How open is the therapeutic horizon for pancreatic cancer patients?

Holger Kalthoff

Institute for Experimental Cancer Research, Christian-Albrechts-University, Kiel, Germany

In the nineties of the last century, it was typical for an article about pancreatic cancer to start with an introductory remark,that this malignancy is a very desperate one and that surgery is the only curative option. But it was also obvious that surgery was not really curative and most importantly, not feasible in the majority of cases. In the last two decades, we observed a dramatic increase in research on pancreatic cancer (e.g. more than 65 0 0 0 articles mentioning pancreatic cancer in PubMed only in the last decade), and clearly, there is scientific progress. This accounts for(i) genetically defined subtypes (for an overview on the various studies [1] ), for (ii) precursor lesions like pancreatic intraepithelial neoplasias (PanINs), intraductal papillary mucinous neoplasm(IPMN) and acinar-ductal metaplasia (ADM) [2] , for (iii) comprehensive characterization of the tumor microenvironment [3] , for(iv) animal models - both genetically engineered mouse models(GEMMs) and xenotransplant systems [4] , for (v) complexinvitrosystems like spheroids, organoids, complex co-cultures, chip-based 3D-models [5] , and for (vi) the impact of the microbiome [6] as well as for (vii) the metabolic implications [7] and finally, for (viii)the biomarker field which has shown progress beyond the tumor marker carbohydrate antigen 19-9 (CA19-9) [8] .

But it will be hard to find any article from today focusing on pancreatic ductal adenocarcinoma (PDAC - the most typical form to be discussed in this editorial), which does not still use introductory statements like “… is a devastating disease…”, “… a malignancy where incidence equals mortality…”, “… a hard to treat cancer with very poor prognosis…”, and even more: “… a killer which will rank second in cancer-related deaths within the end of this decade in USA …” [9] .

So, why is there no progress - seemingly? What real achievements have been reached in the clinic? In this issue, Bai et al. [10] summarized a remarkable number of clinical studies using different approaches and strategies. But before this very comprehensive piece of work will become adequately appreciated, it might be worth to remember the beginning of a new palliative and later adjuvant treatment area for PDAC, i.e. the introduction of gemcitabine. The time before the introduction of gemcitabine was characterized by either therapeutic nihilism or by extremely toxic, mostly arbitrarily assembled cocktails of chemotherapeutic drugs. With gemcitabine, a compound was launched, which helped to establish a new paradigm in PDAC care that improved quality of life [11] . Indeed, the first survival data for the initial gemcitabine approval study were very modest [1.2 months longer compared with that of 5-fluorouracil (5-FU)] even in the PDAC field, and it was merely sufficient for an adequate admission. However, the typical symptoms of advanced PDAC like pain, vomiting, loss of appetite, cachexia were at least moderately attenuated. In this regard,and also combating nihilism in the clinic, gemcitabine operated like a gamechanger. But instead of focusing on mechanistic studies to elucidate the overall poor growth-inhibitory activity of gemcitabine, the next period was characterized again mostly by drug cocktails [12] . For those “l(fā)ucky” patients for whom surgery was regarded as possible, initially no adjuvant treatment was given. It took rather long to reach a common sense, and more importantly,a common or at least comparative technique for tumor resections among surgeons [13] . Importantly, high (or at least medium-high)volume centers with certification of national cancer societies also took time to become established. But now they are in place to perform clinical trials for adjuvant treatment studies with an appropriate basis. Yet, still the average number of patients recruited for most of the studies was not resulting in a real powerful conclusion.

Conventional chemotherapy is still mostly based on gemcitabine “plus”, and the “plus” is mostly represented by albuminbound paclitaxel. One alternative option, which may have originated from the “arbitrary cocktail period” has been more recently launched. The combination chemotherapy regimen consisting of oxaliplatin, irinotecan, fluorouracil, and leucovorin (FOLFIRINOX regimen) [14] has been shown to substantially prolong the survival - yet only of selected patients with an appropriate Karnofsky performance status. As pointed out by Bai et al., we still miss a convincing comparison between both alternative treatment regimens also including their side effects, which are rather frequently hard to tolerate by the patients.

Considering the dramatic increase of knowledge about PDAC genetics, far exceeding the long known four mutations (KRAS,SMAD4,TP53,CDKN2A) elaborated about a decade ago, one might wonder why we still have no better stratification system after having sequenced far more than 10 0 0 PDAC specimens. One point of caution is the fact that our molecular-genetic knowledge of human PDAC is largely based on resectable cases, and this is clearly the minority of patients. One other knowledge-limiting issue here might have been the eagerness of the various consortia to establish their own classification system. It took time to reach at least a certain degree of consensus [ 1 , 15 ]. But regrettably, a specific therapeutic matching of the major PDAC sub-types with different approaches is largely missing. Exceptions like DNA-damage-repair-deficits,BRAFandBRCAmutations are rather seldom, but if given, they provide a better chance for therapeutic success. In line with these positive exceptions are the findings of the “Know-Your-Tumor” program recently achieved in USA [16] . In those cases with a targetable genetic alteration (about 30%), patients who received an appropriate drug lived about one year longer than those who did not. Besides these still modest achievements (in relation to the efforts), limiting factors might be given by the open questions regarding sequencing and data analyses costs covered by the health insurance systems and also by the proper pipeline organization established within Comprehensive Cancer Centers (CCCs). Still, many patients are treated outside such centers and even in the CCCs the typical rapid progression of PDACs does not allow any delay in genetic testing with an appropriate bioinformatic evaluation. There is still much space for improvements in the standard health care system to catch up with the scientific progress having defined molecular options.

In contrast to sophisticated personalization strategies, one might ask why the key genetic alteration in theKRASoncogene,known since decades, did not form a targetable Achilles’ heel for the roughly 90% of patients being affected. Bai et al. [10] described the many attempts, which have passed the translational hurdle between pre-clinicalinvivomodels and clinical trials - still with modest outcome, yet ongoing hope. Along the same line, yet with a lower incidence of about 50%,CDKN2Amight be a good candidate since it is more easily druggable. Clinical trials are ongoing.

In the early phase of monoclonal antibodies - regarded as magic bullets for cancer therapy - EGF-receptor targeting was also tested for PDAC patients [17] ending with a little bit of hope largely based on the assumption that Fc-receptor-mediated antibody-dependent cellular cytotoxicity (ADCC) mechanisms (previously shown to exist in pre-clinical mouse models) may play a major role. Later, cetuximab and other EGFR inhibitors turned out to be not or very moderately successful for PDAC treatment. However, it might be worth to reconsider ADCC as a powerful weapon since manipulation of the Fc portion [18] as well as antibody chimerization for dual targeting of members of the EGFR-family may result in a higher tumor specificity and therapeutic activity.

It is not that long ago that tumor immunologists were regarded as researchers curing mice - if not totally neglected by clinical oncologists. This has changed dramatically with the discovery of immune checkpoints as central regulatory elements for effective Tcell responses. An overwhelming success in treating, for example,end-stage melanoma and many other cancer patients, has stimulated the Nobel Prize committee to award two pioneers in 2018.Yet, as always, there are many frustrating, ineffective clinical results as well - depending either on the individual tumor within a non-responder group, or on the tumor type. PDAC is a striking example for the latter situation and is thus regarded as immunologically “cold” and characterized by a highly immunosuppressive milieu. More than 20 years ago, our group described one possible mechanism (reduced zeta-chain expression in tumor-infiltratinglymphocytes) [19] . Moreover, we seldomly observed any direct Tcell localization next to PDAC cellsinsitu[20] . Meanwhile, many more mechanisms have been found to hamper clinical success of immunotherapies in PDAC patients and are comprehensively as well as critically discussed by Bai et al. [10] . Besides checkpoint inhibitors, also oncolytic viruses, various vaccination strategies,adoptive T-cell therapy includingCART-cell technology have been tested - a breakthrough is still missing. However, some mostly preclinical data on combination therapies - including chemotherapeutics like gemcitabine - may allow some hope.

One indirect attempt to combat the immune suppression is to directly target the tumor microenvironment, in particular the desmoplasia of PDAC [ 10 , 21 ]. Opposing data in the outcome of anti-cancer associated fibroblast (anti-CAF) or anti-stellate cell therapies in various model systems have shown the complexity and revealed the challenge: is the desmoplasia representing a “wound that does not heal” or a “wound that at least tries to heal”?

Overall, the review by Bai et al. [10] summarizes about 20 treatment regimens/modes of action, more than 40 targets, at least 50 drugs/cellular compounds documented in about 90 publications in all these different fields of PDAC research. Phase III studies are rare and it is likely that many more studies have been undertaken -but were not published because of “negative” outcome. The key question still remains: why is there so little progress? From a clinical point of view, many studies look under-powered and in the context of still rather low (although rising) incidence, recruiting enough patients in a reasonable time period remains a challenge -unless oncologists decide to even more team up and strictly try to include all their PDAC patients in clinical trials. Hemato- and pediatric oncologists have shown both, the possibility to do so and the progress-stimulating results with major benefits for the affected patients.

The experimental cancer researchers should avoid any easygoing, fast growing etc.invitroandinvivomodels, but instead squeeze the doctoral students - and themselves - with model systems fully representing the clinical challenges. For example, subcutaneous PDAC tumor models must not be used anymore [22] .In vitrostandard monolayer cultures should be banned unless epithelial barrier functions are addressed. Complex/composite 3D models,organoids, with or without scaffolds, matrices, nano-chip carriers will allow much more long-lasting (translatable) answers - despite the technician-unfriendly high effort s being necessary.

Another possible point of hinderance for therapeutic progress was the long-lasting neglection of the possible impact of Caucasian and Asian genetic background in the past. With the dramatic increase in scientific output in Asia, particularly in China, during the last decade, this hurdle will not further block scientific and clinical progress.

New (therapeutic) kids on the block? A possible new level of treatment options is provided, not directly based on the genetic level considering mutations etc., but more on the cell-biological level. We and others have studied the impact of sub-cellular distributions of a possible target, in this case a death-receptor. Initially regarded as a direct option to kill tumor cells selectively by ligand-induced apoptosis, TRAIL-receptor 2 (TRAIL-R2), which was frequently found to be overexpressed in many tumors including PDAC, turned out to function as an oncogene when it reaches the nucleus [23]. This topology has been long time neglected and was later found to be TRAIL-mediated, particularly inKRAS-mutated cells [24] . Furthermore, TRAIL-R2 turned out to act as a negative regulator of p53 [25] . This “quenching” of the guardian of the genome is another example of an apoptosis-regulator acting oncogenic upon nuclear translocation. On top of this, a chromatin association of TRAIL-R2 has been very recently demonstrated, as well pointing to further gene-regulatory functions [26] .

Together, these findings not only give a plausible interpretation of the long-established, yet surprising data describing an overexpression of an apoptosis machinery (death receptor plus its cognate ligand) in malignant cells like PDAC, but also open a new avenue for therapeutic options like trying to shuttle the nuclear form of TRAIL-R2 back to the surface and thereby rendering the cells more vulnerable towards programmed cell death.

Besides cell-biology, physiology is likely to impact therapeutic outcome as well. The particular physiology of this human gland [ 27, 28] creates an environment for transformed malignant cells, which was mostly neglected [29] . Moreover, some of the ion transport- and channel proteins of the ductal and acinar cells, the“transportome”, have been shown to provide an attractive therapeutic target in PDAC cells [30] , partially also due to pre-existing specific drugs and pharmacological knowledge from other indications [1] . Within the “transportome”, mechanosensitive ion channels are likely to play a particular physical role, especially in desmoplastic tissues like PDAC [31] .

In conclusion, the remarkable increase of knowledge and improved understanding of PDAC genetics still wait for a clinically useful translation. Centers of excellence in oncology, such as CCCs,and highly specialized centers for PDAC may serve as pacemaker with highly interactive disciplines of surgery, pathology, and experimental as well as clinical oncology. Seemingly odd infrastructural problems like biomaterials and bioinformatics must be solved and rolled out in a broad clinical area. Pre-clinical research must adjust theinvitroandinvivoexperimental systems at least as complicated and tough as the clinical reality. Areas of research like physiology and physics which have been out-competed by molecular genetics in the past will yield new options for improved PDAC therapeutic outcome.

Acknowledgments

The critical reading and support of the final editing by Dr.Christian R?der, from Kiel, Germany, is highly appreciated. This editorial comment is dedicated to all the many doctoral students,technicians, clinical colleagues, researchers, and visiting guest scientists whom I had the pleasure and privilege to meet during my active career.

CRediT authorship contribution statement

Holger Kalthoff: Conceptualization, Writing - original draft,Writing - review and editing.

Funding

None.

Ethical approval

Not needed.

Competing interest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.