Prognostic impact of Ki-67 in canine splenic hemangiosarcoma: A preliminary study.

Canine splenic hemangiosarcoma has a high metastatic rate and short survival time. Currently, the main prognostic parameters are tumor stage and therapy, while data on histologic parameters, such as grade and Ki-67 expression, are scarce. The aims of this study were to compare two methods of assessment of Ki-67, verify their prognostic impact, and define a threshold value based on survival. Thirty-one cases of histologically diagnosed canine splenic hemangiosarcoma, which were treated with splenectomy and had full staging and follow-up information, were collected. Three were stage I, 17 stage II, and 11 stage III. The mean mitotic count (MC) was 23.9 (standard deviation [SD]: 22.1) and the median was 15 (range, 1-93). Immunohistochemistry for Ki-67 was performed, the Ki-67 labeling index (Ki-67LI) was assessed as a percentage of positive neoplastic nuclei per ≥500 cell, and the Ki-67 count (KI-67C) was defined as the average number of positive nuclei using a 1 cm2 optical grid performed in 5, 40× fields. The mean Ki-67LI and Ki-67C were 56.4% (SD: 38.7) and 27.2 (SD: 12.9) and medians were 51% (range, 8.2-55.2) and 26 (range, 5.5-148), respectively. Using a cut-off of 56% and 9, respectively, Kaplan-Meier survival curves showed an association of overall survival with Ki-67LI and MC. In addition to clinical stage, Ki-67LI maintained its prognostic value on multivariate analysis, supporting the role of Ki-67LI as an independent prognostic parameter. Based on these results, we propose a diagnostically applicable cut-off value of 56% for Ki-67LI as a prognostic parameter for canine splenic hemangiosarcoma.

Subcutaneous panniculitis-like T-cell lymphoma: Morphological, immunophenotypical and clonality assessment in six cats.

BACKGROUND: Primary cutaneous lymphoma represents 0.2%-3% of all feline lymphomas, with nonepitheliotropic lymphomas being the most common. In humans and dogs, subcutaneous panniculitis-like T-cell lymphoma (SPTCL) is a primary nonepitheliotropic lymphoma with a T-cell phenotype developing in the subcutis and often mimicking inflammation. OBJECTIVE: The aim of this report is to describe pathological, phenotypical and clonal features of SPTCL in cats. ANIMALS: Six cats with SPTCL were included in this study. MATERIALS AND METHODS: Skin biopsies were formalin-fixed, routinely processed and stained. Histological and immunohistochemical investigation for anti-CD18, CD204, CD79a, CD20, CD3, FeLVp27and FeLVgp70 and clonality assessment were performed. RESULTS: Four male and two female domestic shorthair cats, mean age 11.2 years, developed SPTCL in the abdominal (three), inguinal (two) and thoracic (one) regions. Variably pleomorphic neoplastic lymphoid cells were present in the panniculus in percentages, expanding the septa (six of six) and extending into fat lobules in one of six cats. Tumours were associated with elevated numbers of neutrophils (five of six), lesser macrophages (six of six) and variable necrosis (six of six). Neoplastic cells expressed CD3+ (six of six), with clonal T-cell receptor rearrangement detected in five of six cats. CONCLUSIONS AND CLINICAL RELEVANCE: This is the first description of SPTCL in cats. Lesions can be confused with panniculitis, leading to delay in diagnosis and therapy. Awareness of this neoplastic disease is relevant to avoid misdiagnoses and to gain greater knowledge about the disease in cats.

Morphologic, phenotypic, and genotypic similarities between primary tumors and corresponding 3D cell cultures grown in a repeatable system-preliminary results.

BACKGROUND: Three-dimensional (3D) cell cultures are the new frontier for reproducing the tumor micro-environment in vitro. The aims of the study were (1) to establish primary 3D cell cultures from canine spontaneous neoplasms and (2) to demonstrate the morphological, phenotypic and genotypic similarities between the primary canine neoplasms and the corresponding 3D cultures, through the expression of tumor differentiation markers. RESULTS: Seven primary tumors were collected, including 4 carcinomas and 3 soft tissue sarcomas. 3D cell cultures reproduced the morphological features of the primary tumors and showed an overlapping immunophenotype of the primary epithelial tumors. Immunohistochemistry demonstrated the growth of stromal cells and macrophages admixed with the neoplastic epithelial component, reproducing the tumor microenvironment. Mesenchymal 3D cultures reproduced the immunophenotype of the primary tumor completely in 2 out of 3 examined cases while a discordant expression was documented for a single marker in one case. No single nucleotide variants or small indel were detected in TP53 or MDM2 genes, both in primary tumors and in 3D cell cultures specimens. In one sample, MDM2 amplicons were preferentially increased in number compared to TP53 ones, indicating amplification of MDM2, detectable both in the primary tumor and in the corresponding cell culture specimen. CONCLUSION: Here we demonstrate a good cell morphology, phenotype and genetic profile overlap between primary tumors and the corresponding 3D cultures grown in a repeatable system.

PDL1 immunohistochemistry in canine neoplasms: Validation of commercial antibodies, standardization of evaluation, and scoring systems.

Immuno-oncology research has brought to light the paradoxical role of immune cells in the induction and elimination of cancer. Programmed cell death protein 1 (PD1), expressed by tumor-infiltrating lymphocytes, and programmed cell death ligand 1 (PDL1), expressed by tumor cells, are immune checkpoint proteins that regulate the antitumor adaptive immune response. This study aimed to validate commercially available PDL1 antibodies in canine tissue and then, applying standardized methods and scoring systems used in human pathology, evaluate PDL1 immunopositivity in different types of canine tumors. To demonstrate cross-reactivity, a monoclonal antibody (22C3) and polyclonal antibody (cod. A1645) were tested by western blot. Cross-reactivity in canine tissue cell extracts was observed for both antibodies; however, the polyclonal antibody (cod. A1645) demonstrated higher signal specificity. Canine tumor histotypes were selected based on the human counterparts known to express PDL1. Immunohistochemistry was performed on 168 tumors with the polyclonal anti-PDL1 antibody. Only membranous labeling was considered positive. PDL1 labeling was detected both in neoplastic and infiltrating immune cells. The following tumors were immunopositive: melanomas (17 of 17; 100%), renal cell carcinomas (4 of 17; 24%), squamous cell carcinomas (3 of 17; 18%), lymphomas (2 of 14; 14%), urothelial carcinomas (2 of 18; 11%), pulmonary carcinomas (2 of 20; 10%), and mammary carcinomas (1 of 31; 3%). Gastric (0 of 10; 0%) and intestinal carcinomas (0 of 24; 0%) were negative. The findings of this study suggest that PDL1 is expressed in some canine tumors, with high prevalence in melanomas.

Standardization of canine meningioma grading: Validation of new guidelines for reproducible histopathologic criteria.

Canine meningiomas are currently graded using the human grading system. Recently published guidelines have adapted the human grading system for use in dogs. The goal of this study was to validate the new guidelines for canine meningiomas. To evaluate the inter-observer agreement, 5 veterinary surgical pathologists graded 158 canine meningiomas following the human grading system alone or with the new guidelines. The inter-observer agreement for histologic grade and each of the grading criteria (mitotic grade, invasion, spontaneous necrosis, macronucleoli, small cells, hypercellularity, pattern loss and anaplasia) was evaluated using the Fleiss kappa index. The diagnostic accuracy (sensitivity and specificity) was assessed by comparing the diagnoses obtained with the 2 grading systems with a consensus grade (considered the reference classification). The consensus histologic grade was obtained by agreement between 4 experienced veterinary neuropathologists following the guidelines. Compared with the human grading alone, the canine-specific guidelines increased the inter-observer agreement for: histologic grade (κ = 0.52); invasion (κ = 0.67); necrosis (κ = 0.62); small cells (κ = 0.36); pattern loss (κ = 0.49) and anaplasia (κ = 0.55). Mitotic grade agreement remained substantial (κ = 0.63). The guidelines improved the sensitivity in identifying grade 1 (95.6%) and the specificity in identifying grade 2 (96.2%) meningiomas. In conclusion, the new grading guidelines for canine meningiomas are associated with an overall improvement in the inter-observer agreement and higher diagnostic accuracy in diagnosing grade 1 and grade 2 meningiomas.

A case of spermatic cord B-cell lymphoma relapsing to the brain in a dog.

A 13-year-old, intact male mixed-breed dog was referred to our clinic for lethargy and asthenia following an episode of gastroenteritis. As an incidental finding during abdominal ultrasound, a mass on the right spermatic cord was seen. Cytology of the mass revealed a monomorphic population of large, round cells with a lymphoid appearance. A bilateral orchiectomy was conducted, and histopathology revealed the presence of a B-cell lymphoma in the right spermatic cord. Based on clinical staging, which showed no involvement of other sites, no additional treatment was administered. Recheck evaluations were scheduled for every 3 mo thereafter. Five months after surgery, the dog developed left central vestibular syndrome with a paradoxical right-sided head tilt. An MRI of the brain showed multifocal lesions and, due to a rapidly worsening clinical condition, the dog was humanely euthanized. The histopathology of the brain lesions was consistent with B-cell lymphoma. Key clinical message: This is the first report of a primary spermatic cord lymphoma relapsing to the brain in a dog. Although rare, spermatic cord tumors should be included among the differential diagnoses for masses arising from the spermatic cord. If lymphoma is diagnosed, location to other sites, especially to the central nervous system, should be considered.

Un cas de lymphome à cellules B du cordon spermatique récidivant au cerveau chez un chien. Un chien de race mixte mâle intact de 13 ans a été référé à notre clinique pour léthargie et asthénie à la suite d'un épisode de gastro-entérite. Comme découverte fortuite lors d'une échographie abdominale, une masse sur le cordon spermatique droit a été observée. La cytologie de la masse a révélé une population monomorphe de grosses cellules rondes d'aspect lymphoïde. Une orchidectomie bilatérale a été réalisée et l'histopathologie a révélé la présence d'un lymphome à cellules B dans le cordon spermatique droit. Sur la base du stade clinique, qui n'a montré aucune implication d'autres sites, aucun traitement supplémentaire n'a été administré. Des évaluations de contrôle étaient programmées tous les 3 mois par la suite. Cinq mois après la chirurgie, le chien a développé un syndrome vestibulaire central gauche avec une inclinaison paradoxale de la tête du côté droit. Une IRM du cerveau a montré des lésions multifocales et, en raison d'une détérioration rapide de l'état clinique, le chien a été euthanasié sans cruauté. L'histopathologie des lésions cérébrales correspondait à un lymphome à cellules B.Message clinique clé :Il s'agit du premier rapport d'un lymphome primaire du cordon spermatique récidivant au cerveau chez un chien. Bien que rares, les tumeurs du cordon spermatique doivent être incluses dans les diagnostics différentiels des masses provenant du cordon spermatique. Si un lymphome est diagnostiqué, la localisation vers d'autres sites, en particulier vers le système nerveux central, doit être envisagée.(Traduit par Dr Serge Messier).

Primary intranasal perivascular wall tumors in 2 cats.

Perivascular wall tumors (PWTs) are common well-known canine mesenchymal tumors. The term PWT has not yet been applied to cats; only 2 cases of feline soft tissue hemangiopericytomas (HEPs) are available. In human medicine, sinonasal HEP-like tumor/glomangiopericytoma (SHPCL/GP) and intranasal solitary fibrous tumor (SFT) are well-known mesenchymal tumors with staghorn vasculature and low malignant potential; however, these entities have not been described in small animals. We describe here the pathologic and immunohistochemical features of 2 cases of feline intranasal mesenchymal tumors consistent with PWTs and resembling human SHPCL/GP (case 1), and human intranasal SFT (case 2). Both cats developed intranasal, unilateral, polypoid, expansile neoplasms with a mostly patternless growth of spindle cells, minimal stroma, and prominent staghorn vessels. The stroma was PAS negative, which excludes a glomus tumor. Immunohistochemistry identified diffuse vimentin and PDGFRß expression. Case 1 was α-SMA positive (as is human SHPCL/GP); case 2 was negative (as is human intranasal SFT). Both tumors were incompletely excised, leading to recurrence in case 1. Case 2 was lost to follow up. To our knowledge, intranasal PWTs have not been reported previously in cats. The frequency of the lesions is not known, but awareness of these entities may assist in their recognition and better characterization in the future.

Expression of Cell-Cycle Regulatory Proteins pRb, Cyclin D1, and p53 Is Not Associated with Recurrence Rates of Equine Sarcoids.

Sarcoids are among the most common tumors diagnosed in equids; their association with bovine papillomaviruses (BPV) infection has been widely reported, but the mechanism of carcinogenesis has not been fully elucidated. To verify whether BPV infection causes dysregulation of the pRb-Cyclin D1-p16CDKN2A-p53 pathway as reported for human papillomavirus (HPV), the study employed immunohistochemistry to test 55 equine sarcoid biopsies for the expression of pRb, Cyclin D1, and p53 cell cycle regulatory proteins and to evaluate the proliferative rate through Ki67. High Cyclin D1 and pRb expression were observed in 51% and 80% of cases, respectively, while low expression was observed in 49% and 20% of cases, respectively. Significantly higher Ki67 proliferation indexes were observed in fibroblastic, nodular, and mixed sarcoids compared to the occult and verrucous. High proliferation was significantly associated with high Cyclin D1 expression. In contrast with previous studies, p53 positivity was not observed in the cases examined in this study. Moreover, follow-up analysis revealed that fibroblastic, mixed sarcoids were associated with significantly higher local recurrence rates while the verrucous subtype was associated with higher rates of new sarcoid development at distant sites.

Identification of Histopathological Criteria for the Diagnosis of Canine Cutaneous Progressive Angiomatosis.

The term angiomatosis is used to denote a group of well-known to poorly characterized proliferative vascular entities. In animals, cutaneous progressive angiomatosis (CPA) is a disorder with variable prognosis related to the extension and depth of infiltration of the surrounding tissues by vessels. CPA may share some microscopical features with other vascular proliferations such as low-grade well-differentiated capillaritic hemangiosarcoma (HS), making the diagnosis not always straightforward, especially in small biopsies. The aim of this study is to retrospectively assess the most common diagnostic microscopical features of CPA in dogs. In this work, 11 histopathological criteria were analyzed on 31 CPA and 11 primary cutaneous HS in dogs. Features significantly associated with CPA included: lobular growth, interposition of connective tissue and adnexa between the vascular proliferation, presence of nerve fibers, and a mixed vascular proliferative component. Absence of plump/prominent endothelial cells, lack of atypia, and lack of mitoses were also significant factors differentiating CPA from HS. Additional distinctive findings in CPA, although with no statistical association to CPA diagnosis, were vascular shunting, absence of necrosis, and endothelial cell piling up. In conclusion, the combined use of different microscopical clues allowed for the distinction of CPA from HS and was considered useful for the diagnosis of CPA.

Detection of Leishmania spp. in Chronic Dermatitis: Retrospective Study in Exposed Horse Populations.

Leishmania infantum is a protozoan causing human zoonotic visceral leishmaniasis (ZVL) and visceral-cutaneous canine leishmaniosis (CanL) in the Mediterranean Basin. L. infantum is able to infect a large number of wild and domestic species, including cats, dogs, and horses. Since the 1990s, clinical cases of equine leishmaniasis (EL), typically characterized by cutaneous forms, have been increasingly diagnosed worldwide. The aim of the present study was to evaluate the presence of clinical forms of EL in CanL-endemic areas in Italy, where exposure of equine populations was ascertained from recent serological surveys. For this purpose, formalin-fixed and paraffin-embedded skin biopsies of 47 horses presenting chronic dermatitis compatible with EL were retrospectively selected for the study and subjected to conventional and q-PCR. A singular positivity for L. infantum was found; BLAST analysis of sequence amplicons revealed a 99-100% homology with L. infantum sequences. The histological examination revealed a nodular lymphoplasmacytic and histiocytic infiltrate; immunohistochemistry showed rare macrophages containing numerous positive amastigotes. The present retrospective study reports, for the first time, a case of a cutaneous lesion by L. infantum occurring in an Italian horse. Pathological and healthy skin samples should be investigated on a larger scale to provide information on the potential clinical impact of EL in the practice, and to define the role of horses in epidemiological ZVL and CanL scenarios.

Nerve Fiber Immunohistochemical Panel Discriminates between Nerve Sheath and Perivascular Wall Tumors.

Benign and malignant nerve sheath tumors (NST) pose a major challenge in routine diagnostic anatomic pathology because of shared histomorphological features with other soft-tissue tumors (STT). As a result, NST are often diagnosed as STT, a broad category that encompasses various entities including perivascular wall tumors (PWT) and that represents approximately 15% of all skin tumors in dogs. Immunohistochemistry (IHC) can assist the identification of histologic subtypes of STT. This IHC pilot study applies various markers largely expressed by peripheral nerves to twelve benign and six malignant NST and determines the intratumoral protein expression of laminin, periaxin-1, Sox-10 and S-100 in the NST subtypes. Furthermore, this study assesses the usefulness of peripheral nerve markers applied to diagnostic work cases and demonstrates the relevance of laminin expression patterns, periaxin-1 and Sox-10 in assisting the differentiation of NST from other STT, in particular from PWT.

Canine smooth muscle tumors: A clinicopathological study.

Canine smooth muscle tumors (SMTs) commonly develop in the alimentary and female genital tracts and less frequently in soft tissue. The definition of histological criteria of malignancy is less detailed for SMTs in dogs than in humans. This study evaluated the clinicopathologic features of canine SMTs and compared the veterinary and human medical criteria of malignancy. A total of 105 canine SMTs were evaluated histologically and classified according to both veterinary and human criteria. The Ki67 labeling index was assessed in all SMTs. Estrogen receptor (ER) and progesterone receptor (PR) expression was evaluated for soft tissue SMTs. Follow-up data were available in 25 cases. SMTs were diagnosed in the female genital tract (42%), alimentary tract (22%), and soft tissue (20%). Soft tissue SMTs frequently arose in the perigenital area, pelvic cavity, and retroperitoneum. A subset of soft tissue SMTs expressed ER and/or PR, resembling the gynecologic type of soft tissue SMT in humans. SMTs were less frequently malignant when assessed with human criteria than with veterinary criteria, better reflecting their benign behavior, especially in the genital tract where human criteria tolerate a higher mitotic count for leiomyoma. Decreased differentiation was correlated with increased proliferation, necrosis, and reduced desmin expression. Mitotic count, Ki67 labeling index, and necrosis were correlated with metastases and tumor-related death. Further prognostic studies are warranted to confirm the better performance of the human criteria when assessing SMT malignancy, especially genital cases, to confirm their usefulness in ER/PR-expressing soft tissue SMTs, and to better define the most useful prognostic parameters for canine SMTs.

Airway Remodeling in Feline Lungs.

Airway remodeling encompass structural changes that occur as the result of chronic injury and lead to persistently altered airway structure and function. Although this process is known in several human respiratory conditions such as asthma and chronic obstructive pulmonary disease (COPD), airway remodeling is poorly characterized in the feline counterpart. In this study, we describe the spontaneous pulmonary changes in 3 cats paralleling the airway remodeling reported in humans. We observed airway smooth muscle cells (ASMCs) hyperplasia (peribronchial and interstitial), airway subepithelial and interstitial fibrosis, and vascular remodeling by increased number of vessels in the bronchial submucosa. The hyperplastic ASMCs co-expressed α-SMA, vimentin and desmin suggesting that vimentin, which is not normally expressed by ASMCs, may play a role in airway thickening, and remodeling. ASMCs had strong cytoplasmic expression of TGFß-1, which is known to contribute to tissue remodeling in asthma and in various bronchial and interstitial lung diseases, suggesting its involvement in the pathogenesis of ASMCs hyperplasia. Our findings provide histologic evidence of airway remodeling in cats. Further studies on larger caseloads are needed to support our conclusions on the value of this feline condition as an animal model for nonspecific airway remodeling in humans.

The first case of Tyzzer's disease in a young foal in Italy: a case report.

Seizures, coma and death rapidly appeared after admission in a one ­month­old foal with a history of lethargy, fever and anorexia. Severe icterus and necrotizing hepatitis were observed at necropsy. Clinical signs, laboratory and postmortem findings were compatible with a suspect of clostridial hepatitis. Tyzzer's disease was confirmed by the presence of organisms morphologically consistent with Clostridium piliforme in the hepatocytes at the margins of multiple areas of hepatic necrosis. To the authors' knowledge, this is the first reported case of clostridial hepatitis caused by Clostridium piliforme in a horse in Italy.

International Guidelines for Veterinary Tumor Pathology: A Call to Action.

Standardization of tumor assessment lays the foundation for validation of grading systems, permits reproducibility of oncologic studies among investigators, and increases confidence in the significance of study results. Currently, there is minimal methodological standardization for assessing tumors in veterinary medicine, with few attempts to validate published protocols and grading schemes. The current article attempts to address these shortcomings by providing standard guidelines for tumor assessment parameters and protocols for evaluating specific tumor types. More detailed information is available in the Supplemental Files, the intention of which is 2-fold: publication as part of this commentary, but more importantly, these will be available as "living documents" on a website (www.vetcancerprotocols.org), which will be updated as new information is presented in the peer-reviewed literature. Our hope is that veterinary pathologists will agree that this initiative is needed, and will contribute to and utilize this information for routine diagnostic work and oncologic studies. Journal editors and reviewers can utilize checklists to ensure publications include sufficient detail and standardized methods of tumor assessment. To maintain the relevance of the guidelines and protocols, it is critical that the information is periodically updated and revised as new studies are published and validated with the intent of providing a repository of this information. Our hope is that this initiative (a continuation of efforts published in this journal in 2011) will facilitate collaboration and reproducibility between pathologists and institutions, increase case numbers, and strengthen clinical research findings, thus ensuring continued progress in veterinary oncologic pathology and improving patient care.

Canine Placenta Histological Findings and Microvascular Density: The Histological Basis of a Negative Neonatal Outcome?

Placenta is essential for the development of the fetus, and its impaired function can lead to a negative outcome (i.e., neonatal mortality). In dogs, investigations on placenta histology and neonatal outcome in healthy bitches are lacking, and a contribution is provided in this study to emphasize the use of placenta histology in practice. Fifty-one placentas from 11 litters were collected during cesarean section, classified according to the litter size (large (L) or small (S)) and the outcome, this latter as healthy (Group 1) or dead within 7 days (Group 2). The placenta/puppy weight ratio (PPR) was calculated, and specimens were formalin-fixed and paraffin-wax embedded, and on the resulting histological slides, capillary density (CD) was quantified. Among necrosis, calcification, and intravascular leucocytes, only the presence of multifocal-confluent necrosis (significantly more frequent in Group 2) was associated with a higher risk of death within 7 days (odds ratio = 30.7). Mixed logistic regression ruled out the effect on death both of a bitch and cesarean type (programmed vs. emergency). PPR and CD values were associated with litter size; large litters had lower PPR (p < 0.01) and higher CD (p < 0.05) than small litters. The relationship between PPR and CD was negative and significant (p < 0.01). Necrosis was a frequent finding in canine placentas, but only when multifocal-confluent was it associated with a poor outcome. The litter size influenced PPR (lower in L) and CD (higher in L), and this is likely due to the plasticity of placenta adaptation.

Fibroblast-associated protein-α expression and BPV nucleic acid distribution in equine sarcoids.

Sarcoids are the most common cutaneous tumor of equids and are caused by bovine papillomavirus (BPV). Different clinical subtypes of sarcoids are well characterized clinically but not histologically, and it is not known whether viral activity influences the clinical or histological appearance of the tumors. The aim of this study was to verify whether the development of different clinical types of sarcoids or the presence of certain histological features were associated with BPV distribution within the tumor. The presence of BPV was assessed by polymerase chain reaction (PCR) and visualized in histological sections by chromogenic in situ hybridization (CISH) in 74 equine sarcoids. Furthermore, to better characterize the molecular features of neoplastic cells, immunohistochemistry for S100, smooth muscle actin-α (αSMA), and fibroblast-associated protein-α (FAPα) was performed. The presence of BPV was confirmed in all tissues examined by either or both PCR and CISH (72/74, 97% each). Of 70/74 CISH-positive cases, signal distribution appeared as either diffuse (61/70, 87%) or subepithelial (9/70, 13%); the latter was more frequently observed in the verrucous subtype. However, no statistically significant association was found between clinical subtypes and specific histological features or hybridization pattern. Moreover, CISH signal for BPV was not detected in the epidermis overlying sarcoids nor in the tissue surrounding the neoplasms. By immunohistochemistry, αSMA confirmed the myofibroblastic differentiation of neoplastic cells in 28/74 (38%) sarcoids. Using tissue microarrays, FAPα labelling was observed in neoplastic fibroblasts of all sarcoids, suggesting this marker as a potential candidate for the immunohistochemical diagnosis of sarcoids.

Kaplan-Meier Curves, Cox Model, and P-Values Are Not Enough for the Prognostic Evaluation of Tumor Markers: Statistical Suggestions for a More Comprehensive Approach.

The assessment of prognostic markers is key to the improvement of therapeutic strategies for cancer patients. Some promising markers may fail to be applied in clinical practice, or some useless markers may be applied, because of misleading results ensuing from inadequate planning of the study and/or from an oversimplified statistical analysis. This commentary illustrates and discusses the main issues involved in planning an effective clinical study and the subsequent statistical analysis for the prognostic evaluation of a cancer marker. Another aim is to extend the most applied statistical models (ie, those using Kaplan-Meier and Cox) to enable the choice of the best-suited methods for study endpoints. Specifically, for tumor-centered endpoints like tumor recurrence, the issue of competing risks is highlighted. For markers measured on a continuous numerical scale, a loss of relevant prognostic information may occur by setting arbitrary cutoffs; thus, the methods to analyze the original scale are explained. Furthermore, because the P-value is not a sufficient criterion to assess the usefulness of a marker in clinical practice, measures for evaluating the ability of the marker to discriminate between "good" and "bad" prognoses are illustrated. Several tumor markers are considered both in human and veterinary medicine. Given the similarity between markers for human breast cancer and canine mammary cancer, an application of the statistical methods discussed within the article to a public dataset from human breast cancer patients is shown.