Evaluation of cytological diagnostic accuracy for canine splenic neoplasms: An investigation in 78 cases using STARD guidelines.

Cytology represents a useful diagnostic tool in the preliminary clinical approach to canine splenic lesions, and may prevent unnecessary splenectomy. However, few studies have evaluated diagnostic accuracy of cytology in the diagnosis of canine splenic neoplasms. The aim of this study was to determine overall accuracy, sensitivity, specificity, positive and negative predictive values (i.e. diagnostic accuracy indexes) of cytology for canine splenic neoplasms following Standards for the Reporting of Diagnostic Accuracy Studies (STARD) guidelines. A consecutive series of canine splenic cytological samples was retrospectively retrieved from the database of the Diagnostic Pathology Service of the Department of Veterinary Medicine (DIMEVET-University of Milan). Histopathology was set as the diagnostic reference standard. Cytological cases were enrolled when slides were available for review and when the same lesion was submitted for histopathology. Seventy-eight (78) lesions were included in the study. By histopathology, 56 were neoplastic and 22 were non-neoplastic. Cytology had an overall accuracy of 73.08% (95% C.I. 61.84%-82.50%), sensitivity of 64.29% (95% C.I. 50.36%-76.64%), specificity of 95.45% (95% C.I. 77.16%-99.88%), and positive and negative predictive values of 97.3% (95% C.I. 84.01%-99.60%) and 51.22% (95% C.I. 42.21%-60.15%), respectively. Low sensitivity and negative predictive value were balanced by very high specificity and positive predictive value. When positive for neoplasia, cytology represents a useful diagnostic tool to rule in splenic neoplasia, prompting surgery independently from other diagnostic tests. Conversely, a negative cytological result requires additional investigations to confirm the dog to be disease free.

Introduction

Ultrasonographic examination of nodular splenic lesions in dogs is not reliable to differentiate with certainty benign and malignant processes, necessitating the use of additional, ideally minimally invasive, diagnostic tests [1]. Hemangiosarcoma (HES) is the most common primary splenic malignant tumor of dogs [2]. Still, HES represents fewer than 25% of overall splenic lesions [3] with up to 74% of dogs being diagnosed with benign lesions [4] such as hematoma and hyperplasia [2,5]. However, due to the poor prognosis associated with HES, splenectomy is still the routine approach to most canine splenic masses for both diagnostic and therapeutic purposes [6-8].

While the two-month post-splenectomy survival rate is lower in dogs with HES (32%) compared to dogs with hematomas (83%), the general survival rate after splenectomy is 52%, regardless of underlying splenic pathology [3]. While this can be interpreted as fair survival data, a proportion of dogs (7.6%) will develop complications secondary to splenectomy because of thrombotic or coagulopathic syndromes [9]. Additional adverse effects following splenectomy in dogs have included peri- and post-operative ventricular arrhythmias [9-11], reduced blood filtration and renewal [12,13], impairment of humoral immune response [14], reduced immune-surveillance against bacteria and parasites [15-19], and higher incidence of gastric dilatation-volvulus [6,20-22]. For these reasons, any preoperative diagnostic approach to splenic lesions, including cytology, may be beneficial in preventing unnecessary splenectomy. Notwithstanding the common belief that splenic aspiration can be dangerous especially when investigating cavitated masses [6,7,23], complications from splenic aspiration procedures are rarely elicited even in thrombocytopenic animals [8,24-26]. For comparison, in human medicine splenic fine needle aspiration (FNA) cytology is seldom associated with complications [27,28], resulting in 5.2% secondary complications with fewer than 1% considered severe and consisting mostly of controllable hemorrhage [29].

Thus, attempts to minimize unnecessary splenectomy should prompt an increased use of additional diagnostic techniques as preoperative screening tests to characterize splenic disease. Fine needle aspiration cytology can provide diagnostic information useful to distinguish inflammatory, benign and malignant nodular lesions and to assess generalized splenomegaly [7,24,30].

Despite the relatively high frequency of splenic diseases in dogs, data regarding usefulness and validity of diagnostic cytology are fragmentary. In veterinary medicine, no studies have comprehensively assessed overall accuracy, sensitivity, specificity, positive and negative predictive values of canine splenic cytology against histopathology utilizing the Standards for the Reporting of Diagnostic Accuracy Studies (STARD) guidelines [8,23-25,31-33]. STARD guidelines have been created to avoid incomplete reporting in diagnostic accuracy studies and to improve the general quality of the latter, reducing problems related to study identification, critical appraisal, and replication [34]. Overall agreement between cytology and histology of canine splenic lesions is the most frequent index reported, ranging from 38 to 100% [8,23-25,31-33]. Specifically, this index has been evaluated on a limited number of splenic cytological specimens (range 5–40) [8,23-25,31-33]. In most reports, sensitivity, specificity, positive and negative predictive values of splenic cytology are not calculated and cannot be properly estimated [8,23-25,31-33] because caseloads simultaneously include multiple species [24,25,31-33], multiple tissues and organs [31-33], or “equivocal” or “provisional” cytological and histological diagnoses [23,24,33]. Application of STARD guidelines in the current study allowed cross-tabulation of cytological results (i.e. the index test) against those of histopathology (i.e. the reference standard) to generate sensitivity and specificity data [34]. These data have not been included in previous studies and will be useful for future researchers comparing diagnostic methods for canine splenic neoplasms.

To avoid unnecessary splenectomy, a diagnostic test with a high sensitivity and negative predictive value is desirable because these indexes measure the percentage of diseased dogs correctly diagnosed with splenic neoplasia and the probability that dogs with a negative cytology truly do not have a neoplasm, respectively. In this context, the aim of this study was to determine the diagnostic accuracy of cytology in the diagnosis of canine splenic neoplasms utilizing the corresponding histopathology as the diagnostic reference standard [24,31,33,35,36], following STARD guidelines [34]. Additionally, sensitivity of cytology in the diagnosis of specific tumor types and in the identification of nodular versus diffuse neoplasms was evaluated.

Materials and methods

Criteria of selection of cases

In this retrospective study, the electronic cytological database of the Diagnostic Pathology Service of the Department of Veterinary Medicine (DIMEVET) of the University of Milan was searched for splenic samples collected between January 1st 1998 to January 31st 2018. The database was searched for specific key words in the following combinations: 1) “dog” and “spleen”, 2) “dog” and “splenic”. A consecutive series of canine splenic cytologies was obtained. Cytological samples came from external referring private practices or from the Veterinary Teaching Hospital (VTH) of the DIMEVET, or were prepared from fresh surgical biopsies and necropsies. Samples were submitted or collected to evaluate splenomegaly or nodular lesions.

The histopathology database was then searched for the histopathology corresponding to the same lesion examined by cytology. Histopathological samples were obtained from splenic biopsies (nodular lesions) or whole spleens (from splenectomies or necropsies) submitted to the Diagnostic Pathology Service of the DIMEVET, or were represented by slides submitted by external pathologists as second opinion cases. A time interval >45 days between cytological and histological sampling was an exclusion criterion. Histopathological samples collected via needle core biopsies were excluded from the study as they may bear reduced diagnostic reliability compared to incisional and excisional histological samples [8,37]. Cases were included in the study only when cytological and histological slides of the same lesion were available for review.

Additional information collected from the archives for cases included in the study were: sex, age, breed, cytological sampling technique (e.g. fine needle aspiration–FNA, touch imprint, scraping) and gross appearance of the lesion (i.e. diffuse versus nodular lesion).

To improve data completeness, transparency, and reproducibility, the study was conducted following the STARD guidelines [34] to the best of authors’ ability. All canine splenic samples included in the current retrospective work, regardless of the sampling technique applied, were submitted to the Diagnostic Pathology Service of the DIMEVET for diagnostic purposes of spontaneous developing diseases. No animals were sampled or euthanized for research use. The use of animal tissue in the current study was approved by the Ethics Committee in charge for animal welfare of the University of Milan (Organismo Preposto al Benessere degli Animali, OPBA) with protocol number OPBA_86_2019. Sensitive information regarding owners and animals were stored, managed and preserved according to European and Italian laws.

Sample processing

Cytological samples were air dried and stained with May-Grünwald-Giemsa (Merck KGaA, Frankfurt, Germany). Tissue samples for histopathology were fixed in 10% neutral buffered formalin, processed routinely, and embedded in paraffin wax. Sections of 5 μm were stained with Hematoxylin and Eosin. Second opinion cases were provided as Hematoxylin and Eosin stained slides by the referring pathologists.

Case review

All cytological and histopathological samples were independently reviewed in a blinded fashion by three cytologists (M.C.—ECVP, G.G.—ECVCP, M.G.—resident) and by three anatomical pathologists (A.F.–ECVP, P.R.—ECVP, M.T.—resident), respectively. Both cytologists and anatomical pathologists were blinded to signalment information related to each case.

For each cytological case, one slide for each sampling technique was reviewed. First, slides were examined at low-power magnification (i.e. 10x objective lens) to assess the adequacy of the specimen. Poorly cellular samples were those characterized by marked hemodilution in the absence of both stromal elements and a mixed population of leukocytes [30]. Poorly cellular samples, poorly smeared samples (i.e. too thick or where most cells were ruptured), and samples where stain quality impaired adequate definition of the cell type (e.g. formalin-contaminated smears), were considered inconclusive [30]. Inconclusive cases were excluded from the statistical analysis as previously reported [8,24,32,33,35-37].

Cytological diagnoses were expressed according to those reported in the literature [4,30,38,39]. To facilitate comparison of the agreement between cytological and histological results, each cytological sample was further classified as neoplastic or non-neoplastic according to the main pathologic process. Non-neoplastic samples were those characterized by degenerative, reactive (including extramedullary hematopoiesis) [23,30], and inflammatory changes, as well as normal specimens consisting of stromal elements with mixed leukocyte population [30,38]. Reviewing cytologists were not allowed to use diagnostic modifiers such as “probably”, “most likely”, “suggestive of”, as previously reported [37], nor to provide equivocal diagnoses (i.e. reporting more than one differential diagnosis). When a univocal diagnosis was not reached, cytologists reviewed the case collaboratively to find an agreement. Only the definitive diagnosis was included in the consecutive statistical analysis.

Neoplastic cytological samples were further subdivided by tumor type into the following subcategories: benign soft tissue tumor (BSTT) including angioma, angiosarcoma (HES), soft tissue sarcoma (STS) excluding angiosarcoma, lymphoma (LYM), mast cell tumor (MCT), histiocytic tumors including hemophagocytic sarcoma (HS), other round cell tumors including plasma cell tumor, myeloid leukemia and undifferentiated round cell tumor (ORCT), carcinoma (CARC), malignant neoplasm not otherwise specified (MNNOS).

Cases were included only when the three anatomical pathologists were in agreement because histopathology served as the diagnostic reference standard to evaluate cytological diagnostic accuracy. Neoplasms were classified applying the World Health Organization’s histologic classification of tumors in domestic animals [40-48]. To further standardize histopathological diagnoses, anatomic pathologists were invited to classify some specific pathological entities (i.e. lymphomas, histiocytic proliferative disorders, nodular lesions previously classified as “fibrohistiocytic nodules”) according to criteria reported in recent literature [49-52]. For nodular lesions, histological samples were considered conclusive and therefore only included in the statistical analysis if at least one slide containing at least one margin between the nodule and the adjacent splenic parenchyma was available for review [2,3,5].

Histological samples were then classified as neoplastic and non-neoplastic. Neoplastic cases were further subdivided utilizing the same subcategories applied to cytological samples.

Data analysis

For all cases, the cytological diagnosis was compared with its paired histopathological diagnosis. Since histological samples maintain tissue architecture and are not biased by cellularity [8,32,37,53], histopathology was set as the reference standard as previously reported [24,31,33,35,36].

To determine diagnostic accuracy indexes, cytological specimens were classified according to four correlation categories (true positive, true negative, false positive, false negative). Specifically, the True Positive (TP) category included all cytological samples diagnosed as neoplastic with a corresponding neoplastic histopathology. The True Negative (TN) category comprised all cytological samples diagnosed as non-neoplastic with a corresponding non-neoplastic histopathology. The False Positive (FP) category included all cytological samples diagnosed as neoplastic with a corresponding non-neoplastic histopathology. The False Negative (FN) category comprised all non-neoplastic cytological diagnoses with a corresponding neoplastic histopathology.

To evaluate the sensitivity of cytology in the diagnosis of specific tumor types, only those cases histologically confirmed as neoplastic were taken into account. The subcategories assigned to each cytological and corresponding histological sample were then compared. When cytological and histological diagnoses matched for both neoplastic categorization and tumor type subcategorization, the case was defined as “true positive with complete agreement”. When cytological and histological diagnoses matched for the neoplastic categorization but did not match for the tumor type subcategorization, the case was defined as “true positive with partial agreement”. When a histopathological diagnosis categorized as neoplastic corresponded to a cytological diagnosis categorized as non-neoplastic, the case was considered in disagreement and defined as “false negative case”.

Statistical methods

Cytological-histological correlation categories (TP, FP, TN, FN) were included in a 2x2 table and used to calculate point estimates of overall accuracy, sensitivity, specificity, positive and negative predictive values [36,54]. Overall accuracy was defined as the ability of cytology to correctly identify neoplastic and non-neoplastic lesions, and was calculated as the sum of cases in which cytology and histology agreed in diagnosing a lesion as neoplastic (i.e. TP) or non-neoplastic (i.e. TN), divided by the total number of cases included in the study [32,55]. Given that pre-determined acceptability criteria for diagnostic performance of splenic cytology to distinguish between neoplastic and non-neoplastic lesions have not been previously established, overall accuracy, sensitivity, specificity, positive and negative predictive values were considered low if <70%, moderate if ≥70% and <80%, high if ≥80% and <90%, and very high if ≥90% [54]. To increase data comparability with other studies, positive and negative likelihood ratios were calculated [36]. Ninety-five percent (95%) confidence interval was calculated for each of the above mentioned indices of diagnostic test accuracy using a web-based application (MEDCALC—https://www.medcalc.org/calc/diagnostic_test.php).

The level of agreement between cytology and histopathology in the diagnosis of splenic neoplastic conditions was further investigated calculating the Cohen’s kappa coefficient (κ), which was then corrected by the standard error. The value of k can be indicative of no agreement (if k <0), slight agreement (k = 0–0.20), fair agreement (k = 0.21–0.40), moderate agreement (k = 0.41–0.60), substantial agreement (k = 0.61–0.80), almost perfect agreement (k = 0.81–0.99), or perfect agreement (k = 1) [36,54]. Cohen’s kappa and standard error were calculated utilizing GraphPad QuickCalcs Web site (GraphPad Inc.—https://www.graphpad.com/quickcalcs/kappa2/).

Sensitivity of cytology in differentiating splenic tumor types was defined as the ability of cytology to correctly identify as neoplastic a sample belonging to a specific neoplastic subcategory. Therefore, sensitivity for each tumor type was calculated as the sum of cases in complete and partial agreement (i.e. true positive cases) divided by the total number of cases with that specific neoplasm [31].

Similarly, the sensitivity of cytology in the diagnosis of neoplastic lesions according to their distribution pattern (i.e. diffuse or nodular) was evaluated. For each distribution pattern, sensitivity was calculated as the sum of cases in complete and partial agreement (i.e. true positive) divided by the total number of cases with a specific distribution pattern. Sensitivity of cytology according to distribution pattern was calculated for neoplastic lesions in general (i.e. the general sensitivity value obtained in our study), as well as for those specific neoplastic subcategories including cases with either diffuse or nodular distribution pattern.

Chi-square analysis applied to pairwise comparison was performed to evaluate whether statistically significant differences existed in the sensitivity of cytology for different tumor types, as well as in the diagnosis of neoplastic lesions with diffuse or nodular distribution pattern [36,37]. Specifically, the sensitivity of cytology for each neoplastic subcategory was compared with the sensitivity for splenic neoplasms in general, the sensitivity for all other neoplastic subcategories, and the sensitivity for any other neoplastic subcategory. Similarly, the difference between sensitivity for nodular or diffuse lesions among neoplasm in general and for each neoplastic subcategory was statistically investigated. Chi-square analysis was performed only on sensitivity values different from 0% and 100%, using MEDCALC (https://www.medcalc.org/calc/comparison_of_proportions.php). A p-value <.05 was considered statistically significant.

Results

Animals and samples

From a total of 950 splenic cytological samples retrieved between 1998–2018, 92 cytological samples from 91 dogs were included in the study; one dog was sampled for two distinct splenic lesions. A total of 858 splenic cytological cases were excluded for the following reasons: lack of a corresponding histopathological sample (832 cases), unavailable cytological and/or histological samples (16 cases), and needle core biopsies (10 cases). Among the selected cytological cases, 14 were considered inconclusive, and therefore excluded from the consecutive statistical analysis. Detailed evaluation of diagnostic accuracy was performed on 78/92 reviewed cytological samples (retrieval rate: 84.78%) obtained from 77 dogs [37].

Sex was available for 76/77 dogs: 19 were spayed females, 6 neutered males, 21 intact females, and 30 intact males. Mean age was 9.05 years (age range 2 months-16 years; age was not available for 2 cases). Twenty-six (26) breeds other than mongrels were represented; in one case breed was not provided.

The time interval between cytological sampling and corresponding histopathology collection ranged from 0 to 44 days for all cases.

Of the 78 cases included in the study, 81 cytological slides were evaluated (3 cases were sampled with two different techniques, i.e. touch imprinting and scraping). Cytological samples consisted of 43/81 touch imprints (53.09%) collected from both surgical biopsies and necropsies, 28 FNAs (34.57%), of these 21 were ultrasound guided, 1 was CT-scan guided, 1 was obtained during surgery, while in 5 FNA biopsies no additional sampling information was available. In 6 cases scrapings were obtained from surgical and necropsy specimens (7.41%). In 4 cases (4.94%) the sampling technique was not specified.

Complete agreement among anatomical pathologists was reached for all the 78 corresponding histopathological samples. Histopathological specimens were distributed as follows: 51 surgical samples from partial or complete splenectomies (51/78 cases, 65.38%), 24 spleens from necropsies (24/78, 30.77%), and 3 cases submitted as a second opinion (3/78, 3.85%).

Cytological and histological diagnoses

All cytological and corresponding histopathological diagnoses (78 cases) are listed in S1 Table. The diagnoses for the cytological-histological pairs excluded due to inconclusive cytology are listed in S2 Table.

No diagnostic differences were found for samples collected using two different sampling techniques, and therefore they were considered as one case in the consecutive statistical analysis. Cytologically, 37/78 cases were diagnosed as neoplastic (47.44%) and 41/78 as non-neoplastic (52.56%). All cases diagnosed as neoplastic were classified as malignant, and indeed, no benign neoplasms were cytologically observed.

Histologically, 56/78 cases (71.79%) were neoplastic (S1 Table) and 22/78 cases (28.21%) were non-neoplastic. Malignant tumors were 51 (51/56 tumors, 91.07%) and 5 were benign. The prevalence of each tumor type is reported in Table 1. No malignant neoplasm not otherwise specified was included in the study.

Table 1

Prevalence, agreement levels and sensitivity of cytology in the diagnosis of each neoplastic subcategory.

	Prevalence	TP cases with complete agreement	TP cases with partial agreement	FN cases	Sensitivity	Confidence Interval (95%)
TOTAL	71.79% (56/78)	42.86% (24/56)	21.43% (12/56)	35.71% (20/56)	64.29%	50.36%–76.64%
HES	28.57% (16/56)	68.75% (11/16)	6.25% (1/16)	25% (4/16)	75%	47.62%–92.73%
LYM	28.57% (16/56)	37.50% (2/16)	12.50% (6/16)	50% (8/16)	50%	24.65%–75.35%
STS	12.50% (7/56)	42.86% (3/7)	28.57% (2/7)	28.57% (2/7)	71.43%	29.04%–96.33%
BSTT	8.93% (5/56)	0% (0/5)	0% (0/5)	100% (5/5)	0%	0.00%–52.18%
HS	7.14% (4/56)	25% (1/4)	50% (1/4)	25% (1/4)	75%	19.41%–99.37%
MCT	7.14% (4/56)	100% (4/4)	0% (0/4)	0% (0/4)	100%	39.76%–100%
CARC	5.36%(3/56)	66.67% (2/3)	33.33% (1/3)	0% (0/3)	100%	29.24%–100%
ORCT	1.79% (1/56)	0% (0/1)	100% (1/1)	0% (0/1)	100%	2.50%–100%

BSTT, benign soft tissue tumor including angioma; CARC, carcinoma; FN, false negative; HES, angiosarcoma; HS, histiocytic neoplasm (including hemophagocytic syndrome); LYM, lymphoma; MCT, mast cell tumor (MCT); ORCT, other round cell tumor; STS, soft tissue sarcoma other than angiosarcoma; TP, true positive.

Of the 78 splenic lesions, 60 were nodular (76.92%), and 17 were diffuse (21.79%), while no information regarding the distribution pattern was available for 1 case (1.28%). Of the 56 neoplastic lesions, 43/56 cases (76.79%) were nodular and 12 cases (21.43%) were diffuse. The case for which distribution pattern was not provided was a liposarcoma (1.79%). This case was excluded from the evaluation of cytology sensitivity according to neoplasm distribution pattern. The proportion of cases with nodular or diffuse pattern for each tumor type are reported in Table 2.

Table 2

Prevalence, agreement levels and sensitivity of cytology in the diagnosis of each neoplastic subcategory on the basis of distribution pattern.

	Nodular				Diffuse
	# of cases	TP	FN	Sensitivity(95% CI)	# of cases	TP	FN	Sensitivity(95% CI)
TOTAL	43/56 (76.79%)	26/43	17/43	60.47%(44.41%–75.02%)	12/56 (21.43%)	9/12	3/12	75%(42.81%–94.51%)
HES	14/16 (87.50%)	10/14	4/14	71.43%(41.90%–91.61%)	2/16 (12.50%)	2/2	0/2	100%(15.81%–100%)
LYM	10/16 (62.50%)	5/10	5/10	50%(18.71%–81.29%)	6/16 (37.50%)	3/6	3/6	50%(11.81%–88.19%)
MCT	1/4 (25%)	1/1	0/1	100%(2.50%–100%)	3/4 (75%)	3/3	0/3	100%(29.24%–100%)

CI, confidence interval; FN, false negative; HES, angiosarcoma; LYM, lymphoma; MCT, mast cell tumor (MCT); TP, true positive.

Cyto-histological correlation

Following the tabulation of cytological and histological diagnoses (S1 Table), 36 cases (46.15%) were classified as TP, 21 (26.92%) were TN, 20 (25.64%) were FN, and 1 (1.28%) was a FP (Table 3).

Table 3

Cytological-histological correlation categories.

Diagnosis	Histology: neoplastic	Histology: non-neoplastic	Total
Cytology: neoplastic	36 (TP)	1 (FP)	37
Cytology: non-neoplastic	20 (FN)	21 (TN)	41
Total	56	22	78

FN, false negative; FP, false positive; TN, true negative; TP, true positive.

The FP case had a cytological diagnosis of lymphoma that corresponded histologically to a purulent bacterial splenitis (in this case the full spleen was available for analysis and no tumor was found; however, severe marginal zone hyperplasia was present).

Neoplastic and non-neoplastic lesions were correctly identified in 57/78 cases (Table 3), therefore overall accuracy of cytology was 73.08% (Table 4). Sensitivity of cytology in the diagnosis of splenic neoplasms was 64.29%, specificity was 95.45%, positive predictive value was 97.30%, and negative predictive value was 51.22% (Table 4). Positive and negative likelihood ratios were 14.14 and 0.37, respectively (Table 4).

Table 4

Prevalence of neoplastic lesions, with point estimate and 95% confidence interval of diagnostic accuracy indexes, likelihood ratios and Cohen’s k.

Diagnostic accuracy index	Value	Confidence Interval (95%)
Prevalence	71.79%	60.47%–81.41%
Overall accuracy	73.08%	61.84%–82.50%
Sensitivity	64.29%	50.36%–76.64%
Specificity	95.45%	77.16%–99.88%
PPV	97.30%	84.01%–99.60%
NPV	51.22%	42.21%–60.15%
PLR	14.14	2.06–96.94
NLR	0.37	0.26–0.54
K value	0.473	0.304–0.643

NLR, negative likelihood ratio; NPV, negative predictive value; PLR, positive likelihood ratio; PPV, positive predictive value.

According to Cohen’s test the level of agreement was considered as “moderate”, with a κ value of 0.473 corresponding to a standard error of 0.086.

The distribution of TP and FN cases for each tumor type is reported in Table 1. The sensitivity of cytology in the diagnosis of each neoplastic subcategory was 100% for MCT, CARC and ORCT, 75% for HES, 75% for HS, 71.43% for STS, 50% for LYM, and 0% for BSTT included in the study. Further details regarding complete and partial agreement between cytological and histological diagnoses as well as confidence intervals of sensitivity value for each tumor type are listed in Table 1. Chi-square analysis of cytological sensitivity was applicable only to HES, HS, STS, and LYM. No statistically significant sensitivity differences were observed (p-value ranging from 0.1506 to 1.0).

The proportion of TP and FN cases with nodular or diffuse pattern for each tumor type are reported in Table 2. Sensitivity of cytology in the diagnosis of neoplastic lesions in general according to their distribution pattern was 60.47% for nodular and 75% for diffuse neoplasms, with no statistically significant difference between the two values (p = 0.3593). For some tumor types the sensitivity of cytology on the basis of the distribution pattern was not calculated, given that only nodular (BSTT, STS, HS, CARC) or diffuse (ORCT) neoplastic lesions were represented in these categories. Sensitivity in the diagnosis of nodular and diffuse lymphomas was for both 50%, with no statistically significant difference between the two values (p = 1.0). Sensitivity for nodular angiosarcomas was 71.43% and 100% for diffuse angiosarcomas, while sensitivity for both nodular and diffuse mast cell tumors was 100%. Considering these results, Chi-square analysis of sensitivity on the basis of the distribution pattern was not performed for angiosarcomas and mast cell tumors.

Discussion

In this study we report overall accuracy, sensitivity, specificity, positive and negative predictive values of cytology for the diagnosis of canine splenic neoplasms. Similar studies [8,23-25,31-33] have limited the evaluation of cytological diagnostic accuracy to overall agreement with histopathology, hampering comparison with our results. Our study has evidenced a moderate overall accuracy of cytology. Specifically, although this technique had a high specificity and positive predictive value for the diagnosis of splenic neoplasia, sensitivity and negative predictive value were lower, indicating that cytological diagnosis of splenic neoplasia is reliable, but a negative result cannot be used to exclude the possibility of splenic neoplasia.

According to overall accuracy and Cohen’s k values, cytology is not a reliable alternative to histopathology in the definitive diagnosis of splenic tumors in most cases. When compared with previous studies, our overall accuracy value (73.08%) laid in between the higher range of 83.87–100% [23-25] and the lower 38–69.7% range [8,31-33] reported in other studies. To allow comparison, the overall accuracy (intended as the sum of complete and partial diagnostic agreements) was calculated from the raw data of previously published caseloads [8,23-25,31-33] when not made explicit in the corresponding manuscript.

Low sensitivity and negative predictive value of this study indicate that a cytology negative for neoplasia should prompt further investigations to confirm a dog to be truly free from neoplastic disease. This contrasts with our initial hypothesis that cytology may represent a useful tool to avoid unnecessary splenectomy. Instead, high specificity and positive predictive value identify cytology as a good and reliable tool to rule in the diagnosis of splenic neoplasia with a high degree of confidence. In practical terms, a cytology positive for neoplasia may lead to a faster surgical treatment, avoiding lag times and higher costs associated with application of diagnostic imaging techniques such as contrast-enhanced ultrasound and computed tomography (CT) [1,56]. Our results are in line with studies evaluating diagnostic accuracy of cytology applied to various organs in dogs [31,32,35,36,54], with sensitivity and negative predictive value generally lower than specificity and positive predictive value, respectively.

Regarding the reliability of cytology in the diagnosis of specific tumor types, the lack of statistically significant differences between subcategories may be related to an imbalance in the number of cases for each tumor type. Also, our results may be influenced by the tumor cell type evaluated, since exfoliation rate varies substantially between round cell, epithelial and mesenchymal tumors [30,31,35,53]. Specifically, mesenchymal tumors have the lowest tendency to exfoliate [30,31,35,53] explaining the low sensitivity in diagnosing benign mesenchymal tumors. Moreover, identification of vascular tumors (i.e. angiomas and HES) among false negative cases is not surprising since the architecture of these tumors often leads to significant peripheral blood contamination in aspirates [4,30,32,57]. The low sensitivity of cytology in the diagnosis of splenic lymphomas relates to the specific distribution of tumor types in the spleen, where indolent nodular lymphomas (i.e. mantle cell lymphoma and marginal zone lymphoma) are frequent as was in this caseload. These are nodular lymphomas composed of small to medium sized cells with minimal atypia and a low mitotic rate [49,58]. Thus, mantle cell lymphoma and marginal zone lymphoma can be easily misinterpreted as reactive lymphoid hyperplasia on cytology, and histopathology is often necessary for a definitive diagnosis that relies on the evaluation of tumor architecture [49].

Although not statistically significant, our results paralleled those of previous reports identifying higher cytological accuracy in the diagnosis of diffuse compared to focal lesions [23,32,57].

One false positive diagnosis of neoplasia (i.e. lymphoma) was included in this study. Splenic marginal zone hyperplasia is a common finding in dogs [49,59], and cytological sampling from these areas may result in a monomorphic specimen mimicking marginal zone lymphoma. This is a risk that pathologists have to bear in mind; thus the diagnosis of nodular low-grade lymphoma should be supported by histological evaluation of architectural changes, especially in dogs. A recent report [60] has demonstrated a high overall concordance between histopathology, immunohistochemistry and PCR for antigen receptor rearrangement (PARR) in the diagnosis of marginal zone lymphoma, mantle cell lymphoma and lymphoid or complex nodular hyperplasia. Therefore, further development of combined methods also applicable to cytological specimens may provide a less invasive and more valuable diagnostic approach to the diagnosis of splenic nodular lymphoid lesions.

Although histopathology is generally considered the diagnostic reference standard [24,31,33,35,36], several limitations should also be considered for this technique in the diagnosis of splenic tumors. Specifically, the diagnosis of splenic hematomas and hemangiosarcomas is considered difficult, especially if spleens are not submitted entirely and if adequate samples from the margin of the lesion are not collected [2,3,5]. Noteworthy, splenic hematomas and HES may not be grossly distinguishable [2,3,5,23,61], and the first may represent a component of the latter [61].

The current study is characterized by several limitations, mainly due to its retrospective nature. One major limit was the inclusion of specimens obtained by different sampling techniques, with a high number of impression smears collected from both surgical biopsies and necropsies. Additionally, the inclusion of splenic cytological samples from necropsies and the university setting of this work may have further biased the study toward cases with a more aggressive behavior and with features of malignancy easier to diagnose. This may not reflect daily clinical practice in which FNA is the most common sampling technique to pre-operatively assess splenic lesions. Also, as previously observed [31,53], different sampling methods may have resulted in an improvement of sensitivity of cytology in this study, especially for those neoplasms characterized by low exfoliation rate. On the other hand, this observation can be viewed also in positive terms. Indeed, in a practical setting the preliminary evaluation of surgical biopsies or entire spleens by cytology prior to fixation could be implemented to facilitate the diagnosis and to reduce turnaround time. Additionally, this approach can provide pathologists with material useful not only for a preliminary diagnosis, but also for immunocytochemistry and for PARR on fresh specimens.

Despite this caseload being larger than previously reported ones, the small number of cases evaluated may explain the relatively wide confidence intervals observed around point estimates of our diagnostic accuracy indexes. Results may have been further biased by the inclusion criteria applied in the current study, leading to the exclusion of more than 90% cytological samples of canine spleen in our archives.

Unfortunately, full agreement with STARD guidelines could not be obtained in this study since the type of treatment administered between cytological and histological sampling, and the incidence of adverse events following splenic sampling, could not be retrieved from our electronic archives.

In conclusion, to the best of our knowledge, this is the first study conjunctively reporting overall accuracy, sensitivity, specificity, positive and negative predictive values of cytology in the diagnosis of canine splenic neoplasms compared to histopathology. Diagnostic accuracy indexes identified limitations of negative cytological results in excluding a dog to be truly free from neoplasia; however, high specificity and positive predictive value highlighted cytology as a valuable tool in the diagnostic approach to splenic neoplasms.

Cytological and corresponding histological diagnosis for each case included in the statistical analysis, with correlation category and lesion distribution pattern.

Neoplastic subcategory and level of agreement are provided when applicable. Dogs from which more than one sample was obtained are marked with (*). The age of dogs is reported in years, if not otherwise specified. Abbreviations: B, biopsy; BSTT, benign soft tissue tumor including angioma; CA, complete agreement; CARC, carcinoma; DA, disagreement; EMH, extramedullary hematopoiesis; F, female; FN, false negative; FNA, fine needle aspirate; FP, false positive; HES, angiosarcoma; HS, histiocytic neoplasm (including hemophagocytic syndrome); LYM, lymphoma; M, male; MCT, mast cell tumor (MCT); mm, months; MNNOS, malignant neoplasm not otherwise specified; N, necropsy; NF, neutered female; NM, neutered male; NPL, neoplastic; NON-NPL, non-neoplastic; NOS, not otherwise specified; n/a, not applicable; n/d, not determined; ORCT, other round cell tumor; PA, partial agreement; RLH, reactive lymphoid hyperplasia; SC, scraping; STS, soft tissue sarcoma other than angiosarcoma; TN, true negative; TP, true positive; 2OP, second opion case.

(XLSX)

Click here for additional data file.

Cytological and corresponding histological diagnosis for each cytologically inconclusive case, with lesion distribution pattern.

Neoplastic subcategory is provided when applicable. Abbreviations: B, biopsy; BSTT, benign soft tissue tumor including angioma; CARC, carcinoma; EMH, extramedullary hematopoiesis; F, female; FNA, fine needle aspirate; HES, angiosarcoma; LYM, lymphoma; M, male; N, necropsy; NF, neutered female; NM, neutered male; NPL, neoplastic; NON-NPL, non-neoplastic; n/a, not applicable; STS, soft tissue sarcoma other than angiosarcoma; 2OP, second opinion case.

(XLSX)

Click here for additional data file.

7 Aug 2019

PONE-D-19-20551

Evaluation of cytological diagnostic accuracy for canine splenic neoplasms: an investigation in 78 cases on adherence to STARD guidelines.

PLOS ONE

Dear Dr. Gambini,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

Please address all Reviewer comments.

==============================

We would appreciate receiving your revised manuscript by Sep 21 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Douglas H. Thamm, V.M.D.

Academic Editor

PLOS ONE

Journal Requirements:

1.    When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Financial Disclosure section:

The authors received no specific funding for this work.

We note that one or more of the authors are employed by a commercial company: IDEXX Laboratories, Wetherby, West Yorkshire, UK

Please provide an amended Funding Statement declaring this commercial affiliation, as well as a statement regarding the Role of Funders in your study. If the funding organization did not play a role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript and only provided financial support in the form of authors' salaries and/or research materials, please review your statements relating to the author contributions, and ensure you have specifically and accurately indicated the role(s) that these authors had in your study. You can update author roles in the Author Contributions section of the online submission form.

Please also include the following statement within your amended Funding Statement.

“The funder provided support in the form of salaries for authors [insert relevant initials], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The specific roles of these authors are articulated in the ‘author contributions’ section.”

If your commercial affiliation did play a role in your study, please state and explain this role within your updated Funding Statement.

2. Please also provide an updated Competing Interests Statement declaring this commercial affiliation along with any other relevant declarations relating to employment, consultancy, patents, products in development, or marketed products, etc.

Within your Competing Interests Statement, please confirm that this commercial affiliation does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests) . If this adherence statement is not accurate and  there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include both an updated Funding Statement and Competing Interests Statement in your cover letter. We will change the online submission form on your behalf.

Please know it is PLOS ONE policy for corresponding authors to declare, on behalf of all authors, all potential competing interests for the purposes of transparency. PLOS defines a competing interest as anything that interferes with, or could reasonably be perceived as interfering with, the full and objective presentation, peer review, editorial decision-making, or publication of research or non-research articles submitted to one of the journals. Competing interests can be financial or non-financial, professional, or personal. Competing interests can arise in relationship to an organization or another person. Please follow this link to our website for more details on competing interests: http://journals.plos.org/plosone/s/competing-interests

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The paper is well-written, but the discussion should be shortened and rewritten to support the conclusions of the study rather than to reiterate results.

Minor suggestions:

Define STARD and DIMEVET in the abstract.

Add a space on line 39 of the abstract between of and 95.45%.

Line 48: Delete "fully".

Line 62: Delete "all".

Lines 69-70 Change to "...consisting mostly of controllable hemorrhage".

Line 90 Change to "...test with a high sensitivity..."

Line 141 Change to "Tissue samples for histopathology were fixed..."

Line 153 Change to "...(i.e., 10x objective lens) to..."

Line 171 Change "revised" to "reviewed"

Line 178 Change to "...other round cell tumors..."

Lines 229-230 Change to '...predetermined acceptability..."

Line 372 Should "p" be "p-value"?

Line 385 Change "Considered" to "Considering"

Line 417 Define "TC"

Line 488 Change "numerosity" to "number of cases evaluated"

Major question to consider:

Why were partial agreements included as true positives for statistical purposes when determining sensitivity of cytology in differentiating splenic tumor type subcategories?

Reviewer #2: The goals of this manuscript are to 1) systematically determine the diagnostic sensitivity and specificity of splenic aspiration in the diagnosis of neoplasia and 2) to ensure, to the best of the authors’ ability, that the study is carried out according to standardized guidelines for diagnostic accuracy studies (STARD). The authors have done a nice job of outlining existing weaknesses that exist in previous studies of canine splenic neoplasms and the role of cytology in making diagnoses. Original data are provided and methodology is provided in sufficient detail to understand how the study was completed.

Overall, this material provides some valuable data not specifically explored in previous manuscripts, and I think there are some interesting findings, such as differences in cytological diagnoses that are made from nodular or diffuse lesions. Conclusions are supported by provided data, and there are clear weaknesses in previous studies that are addressed by application of STARD criteria in this data set.

I have suggestions for some concepts and word choices that could benefit from additional clarification and/or expansion, as outlined below:

Concepts to clarify:

Line 59: I think this section could be worded a bit more strongly in order to emphasize the initial thought process behind your study. I’m not sure that a 52% survival post-splenectomy is “high.” I might say “fair” or simply list the overall survival rate, then go on to discuss complications related to splenectomy.

Paragraph, lines 76-89: This section could benefit from some reorganization for clarity. A few more details on the purpose of STARD guidelines would help readers understand what is unique about your study. For example, I might state the purpose of STARD and how guidelines came into existence (what are common weaknesses in studies of this kind that STARD addresses and improves?). Next, I would discuss where existing veterinary studies fall short. Finally, I would state that you are following recommended STARD criteria to perform cross-tabulation of test results against those of the reference standard to generate sensitivity and specificity data, absent in other studies.

Line 89: I’m not sure that I understand what a “deferred” cytological diagnosis is. Is that when cytology is equivocal until a histological diagnosis is available? If so, I think equivocal or provisional cover the spectrum of what is intended. If you mean “non-diagnostic,” say that instead.

Line 108: When you say “continuous series,” do you mean a series of diagnostic samples collected over 20 years, or a representative collection of splenic aspirate samples that include all pathologic states (neoplasia, hyperplasia, atypical hyperplasia, EMH, etc.)?

Line 128: Again, some expansion of exactly how your study is an improvement over others based on STARD guidelines would strengthen this section.

Paragraph, lines 161-173, Cytological diagnoses: Can you comment on the real world realism of removing diagnostic modifiers for these kinds of cases? I use diagnostic modifiers to improve the clarity of cytological reports, and in some cases, you might actually lose diagnostic information by being so stringent with reporting guidelines. I understand why the guidelines were set in this way for the study. However, readers may wonder what would happen if you had considered cases that originally included “most likely,” “suggestive of,” or “probable” HES at the time of diagnosis. Did you consider this, and based on the data you were able to evaluate, do you think this would this likely increase diagnostic sensitivity in “the real world” or would not actually have much of an effect?

Line 181: How often were all four anatomical pathologists in agreement for the reference standard, and how many cases were discarded? Did the discarded cases have interesting features pathologists should be aware of?

Line 220: You list a false positive in the conclusions, so it would be helpful to define “false positive” in this section alongside your other categories.

Line 492: You discuss that you were not able to completely follow STARD guidelines, but line 128 definitively states “The study was conducted following STARD guidelines.” Your disclaimer should probably come earlier in the manuscript so that there is not an apparent conflict in language.

Suggested language: Some language changes may improve readability. Here are some suggestions that may be helpful.

• Title: The way this is worded, it makes it sound like you are trying to determine how well your cases adhere to STARD guidelines, rather than determining diagnostic accuracy of cytology utilizing the STARD guidelines. Consider “Evaluation of cytological diagnostic accuracy for 78 canine splenic neoplasms using STARD guidelines.”

• Line 25: Cytology represents a useful diagnostic tool in the preliminary clinical 26 approach to canine splenic lesions, and may prevent unnecessary splenectomy.

• Line 42: When positive for neoplasia, cytology represents a useful diagnostic tool to rule in splenic neoplasia, prompting surgery independently from other diagnostic tests.

• Line 52: Use “fewer” instead of “less,” since the total quantity of splenic lesions diagnosed as HES or other is actually known.

• Line 69: Again use “fewer” instead of “less” since the number of total complications is known.

• Line 78: Word choice for clarity – consider “comprehensively assessed” instead of “addressed conjunctively”

• Line 166: Are reactive and inflammatory separate categories? If so, we need a comma after “reactive (including extramedullary hematopoiesis”)

• Line 169: remove the word “with.”

• Line 190: Move “only” to before “included” to improve flow.

• Line 237: Use “were” instead of “was”

• Line 230: Use “pre-determined” instead of “pre-determine”

• Line 312: Use “and indeed,”

• Line 385: Use “Considering”

• Line 426: Use “imbalance” instead of “unbalance”

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Lisa J. Schlein

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

17 Sep 2019

REBUTTAL LETTER – Manuscript PONE-D-19-20551

RESPONSES TO REVIEWER 1

1) “The paper is well-written, but the discussion should be shortened and rewritten to support the conclusions of the study rather than to reiterate results.”

As suggested, the discussion has been shortened and reworded to avoid as possible reiteration of results. Major changes were avoided to try to balance Reviewer 1 and Reviewer 2 observations regarding the discussion.

The following changes have been implemented:

- Former lines 399-408 (current lines 416-423) now run as follows: “According to overall accuracy and Cohen’s k values, cytology is not a reliable alternative to histopathology in the definitive diagnosis of splenic tumors in most cases. When compared with previous studies, our overall accuracy value (73.08%) laid in between the higher range of 83.87-100% [23-25] and the lower 38-69.7% range [8,31-33] reported in other studies. To allow comparison, the overall accuracy (intended as the sum of complete and partial diagnostic agreements) was calculated from the raw data of previously published caseloads [8,23-25,31-33] when not made explicit in the corresponding manuscript.”

- Former lines 421-427 (current lines 437-439) now run as follows: “Regarding the reliability of cytology in the diagnosis of specific tumor types, the lack of statistically significant differences between subcategories may be related to an imbalance in the number of cases for each tumor type.”

- Former lines 443-447 (current lines 455-457) now run as follows: “Although not statistically significant, our results paralleled those of previous reports identifying higher cytological accuracy in the diagnosis of diffuse compared to focal lesions [23,32,57].“

- Former lines 448-450 (current lines 458-459) now run as follows: “One false positive diagnosis of neoplasia (i.e. lymphoma) was included in this study.”

2) “Define STARD and DIMEVET in the abstract.”

Changed as suggested (current lines 31 and 33-34). Additionally, to maintain consistency with the Material and Methods section (current line 118), “Università degli Studi di Milano” was changed in “University of Milan” (current line 34).

3) “Add a space on line 39 of the abstract between of and 95.45%.”

Changed as suggested (current line 40).

4) “Line 48: Delete "fully".”

Changed as suggested (current lines 50-51).

5) “Line 62: Delete "all".”

Changed as suggested (current line 68).

6) “Lines 69-70 Change to "...consisting mostly of controllable hemorrhage".”

Changed as suggested (current line 76).

7) “Line 90 Change to "...test with a high sensitivity..."”

Changed as suggested (current line 104).

8) “Line 141 Change to "Tissue samples for histopathology were fixed..."”

Changed as suggested (current line 155).

9) “Line 153 Change to "...(i.e., 10x objective lens) to..."”

Changed as suggested (current lines 167-168).

10) “Line 171 Change "revised" to "reviewed"”

Changed as suggested (current lines 185).

11) “Line 178 Change to "...other round cell tumors..."”

Changed as suggested (current lines 192).

12) “Lines 229-230 Change to '...predetermined acceptability..."”

Changed as suggested (current lines 245-246).

13) “Line 372 Should "p" be "p-value"?”

“p” has been changed to “p-value” (current line 287). Additionally, the letter “p” has been changed to italic throughout the manuscript (current lines 389, 395 and 400) in agreement with papers previously published by PLOS ONE.

14) “Line 385 Change "Considered" to "Considering"”

Changed as suggested (current line 402).

15) “Line 417 Define "TC"”

“TC” was a typo. The sentence (current lines 431-433) now runs as follows: “…avoiding lag times and higher costs associated with application of diagnostic imaging techniques such as contrast-enhanced ultrasound and computed tomography (CT) [1,56].”

16) “Line 488 Change "numerosity" to "number of cases evaluated"”

Changed as suggested (current lines 495-496).

17) “Why were partial agreements included as true positives for statistical purposes when determining sensitivity of cytology in differentiating splenic tumor type subcategories?”

This approach is consistent with criteria applied in previous studies (see: Cohen M, Bohling MW, Wright JC, Welles EA, Spano JS. Evaluation of sensitivity and specificity of cytologic examination: 269 cases (1999-2000). J Am Vet Med Assoc. 2003;222: 964–7). Indeed, the major aim of the study was to investigate the diagnostic accuracy of cytology in the diagnosis of splenic neoplasm. Based on this aim, the main threshold set was the ability of cytology to correctly identify the neoplastic nature of a lesion, against degenerative and inflammatory conditions. Since the cytological diagnosis of a neoplasm is considered as sufficient to elect for splenectomy, cases with “partial agreement” (defined as agreement on the main ongoing pathological process, intended as neoplastic versus non-neoplastic) were included in the statistical analysis as true positives.

To improve clarity, the following changes have been made in the Material and methods section:

- Former lines 212-220 (current lines 226-236) now run as follows: “To evaluate the sensitivity of cytology in the diagnosis of specific tumor types, only those cases histologically confirmed as neoplastic were taken into account. The subcategories assigned to each cytological and corresponding histological sample were then compared. When cytological and histological diagnoses matched for both neoplastic categorization and tumor type subcategorization, the case was defined as “true positive with complete agreement”. When cytological and histological diagnoses matched for the neoplastic categorization but did not match for the tumor type subcategorization, the case was defined as “true positive with partial agreement”. When a histopathological diagnosis categorized as neoplastic corresponded to a cytological diagnosis categorized as non-neoplastic, the case was considered in disagreement and defined as “false negative case”.”

- Former lines 248-251 (current lines 264-268) now run as follows: “Sensitivity of cytology in differentiating splenic tumor types was defined as the ability of cytology to correctly identify as neoplastic a sample belonging to a specific neoplastic subcategory. Therefore, sensitivity for each tumor type was calculated as the sum of cases in complete and partial agreement (i.e. true positive cases) divided by the total number of cases with that specific neoplasm [31].”

Note that percentage of cases in complete agreement (grouped by each tumor type in the third column of Table 1 (labeled as “TP cases with complete agreement“) completely overlaps with the value of sensitivity of cytology in recognizing a neoplasm as belonging to a specific neoplastic subcategory.

RESPONSES TO REVIEWER 2

1) “Line 59: I think this section could be worded a bit more strongly in order to emphasize the initial thought process behind your study. I’m not sure that a 52% survival post-splenectomy is “high.” I might say “fair” or simply list the overall survival rate, then go on to discuss complications related to splenectomy.”

According to Reviewer 2’s request, former lines 57-62 have been rearranged and now run as follows (current lines 59-64): “While the two-month post-splenectomy survival rate is lower in dogs with HES (32%) compared to dogs with hematomas (83%), the general survival rate after splenectomy is 52%, regardless of underlying splenic pathology [3]. While this can be interpreted as fair survival data, a proportion of dogs (7.6%) will develop complications secondary to splenectomy because of thrombotic or coagulopathic syndromes [9].”

As suggested, complications following splenectomy have been included as follows (current lines 64-68): “Additional adverse effects following splenectomy in dogs have included peri- and post-operative ventricular arrhythmias [9-11], reduced blood filtration and renewal [12,13], impairment of humoral immune response [14], reduced immune-surveillance against bacteria and parasites [15-19], and higher incidence of gastric dilatation-volvulus [6,20-22].”

Reference list has been updated accordingly.

2) “Paragraph, lines 76-89: This section could benefit from some reorganization for clarity. A few more details on the purpose of STARD guidelines would help readers understand what is unique about your study. For example, I might state the purpose of STARD and how guidelines came into existence (what are common weaknesses in studies of this kind that STARD addresses and improves?). Next, I would discuss where existing veterinary studies fall short. Finally, I would state that you are following recommended STARD criteria to perform cross-tabulation of test results against those of the reference standard to generate sensitivity and specificity data, absent in other studies.”

According to Reviewer 2’s request, current lines 87-90 now run as follows: “STARD guidelines have been created to avoid incomplete reporting in diagnostic accuracy studies and to improve the general quality of the latter, reducing problems related to study identification, critical appraisal, and replication [34].“

According to Reviewer 2’s suggestion, the section was further expanded and current lines 98-102 now run as follows: “Application of STARD guidelines in the current study allowed cross-tabulation of cytological results (i.e. the index test) against those of histopathology (i.e. the reference standard) to generate sensitivity and specificity data [34], not retrieved from other former studies and useful to facilitate data comparability by potential future studies on the same topic.”

3) “Line 89: I’m not sure that I understand what a “deferred” cytological diagnosis is. Is that when cytology is equivocal until a histological diagnosis is available? If so, I think equivocal or provisional cover the spectrum of what is intended. If you mean “non-diagnostic,” say that instead.”

The terms came from previous reports. “Deferred” means that diagnoses were postponed awaiting histopathology. This does not mean neither equivocal nor provisional, but pending histology. This is somewhat an unusual way of dealing with cases, but this is how they were reported in previous reports. The closes term would be “provisional”. Thus “deferred” was erased and according to Reviewer 2’s suggestion, current lines 94-98 now run as follows: “In most reports, sensitivity, specificity, positive and negative predictive values of splenic cytology are not calculated and cannot be properly estimated [8,23-25,31-33] because caseloads simultaneously include multiple species [24,25,31-33], multiple tissues and organs [31-33], or “equivocal” or “provisional” cytological and histological diagnoses [23,24,33].”

4) “Line 108: When you say “continuous series,” do you mean a series of diagnostic samples collected over 20 years, or a representative collection of splenic aspirate samples that include all pathologic states (neoplasia, hyperplasia, atypical hyperplasia, EMH, etc.)?”

Using the term “continuous series” we meant a series of diagnostic samples collected over 20 years. Seen this observation and and according with STARD guidelines (including the specific explanation and elaboration document that can be found at: https://www.ncbi.nlm.nih.gov/pubmed/28137831), we substituted “continuous” with “consecutive” to improve clarity. Current lines 120-121 now run as follows: “A consecutive series of canine splenic cytologies was obtained.”

5) “Line 128: Again, some expansion of exactly how your study is an improvement over others based on STARD guidelines would strengthen this section.”

According to Reviewer 2’s suggestion, current lines 141-142 now run as follows: “To improve data completeness, transparency, and reproducibility, the study was conducted following the STARD guidelines [34] to the best of authors’ ability.” To avoid excessive expansion of the Materials and methods section, the explanation of usefulness and improvement provide by STARD guidelines were included in the Introduction as Reviewer 2 suggested (see answer to comment #2 by Reviewer 2).

6) “Paragraph, lines 161-173, Cytological diagnoses: Can you comment on the real world realism of removing diagnostic modifiers for these kinds of cases? I use diagnostic modifiers to improve the clarity of cytological reports, and in some cases, you might actually lose diagnostic information by being so stringent with reporting guidelines. I understand why the guidelines were set in this way for the study. However, readers may wonder what would happen if you had considered cases that originally included “most likely,” “suggestive of,” or “probable” HES at the time of diagnosis. Did you consider this, and based on the data you were able to evaluate, do you think this would this likely increase diagnostic sensitivity in “the real world” or would not actually have much of an effect?”

This is a truly interesting point. Despite the setting of a scientific manuscript requires rigid categorization to perform clear cut statistics, we agree with Reviewer 2, and we usually include diagnostic modifiers in our routine cytological reports to express the degree of certainty in the diagnosis and/or the interpretation of the examined cytological samples. Removing such modifiers might, in some cases, implicate a loss of diagnostic information, intended as information that might have a considerable influence on clinical decision-making process following the cytological report. Therefore, diagnostic modifiers are useful and for sure they should not be avoided in real life scenarios.

On the other hand, according to different reports, the use of modifiers and their meaning are not standardized, becoming subject to interpretation (see Sharkey LC, Dial SM, Matz ME. Maximizing the Diagnostic Value of Cytology in Small Animal Practice. Vet Clin North Am - Small Anim Pract. 2007;37: 351–372; and Christopher MM, Hotz CS. Cytologic diagnosis: expression of probability by clinical pathologists. Vet Clin Path. 2004;33(2): 84-95). In this specific context, the use of non-standardized parameters would introduce variability that could reduce or void the strength of the statistical analysis. Therefore, considered that the aim our work was to investigate the agreement of cytology (i.e. the index test) with histopathology (i.e. the reference test) disregarding the clinical history (another unreal scenario that we are asked for in papers and in board exams), and considered that an objective statistical analysis could not be performed if modifiers were included, we opted to remove modifiers during the review process of cytological samples. This choice is in line with other studies concerning diagnostic accuracy of cytology following STARD guidelines (see Berzina I, Sharkey LC, Matise I, Kramek B. Correlation between cytologic and histopathologic diagnoses of bone lesions in dogs: a study of the diagnostic accuracy of bone cytology. Vet Clin Pathol. 2008;37: 332–338.).

Regarding the sentence stating that “readers may wonder what would happen if you had considered cases that originally included “most likely,” “suggestive of,” or “probable” HES at the time of diagnosis”, we fear that some kind of misunderstanding happened. Indeed, as stated in the Materials and methods section (current lines 116-121), samples included in the study belong to a consecutive series of archived cytological slides, mined using keywords only concerning the species and the organ from which cytological samples were obtained. This means that the original cytological diagnosis reported in our electronic archives was not a selection criterion for inclusion of cases in the study. Therefore, all the cytological samples conforming to the inclusion criteria (i.e. availability for review of both cytological slides and histopathological sample of the same lesion) were included in the study, disregarding the fact that the original diagnosis might be reported as “most likely”, “suggestive of”, or “probably” indicative of a lesion. Therefore, seen our selection criteria, the sensitivity should probably not change substantially.

7) “Line 181: How often were all four anatomical pathologists in agreement for the reference standard, and how many cases were discarded? Did the discarded cases have interesting features pathologists should be aware of?”

Agreement among anatomical pathologists was reached for all the cases included in the study, as already stated in former lines 298-299 (current lines 315-316).

Thanks to comment #7 by Reviewer 2, we found an error that has now been corrected throughout the manuscript. Specifically, former lines 149-150 reported that a fourth anatomical pathologist was involved in the study, but pathologists were actually three The following changes have been applied to correct the mistake:

- Current lines 161-164 (former lines 147-150): “All cytological and histopathological samples were independently reviewed in a blinded fashion by three cytologists (MC - ECVP, GG - ECVCP, MG - resident) and by three anatomical pathologists (AF – ECVP, PR - ECVP, MT - resident), respectively.”

- Current lines 195-197 (former lines 181-183): “Cases were included only when the three anatomical pathologists were in agreement because histopathology served as the diagnostic reference standard to evaluate cytological diagnostic accuracy.”

8) “Line 220: You list a false positive in the conclusions, so it would be helpful to define “false positive” in this section alongside your other categories.”

The definition of false positive cases is already available in the manuscript in the previous sentence lines (current lines 222-223). For better clarity, the text was modified (current lines 226-236) and now run as follows: “To evaluate the sensitivity of cytology in the diagnosis of specific tumor types, only those cases histologically confirmed as neoplastic were taken into account. The subcategories assigned to each cytological and corresponding histological sample were then compared. When cytological and histological diagnoses matched for both neoplastic categorization and tumor type subcategorization, the case was defined as “true positive with complete agreement”. When cytological and histological diagnoses matched for the neoplastic categorization but did not match for the tumor type subcategorization, the case was defined as “true positive with partial agreement”. When a histopathological diagnosis categorized as neoplastic corresponded to a cytological diagnosis categorized as non-neoplastic, the case was considered in disagreement and defined as “false negative case”.”

9) “Line 492: You discuss that you were not able to completely follow STARD guidelines, but line 128 definitively states “The study was conducted following STARD guidelines.” Your disclaimer should probably come earlier in the manuscript so that there is not an apparent conflict in language.”

According to Reviewer 2’s request, current lines 141-142 now run as follows: “To improve data completeness, transparency, and reproducibility, the study was conducted following the STARD guidelines [34] to the best of authors’ ability.”

10) “Title: The way this is worded, it makes it sound like you are trying to determine how well your cases adhere to STARD guidelines, rather than determining diagnostic accuracy of cytology utilizing the STARD guidelines. Consider “Evaluation of cytological diagnostic accuracy for 78 canine splenic neoplasms using STARD guidelines.””

The title (current lines 4-6) has been changed as suggested.

11) “Line 25: Cytology represents a useful diagnostic tool in the preliminary clinical 26 approach to canine splenic lesions, and may prevent unnecessary splenectomy.”

Changed as suggested (current lines 25-27).

12) “Line 42: When positive for neoplasia, cytology represents a useful diagnostic tool to rule in splenic neoplasia, prompting surgery independently from other diagnostic tests.”

Changed as suggested (current lines 43-45).

13) “Line 52: Use “fewer” instead of “less,” since the total quantity of splenic lesions diagnosed as HES or other is actually known.”

Changed as suggested (current lines 54).

14) “Line 69: Again use “fewer” instead of “less” since the number of total complications is known.”

Changed as suggested (current lines 75).

15) “Line 78: Word choice for clarity – consider “comprehensively assessed” instead of “addressed conjunctively””

Changed as suggested (current lines 84).

16) “Line 166: Are reactive and inflammatory separate categories? If so, we need a comma after “reactive (including extramedullary hematopoiesis”)”

Changed as suggested (current lines 179-180).

17) “Line 169: remove the word “with.””

Changed as suggested (current line 183).

18) “Line 190: Move “only” to before “included” to improve flow.”

Changed as suggested (current lines 203-204).

19) “Line 237: Use “were” instead of “was””

To improve clarity, the sentence was reworded as follows (current lines 251-254): “Ninety-five percent (95%) confidence interval was calculated for each of the above mentioned indices of diagnostic test accuracy using a web-based application (MEDCALC - https://www.medcalc.org/calc/diagnostic_test.php).”

20) “Line 230: Use “pre-determined” instead of “pre-determine””

Changed as suggested (current lines 245-246).

21) “Line 312: Use “and indeed,””

Changed as suggested (current line 329).

22) “Line 385: Use “Considering””

Changed as suggested (current lines 402).

23) “Line 426: Use “imbalance” instead of “unbalance””

Changed as suggested (current lines 439).

Additional typos evidenced during the review were corrected as follows:

- Former line 59 (current line 61): double space preceding “While” deleted.

- Former line 105 (current line 118): “University of Milano” changed to “University of Milan”

- Former line 153 (current line 167): comma between “i.e.” and “10x” deleted.

- Former line 287 (current line 304): comma between “represented” and “in” substituted by semicolon.

- Former lines 340-342 (current lines 357-359): percentages were wrong, deriving from a former draft of the manuscript, including inconclusive cases and lacking reference to Table 3. The sentence now runs as follows: “Following the tabulation of cytological and histological diagnoses (S1 Table), 36 cases (46.15%) were classified as TP, 21 (26.92%) were TN, 20 (25.64%) were FN, and 1 (1.28%) was a FP (Table 3).”

- Former line 391 (current line 408): “of” added between “diagnosis” and “canine”.

- Former line 492 (current line 501): additional space preceding “Unfortunately” deleted.

Submitted filename: PONE-D-19-20551 - Response to Reviewers.doc

Click here for additional data file.

4 Oct 2019

PONE-D-19-20551R1

Evaluation of cytological diagnostic accuracy for canine splenic neoplasms: an investigation in 78 cases on adherence to STARD guidelines.

PLOS ONE

Dear Dr. Gambini,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

Please address the Reviewer's additional minor grammatical comments.

==============================

We would appreciate receiving your revised manuscript by Nov 18 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Douglas H. Thamm, V.M.D.

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript is improved. No further comments. The manuscript is improved. No further comments. The manuscript is improved. No further comments. The manuscript is improved. No further comments.

Reviewer #2: Thank you for your thoughtful revision of this manuscript. The clarity of your study is markedly improved and I think the study is interesting and important.

I have some minor grammatical suggestions that I think will improve the readability of the manuscript, and there are two statements that I think could be clarified:

1. Remove "of" in line 57.

2. Lines 98-102 are not clear. Consider:

"Application of STARD guidlines in the current study allowed cross-tabulation of cytological results (i.e. the index test) against those of histopathology (i.e. the reference standard) to generate sensitivity and specificity data. These data have not been included in previous studies and will be useful for future researchers comparing diagnostic methods for canine splenic neoplasms."

3. Line 106, remove "a" before neoplasia (or say "a neoplasm")

4. Line 132, remove "considered as"

5. Lines 218, 220, 222: add "the" before each category name (the True Positive (TP), etc.)

6. Line 305, add "The" before "Time interval"

7. Line 326: Either add an "and" before therefore, or change the comma to a semicolon.

8. Line 332: a dash is missing in "non-neoplastic."

9. Line 368: change "ratio" to "ratios"

10. Line 444: This is not clear. Consider "Moreover, identification of vascular tumors (angiomas and HES) among false negative cases is not surprising since the architecture of these tumors often leads to significant peripheral blood contamination in aspirates."

11. Line 450: Change "by" to "of"

12: Line 451: Add "a" before low mitotic rate.

13: Line 462: Change the comma to a semicolon.

14: Line 471: Change "be considered also" to "also be considered."

15: Line 509: change "neoplasia, however high" to "neoplasia; however, high"

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

22 Oct 2019

REBUTTAL LETTER – Manuscript PONE-D-19-20551

RESPONSES TO REVIEWER 2

1) “Remove "of" in line 57.”

Changed as suggested (current line 57).

2) “Lines 98-102 are not clear. Consider: "Application of STARD guidlines in the current study allowed cross-tabulation of cytological results (i.e. the index test) against those of histopathology (i.e. the reference standard) to generate sensitivity and specificity data. These data have not been included in previous studies and will be useful for future researchers comparing diagnostic methods for canine splenic neoplasms."”

Changed as suggested (current lines 98-103).

3) “Line 106, remove "a" before neoplasia (or say "a neoplasm")”

Changed as suggested (current line 107).

4) “Line 132, remove "considered as"”

Changed as suggested (current lines 132-133).

5) “Lines 218, 220, 222: add "the" before each category name (the True Positive (TP), etc.)”

Changed as suggested (current lines 219, 221, 223, 225).

6) “Line 305, add "The" before "Time interval"”

Changed as suggested (current line 306).

7) “Line 326: Either add an "and" before therefore, or change the comma to a semicolon.”

Changed as suggested (current line 327).

8) “Line 332: a dash is missing in "non-neoplastic."”

Changed as suggested (current line 333).

9) “Line 368: change "ratio" to "ratios"”

Changed as suggested (current line 369).

10) “Line 444: This is not clear. Consider "Moreover, identification of vascular tumors (angiomas and HES) among false negative cases is not surprising since the architecture of these tumors often leads to significant peripheral blood contamination in aspirates."”

Changed as suggested (current lines 445-447).

11) “Line 450: Change "by" to "of"”

Changed as suggested (current line 451).

12) “Line 451: Add "a" before low mitotic rate.”

Changed as suggested (current line 452).

13) “Line 462: Change the comma to a semicolon.”

Changed as suggested (current line 463).

14) “Line 471: Change "be considered also" to "also be considered."”

Changed as suggested (current line 472).

15) “Line 509: change "neoplasia, however high" to "neoplasia; however, high"”

Changed as suggested (current line 510).

Additional typos evidenced during the review were corrected as follows:

- Former line 45 (current line 45): double space preceding “Conversely” deleted.

- Former line 50 (current line 50): extra space preceding “Ultrasonographic” deleted.

- Former line 53 (current line 53): double space preceding “is” deleted.

- Former line 146 (current line 145): double space preceding “were” deleted.

- Former line 174 (current line 173): extra comma following “e.g.” deleted.

- Former line 182 (current line 181): double space preceding “specimens” deleted.

- Former line 190 (current line 189): extra space preceding “Neoplastic” deleted.

- Former line 347 (current line 346): double space preceding “Of” deleted.

- Former line 364 (current line 363): “was found, however severe marginal” reworded as “was found; however, severe marginal”.

- Former line 418 (current line 417): double space preceding “values” deleted.

- Former line 424 (current line 423): double space preceding “when” deleted.

- Former line 475 (current line 473): extra comma following “tumors” deleted.

- Former line 747 (current line 745): double space preceding “n/a” deleted.

Submitted filename: PONE-D-19-20551R2 - Response to Reviewers.doc

Click here for additional data file.

25 Oct 2019

Evaluation of cytological diagnostic accuracy for canine splenic neoplasms: an investigation in 78 cases using STARD guidelines.

PONE-D-19-20551R2

Dear Dr. Gambini,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Douglas H. Thamm, V.M.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

31 Oct 2019

PONE-D-19-20551R2

Evaluation of cytological diagnostic accuracy for canine splenic neoplasms: an investigation in 78 cases using STARD guidelines.

Dear Dr. Gambini:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Douglas H. Thamm

Academic Editor

PLOS ONE