An international multi-institutional validation study of the algorithm for prostate cancer detection and Gleason grading

Pathologic examination of prostate biopsies is time consuming due to the large number of slides per case. In this retrospective study, we validate a deep learning-based classifier for prostate cancer (PCA) detection and Gleason grading (AI tool) in biopsy samples. Five external cohorts of patients with multifocal prostate biopsy were analyzed from high-volume pathology institutes. A total of 5922 H&E sections representing 7473 biopsy cores from 423 patient cases (digitized using three scanners) were assessed concerning tumor detection. Two tumor-bearing datasets (core n = 227 and 159) were graded by an international group of pathologists including expert urologic pathologists (n = 11) to validate the Gleason grading classifier. The sensitivity, specificity, and NPV for the detection of tumor-bearing biopsies was in a range of 0.971-1.000, 0.875-0.976, and 0.988-1.000, respectively, across the different test cohorts. In several biopsy slides tumor tissue was correctly detected by the AI tool that was initially missed by pathologists. Most false positive misclassifications represented lesions suspicious for carcinoma or cancer mimickers. The quadratically weighted kappa levels for Gleason grading agreement for single pathologists was 0.62-0.80 (0.77 for AI tool) and 0.64-0.76 (0.72 for AI tool) for the two grading datasets, respectively. In cases where consensus for grading was reached among pathologists, kappa levels for AI tool were 0.903 and 0.855. The PCA detection classifier showed high accuracy for PCA detection in biopsy cases during external validation, independent of the institute and scanner used. High levels of agreement for Gleason grading were indistinguishable between experienced genitourinary pathologists and the AI tool.

Genetic polymorphisms in interleukin-1β (rs1143634) and interleukin-8 (rs4073) are associated with survival after resection of intrahepatic cholangiocarcinoma

Intrahepatic cholangiocarcinoma (iCCA) is a rare, understudied primary hepatic malignancy with dismal outcomes. Aiming to identify prognostically relevant single-nucleotide polymorphisms, we analyzed 11 genetic variants with a role in tumor-promoting inflammation (VEGF, EGF, EGFR, IL-1B, IL-6, CXCL8 (IL-8), IL-10, CXCR1, HIF1A and PTGS2 (COX-2) genes) and their association with disease-free (DFS) and overall survival (OS) in patients undergoing curative-intent surgery for iCCA. Genomic DNA was isolated from 112 patients (64 female, 48 male) with iCCA. Germline polymorphisms were analyzed with polymerase chain reaction-restriction fragment length polymorphism protocols. The IL-1B +3954 C/C (73/112, hazard ratio (HR) = 1.735, p = 0.012) and the IL-8 -251 T/A or A/A (53/112 and 16/112, HR = 2.001 and 1.1777, p = 0.026) genotypes were associated with shorter OS in univariable and multivariable analysis. The IL-1B +3954 polymorphism was also associated with shorter DFS (HR = 1.983, p = 0.012), but this effect was not sustained in the multivariable model. A genetic risk model of 0, 1 and 2 unfavorable alleles was established and confirmed in multivariable analysis. This study supports the prognostic role of the IL-1B C+3954T and the IL-8 T-251A variant as outcome markers in iCCA patients, identifying patient subgroups at higher risk for dismal clinical outcomes.

[Urothelial carcinoma of the upper urinary tract]

Urothelial carcinoma of the upper urinary tract (upper tract urothelial carcinoma, UTUC) is less common than bladder carcinoma with nearly identical risk factors and has a poorer prognosis. The standard diagnostic procedure is imaging of the upper urinary tract by computed tomography urography. In cases of diagnostic uncertainty, a diagnostic ureterorenoscopy with biopsy sampling can be performed in addition to urine cytology. Treatment depends primarily on the stage and grading of the tumor. Depending on the extent and localization, organ-preserving treatment or radical nephroureterectomy is indicated. Perioperative systemic treatment in high-risk UTUC can be performed in both neoadjuvant and adjuvant settings, although the current data on neoadjuvant chemo- and immunotherapy do not yet allow standard application. For metastatic disease, a multimodal treatment approach consisting of cisplatin-based or carboplatin-based chemotherapy, immunotherapy, and treatment with enfortumab vedotin can be considered. Salvage surgery, radiotherapy and metastasectomy are available for rare individual cases.

[New WHO classification 2022: urinary bladder cancer]

The new World Health Organization (WHO) classification of urogenital tumors is still primarily based on anatomic location, but is also a hierarchical taxonomic classification without separate chapters for tumors of the upper urinary tract and the urethra. It clarifies aspects regarding grading and noninvasive entities. It consolidates the use of the Paris system for urinary cytology as well as various subtypes/special types of neoplasms, and incorporates general concepts of the 5th edition of the WHO blue book. In addition to mesenchymal tumors, well-differentiated neuroendocrine tumors and neuroendocrine carcinomas are addressed in separate chapters. Papillary non-invasive low- and high-grade carcinomas and carcinoma in situ remain, while dysplasia and urothelial proliferation of unknown malignant potential (UPUMP) are no longer treated as separate entities. Former variants of urothelial carcinoma are now called subtypes and aberrant differentiation and special types are more precisely defined.

Artificial intelligence for detection of microsatellite instability in colorectal cancer-a multicentric analysis of a pre-screening tool for clinical application

BACKGROUND

Microsatellite instability (MSI)/mismatch repair deficiency (dMMR) is a key genetic feature which should be tested in every patient with colorectal cancer (CRC) according to medical guidelines. Artificial intelligence (AI) methods can detect MSI/dMMR directly in routine pathology slides, but the test performance has not been systematically investigated with predefined test thresholds.

METHOD

We trained and validated AI-based MSI/dMMR detectors and evaluated predefined performance metrics using nine patient cohorts of 8343 patients across different countries and ethnicities.

RESULTS

Classifiers achieved clinical-grade performance, yielding an area under the receiver operating curve (AUROC) of up to 0.96 without using any manual annotations. Subsequently, we show that the AI system can be applied as a rule-out test: by using cohort-specific thresholds, on average 52.73% of tumors in each surgical cohort [total number of MSI/dMMR = 1020, microsatellite stable (MSS)/ proficient mismatch repair (pMMR) = 7323 patients] could be identified as MSS/pMMR with a fixed sensitivity at 95%. In an additional cohort of N = 1530 (MSI/dMMR = 211, MSS/pMMR = 1319) endoscopy biopsy samples, the system achieved an AUROC of 0.89, and the cohort-specific threshold ruled out 44.12% of tumors with a fixed sensitivity at 95%. As a more robust alternative to cohort-specific thresholds, we showed that with a fixed threshold of 0.25 for all the cohorts, we can rule-out 25.51% in surgical specimens and 6.10% in biopsies.

INTERPRETATION

When applied in a clinical setting, this means that the AI system can rule out MSI/dMMR in a quarter (with global thresholds) or half of all CRC patients (with local fine-tuning), thereby reducing cost and turnaround time for molecular profiling.

Genetic Variant of CXCR1 (rs2234671) Associates with Clinical Outcome in Perihilar Cholangiocarcinoma

BACKGROUND

Perihilar cholangiocarcinoma (pCCA) is a rare primary liver malignancy. Even in patients amenable to surgery, outcomes are often dismal. Here, we aimed to identify prognostic markers for patient outcomes by analyzing functionally relevant single-nucleotide polymorphisms (SNPs) in genes with a role in tumor inflammation and angiogenesis. We analyzed 11 polymorphisms in the inflammation-angiogenesis axis (VEGF, EGF, EGFR, IL-1b, IL-6, CXCL8 (IL-8), IL-10, CXCR1, HIF1A, and COX2 genes) for their prediction of tumor recurrence and survival in pCCA patients undergoing surgery in a curative intent.

METHODS

Samples were obtained from 111 patients with pCCA undergoing liver resection in curative intent. DNA was extracted and analyzed using polymerase chain reaction-restriction fragment length polymorphism protocols and correlated with patients' outcomes.

RESULTS

Out of the assessed variants, only the CXCR1 (also: interleukin-8-receptor alpha - IL-8RA) +860C>G heterozygous polymorphism (rs2234671) was associated with decreased disease-free survival (DFS), cancer-specific survival (CSS), and overall survival (OS) (18/111 (16.2%), median DFS 14 months, log-rank p = 0.007; median CSS 31 months, log-rank p = 0.007; and median OS 6 months, log-rank p = 0.002), compared to the GG genotype (92/111 (82.9%), median DFS 55 months, median CSS 63 months, and median OS 33 months). In the multivariate analysis, +860C>G remained an independent prognostic factor for DFS (adjusted p = 0.008), CSS (adjusted p = 0.001), and OS (adjusted p = 0.001).

CONCLUSION

Genetic variant of CXCR1 +860C>G may serve as a molecular marker for DFS, CSS, and OS in patients undergoing curative-intent surgery for pCCA, indicating that the analysis of SNPs in genes involved in immune-mediated angiogenesis may help to identify patient subgroups at high risk for dismal oncological and overall outcome.

Development and validation of deep learning classifiers to detect Epstein-Barr virus and microsatellite instability status in gastric cancer: a retrospective multicentre cohort study

BACKGROUND

Response to immunotherapy in gastric cancer is associated with microsatellite instability (or mismatch repair deficiency) and Epstein-Barr virus (EBV) positivity. We therefore aimed to develop and validate deep learning-based classifiers to detect microsatellite instability and EBV status from routine histology slides.

METHODS

In this retrospective, multicentre study, we collected tissue samples from ten cohorts of patients with gastric cancer from seven countries (South Korea, Switzerland, Japan, Italy, Germany, the UK and the USA). We trained a deep learning-based classifier to detect microsatellite instability and EBV positivity from digitised, haematoxylin and eosin stained resection slides without annotating tumour containing regions. The performance of the classifier was assessed by within-cohort cross-validation in all ten cohorts and by external validation, for which we split the cohorts into a five-cohort training dataset and a five-cohort test dataset. We measured the area under the receiver operating curve (AUROC) for detection of microsatellite instability and EBV status. Microsatellite instability and EBV status were determined to be detectable if the lower bound of the 95% CI for the AUROC was above 0·5.

FINDINGS

Across the ten cohorts, our analysis included 2823 patients with known microsatellite instability status and 2685 patients with known EBV status. In the within-cohort cross-validation, the deep learning-based classifier could detect microsatellite instability status in nine of ten cohorts, with AUROCs ranging from 0·597 (95% CI 0·522-0·737) to 0·836 (0·795-0·880) and EBV status in five of eight cohorts, with AUROCs ranging from 0·819 (0·752-0·841) to 0·897 (0·513-0·966). Training a classifier on the pooled training dataset and testing it on the five remaining cohorts resulted in high classification performance with AUROCs ranging from 0·723 (95% CI 0·676-0·794) to 0·863 (0·747-0·969) for detection of microsatellite instability and from 0·672 (0·403-0·989) to 0·859 (0·823-0·919) for detection of EBV status.

INTERPRETATION

Classifiers became increasingly robust when trained on pooled cohorts. After prospective validation, this deep learning-based tissue classification system could be used as an inexpensive predictive biomarker for immunotherapy in gastric cancer.

FUNDING

German Cancer Aid and German Federal Ministry of Health.

The Presence of Small Nerve Fibers in the Tumor Microenvironment as Predictive Biomarker of Oncological Outcome Following Partial Hepatectomy for Intrahepatic Cholangiocarcinoma

Simple Summary

Nerve fibers in the microenvironment of malignant tumors have been shown to be an important prognostic factor for long-term survival in various cancer types; however, their role in intrahepatic cholangiocarcinoma remains to be determined. Therefore, the impact of nerve fibers on long-term survival was investigated in a large European cohort of patients with intrahepatic cholangiocarcinoma who were treated by curative-intent surgical resection. By univariate and multivariate statistics, the absence of nerve fibers was determined to be an independent predictor of impaired long-term survival. A group comparison between patients with and without nerve fibers showed a statically significant difference with a cancer-specific 5-year-survival of 47% in patients with nerve fibers compared to 21% in patients without nerve fibers. Thus, the presence of nerve fibers in the microenvironment of intrahepatic cholangiocarcinoma is revealed as a novel and important prognostic biomarker in these patients.

Abstract

The oncological role of the density of nerve fibers (NFs) in the tumor microenvironment (TME) in intrahepatic cholangiocarcinoma (iCCA) remains to be determined. Therefore, data of 95 iCCA patients who underwent hepatectomy between 2010 and 2019 was analyzed regarding NFs and long-term outcome. Extensive group comparisons were carried out and the association of cancer-specific survival (CSS) and recurrence-free survival (RFS) with NFs were assessed using Cox regression models. Patients with iCCA and NFs showed a median CSS of 51 months (5-year-CSS = 47%) compared to 27 months (5-year-CSS = 21%) in patients without NFs (p = 0.043 log rank). Further, NFs (hazard ratio (HR) = 0.39, p = 0.002) and N-category (HR = 2.36, p = 0.010) were identified as independent predictors of CSS. Patients with NFs and without nodal metastases displayed a mean CSS of 89 months (5-year-CSS = 62%), while patients without NFs or with nodal metastases but not both showed a median CCS of 27 months (5-year-CSS = 25%) and patients with both positive lymph nodes and without NFs showed a median CCS of 10 months (5-year-CSS = 0%, p = 0.001 log rank). NFs in the TME are, therefore, a novel and important prognostic biomarker in iCCA patients. NFs alone and in combination with nodal status is suitable to identify iCCA patients at risk of poor oncological outcomes following curative-intent surgery.

Impact of Angiogenesis- and Hypoxia-Associated Polymorphisms on Tumor Recurrence in Patients with Hepatocellular Carcinoma Undergoing Surgical Resection

Simple Summary

Hepatocellular carcinoma remains a leading cause of cancer-related death and the most common primary hepatic malignancy in the Western hemisphere. Previous research found that angiogenesis-related cytokines and elevated levels of interleukin 8 and vascular endothelial growth factor (VEGF) shorten the expected time of survival. Moreover, factors of tumor angiogenesis- and hypoxia-driven signaling pathways are already associated with worse outcome in disease-free survival in several tumor entities. Our study investigates the prognosis of hepatocellular carcinoma patients based on a selection of ten different single-nucleotide polymorphisms from angiogenesis, carcinogenesis, and hypoxia pathways. Our study with 127 patients found supporting evidence that polymorphisms in angiogenesis-associated pathways corelate with disease-free survival and clinical outcome in patients with hepatocellular carcinoma.

Abstract

Tumor angiogenesis plays a pivotal role in hepatocellular carcinoma (HCC) biology. Identifying molecular prognostic markers is critical to further improve treatment selection in these patients. The present study analyzed a subset of 10 germline polymorphisms involved in tumor angiogenesis pathways and their impact on prognosis in HCC patients undergoing partial hepatectomy in a curative intent. Formalin-fixed paraffin-embedded (FFPE) tissues were obtained from 127 HCC patients at a German primary care hospital. Genomic DNA was extracted, and genotyping was carried out using polymerase chain reaction (PCR)–restriction fragment length polymorphism-based protocols. Polymorphisms in interleukin-8 (IL-8) (rs4073; p = 0.047, log-rank test) and vascular endothelial growth factor (VEGF C + 936T) (rs3025039; p = 0.045, log-rank test) were significantly associated with disease-free survival (DFS). After adjusting for covariates in the multivariable model, IL-8 T-251A (rs4073) (adjusted p = 0.010) and a combination of “high-expression” variants of rs4073 and rs3025039 (adjusted p = 0.034) remained significantly associated with DFS. High-expression variants of IL-8 T-251A may serve as an independent molecular marker of prognosis in patients undergoing surgical resection for HCC. Assessment of the patients’ individual genetic risks may help to identify patient subgroups at high risk for recurrence following curative-intent surgery.

Clinical-Grade Detection of Microsatellite Instability in Colorectal Tumors by Deep Learning

BACKGROUND & AIMS

Microsatellite instability (MSI) and mismatch-repair deficiency (dMMR) in colorectal tumors are used to select treatment for patients. Deep learning can detect MSI and dMMR in tumor samples on routine histology slides faster and less expensively than molecular assays. However, clinical application of this technology requires high performance and multisite validation, which have not yet been performed.

METHODS

We collected H&E-stained slides and findings from molecular analyses for MSI and dMMR from 8836 colorectal tumors (of all stages) included in the MSIDETECT consortium study, from Germany, the Netherlands, the United Kingdom, and the United States. Specimens with dMMR were identified by immunohistochemistry analyses of tissue microarrays for loss of MLH1, MSH2, MSH6, and/or PMS2. Specimens with MSI were identified by genetic analyses. We trained a deep-learning detector to identify samples with MSI from these slides; performance was assessed by cross-validation (N = 6406 specimens) and validated in an external cohort (n = 771 specimens). Prespecified endpoints were area under the receiver operating characteristic (AUROC) curve and area under the precision-recall curve (AUPRC).

RESULTS

The deep-learning detector identified specimens with dMMR or MSI with a mean AUROC curve of 0.92 (lower bound, 0.91; upper bound, 0.93) and an AUPRC of 0.63 (range, 0.59-0.65), or 67% specificity and 95% sensitivity, in the cross-validation development cohort. In the validation cohort, the classifier identified samples with dMMR with an AUROC of 0.95 (range, 0.92-0.96) without image preprocessing and an AUROC of 0.96 (range, 0.93-0.98) after color normalization.

CONCLUSIONS

We developed a deep-learning system that detects colorectal cancer specimens with dMMR or MSI using H&E-stained slides; it detected tissues with dMMR with an AUROC of 0.96 in a large, international validation cohort. This system might be used for high-throughput, low-cost evaluation of colorectal tissue specimens.

Variant morphology and random chromosomal integration of BK polyomavirus in posttransplant urothelial carcinomas

BK polyomavirus (BKPyV) causes major complications in solid organ transplant recipients but little is known about its role in the development of urothelial carcinoma (UC) during immunosuppression. Immunohistochemistry (IHC) screening for polyomavirus large T antigen (LTag) was performed in 94 micropapillary UC (MPUC), 480 unselected UC, 199 muscle invasive UC (including 83 UC with variant differentiation), 76 cases of plasmocytoid, nested and large nested UC and 15 posttransplant UC. LTag expressing UC were reevaluated regarding their histomorphological features and characterized by IHC for p53 and HER2, chromogenic in situ hybridization for HER2 and SNaPshot analysis of the TERT promoter and HRAS. Real-time PCR and next generation sequencing (NGS) were performed to search for BKPyV-DNA and for variants in the tumor and viral genomes. We detected five LTag expressing UC which were diagnosed between 2 and 18 years after kidney (n = 4) or heart (n = 1) transplantation. 89 MPUC without history of organ transplantation and overall 755 UC (including cases with variant histology) were LTag negative. Of the five LTag expressing UC, three were MPUC, one showed extensive divergent differentiation with Mullerian type clear cell carcinoma, and one displayed focal villoglandular differentiation. All five tumors had aberrant nuclear p53 expression, 2/5 were HER2-amplified, and 3/5 had TERT promoter mutations. Within the 50 most common cancer related genes altered in UC we detected very few alterations and no TP53 mutations. BKPyV-DNA was present in 5/5 UC, chromosomal integration of the BKPyV genome was detectable in 4/5 UC. Two UC with BKPyV integration showed small deletions in the BKPyV noncoding control region (NCCR). The only UC without detectable BKPyV integration had a high viral load of human herpesvirus 6 (HHV-6). Our results suggest that LTag expression of integrated BKPyV genomes and resulting p53 inactivation lead to aggressive high-grade UC with unusual, often micropapillary morphology.

[Status of the availability and use of next generation sequencing (NGS) in bladder cancer-a questionnaire within the uropathology working group]

BACKGROUND

Technical advancement and availability of high-throughput analysis has advanced molecular subtyping of most cancers. Thus, new possibilities for precision oncology have emerged.

AIM

Therefore, we aimed to collect data regarding availability and use of next generation sequencing (NGS) for urothelial cancer within the uropathology working group of the German Society of Pathology.

METHODS

We collected data by questionnaires and additionally asked for sequencing results of bladder cancers in the participating institutions.

RESULTS

A total of 13 university-affiliated institutes of pathology took part in the survey. All university institutes offer NGS-based molecular panel diagnostics and provide panels covering between 15 and 170 genes. Altogether, only 20 bladder cancers were sequenced in routine diagnostics and for 10 cancers potential targeted treatment options were available.

DISCUSSION

So far, despite availability of NGS diagnostics at university institutes of pathology, only few bladder cancer samples have been sequenced. Based on current data from the molecular subtyping of bladder cancers, we recommend a step-by-step protocol with basic immunohistochemistry analysis and subsequent subtype-dependent analyses, e.g., alterations of the fibroblast growth factor receptors (FGFR) or comprehensive gene panel analyses.

[Molecular tumor board in uro-oncology-from a pathologist's view]

Besides evidence- and guideline-based tumor therapy, personalized targeted therapies within study settings or individual experimental settings in advanced cancers without further therapeutic options are emerging. A comprehensive molecular analysis of the tumor in a molecular pathology laboratory is important for all targeted therapy approaches. However, the interpretation of the molecular results is crucial and potential therapeutic conclusions can only be drawn by considering the clinical situation and within a setting of oncological experience. Therefore, the molecular results and their potential impact have to be discussed at a molecular tumor board, an interdisciplinary expert team consisting of clinicians, oncologists, (molecular) pathologists, systems physicians, study teams and where required geneticists. If the molecular tumor board decides a targeted therapeutic approach is appropriate, patients should be enrolled in studies or registries with controlled settings and documentation in order to evaluate the therapeutic concepts. Furthermore, molecular-based individual experimental therapies are possible within extreme clinical situations.

[Specific types of bladder cancer]

Bladder cancer shows rare variants and special subtypes with diverse prognostic importance and therefore may necessitate different therapeutic approaches. For pathologists it is important to histologically diagnose and specify such variants. Nested variants of urothelial carcinoma with inconspicuous, well-formed tumor cell nests present with an aggressive course. The plasmacytoid variant, which morphologically resembles plasma cells is associated with a shorter survival time and a high frequency of peritoneal metastasis. Micropapillary urothelial carcinoma with small papillary tumor cell islands within artificial tissue retraction spaces and frequent lymphovascular invasion also has a poor prognosis. Other important rare differential variants listed in the World Health Organization (WHO) classification are microcystic, lymphoepithelioma-like, sarcomatoid, giant cell and undifferentiated urothelial carcinomas. Additionally, there are three special types of bladder cancer: squamous cell carcinoma, adenocarcinoma and small cell neuroendocrine carcinoma of the bladder. These tumors are characterized by pure squamous cell or glandular differentiation and are sometimes less responsive to adjuvant (chemo)therapy. Small cell carcinoma of the bladder mimics the neuroendocrine features of its pulmonary counterpart, shows an aggressive course but is sensitive to (neo-)adjuvant chemotherapy. The morphology and histology of the most important variants and special types are discussed in this review.

[Preneoplastic lesions and precursors of urothelial cancer]

As even a mere thickening of the urothelium can harbor genetic changes identical to that of low grade papillary urothelial tumors, it is not always possible to clearly recognize a precursor lesion of urothelial carcinoma by routine histological diagnostics. Complementary immunohistochemical and molecular diagnostic methods assist the recognition of these entities. These methods especially help to identify clinically important genetically unstable cells as the hallmark of carcinoma in situ (CIS). Little is known about the clinical significance of the morphological subtypes of CIS, which range from large cell to micropapillary variants. For a better understanding of special types of bladder cancer (e.g. adenocarcinoma and squamous cell carcinoma), it seems to be important to define the phenotype and the molecular pattern of non-urothelial lesions, such as intestinal metaplasia and squamous metaplasia, better and more precisely.

[Non-invasive and invasive urothelial tumours: special challenges in uropathological diagnostics]

The current 2004 WHO classification of bladder tumors categorizes non-invasive and invasive urothelial neoplasms into prognostically relevant groups according to the histopathological cell morphology and underlying genetic changes. Although many parts of the classification have not been changed dramatically, even small changes have caused uncertainty and scepticism among urologists and pathologists in recent years. The following review article is structured into various challenges for urologists and pathologists and provides an overview of rare but clinically relevant subgroups and diagnostics, interpretation of diagnoses and pathological findings with respect to consequences for the daily clinical routine (extended diagnosis, therapy and prognosis).

Expression of acyl-CoA synthetase 5 in human epidermis

The human epidermis is characterized by a constant renewal of keratinocytes embedded in a matrix enriched with lipids. Numerous proteins involved in lipid metabolism are found in human epidermis, especially in keratinocytes. Long-chain acyl-CoA derivatives, which are catalyzed by human ACSL5, are important metabolites in several biochemical pathways, including ceramide de novo synthesis. The aim of the present study was to investigate expression of acyl-CoA synthetase isoform 5 (ACSL5) in human epidermis by an in situ, as well as a molecular approach. We show that ACSL5 mRNA and protein are found in human epidermis, as well as in non-differentiated and differentiated HaCaT cells. Keratinocytes of stratum spinosum are the main source for ACSL5 expression in both meshed facial or abdominal skin and ridged skin of upper or lower extremities including TUNEL-positive cells in upper cellular layers. Single keratinocytes of chronic solar-exposed meshed facial epidermis occasionally display a stronger ACSL5 immunostaining. In conclusion, our study indicates that epidermal ACSL5 expression might be involved in differentiation and the stress response of keratinocytes.