Clinico-Biochemical Parameters, Treatment, and Prognostic Indicators of Peritonitis in Buffaloes

.


INTRODUCTION
Peritonitis means inflammation of the peritoneal cavity and is one of the main causes of animal deaths. Peritonitis may be localized or generalized and may occur as a primary or secondary part of a specific disease. As a primary disease, it commonly results from injury of the serosal surface of the alimentary tract [1]. Less commonly, there is perforation of the abdominal wall from the exterior by penetrating foreign bodies or intra-peritoneal injection of non-sterile solutions/irritating substances, or exploratory laparotomy [1,2]. The perforated abomasal ulcers may outcome of this disease has not been evaluated previously, and the data about prognostic indicators are lacking. Thus, the present study described signalment, clinical findings, hemato-biochemical changes, response to medical treatment, and outcome in buffalo cases suffering from peritonitis. Another objective of this study was to determine prognostic factors associated with the outcome of the disease, fetal survival, and milk production in the current and subsequent lactation.

Selection of Animals and Criteria for Diagnosis of Peritonitis
The study was conducted on 25 buffalo cases suffering from peritonitis. The study was performed in compliance with the institutional ethical guidelines. All owners gave their consent for the animals to be included in the study and undergo the testing procedures. The diagnosis of peritonitis was made based on clinical examination and peritoneal fluid examination [5,11]. Before predicting the diagnosis, traumatic reticuloperitonitis and perforated abomasal ulcer were ruled out by physical examination and other ancillary tests like radiography and ultrasonography.

Signalment, Anamnesis, and Physical Examination
Data related to age, sex, parity, calving, pregnancy status, and duration of illness were recorded in all the buffaloes, along with the detailed history of the disease. Each animal was subjected to a general physical examination and a special examination of the gastrointestinal system [12,13]. As previously described by us, each animal was subjected to rectal examination [14]. The rectal examination was performed for rectal mucosa, the quantity of feces in the rectum, consistency of feces, fecal color, rumen size and fill, movement of the hand, intestines, and other abnormal findings. Fecal samples were examined for fecal occult blood by commercially available HEMOSPOT test cards (Crest Biosystems, A Division of Coral Clinical Systems, Goa, India) based on the standard Guaiac method. The fecal samples negative for occult blood were re-examined after 24 and 48 hours [15].

Laboratory Analyses and Diagnostic Procedures
The blood samples (2ml) collected in EDTA coated vials (Accuvete-PLUS, Quantum Biologicals Pvt. Ltd.) were used for the determination of hematological parameters by automatic hematology analyzer (ADVIA ® 2120 Hematology system, Siemens Healthcare Diagnostics Inc., USA). The differential leukocytic count was performed manually in a blood smear stained by Leishman stain [16]. A thorough examination of stained blood smear was also done to determine left shift and toxic changes in neutrophils if any. Serum was used to estimate biochemical parameters except for glucose and fibrinogen. For glucose and fibrinogen estimation, blood samples were collected in sodium fluoride and sodium citrate coated vials (Accuvote Disposables), respectively. The concentration of total globulin was calculated by subtracting the albumin concentration from the total protein concentration. Fibrinogen was estimated with the help of a handheld refractometer [16]. The fibrinogen ratio was calculated by subtracting the fibrinogen value from the total protein concentration and then dividing that value by fibrinogen concentration. VITROS DT-II Chemistry system (Ortho-Clinical Diagnostics, Johnson and Johnson Company) was used to estimate other biochemical parameters.
Rumen fluid samples were collected as described by Hussain et al. [13]. Rumen fluid samples were evaluated for pH and chloride concentration. Rumen chloride was estimated after filtering the rumen fluid through a double layer muslin cloth and then centrifuging. Abdominocentesis was done in all the buffaloes [11]. Representative peritoneal fluid samples were collected in EDTA coated vials by free hand centesis using a 16-18 gauge, 1.5-inch-long hypodermic needle. The peritoneal fluid was evaluated for volume, color, consistency/turbidity, and specific gravity. Specific gravity was measured with a handheld refractometer. Peritoneal fluid samples were subjected to cytological examination [11]. White blood cell count (WBC), neutrophil count, and total protein were estimated by the same methods as used for blood. Reticular radiographs were taken in right lateral recumbency as described [17]. Ultrasonography of the abdomen was performed with Wipro GE Logiq III Expert/ Aloka 500V in real time B mode and B+M mode with a 3.5/5.0 MHz convex transducer by standard procedure [18].

Treatment and Follow-Up
Out of 25 buffaloes, one died on the day of presentation, while the owner of another buffalo was not willing to treatment due to a poor prognosis. The rest 23 buffaloes were treated with intravenous fluids, antibiotics, and other supportive treatments. Depending upon the severity of hemato-biochemical and peritoneal fluid alterations, the antibiotics were administered as per our clinic's protocol [3,19,20]. Treatment included intravenous administration of 5-10 liters of normal saline and 5 liters of dextrose normal saline once daily for 3-5 days, one dose of intravenous calcium therapy (450ml of MIFEX, Novartis India Limited, India), 200mL of liquid POTKLOR (containing 20 g of potassium chloride for 3 days) orally, 8-10 ml intramuscular injection of belamyl (B complex liver extract with vitamin B12 injection, Zydus AH, India) for 3-8 days and 100 g charcoal (as anti-bloat agent) once daily till resolution of tympany. Meloxicam (0.5mg/kg, injection Melonex, Intas Pharmaceuticals Ltd, India), administered once daily for three days. Ceftiofur (2.2mg/kg once daily, injection Xnel, Zoetis India Ltd), ampicillin (20mg/kg twice daily, injection Roscillin, Ranbaxy Laboratories Limited), gentamicin (2.2mg/kg once daily, injection E-Gmicin Vet, Facmed Pharmaceuticals Private Ltd, India) and enrofloxacin (5mg/kg once daily, Injection Fortivir, Virbac Animal Health, India) were administered intramuscularly while metronidazole (20mg/kg twice daily, injection Flagyl, Hindustan Agencies, India) was administered intravenously. Follow-up information was obtained from the owners by telephonic conversation. The owners were contacted at 3 days intervals for 15 days, weekly till 2 months, then at regular intervals of three months for a period of 24 months. The owners were questioned about the general health status of the buffaloes and the born calves (for pregnant buffaloes), any effect on milk yield in current and subsequent lactation, and any recurrence of disease in the next pregnancy.

Statistical Analysis
The categorical/qualitative data were presented as frequency. The quantitative data were expressed as mean ± standard deviation. The student's t-test was used to analyze the significance of the difference between survivors and non-survivors. A Chi-square test was used to compare binary variables between survivors and non-survivors. The significance level was kept at P<0.05 for all statistical procedures.

History and Clinical and Diagnostic Imaging Findings
The buffaloes suffering from peritonitis were 2-10year-old females (mean=6.16±1.88). Sixteen buffaloes (64 %) were in 2 nd -4 th lactation, two (8%) were heifers, four (16%) were in the third lactation, and three (12%) were in 5 th -7 th lactation. The majority of buffaloes (60%) were in different stages of pregnancy (12 were > 6month pregnant), and 40% were non-pregnant (six had calved <3 months back, three had calved 3.5-10 months back, and one was heifer). The duration of the illness ranged from 2 to 30 days (mean 11.92±8.1, median=10 days), being less than a week in 32%, 1-2 weeks in 32%, >2-3 weeks in 28%, and 30 days in 8% buffaloes. The history and clinical examination findings are presented in Tables 1-2. Sixty percent of buffaloes had complete anorexia, and others had chronic inappetence. The majority of buffaloes had a history of fever followed by reduced or scanty defecation. The fall in milk yield was sudden in 86% of lactating buffaloes, and 40% of buffaloes had at least one episode of tympany. Colic was observed in six (24%) buffaloes. Before referral to our clinic, all except one buffalo had been treated symptomatically. Clinically, the majority of buffaloes were depressed, dehydrated, and had congested mucous membranes, tachypnea, and atonic or hypomotile rumen with variable rumen consistency ( Table  2). Pneumoperitoneum was observed in three buffaloes, and fluid splash on swinging palpation of the abdomen was recorded in three buffaloes. Rectal examination revealed reduced or scanty pasty consistency feces, sticky rectal mucosa, and normal intestines in the majority of buffaloes ( Table 3). On rectal examination, peritoneal fluid was appreciable in two buffaloes, and rough peritoneum was palpable in one buffalo. Vomiting was observed in two buffaloes (8%), and a fecal occult blood test (FOBT) was positive in ten (40%) buffaloes.

Hematobiochemical Findings
The hematological and biochemical parameters are presented in Table 4. The mean hemoglobin and lymphocyte count were within the normal reference range, while hematocrit, WBC, and neutrophil count were higher than the reference values. The WBC was 7400-10000 in six (24%), 10001-12000 /µL in three (12%), 12001-16000/µL in ten (40%), 16001-20000 in four (16%) and >20000/µL in two (8%) buffaloes. The toxic changes in neutrophils were observed in 76% of cases (mild in eight, moderate in eight, and severe in three), and left shift was observed in 20 cases (80%), being mild in six (24%), moderate in three (12%) and marked in eleven (44%). The mean levels of cholesterol, triglycerides, glucose, total bilirubin, aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma-glutamyl transferase (GGT), fibrinogen, blood urea nitrogen (BUN), creatinine, lactate, and rumen chloride were higher than the reference values, while albumin, fibrinogen ratio, potassium, chloride, calcium and phosphorus were lower than the reference values. The other biochemical parameters were within the normal reference range values.

Peritoneal Fluid Examination Findings
The collected peritoneal fluid ranged from 4 ml to >2 liters. More than one liter of turbid fluid was drained from the post xiphoid site in two buffaloes. In another buffalo, foul smelling >2litre pus, like fluid was obtained from every site tapped in the abdomen. In one more buffalo that became recumbent soon after presentation, there was massive peritoneal fluid, and free-flowing peritoneal fluid could be collected on paracentesis of even lateral abdomen. The peritoneal fluid was crystal clear in 44% and yellow in 32% cases ( Table 5). The consistency was watery in 40% and turbid in 60% of cases. Peritoneal fluid was off-smelling in three (12%) buffaloes. The specific gravity was increased in 80% of cases, while WBC was <5000 in 60% of cases. The total cell count of peritoneal fluid ranged from 350 to >100000/µL. The majority (80%) of buffaloes had a neutrophil count of >50%, but the total protein was > 3g/dL in 60% of cases only. The abnormal findings on cytological examination of peritoneal fluid included neutrophilia, variable numbers of degenerated neutrophils, the aggregate of macrophages, bacteria, and gut contents (Figures 1-3). In some cases, fibroblasts and mesothelial cells were also observed. Microscopic examination of peritoneal fluid smear revealed bacteria in 60% of cases. Bacteria were cocci in six (24%) samples, bacilli in three (12%), both cocci and bacilli in six (24%), and additionally other shapes in two samples. The bacteria were seen as free or phagocytosed by neutrophils (Figures 4-5). Furthermore, bacteria were observed in isolated fields,  as chains or as aggregates. Severe sepsis (n=4) was characterized by massive neutrophilia and markedly degenerated neutrophils with engulfed bacteria along with a high count of activated macrophages and mesothelial cells (Figure 6). The presence of gut contents in the peritoneal fluid was observed in three cases. The gut contents included plant fiber and actinomycete-like organisms (Figures 7-8). Adhesive peritonitis was diagnosed in four (Figure 9) and overwhelming peritonitis in two cases (Figure 10).  The buffalo had anorexia, fever, and a sudden reduction in milk yield from 2 days. There was neutrophilic leukocytosis, hypoproteinemia, and hyperlactatemia. There was a slight response to treatment, but the buffalo died after 32 days of treatment.  . The buffalo had anorexia, fever, and colic from 12days, atonic rumen, heart rate of 71bpm, WBC of 13260/µL, a moderate degree of toxic changes in neutrophils and marked left shift, low albumin (2.9g/dL), azotemia (BUN=68 mg/dl, creatinine=2.6mg/dL), hypokalemia (2.4mmol/L) and severe hyperlactatemia (9.4mmol/L). The buffalo died after 5 days of treatment.

Treatment and Follow Up
Medical treatment was initiated in 23 buffaloes, which included the administration of intravenous fluids, antibiotics, and other supportive therapy. Four (17%) buffaloes died within three days of the treatment, six (26%) after 4-7 days, and one after 32 days of treatment. The buffalo, whose owner was not willing to treatment, died after a period of 14 days. Six (26%) buffaloes recovered after 6-9 days, five (22%) after 11-15 days, and one buffalo recovered after 23 days ( Table 6). All the survivors were healthy at the time of the last feedback. The pregnant buffaloes had calved normally, and some non-pregnant buffaloes had become pregnant and then calved normally during follow-up. After recovery, there was no significant effect on the milk yield in the current or subsequent lactation.

Survival Analysis and Prognostic Indicators
The age, duration of illness, clinical signs, and clinical examination parameters did not differ significantly (p<0.05) between survivors and nonsurvivors. None of the buffaloes with pneumoperitoneum survived. Hence peritoneal tympany could be regarded as a poor prognostic sign. Rumen motility was absent in 13 buffaloes, and only two of them recovered after a period of 12 and 23 days,    The physical characteristics and quantitative parameters of peritoneal fluid did not differ significantly (P<0.05) between survivors and non-survivors. Out of 15 buffaloes with septic peritonitis, only five survived. The non-survivors had a greater degree of degenerative changes in peritoneal fluid neutrophils than the survivors. None of the buffaloes with diffuse peritonitis, severe sepsis, overwhelming peritonitis, and the presence of gut contents in peritoneal fluid survived.

DISCUSSION
Peritonitis was diagnosed based on peritoneal fluid examination supported by clinical examination [11,21]. Due to melena, loss of defection, and colic, the important abdominal disorders to be ruled out were perforated abomasal ulcer, abomasal volvulus, intussusception, and hemorrhagic bowel syndrome [1,22]. Hemorrhagic bowel syndrome and abomasal volvulus have not been reported in the area of this study, but intussusception and abomasal ulceration are prevalent [4,6,23]. These disorders were ruled out by clinical evaluation and other ancillary tests like ultrasonography. However, type 1 abomasal ulceration could not be ruled out with certainty, and it was assumed to be secondary ulceration. Type 1 ulcers are reported to cause no or mild clinical signs [15,24] and hence were less likely to result in the clinical signs observed in the present study. Similar to an earlier study [15], retesting of fecal samples after storage did not increase their positivity for occult blood. Like the present study, other abdominal disorders have been reported to result in secondary melena like traumatic reticulo-peritonitis [20], traumatic pericarditis [25], and reticular diaphragmatic hernia [26]. These disorders were also ruled out in the present study. Type 3 abomasal ulcer could not be ruled out as we did not examine the abomasum directly.
The majority of buffaloes were pregnant (60%), and 16% had calved <2 months back. So, pregnancy seemed to be a greater risk factor than recent calving in the present study. The clinical findings were nonspecific and resembled other abdominal disorders like traumatic reticuloperitonitis, abomasal ulceration, and reticular diaphragmatic hernia [4,26,27], but were similar to published findings in cattle and buffaloes [5,11,28]. It is important to mention that the signs of complete anorexia, colic, tympany, loss of defecation, depression, dehydration, tachycardia, and tachypnoea were less pronounced than reported for peritonitis in cattle [11]. Further, in contrast to peritonitis in cattle [11] higher percentage of buffaloes in this study had normal rumen consistency and normal abdominal contour. The main complaint by the owners in the majority of cases was fever and scanty defecation. History of fever, a common sign in the initial stages of peritonitis, corroborates with observations of Hosgood and Salisbury [29]. The fecal output can be normal or reduced and stale in peritonitis because of prolonged transit [10,11]. Before referral to our clinic, the majority of animals had been treated with antibiotics along with rumen tonics, liver tonics, anti-bloat compounds, calcium infusion, and/or purgatives. It is also worthwhile to mention that in some cases, previous treatment had resulted in mild improvement in appetite and fecal output, but clinical recovery was not complete. Signs of abdominal pain were observed less frequently than other abdominal disorders in bovines like traumatic reticuloperitonitis, intestinal obstruction, caecal dilatation, omasal impaction, abomasal ulcers, and abomasal volvulus [4,6,10,13,22,27,30]. Pneumoperitoneum was characterized by the presence of gas on tapping of the paralumbar fossa without puncturing the rumen. Pneumoperitoneum not associated with surgery has been reported to be indicative of bacterial peritonitis and the presence of gas-producing bacteria [10]. Like peritonitis, vomiting has been reported to be an infrequent sign in other abdominal disorders like intestinal obstruction [6] and reticular diaphragmatic hernia [26].
The rumen-reticular motility depends upon the excitatory and inhibitory inputs to the gastric center [1,10]. The suspended or reduced rumen motility may be attributed to toxemia, peritoneal adhesions, or generalized gastrointestinal tract atony. Reduced ruminal motility may also be attributed to the observed hypocalcemia. Hypomotility corroborates the earlier findings [11]. In contrast to the present study, Kumar et al. [31] reported normal rumen motility in bovines with diffuse peritonitis. The mushy rumen consistency may be attributed to prolonged ingesta retention in the rumen due to ineffective or absent rumen contractions. Rectal examination is a valuable diagnostic procedure in bovine gastroenterology and helps to diagnose many gastrointestinal disorders in bovines like rumen impaction intestinal obstruction, and caecal dilatation [6,14,30], and could also help in the subjective assessment of omasal impaction [13]. In this study, the rectal findings were not diagnostic except in three buffaloes.
The main hematological findings were increased hematocrit and neutrophilia or neutrophilic leucocytosis with left shift and the presence of toxic neutrophils and were similar to earlier reported findings in peritonitis [1,11,32]. The changes in hemogram may be attributed to an immunological response to peritonitis [33]. The mean WBC count was higher than reported for other bovine gastrointestinal disorders like rumen impaction, reticular diaphragmatic hernia, intestinal obstruction, caecal dilatation, and vagal indigestion [6,14,26,30,34] but comparable to that reported for omasal impaction, peritonitis, and traumatic pericarditis [11,13,25].
The increased total bilirubin and other liver enzymes indicated impaired hepatic function and/or hepatobiliary circulation. The possible cause of higher liver enzymes could be degenerative lesions in the peritoneum and liver. The probable explanation for increased glucose is that all these animals were transported to our clinic, which could have triggered the release of stress-related hormones such as cortisol and adrenalin, increasing blood concentrations of glucose [35]. In addition, the stress of the disease could have also resulted in hyperglycemia. The lower cholesterol level may be attributed to inflammatory response and hepatic malfunction impairing the lipid transport [36][37][38].
Change in serum proteins following tissue destruction and inflammation is expected [39]. Leukogram and peritoneal fluid alterations indicated an inflammatory response in our study. So the expected change in total protein would have been an increase in its concentration. But on the contrary, the total protein concentration was within the normal reference range (5.7-8.1 g/dL). Albumin level was towards the lower side of the reference range and resembled earlier findings in diffuse peritonitis of buffaloes [5]. There could be two possible reasons for lower total protein and albumin. The First could be third space loss, and another cause could be reduced synthesis of albumin (negative acute phase protein) by the liver. The noted hemoconcentration, yet a tendency for normal total protein, suggested significant third space losses [6]. It has been shown that during inflammation, the synthesis of positive acute phase proteins in the liver increases while the synthesis of negative acute phase proteins such as albumin decreases, and hence albumin concentration may fall in inflammatory and infectious diseases [39,40].
The fibrinogen was elevated due to peritoneal inflammation. The fibrinogen ratio in bovines gives a more realistic picture of increased fibrinogen in inflammatory conditions and rules out any alterations due to dehydration. The lower fibrinogen ratio indicated a marked increase in fibrinogen [16]. In some acute diseases like intestinal obstruction, the fibrinogen and fibrinogen ratio may be misleading and should be interpreted with caution [6]. However, we believe that the fibrinogen and fibrinogen ratio reflected true changes in the disease in the present study.
The main causes for the increase in blood lactate are increased anaerobic metabolic activity and a decrease in hepatic utilization. Under clinical situations, the most common cause of increased lactate is dehydration [6,13,30]. Furthermore, not only hypovolemia or oxygen debt but also metabolic conditions such as hepatic malfunction with decreased lactate uptake may induce hyperlactatemia [41]. In the present study, most cases had multiple organ dysfunctions, as indicated by biochemical analysis. So, the increased lactate had multi-factorial etiology in the present study. The important probable causes could be hypovolemia, reduced hepatic utilization, and decreased renal clearance due to azotemia. Abomasal reflux (indicated by increased rumen chloride concentration) could be the cause of dehydration, hypochloremia and azotemia [1,26,34]. As forage is the main source of potassium, anorexia could be the cause of hypokalemia [42]. Although lower albumin may have caused alterations in the ratios of chloride and potassium, it seems more probable that in peritonitis, abomasal reflux precedes the third space loss. Third space losses may have resulted in low sodium levels, but the compensating renal responses causing water retention may have resulted in about normal sodium concentration [10]. The lower calcium concentration was ascribed to decreased absorption and anorexia [1,10].
Peritoneal fluid analysis has been considered a supplementary aid to the hematological and clinical examination in the diagnosis of bovine abdominal disorders [1,17]. Assessing the volume, cellularity, and protein concentration of peritoneal fluid can give an indication of inflammatory changes in the peritoneal cavity [43,44]. However, there are no definite ranges for WBC and total protein in peritoneal fluid for classifying samples as inflammatory or noninflammatory, and the values may overlap between normal and severely infected samples [1]. An increased number of peritoneal fluid leukocytes is commonly associated with inflammation [45], but the sensitivity and specificity of this parameter are low due to numerous false positive and false negative results [46]. Peritoneal fluid nucleated cell count of greater than 6000/µLhas been reported to be consistent with a diagnosis of peritonitis in cattle [47][48][49][50]. The nucleated cell count criterion did not hold well in the present study. Rather differential cell count of peritoneal fluid was more useful in diagnosing peritonitis as WBC was ≤5000 in 68%cases. The present study's consistent findings on peritoneal fluid examination were increased specific gravity and >50% neutrophils. A total protein concentration of>3g/dL in the peritoneal fluid has normally been used as the cut-off value for exudates established in the classical transudate-exudate system [45]. The reference range for total protein has been redefined to 0.56-4.18 g/dL in peritoneal fluid of healthy cows [51]. However, higher cut-off values might increase the risk of false negatives and therefore lower the specificity of total protein in peritoneal fluid as a measure of peritoneal inflammation, as demonstrated in the present study.
The common cytological abnormalities of peritoneal fluid were an increased number of neutrophils with degenerative or toxic changes and resembled earlier findings of Ziemer [52]. The samples with the presence of bacteria were categorized under septic peritonitis as per Hirsch and Townsend [48]. A probable cause for the presence of gut contents in the peritoneal fluid could be trocarisation of rumen done at the farm level. Overwhelming peritonitis was characterized by the presence of a massive number of bacteria, greater than the cells in peritoneal fluid with markedly degenerated neutrophils. Overwhelming peritonitis, along with the presence of gut contents in the peritoneal fluid, was indicative of a poor prognosis.
In the present study, the treatment of peritonitis was a conservative medical approach with systemic antibiotics, fluid therapy, and supportive care. The individual buffaloes were treated with different antibiotic combinations based on inflammatory changes in leukogram and peritoneal fluid. All the antibiotics were administered intramuscularly at the standard dose rates. Depending upon the severity of dehydration, all these animals received intravenous fluids. The replacement of fluids is necessary so that adequate tissue perfusion could begin again, kidneys can adequately excrete the metabolic by-products again, and anaerobic metabolic processes can be curtailed. The controlled trials for determining different drug concentrations in peritonitis are not available, and peritoneal lavage is not of much use [1]. Also, the intraperitoneal administration of antimicrobials has been discouraged as it had no advantage over the use of parental antibiotics. Moreover, it is associated with the risk of local adhesion formation [1].
Smith et al. [53] have advocated stabilization of animals with supportive therapy combined with the use of antimicrobials in generalized peritonitis. The specific treatment directed to contain the infection is the use of antimicrobials, and third generation cephalosporin or synthetic penicillin have been reported to be good choices in peritonitis [32]. So, both were used in the present study and were administered parentally for a period of ≤12 days in the majority of animals. From Table 6, it can be inferred that antibiotic combination did not seem to affect the final outcome, e.g., an equal number of animals treated with ceftiofur, ampicillin, and metronidazole recovered and died. The electrolyte and mineral abnormalities should be identified and corrected. Meloxicam was administered to prevent the synthesis of more inflam-matory mediators and help pain management [32].
The findings with respect to the use of antimicrobials in buffalo peritonitis can be further confirmed by a clinical study with the randomized allocation of antibiotics. Whether supportive therapy had or had not had any effect on the improvement of the animals could not be concluded. The recovery period was 1-2 weeks in the majority of buffaloes. The treatment success rate in the present study was low, but it reflects the clinical reality. However, it is worth mentioning that the treatment success rate was far higher than reported for peritonitis in cattle [11]. Maybe a greater number of animals could have recovered after treatment if they had been diagnosed and treated at an early stage of the disease.
The duration of illness and other clinical parameters did not seem to play a role in prognosis. Rather, the absence of rumen motility was the important negative prognostic sign. Barely detecting rumen motility was associated with a poor prognosis. Marked left shift and toxic changes in neutrophils, higher BUN, and ≤3.6 mmol/L potassium were indicative of poor prognosis. The abnormalities of peritoneal fluid of non-survivors were diffuse and septic peritonitis, overwhelming bacterial infection, massive neutrophilia along with degenerated neutrophils, and the presence of gut contents. Five survivors had bacteria in peritoneal fluid (septic peritonitis), but degenerative changes in neutrophils of peritoneal fluid were mild or absent. In buffaloes with mild inflammatory leukogram and less severe peritoneal fluid changes, the severity of infection could have increased with the passage of time, resulting in death. It is always advisable to take multiple samples at different time intervals or stages of the disease to predict the exact prognosis [26]. Prognosis, as predicted on a single sample basis, holds true for that animal at that point in time and may change subsequently with improvement or worsening of condition and constitution of therapeutic interventions or response to therapy. Therefore, a good prognosis may worsen with time, and a guarded prognosis may improve [26]. No recurrence was recorded during the follow-up period, and there was no significant effect on the milk yield in the current or subsequent lactation.

CONCLUSION
Peritonitis in buffaloes usually occurs during pregnancy but can also occur at any stage of lactation. The main clinical features of peritonitis were fever, anorexia, dehydration and scanty defecation. The hematological analysis revealed hemoconcentration and neutrophilia. Biochemical alterations revealed hepatic and renal malfunction, hypocalcemia, hypokalemia, hypochloremia, hyperlactatemia, and increased rumen chloride concentration. In treatment, the deficient electrolytes and minerals should be corrected in addition to the use of antibiotics and liver tonics. The response to medical treatment was fair. However, for substantiation of the use of antibiotics, studies on a large number of clinical cases with proper design about randomization should be conducted. Long time survival rate was good, and there was no recurrence. The absence of rumen motility marked left shift, toxic changes in neutrophils, higher BUN, lower potassium (≤3.6 mmol/L), and unfavorable peritoneal fluid changes were the negative prognostic signs.

ETHICS APPROVAL
All procedures have been conducted as per the guidelines by Guru Angad Dev Veterinary and Animal Science University Institutional Animal Ethics Committee (IAEC), constituted as per article number 13 of the CPCSEA rules laid down by the Government of India. The procedures being performed were routine clinical procedures, and all owners gave consent for such procedures. Ethical approval was waived by the IAEC as handling and measurements of the animals were done by qualified veterinarians, and the study was on clinical cases.