3 mutations in PTCL, NOS in comparison with ALK^- ALCL (17% vs. 13%), and a lower rate of JAK1/3 mutations (0 vs. 15%) (14). In our study, five PTCL, NOS and one ALK^- ALCL cases were subjected to DNA sequencing. All patients, including those who were positive for pSTAT3-Y705 and/or pSTAT3-S727 were negative for JAK1/STAT3 mutations. The low prevalence of JAK1/STAT3 mutations makes the positivity rate vary in different studies. As of now, there is still a lack of positive association between pSTAT3 expression and JAK1/STAT3 mutations. Other signaling pathway was suggested to contribute to phosphorylation of STAT3. For example, U. Rozovski et al. reported that in chronic lymphocytic leukemia (CLL), the constitutive phosphorylation of STAT3 was induced by CK2-BLNK-CD5 complex (34). Further studies are needed to explore the mechanisms of STAT3 phosphorylation other than JAK1/STAT3 mutations.

Our results showed that pSTAT3-Y705 and -S727 in tumor cells were not prognostically predictive in ALCL patients. For PTCL, NOS patients, only the expression of S727_TILs was associated prognosis: those with higher S727_TILs H score tended to have a higher 3-year OS rate than those with a lower S727_TILs H score (43% vs. 0%, p=0.013). This result is plausible, given that PTCL, NOS is characterized by abundant background inflammatory cells including TILs. The interplay between tumor cells and the background immune cells mediated by cytokines release or exosome intake might contribute to the poor prognosis of PTCL, NOS patients. Our results suggested that the activation of STAT3 might contribute to the prognostic significance of TILs in PTCL, NOS patients. Furthermore, we found that high S727_TILs H score was enriched in PTCL, NOS patients with low MPIT risk factors compared to the low S727_TILs group (data not shown, p<0.05). MPIT model had a good performance in predicting patient prognosis in our study. This suggests that S727_TILs could help stratify PTCL, NOS patients with low MPIT index. However, the small sample size of the study warrants cautious interpretation of the difference in OS between the subgroups based on S727_TILs H score. Larger studies are needed to confirm findings and provide more robust evidence.

The discrepancy of pSTAT3-Y705 and pSTAT3-S727 was observed in our study, with pSTAT3-Y705 expressed mainly by lymphoma cells and pSTAT3-S727 also positive in background besides lymphoma cells. This was verified in our flowcytometric analysis, where pSTAT3-S727 were constitutively positive in most T/B lymphocytes of normal tonsil and PTCL, NOS samples, and pSTAT3-Y705 were absent in all cell populations. There was still a lack of knowledge about this phenomena. As we know, the canonical IL-6/JAK signaling pathway activates STAT3 at Y705 and S727 simultaneously, and S727 phosphorylation is indispensable for STAT3 activation to its fullest extent (35). However, unlike Y705, S727 was activated also by alternative upstream signaling pathways. For example, the TLR-4 recruited TBK-1 to phosphorylate STAT3 S727 in macrophage cells (36). Additionally, S727 was phosphorylated by MAPK in embryonic stem cells (37), correlated with YAP1 in glioma (38) and PIK3K-AKT-mTOR in colorectal cancer cells (38). In summary, unlike Y705, S727 was more frequently activated by alternative upstream signaling instead of IL-6/JAK signaling, and represented an non-canonical way of STAT3 activation. Furthermore, pSTAT3-Y705 and pSTAT3-S727 might play different roles in various conditions. Liang et al. reported that only STAT3 Y705 but not S727 promoted cancer cell EMT and metastasis through slug (39). Mandal et al. reported that overexpression of CK2 reduced Stat3 S727 phosphorylation enhanced tumorigenicity, but conversely phosphorylated Stat3 Y705 in similar conditions (40). The uncoupling of Y705 and S727 was observed in Kaposi sarcoma, in which case, the viral infection exclusively induced S727 phosphorylation to foster a unique chronic inflammatory environment by activating mitogen-activated kinase-activated protein (MAPKAP) kinase 2 (MK2) chronic inflammatory (41). We still have no evidence to explain the discrepancy of Y705 and S727 in our case, however we suggested the unique tumor microenvironment consisting of unknown released cytokines and microorganisms infection like Epstein-Barr virus (EBV), could be the reasons.

In summary, we found that ALCL and PTCL, NOS show different activation status of STAT3. Based on this difference, we established an IHC classifier to help differential diagnosis between ALK^- ALCL and PTCL, NOS with diffuse CD30 expression. Furthermore, we found that a higher expression of pSTAT-S727 by TILs in PTCL, NOS predicts a better prognosis.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by the ethics committee of Affiliated Hospital of Xuzhou Medical University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

HL designed the study. CX and WW designed the study, performed the experiments, analyzed the data and drafted this manuscript. ZW, XF, CL, and JL reviewed part of the slides. LX, GL, and HS performed the IHC assays. YG, NL, DL, MF, and YW collected the clinical information of the patients. SH helped to improve the manuscript. All authors contributed to the article and approved the submitted version.