Clinical/research abstracts accepted for presentation ...

845_859_Sorrento abstracts.qxp_FAB 07/08/2018 09:48 Page 845

Journal of Feline Medicine and Surgery (2018) 20, 845–859

CONGRESS abstracts

Clinical/research abstracts accepted for presentation at ISFM Congress 2018

ISFM Congress poster session A total of 28 clinical/ research abstracts were accepted for presentation at the poster session held during the 2018 ISFM Congress in Sorrento, Italy, 27 June to 1 July.

OUTBREAK OF TUBERCULOSIS CAUSED BY MYCOBACTERIUM BOVIS IN A CATTERY OF ABYSSINIAN CATS IN ITALY

INFLUENCE OF DIFFERENT ANESTHETIC PROTOCOLS ON ELECTROPHYSIOLOGICAL TESTING: PRELIMINARY STUDIES IN CATS

Petra Černá1, Conor J O’Halloran2, Olga Sjatkovskaj3, Danièlle A Gunn-Moore2 1 University of Veterinary and Pharmaceutical Sciences Brno, Czech Republic 2 Royal (Dick) School of Veterinary Studies and The Roslin Institute, The University of Edinburgh, Easter Bush Campus, Roslin, Scotland 3 Haabersti Loomakliinik, Tallinn, Estonia Email: [email protected]

Ana C Mortari, Ana C Ricarte, Weslley S Barbosa, Aline Araki, Ricardo M Almeida Faculty of Agronomy and Veterinary Medicine, University of Brazil, Brazil Email: [email protected]

Mycobacterium bovis is a re-emerging zoonosis. It was diagnosed in five Abyssinian cats in a breeding cattery in Italy. The infection entered the cattery with an imported kitten (cat A); it had a suspected bite wound on its leg that had been treated at a veterinary clinic in Kiev, Ukraine, which is likely where it became infected with M bovis. When the kitten arrived in Italy there were four cats in the cattery; an adult female (cat B), this cat’s two kittens (cats C and D), and a kitten imported from Russia (cat E). These were all healthy, and had no outdoor access. All five cats developed tuberculous interstitial pneumonia; in cat A this occurred 6 weeks after importation, and the others were diagnosed 4–6 weeks later. Three cats were euthanased with deteriorating pneumonia while two cats remained clinically well on antibiotic therapy (marbofloxacin, doxycycline and azithromycin). The latter cases were euthanased after 5 weeks, as required by Italian law, once M bovis infection was suspected. Changes consistent with tuberculosis on gross post-mortem examination included mesenteric and mesenchymal lymphadenopathy, splenomegaly and hepatomegaly, and disseminated focal white lesions on the cut surface of the spleen, liver and lungs. Visible acid-fast bacteria (cats A, B and C) were confirmed as Mycobacterium tuberculosis complex by PCR (cats A, B, C, D and E), speciated to M bovis (cats A, B and D), spoligotype SB0950 (cats A and D).

doi: 10.1177/1098612X18791875 © The Author(s) 2018

Electrophysiological testing is always conducted in cats under anesthesia; therefore, the effects of the anesthetic drugs should be considered. The aim of this study was to evaluate the influence of different anesthetic protocols on electromyography studies. Motor nerve conduction, F-wave study, repetitive stimulation and resting electromyography were carried out in the hindlimbs of nine clinically healthy cats (1–5 years of age), five females and four males, weighing from 2.5–5 kg. Animals were randomly allocated to two different anesthetic protocols: the PROP group was premedicated with acepromazine (0.02 mg/kg IM), and anesthesia was induced with propofol (4.0 mg/kg IV) and maintained with the same drug by IV continuous infusion (0.35 mg/kg/min). The ISO group was premedicated with fentanil (5.0 μg/kg IV), induced with propofol (4.0 mg/kg IV) and maintained with isoflurane (1.0 MAC [minimum alveolar concentration]). Electrophysiology tests were obtained from the sciatic-tibial nerves using surface electrodes. Resting electromyography activity was performed with needle electrodes in semitendinosus, vastus lateralis and cranial tibial muscles. Data were analyzed with unpaired t test (P 350 mg/dl. The UPC value measured at T0 ranged between 9.76 and 11 samples scored ⩽0.2, 10 samples ranged between 0.2 and 0.4, and 17 samples were >0.4. A significant difference was found between urine protein values at T0 and at 12, 24 and 48 h (both at RT and 4°C) and between T0 and values measured at 7 and 90 days. A significant difference was not found between urine creatinine values at T0 and concentrations measured at the subsequent evaluations on the stored samples. However, the UPC values calculated for samples stored at 4°C for 48 h and at –20°C (both for 7 and 90 days) were significantly different from those found at T0. Moreover, misclassification of five urine samples that had UPC ratios close to the threshold values between consecutive IRIS proteinuria classes was found at different storage times and temperature. Additionally, storage at –20°C for 90 days yielded up to a 50% decrease in UPC values in four highly proteinuric samples (UPC range 1.11–8.0). In conclusion, preanalytic variability due to storage of feline urine samples at RT, 4°C or –20°C can significantly change UPC values from a diagnostic and prognostic standpoint. This source of error should be considered when interpreting UPC data obtained from samples shipped to external laboratories or from stored samples analysed in retrospective studies.

2018

A 7-year-old male neutered domestic shorthair cat was referred to our centre for evaluation of a mass on the right maxillary region first noted 2 months earlier. On physical examination a firm, non-painful subcutaneous nodule of 20 mm diameter was detected at the level of the upper portion of the right lip extending to the right ventral eyelid. Diagnostic work-up included a total body CT scan followed by fine-needle aspirates of the mass. CT findings showed a 16 mm mass arising from the soft tissues in the right maxillary region with mild to moderate local invasiveness, without lytic or proliferative changes of the bone structures. No changes suggestive of local or distant metastases were observed. Cytological findings were compatible with a mesenchymal neoplasia and an incisional biopsy was performed. Histopathology confirmed a fibrosarcoma. The extension of the tumour and its location did not allow a curative-intent surgery with acceptable surgical margins; therefore, neoadjuvant doxorubicin was started to try to achieve cytoreduction. Complete blood counts and renal parameters were checked prior to each treatment. Doxorubicin was administered at a dose of 25 mg/m2 as a slow intravenous (IV) infusion, following premedication with IV maropitant at 1 mg/kg. Treatments were administered every 2 weeks up to a total of three. Chemotherapy was well tolerated overall; a sporadic, self-limiting grade I (Veterinary Cooperative Oncology Group) gastrointestinal toxicity (nausea) was reported. Following a 39% reduction in volume, the mass was surgically excised using an advancement flap. Histopathology confirmed a soft tissue sarcoma with small mature lymphocyte infiltrates mainly observed at the periphery of the tumour, suggesting a traumatic aetiology. Surgical margins were incomplete on histopathology, so adjuvant doxorubicin was started. Three treatments were administered in total at 3-weekly intervals. The patient was clinically re-evaluated 3 months following completion of chemotherapy; no evidence of local recurrence was detected and thoracic radiographs did not show any changes compatible with distant metastases. This case report suggests that neoadjuvant doxorubicin can be considered to try to achieve cytoreduction for feline soft tissue sarcomas, as similarly reported for feline injection-site sarcoma, with minimal side effects. The role of adjuvant treatment is unclear and needs further investigation.

JFMS CLINICAL PRACTICE

847


A B S T R A C T S / ISFM Congress 2018

MEAN CELL VOLUME, MEAN CELL HAEMOGLOBIN CONCENTRATION AND RED CELL DISTRIBUTION WIDTH LIMITATIONS IN ASSESSING BONE MARROW RESPONSE IN ANAEMIC CATS Aline Barry1, Julia Chase2, Peter Hendrickson2, Jeremy Hammond2, Dennis B DeNicola2 1 IDEXX Laboratories, Hoofddorp, Netherlands 2 IDEXX Laboratories, Westbrook, ME, USA Email: [email protected] The absolute reticulocyte count (RETIC) is a well established gold standard haematological parameter used to assess the bone marrow response (BMR) by quantifying immature erythrocytes circulating in the peripheral blood. Only a few in-house haematology analysers offer the possibility of direct measurement of BMR. Mean cell volume (MCV), mean cell haemoglobin concentration (MCHC) and red cell distribution width (RDW) are red blood cell (RBC) population statistics describing cell size, cell haemoglobin concentration and overall RBC size variability, respectively. All have been proposed as indirect indices of BMR and suggested as a replacement for RETIC. The objective of this study was to evaluate if the use of indirect parameters is sufficient to provide accurate assessment of BMR in anaemic cats. One million randomly identified complete blood count (CBC) results from domestic cats performed in 2017 were used in this analysis. Whole blood samples were run on the ProCyte Dx Hematology Analyzer (PDx; IDEXX Laboratories) from veterinary practices worldwide. Anaemia was defined as haematocrit below the lower limit of the PDx reference interval of 30.3%. BMR was categorized as regenerative when RETIC was above 50 K/µl. Analyses were performed using JMP and Microsoft Excel. Of the 1,000,000 CBCs, 25.48% (254,787) met the criterion for anaemia. Of these, 88.04% (224,326) were non-regenerative and 11.96% (30,461) were regenerative. In marked regeneration, the MCV and RDW should increase and MCHC decrease. The percentages of the alterations in the indirect parameters are showed in the table.

Comparison of red blood cell indices in blood samples from cats with non-regenerative and regenerative anaemia Type of anaemia

MCV >53.1 fl

MCHC 27%

Non-regenerative (⩽50 K/µl)

4.86% (10,912/224,326)

1.26% (2828/224,326)

16.86% (37,819/224,326)

Regenerative (>50 K/µl)

19.70% (6000/30,461)*

13.50% (4112/30,461)*

44.97% (13,697/30,461)*

*P 14 mcg% (one >19 mcg%). Based on serum creatinine 21 cats were diagnosed with CKD: 10 cats were IRIS stage 1 (seven had SDMA >14 mcg%, three >19 mcg%). Two cats were IRIS stage 2 (one had SDMA >14 mcg%, one >19 mcg%). Nine cats were IRIS stage 3 or 4 (eight had SDMA >21 mcg%). The number of RT-PCR-positive cats was significantly higher in Sicily (10/11 = 91%) compared with Calabria (1/11 = 9%), and from suburban/rural areas (9/11 = 82%) compared with urban areas (2/11 = 18%). Cats affected by CKD were not more frequently found to be positive for FeMV. Increased SDMA values were found in cats without urinary or serum abnormalities suggestive of CKD and these individuals should be further evaluated and followed up.

Spontaneous feline hypothyroidism is described as an uncommon endocrinopathy and it results from deficient secretion of thyroid hormones. The congenital form is reported more frequently and is caused by different conditions including dysgenesis and dyshormonogenesis of the thyroid gland. Clinical signs appear at 4–8 weeks of age and include disproportionate dwarfism, retained deciduous teeth, lethargy, constipation, poor haircoat and goiter, among others. During a period of 18 months, six young cats (5–12 months old) were presented with a history related to one or more of the following signs: lethargy (2), obesity (5), constipation (2) and small size (6). Physical examination showed poor haircoat (6), overweight (6), disproportionate dwarfism (6), puffy facial features (6), retained deciduous teeth (6), bilateral goiter (6), bradycardia (6), hypothermia (6), alopecia in pressure points (3) and mental dullness (2). In all cases the suspicion was hypothyroidism and total thyroxine (TT4) and thyroid-stimulating hormone (TSH) were requested. The values of TT4 were between 0.4 µg/dl and 0.5 µg/dl (reference interval 0.6–3.6 µg/dl), but one cat had a value of 1.4 µg/dl. The values of TSH were between 2.7 ng/ml and 12 ng/ml (reference interval 0.06–0.32 ng/ml). The diagnosis of hypothyroidism was confirmed in the six cats and levothyroxine was administered at 10 µg/kg q12h. After 4–8 weeks of treatment all owners reported an improvement in haircoat quality and greater physical activity. Heart rate and temperature were normalized at 4–8 weeks. Alopecia disappeared after 8–12 weeks. A better physical appearance, normal body size and permanent teeth were observed after 16 weeks. The thyroid gland decreased in size, but remained palpable in all cases after reaching normal hormone values. In all cats the values of TT4 were normalized at 3–4 weeks, but TSH remained high. At 8 weeks of treatment TSH values were within the reference interval. History, clinical signs and hormone values indicated that all six cats had primary goitrous congenital hypothyroidism, probably caused by thyroid dyshormonogenesis (due to the presence of goiter), although one cat was 12 months old at the time of diagnosis. All cats responded to levothyroxine supplementation, reaching normal values of TT4 and TSH with resolution of clinical signs after 16 weeks.


849



DENTAL HEALTH IN MATURE CATS

Initial data from CatPAWS indicate that dental disease is a significant problem in mature cats; the findings are similar to previous work showing that FORLs are more prevalent on premolars 307 and 407. In the authors’ opinion, a thorough dental examination should be part of every annual health check in cats and should involve retraction of the lower lip to view 307 and 407 adequately.

Nathalie J Dowgray1, Gina L Pinchbeck2, Eithne J Comerford1, Kelly J Eyre1, Julie Bayle3, Alexander J German1 1 Institute of Ageing and Chronic Disease, University of Liverpool, UK 2 Institute of Infection and Global Health, University of Liverpool, UK 3 Royal Canin Research Centre, Aimargues, France Email: [email protected] The Cat Prospective Ageing and Welfare Study (‘CatPAWS’) is a recently established longitudinal study into the ageing of cats, based at the University of Liverpool, UK. Cats are enrolled from 7–10 years of age and then followed prospectively by health evaluations on a biannual basis. Seventy-two cats were enrolled in 2017 and the current study sought to determine the prevalence of dental disease in mature cats. Data were collected from an owner questionnaire and visual oral examination. The questionnaire asked about previous dental treatment and veterinary recommendations. The examination comprised scoring each quadrant for tartar/gingivitis and recording fractured teeth and feline odontoclastic resorptive lesions (FORLs). Data on previous dental treatment were available for 71 cats. Examination data were available for all 72 cats. From the questionnaire data, 17/71 (24%) owners had been recommended dental treatment for their cat previously: 13 (76%) had been advised to have dental work performed; two (12%) were advised to have dental work as well as advice on additional preventive measures; and two (12%) were given advice on preventive measures only. Fourteen of the 15 (93%) cats that had been advised dental work had this performed, and 13 of these required extractions in addition to a scale and polish. At the examination, 27/72 (37.5%) cats had current dental disease, nine (33%) of which had undergone prior dental treatment. FORLs were identified in 18 (66%) cats, two (11%) of which had not had prior dental work and were missing multiple teeth, with FORLs present on all remaining molars. The table summarises the distribution of FORLs and missing teeth in the other 16 cats.

Distribution of dental disease (feline odontoplastic resorptive lesions [fo] and missing teeth [mi]) in 16 mature cats Cat Tooth

1

107

fo

207

fo

307

mi

fo

407

fo

fo

850

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16 fo

fo fo fo

fo

fo

fo


fo fo

fo

fo fo

fo

mi

fo

fo

fo

fo

AT LEAST 3 YEARS’ DURATION OF IMMUNITY AGAINST RABIES IN CATS VACCINATED WITH THE INACTIVATED VP12 STRAIN IN FIELD CONDITIONS Cynthia Lesbros, Christelle Fontaine, Sylvie Gueguen Virbac, Carros, France Email: [email protected] This open-label, multicentric field study aimed to evaluate the duration of immunity against rabies in cats vaccinated with the inactivated VP12 strain. Ten veterinary clinics, five in France and five in Germany, both countries with a free-of-rabies status, participated in this study. In these clinics, cats were recruited if they had received their last anti-rabies vaccine injection containing the VP12 strain, administered alone or in association with the feline calicivirus, feline herpesvirus and feline panleukopenia virus (RCP) valences, at least 3 years before the day of sampling. Cats under immunosuppressive or corticosteroid treatment were not included. Appropriate documents were collected to ensure that the cats fulfilled the inclusion criteria. Individual serum anti-rabies antibody titres were assessed by fluorescent antibody virus neutralisation assay performed at a French reference laboratory for rabies. Serum anti-rabies antibody titers >0.5 IU/ml were correlated with protection. Fifty-six cats of 12 different breeds, males and females from 3–18 years of age, were recruited. Among the 56 cats, 25 had not been vaccinated against rabies for 3 years and 31 cats for 4–10 years. In the cats that received their last anti-rabies vaccine injection 3 years ago, serum anti-rabies antibody titers above 0.5 IU/ml were found in 24/25 (96%). The protection rate reached 100% (11/11 cats) when the cats had received a primary vaccination course and at least one annual booster injection, with serum anti-rabies antibody titres ranging from 1.12–70.63 IU/ml (n = 11). In the cats that received their last anti-rabies vaccine injection ⩾4 years ago, the protection rate was 77% (24/31 cats), with serum anti-rabies antibody titers ranging from 0.07–93.1 IU/ml (n = 31). A primary vaccination course with the inactivated VP12 rabies vaccine strain, administered alone or in combination with the RCP feline vaccine valences, induced and maintained a protective immune response against rabies in the majority of cats for more than 3 years in field conditions. The administration of at least one anti-rabies vaccination booster should help to maintain a serum anti-rabies antibody titre >0.5 IU/ml for 3 years or more.



DEVELOPMENT OF A RAPID PCR DIAGNOSTIC TEST FOR FELINE MYCOBACTERIAL DISEASE USING FORMALIN-FIXED PARAFFIN-EMBEDDED TISSUE BIOPSIES

PRIORITIES IN TREATMENT AND MONITORING OF DIABETIC CATS – FROM THE OWNERS’ POINT OF VIEW

Identifying the causative species of feline mycobacterial infections is essential to evaluate zoonotic risk. Current diagnostics are limited in sensitivity, timeliness and/or cost. We have therefore developed, and report here, a rapid, cost-effective and sensitive PCR assay for use on histopathological samples. A test population of 64 feline formalin-fixed paraffin-embedded (FFPE) tissue samples with confirmed mycobacterial disease (ie, granulomatous inflammation containing acid-fast bacilli with mycobacterial morphology) were examined. Sections were dewaxed in xylene and rehydrated through an ethanol gradient. Tissues were then digested in an enzyme buffer containing lysozyme for a minimum of 24 h at 37ºC. The resulting lysates were mechanically disrupted using silica beads in a tissue homogeniser. DNA was extracted using a DNeasy Blood & Tissue Kit (Qiagen). PCR primers for use in this study were designed to target the highly conserved gene encoding mycobacterial 16S rDNA using Primer3 software. To assess the sensitivity of this assay, it was compared with an existing, diagnostically utilised primer pair that targets the 65 KDa heat shock protein (hsp-65) encoding gene, which is also common to all mycobacteria. Both PCR reactions were run using the same cycle conditions: 95ºC for 3 mins, followed by 45 cycles of 95ºC for 30 s, 60ºC for 30 s and 72ºC for 45 s, with a final step of 72ºC for 5 mins. PCR products were visualised using 2% agarose gel electrophoresis. Positive (Mycobacterium bovis gDNA) and negative (feline gDNA and no template) controls were included for each gene targeted. All positive controls gave a single product of the correct predicted nucleotide length. No product was visualised in the negative control reactions. In the 64 tissue samples tested: mycobacterial gDNA from the hsp-65 gene was amplified from 11 samples (17% sensitivity), and mycobacterial gDNA from the 16s rDNA gene targeted by this study was amplified in 50 samples (78% sensitivity). In this study we demonstrate for the first time that it is both practical to isolate, and possible to amplify with high sensitivity, mycobacterial gDNA from feline FFPE biopsy material for the diagnosis of mycobacterial infection. This project was funded by BBSRC PhD Studentship funding.

There is scarce information on the effect that having a cat with diabetes mellitus has on its owner. We know little about how well veterinarians advise owners when making decisions about their cat’s diabetes, about how having a diabetic cat affects owners’ lives, or what methods they find most helpful when managing their cat’s disease. The aims of this study were to evaluate: i) owner–veterinarian communication; ii) owners’ perceptions on and priorities for treating and monitoring their cat’s diabetes; and the impact the disease has on (iii) the owners’ everyday lives and iv) the human–pet bond. This study used a questionnaire made available to owners of diabetic cats throughout the world on VetProfessionals.com. A total of 748 completed questionnaires were collected; 76% of the respondents stated that websites/online forums that they had found themselves were the most useful resource to learn about their cat’s diabetes; this exceeded slightly the percentage of respondents who found discussions with the veterinarian useful (74.5%). Only 29% of respondents had been given advice on websites/forums by their veterinary clinic. When asked for the most important factors when decision making regarding treatment, ‘what is best for my cat’ and ‘veterinarian recommendations’ were selected by 100% and 86% of owners, respectively; while ‘ease of treatment’ and ‘fear of injections’ played a much smaller role (65% and 26.2% of owners selected these factors, respectively). Perhaps disappointingly, 39% of cats were not fed a special diet to help diabetic control but, impressively, 71% of owners used home blood glucose monitoring. Initial concerns about costs, the effect on their daily lives and a potential negative impact on the human–pet bond, reduced significantly after initiating treatment (P 15 red blood cells (RBCs)/high-power field (HPF). The test sensitivity (Se), specificity (Sp), and positive and negative predictive values (PPV, NPV, respectively) were determined at each level of the color scale, and its response modifiers were assessed with a generalized linear mixed model for ordinal response variables (α = 0.05).



The interim statistical analysis performed on the first 34 cats (four clinics) revealed that the hematuria threshold was too conservative and could be decreased to 5 RBCs/HPF. Also, the second study arm was under-represented; a cat shelter was, therefore, recruited to increase the number of cats. To detect at least one false negative and false positive and perform the final statistical analysis, 80 cases were required. The final statistical analysis revealed that the number of RBCs/HPF, the blood estimation from the dipstick, the proteinuria and the pH significantly increased the cumulative odds of a positive screening result (P 19,500/μl or 5%. Cats were excluded if they were vaccinated against FPV within 3 weeks of hospital admission. Differences between survivors and nonsurvivors with regards to the presence of SIRS were evaluated by the Fisher’s exact test; all the other data were analysed by the Mann-Whitney U-test. Sixty-one cats entered the study. Forty-one cats were discharged alive from the hospital (survivors), 20 died or were euthanased (non-survivors). Median length of hospitalisation was 5 days (range 0–21 days) in survivors and 3 days (range 1–9 days) in nonsurvivors (P = 0.4). SIRS was present in 6/41 (14.6%) survivor cats and in 13/20 (65%) non-survivor cats (P = 0.0004). Median serum TT4 concentration was significantly higher (P = 0.0003) in survivors (median 1.25 µg/dl, range 0.49–3.49 µg/dl) vs non-survivors (median 0.49 µg/dl, range 0.49–1.98 µg/dl). Median serum cholesterol concentration was not significantly different (P = 0.4) between survivors (median 141 mg/dl, range 60–240 mg/dl) and nonsurvivors (median 113 mg/dl, range 79–215 mg/dl). SIRS and serum TT4 on admission are prognostic in cats with FPV infection.

DENTAL PATHOLOGY IN CATS: ASSOCIATION BETWEEN CLINICAL AND RADIOGRAPHIC DIAGNOSIS Inês M Neves1, Inês Viegas1, Michèle M Venturini2, Daniel Ferro2, Herbert Corrêa2, Joana T de Oliveira1, João F Requicha3 1 Faculty of Veterinary Medicine, Lusófona University, Lisbon, Portugal 2 Odontovet – Veterinary Dentistry Center, São Paulo, Brazil 3 Department of Veterinary Sciences, University of Trás-os-Montes e Alto Douro, Vila Real, Portugal Email: [email protected] Dental pathology is highly prevalent in cats although apparently asymptomatic abnormalities are frequently present, hindering correct and prompt diagnosis. To date, few studies have compared the results of clinical and imaging evaluation of cats’ teeth. The present study aimed to compare clinical examination and intraoral radiology as diagnostic approaches; thus, estimating the percentage of dental pathology missed to clinical-only diagnosis. Ninety-five cats that attended Odontovet – Veterinary Dentistry Center (São Paulo, Brazil) were included. A total of 1520 teeth were analysed, based on dental charts, photographs and dental radiographs. The following clinically observed parameters were recorded: periodontal disease, dental calculus index, changes in gingival morphology, tooth fractures, tooth resorption, absence of teeth, dental malformations and alveolar bone expansion. Additionally, radiographically observed parameters were recorded as follows: periodontal disease stage, tooth resorption stage and type, types of tooth fractures, degree of alveolar bone

856


expansion, tooth absence, tooth impaction, dental malformations and periapical lesions. We found the most prevalent condition to be periodontal disease, clinically observed in 72.6% of cats and radiologically confirmed in 92.6%. Tooth resorption was the second-most prevalent disease, although only 48.4% of the animals were clinically diagnosed; 76.8% of the cats had it confirmed by radiography. In both diseases, statistically significant differences (P