Immunohistochemistry in the workup of prostate biopsies: Frequency, variation and appropriateness of use among pathologists practicing at an academic center

Prospective Clinical Implementation of Paige Prostate Detect Artificial Intelligence Assistance in the Detection of Prostate Cancer in Prostate Biopsies: CONFIDENT P Trial Implementation of Artificial Intelligence Assistance in Prostate Cancer Detection

PURPOSE

Pathologists diagnose prostate cancer (PCa) on hematoxylin and eosin (HE)-stained sections of prostate needle biopsies (PBx). Some laboratories use costly immunohistochemistry (IHC) for all cases to optimize workflow, often exceeding reimbursement for the full specimen. Despite the rise in digital pathology and artificial intelligence (AI) algorithms, clinical implementation studies are scarce. This prospective clinical trial evaluated whether an AI-assisted workflow for detecting PCa in PBx reduces IHC use while maintaining diagnostic safety standards.

METHODS

Patients suspected of PCa were allocated biweekly to either a control or intervention arm. In the control arm, pathologists assessed whole-slide images (WSI) of PBx using HE and IHC stainings. In the intervention arm, pathologists used the Paige Prostate Detect AI algorithm on HE slides, requesting IHC only as needed. IHC was requested for all morphologically negative slides in the AI arm. The main outcome was the relative risk (RR) of IHC use per detected PCa case at both patient and WSI levels.

RESULTS

Overall, 143 of 237 (60.3%) slides of 64 of 82 patients contained PCa (78.0%). AI assistance significantly reduced the risk of IHC use per detected PCa case at both the patient level (RR, 0.55; 95% CI, 0.39 to 0.72) and slide level (RR, 0.41; 95% CI, 0.29 to 0.52). Cost reductions on IHC were €1,700 for the trial, at €50 per IHC stain. AI-assisted pathologists reported higher confidence in their diagnoses (80% v 56% confident or high confidence). The median assessment time per HE slide showed no significant difference between the AI-assisted and control arms (139 seconds v 112 seconds; P = .2).

CONCLUSION

This study demonstrates that AI assistance for PCa detection in PBx significantly reduces IHC costs while maintaining diagnostic safety standards, supporting the business case for AI implementation in PCa detection.

Diagnostic immunohistochemistry use in Belgian laboratories

OBJECTIVES

In Belgium, the use of IHC testing has grown in the last decade. However, there is a lack of information on the specific indications for which it is reimbursed. The aim of the study is to offer an overview on the use of diagnostic inmunohistochemistry (IHC) testing and its recent trends.

METHODS

Our analysis is limited to reimbursed use, which in Belgium is restricted to a maximum of 4 different IHC stains per sampling session for diagnostic IHC. Consulted sources included data from the compulsory health insurance, and data extracted from a sample of pathology reports gathered from Belgian laboratories for the year 2019.

RESULTS

Over the last 10 years, the use of IHC in Belgium grew from 729 030 stains in 2012 to 1,194,331 in 2019, an increase of 63.8 % while the increase in the number of histological or cytological examinations was 13.3 %. The main stains used in 2019 were H. pylori, Ki-67 and broad spectrum CK, which were used in multiple body sites, reflecting the difficulties to identify specific indications. The gastro-intestinal tract is the body site with the highest number of IHC stains (38.2 % of all stains performed), and the most frequently used stain in gastro intestinal biopsies were H. pylori (43.1 %), and CD3 (6.8 %).

CONCLUSION

This study offers an overview of the most frequent indications for which diagnostic IHC staining is used in Belgium, and reflects the evolving nature of this field, highlighting the importance to increase clarity and improve data collection.

Pros and cons of artificial intelligence implementation in diagnostic pathology

The rapid introduction of digital pathology has greatly facilitated development of artificial intelligence (AI) models in pathology that have shown great promise in assisting morphological diagnostics and quantitation of therapeutic targets. We are now at a tipping point where companies have started to bring algorithms to the market, and questions arise whether the pathology community is ready to implement AI in routine workflow. However, concerns also arise about the use of AI in pathology. This article reviews the pros and cons of introducing AI in diagnostic pathology.

Immunohistochemistry and immunofluorescence utilization audit by subspecialty in an academic setting: A step toward stewardship

There are immunohistochemistry (IHC) and immunofluorescence (IF) panels described in the literature and established by personal and institutional experiences that are in common use by pathologists in their daily practice. Stewardship is a difficult discussion because IHC utilization is influenced by many factors including the pathologist's experience, background, practice setting, personal bias, and medicolegal culture. We developed the methodology to audit the IHC/IF utilization in our academic subspecialty practice. We aim to share this methodology and to provide our data that can be used for consideration by other subspecialized academic practices. This analysis included a total of 63,157 specimens that were accessioned during 2022, representing 38,612 cases. The likelihood of ordering IHC/IF ranged from 1 % (in genitourinary pathology) to 59 % (in renal pathology). The average percentage of specimens with IHC/IF was 21 % for the entire practice. In cases where IHC/IF was ordered, the number of stained slides averaged 4.9 per specimen for the entire practice. The number of IHC/IF slides per specimen ranged from 1.9 (in gastrointestinal pathology) to 12.2 (in renal pathology). The highest number of antibodies ordered for a single specimen by subspecialty ranged from 11 (in cardiac pathology) to 63 (in dermatopathology). Renal pathology was the only subspecialty that had an average number of IHC/IF slides that was statistically significantly different from all other subspecialties. We described the various patterns of utilization by subspecialty and rationalized their subtle differences. We also analyzed the types of cases that exceeded the reimbursement limits set by the Centers for Medicare and Medicaid Services (CMS).

Artificial intelligence for advance requesting of immunohistochemistry in diagnostically uncertain prostate biopsies

The use of immunohistochemistry in the reporting of prostate biopsies is an important adjunct when the diagnosis is not definite on haematoxylin and eosin (H&E) morphology alone. The process is however inherently inefficient with delays while waiting for pathologist review to make the request and duplicated effort reviewing a case more than once. In this study, we aimed to capture the workflow implications of immunohistochemistry requests and demonstrate a novel artificial intelligence tool to identify cases in which immunohistochemistry (IHC) is required and generate an automated request. We conducted audits of the workflow for prostate biopsies in order to understand the potential implications of automated immunohistochemistry requesting and collected prospective cases to train a deep neural network algorithm to detect tissue regions that presented ambiguous morphology on whole slide images. These ambiguous foci were selected on the basis of the pathologist requesting immunohistochemistry to aid diagnosis. A gradient boosted trees classifier was then used to make a slide-level prediction based on the outputs of the neural network prediction. The algorithm was trained on annotations of 219 immunohistochemistry-requested and 80 control images, and tested by threefold cross-validation. Validation was conducted on a separate validation dataset of 222 images. Non IHC-requested cases were diagnosed in 17.9 min on average, while IHC-requested cases took 33.4 min over multiple reporting sessions. We estimated 11 min could be saved on average per case by automated IHC requesting, by removing duplication of effort. The tool attained 99% accuracy and 0.99 Area Under the Curve (AUC) on the test data. In the validation, the average agreement with pathologists was 0.81, with a mean AUC of 0.80. We demonstrate the proof-of-principle that an AI tool making automated immunohistochemistry requests could create a significantly leaner workflow and result in pathologist time savings.

The utilization and utility of immunostains in body fluid cytology

BACKGROUND

Body fluid cytology (BFC) is an important tool in the diagnosis and staging of malignancy and is aided by the judicious use of immunohistochemistry (IHC). The aim of this study was to determine the usage rates of IHC stains in BFC, their type and indications, and their diagnostic impact. We also attempted to estimate the optimal rate of IHC use in BFC by comparing the entire laboratory's and each individual cytopathologist's IHC use rates with their respective indeterminate and malignant diagnosis rates.

METHODS

We conducted a retrospective study of IHC stain use in BFC during a 5.5-year interval (2013-2018) and determined the laboratory's and each individual cytopathologist's IHC usage patterns according to the final diagnosis, site, and indications for their use.

RESULTS

A total of 477 out of 4144 (11.5%) BFC cases had 2128 individual immunostains performed, with an average of 4.5 immunostains per case. Individual cytopathologists used IHC stains on 6.7% to 22% of their BFC cases. Pathologists with higher rates of IHC stain use than the laboratory's mean were less experienced and had higher rates of indeterminate but not of malignant diagnoses. The most common indication for the use of IHC stains was differentiating mesothelial from malignant cells. MOC31, calretinin, Ber-EP4, CD68, and D2-40 were the most commonly used of the 67 different IHC stains used in BFC.

CONCLUSIONS

The laboratory's mean may represent the optimal IHC use rate, as higher IHC use rates did not lead to more diagnostic certainty or higher pickup rates of malignant cells.

Differences in immunohistochemistry utilization by general and breast subspecialty pathologists at a large academic institution

OBJECTIVES

Immunohistochemistry (IHC) can be a useful adjunct in diagnostic breast pathology, but best practices have not been clearly established. We sought to compare the patterns of p63 utilization between general pathologists (GP) and subspecialized breast pathologists (BP), analyze diagnostic discrepancy rates, and identify types of lesions requiring immunohistochemistry.

METHODS

The pathology database was searched over 6-year period to identify breast needle core biopsy cases utilizing p63 and subsequent excision results.

RESULTS

P63 was ordered more frequently by BP (2.3% of cores) compared to GP (1.1% of cores, p = 0.0005). The most frequent utilization of p63 by GP for benign lesions (44.0%) followed by invasive carcinomas (36.0%) while BP most frequently ordered p63 on invasive carcinomas (49.5%) and DCIS (26.6%).

CONCLUSIONS

While IHC use may be thought to be most helpful to those with less experience or knowledge in breast pathology, these results suggest that utilization is increased with subspecialty training.

Three-dimensional architecture of common benign and precancerous prostate epithelial lesions

Aims

Many glandular lesions can mimic prostate cancer microscopically, including atrophic glands, adenosis and prostatic intraepithelial neoplasia. While the characteristic histopathological and immunohistochemical features of these lesions have been well established, little is known about their three‐dimensional architecture. Our objective was to evaluate the three‐dimensional organisation of common prostate epithelial lesions.

Methods and results

500 μm‐thick punches (n = 42) were taken from radical prostatectomy specimens, and stained with antibodies targeting keratin 8–18 and keratin 5 for identification of luminal and basal cells, respectively. Tissue samples were optically cleared in benzyl alcohol:benzyl benzoate and imaged using a confocal laser scanning microscope. The three‐dimensional architecture of peripheral and transition zone glands was acinar, composed of interconnecting and blind‐ending saccular tubules. In simple atrophy, partial atrophy and post‐atrophic hyperplasia, the acinar structure was attenuated with branching blind‐ending tubules from parental tubular structures. Three‐dimensional imaging revealed a novel variant of prostate atrophy characterised by large Golgi‐like atrophic spaces parallel to the prostate surface, which were represented by thin, elongated tubular structures on haematoxylin and eosin (H&E) slides. Conversely, adenosis lacked acinar organisation, so that it closely mimicked low‐grade prostate cancer. High‐grade prostatic intraepithelial neoplasia displayed prominent papillary intraluminal protrusions but retained an acinar organisation, whereas intraductal carcinoma predominantly consisted of cribriform proliferations with either spheroid, ellipsoid or complex interconnecting lumens.

Conclusions

While various prostate epithelial lesions might mimic malignancy on H&E slides, their three‐dimensional architecture is acinar and clearly different from the tubular structure of prostate cancer, with adenosis as an exception.

Circulating HER-2 mRNA in the peripheral blood as a potential diagnostic and prognostic biomarker in females with breast cancer

Breast cancer is a prevalent malignant cancer worldwide, and a lack of defined biomarkers for early prognostication contributes to its high associated mortality rate, especially in human epidermal growth factor receptor 2 (HER-2)-positive breast cancer. In the present study, HER-2 mRNA levels in patients were detected prior to surgery and during neoadjuvant chemotherapy to explore its potential diagnostic and prognostic value. Blood samples were collected from 70 patients with breast cancer, including 50 HER-2-negative and 20 HER-2-positive patients, prior to and following surgery (postoperative, n=13; neoadjuvant chemotherapy, n=5); the control group included 35 samples from healthy individuals. The relative mRNA level of HER-2 in blood was determined by one-step reverse transcription-quantitative polymerase chain reaction. HER-2 expression curves of measurements taken during neoadjuvant chemotherapy were compared with the tumor size. A significant difference in the blood HER-2 mRNA level was observed between healthy women and patients with breast cancer (P<0.0001). A cutoff value of 1.512 was established for the circulating HER-2 level in healthy subjects based on the upper 95% confidence interval value of samples from the control group. The level of HER-2 mRNA in blood was associated with the HER-2 status, Ki-67 expression, and lymphovascular invasion in primary tumor tissue samples; however, there was no association with the lymph node status, tumor stage, tumor grade, tumor size, patient age, estrogen or progesterone receptor status of the primary tumor. HER-2 mRNA levels were associated with the response rate, as determined by primary tumor size, in patients who received neoadjuvant chemotherapy. In conclusion, baseline and early changes in peripheral blood HER-2 mRNA indicated that HER-2 mRNA may be a potential diagnostic biomarker for breast cancer and a prognostic marker for predicting the efficacy of neoadjuvant therapy.

Impact of the Pathologist on Prostate Biopsy Diagnosis and Immunohistochemical Stain Usage Within a Single Institution

OBJECTIVES

To determine whether pathologists in a tertiary care institution vary in diagnosis and immunohistochemical stain usage in prostate biopsy specimens.

METHODS

Men who underwent prostate needle biopsies between 2008 and 2013 were included.

RESULTS

In total, 1,777 prostate biopsy specimens diagnosed by nine pathologists showed variation in diagnostic reporting (atypical small acinar proliferation, 2.0%-8.0%; high-grade prostatic intraepithelial neoplasia, 2.0%-8.5%; nonneoplastic, 30.2%-48.3%; adenocarcinoma, 46.2%-55.3%; P < .001). Variation in Gleason scoring was observed (P < .001), with the 4 + 3 = 7 category having the greatest variability (6.9%-30.3%). A blinded review from the most outlying pathologist in this category revealed 45% grading discrepancies. The mean number of immunostains performed per case (0.3-1.2) differed between pathologists (P < .001), and one pathologist used immunostains at twice the rate of the remaining cohort.

CONCLUSIONS

Case pathologist significantly affects prostate biopsy diagnosis and immunohistochemical workup. We recommend evaluation for outlying practice patterns to provide consistent and efficient patient care.

Classification of pulmonary neuroendocrine tumors: new insights

Neuroendocrine tumors of the lung (Lu-NETs) embrace a heterogeneous family of neoplasms classified into four histological variants, namely typical carcinoid (TC), atypical carcinoid (AC), large cell neuroendocrine carcinoma (LCNEC) and small cell lung carcinoma (SCLC). Defining criteria on resection specimens include mitotic count in 2 mm2 and the presence or absence of necrosis, alongside a constellation of cytological and histological traits including cell size and shape, nuclear features and overall architecture. Clinically, TC are low-grade malignant tumors, AC intermediate-grade malignant tumors and SCLC/LCNEC high-grade malignant full-blown carcinomas with no significant differences in survival between them. Homologous tumors arise in the thymus that occasionally have some difficulties in differentiating from the lung counterparts when presented with large unresectable or metastatic lesions. Immunohistochemistry (IHC) helps refine NE diagnosis at various anatomical sites, particularly on small-sized tissue material, in which only TC and small cell carcinoma categories can be recognized easily on hematoxylin & eosin stain, while AC and LCNEC can only be suggested on such material. The Ki-67 labeling index effectively separates carcinoids from small cell carcinoma and may prove useful for the clinical management of a metastatic disease to help the therapeutic decision-making process. Although carcinoids and high-grade neuroendocrine carcinomas in the lung and elsewhere make up separate tumor categories on molecular grounds, emerging data supports the concept of secondary high-grade NETs arising in the preexisting carcinoids, whose clinical and biological relevance will have to be placed into the proper context for the optimal management of these patients. In this review, we will discuss the selected, recent literature with a focus on current issues regarding Lu-NET nosology, i.e., classification, derivation and tumor evolution.