Immunotherapy has significantly revolutionized the treatment and management of malignant tumors, particularly in malignant melanoma, lung, and renal cancers [13, 14]. However, pancreatic cancer remains challenging due to its immunosuppressive tumor microenvironment (TME) and dense stroma [15]. Research on ICIs, such as PD-1/PD-L1 and CTLA-4 inhibitors, has shown limited efficacy in clinical trials. For instance, the clinical trial NCT02798536, which combined ICIs with chemotherapy or radiation, showed partial responses but low overall response rates (ORRs). Promising results have emerged from combing ICIs with vaccines. The GVAX vaccine, derived from irradiated allogeneic pancreatic cancer cells, upregulates PD-1/PD-L1, indicating a potential synergistic effect with ICIs [16]. Clinical trials combining GVAX and ipilimumab have shown improved median overall survival compared to ipilimumab alone, though not statistically significant. Additionally, new targets like indoleamine 2,3-dioxygenase (IDO) and CCR2 inhibitors are being explored. A phase I trial with an IDO inhibitor plus gemcitabine/nab-paclitaxel in metastatic pancreatic cancer patients showed a 37% ORR. Another trial with the CCR2 inhibitor PF-04136309 and FOLFIRINOX for borderline resectable and locally advanced pancreatic cancer achieved a 49% ORR. Currently, pancreatic cancer immunotherapy focuses on combination strategies, with optimal regimens still under investigation. In addition to exploring treatment combinations, integrating immunonutrition and addressing treatment-induced hypertransaminasemia are emerging areas of interest that may enhance the effectiveness of immunotherapy in pancreatic cancer [17, 18].

Our patient was a 59-year-old male with pancreatic cancer and multiple metastases. The next generation sequencing report suggested potential effectiveness for PD-1 immunotherapy, so the initial treatment strategy was gemcitabine/nab-paclitaxel plus serplulimab immunotherapy. This is the first case of HPD in the context of immunotherapy combined with chemotherapy.

In this case, we mainly replenish the definitions of HPD. Several definitions of HPD are widely accepted. Saâda-Bouzid et al. defined HPD as an increase of at least twofold in the TGK during PD-1/PD-L1 inhibitor therapy (post-TGK) compared with the TGK before PD-1/PD-L1 inhibitors (pre-TGK) [19]. Similarly, Aoki et al. defined HPD as post-TGR/pre-TGR ≥ 2. Such definitions based on tumor growth acceleration require at least three radiologic examinations (pre-baseline, baseline, and posttreatment) [20]. Kato et al. considered TTF < 2 months as one condition of HPD [7]. In general, these definitions rely on radiographic support. In this case, pre-baseline imaging data were lacking, and the interval between CT scans was more than two months. Therefore, currently recognized indicators for identifying HPD such as TGR/TGK and TTF are unavailable. Thus, we used a more comprehensive definition of HPD, which considered the overall tumor burden (including new lesions or nontarget lesions) and clinical presentation (ECOG score). Despite the differences in methods, all definitions highlighted the importance of quantifying tumor burden kinetics.

The patient initially responded favorably to the combination regimen, although he ultimately developed HPD. We considered that our patient showed clinical benefits from the initial three treatment cycles according to irRC and iRECIST, which proposed that patients with SD belong to the clinical benefit group [21, 22]. This benefit effect may be due to the immunomodulatory effects of chemotherapy. Research indicated that chemotherapy can enhance immunity by increasing the antigenicity and immunogenicity of tumor cells, upregulating major histocompatibility complex molecules and PD-L1 expression, increasing CD8 +cell numbers, and depleting tumor-infiltrating Treg cells [23,24,25]. These findings provide a rationale for the combination of chemotherapy and immunotherapy. Results of clinical trial (NCT03214250, NCT03611556) indicated that chemotherapy combined with PD-1 immunotherapy is a promising approach for pancreatic cancer with multiple metastases. [26, 27]

After continuing two cycles of treatment, our patient showed HPD and died shortly. Retrospectively, he has MDM4 amplification, which has been shown to be associated with HPD. [7] (Figs. 4, 5, Table 1) MDM4 can act alone or with MDM2 to negatively regulate p53 in multiple ways [28,29,30,31,32]. Peng et al. assumed the increase of IFN-γ levels after immunotherapy may trigger JAK/STAT signaling, which upregulates the interferon regulatory factor (IRF)-8 gene. This gene binds to the promoter of MDM4, favoring MDM4 expression [33,34,35]. Furthermore, MDM4 can inhibit members of E2F and SMAD transcription-factor families and induce chromosomal instability [36, 37]. These MDM4 activities may contribute to HPD. Interestingly, higher level of IFN-γ seem to promote cancer [38, 39]. Recently, MDM4 has been identified as a promising target for cancer treatment, and effective inhibitors have been developed, mainly through three pathways: direct inhibition of p53-MDM4 interaction, inhibition of MDM4 expression, and degradation of MDM4 protein [40,41,42]. In a preclinical trial, degrading MDM4 to synergize anti-PD-1 immunotherapy was a potentially viable therapeutic strategy [43]. Overall, the amplification of MDM4 may be the main cause of HPD in this patient.

FISH results show MDM4 is amplified. Interphase FISH analysis indicates that MDM4 DNA is expressed in the nucleus. MDM4 in red. FISH fluorescence in situ hybridization

IHC results indicate MDM4 is positive. The results of immunohistochemistry showed that MDM4 protein is expressed in the nucleus and cytoplasm, mainly in the nucleus. Nuclei in blue and MDM4 in brown

Laboratory data including absolute neutrophil count (ANC) and lactate dehydrogenase (LDH) appear to be predictive factors of HPD. [44,45,46] Among 34 melanoma patients, 11 patients with partial response had a mean reduction of − 27.3%, and 15 patients with progressive disease had a mean increase of + 39% compared to elevated baseline LDH. In our case, the baseline level of ANC and LDH and their changes during treatment appear similar to those of melanoma patients mentioned above (Fig. 2), with increase of 140% and 120%, respectively. ANC was found to induce the release of premature myeloid cells, and more importantly, resistance to ICI treatment is related to the recruitment of myeloid-derived suppressor cells (MDSCs) [47]. Similarly, LDH-associated lactate can also promote the recruitment of MDSCs, inhibiting both innate and adaptive immunity [48]. Notably, the serum CA-199 decreased throughout the course of treatment, even during hyperprogressive periods (supplemental material).

Altogether, we may use a modified definition of HPD to evaluate similar cases in pancreatic cancer. This also reflects that the current definition of HPD is insufficient for clinical application. This report raises some new questions: Firstly, we do not know whether the phenomenon of HPD is due to the combination therapy pattern; Secondly, we do not know what role MDM4 plays in HPD. There is still much research needed in the field of immunotherapy for pancreatic cancer, and research into HPD is still in its infancy. We believe that with the deepening of related research, the above questions will be answered, and the development of immunotherapy for pancreatic cancer will be promoted.