Etiopathology and Therapy of Retained Fetal Membranes and Postpartum Uterine Infection in Buffaloes

1. Retained After Birth

Retained fetal membranes (RFM) is a frequent postpartum disorder affecting buffaloes [1]. The incidence of RFM is known to increase with parity with maximum incidence at the fifth calving [2,3]. Bacteria frequently invade the uterus of buffaloes causing inflammation of the retained fetal membranes and subsequently uterine infections [4]. If left untreated, the disorder culminates in an acute systemic reaction in affected buffaloes with loss of appetite, hyperthermia and decreased milk production [5,6] and increased services per conception, delayed resumption of postpartum ovarian cyclicity, increased days open and delayed pregnancy [7,8,9] as the long term negative effects of placental retention in buffaloes. The disorder has been identified as a major risk factor for postpartum uterine infections [4,10,11] and poor reproductive efficiency in buffaloes [6,9,12].

1.1.        The Bubaline Placenta

The bubaline placenta is polycotyledonary with most cotyledons being oval or dome shaped [13,14]. The placentomes are arranged in four rows varying in number from 102-229 placentomes in gravid uterine horns [14,15] and 91 - 153 in non-gravid uterine horns [14,15]. The maximum size of a placentome in a gravid uterine horn up to 210 days of pregnancy was 10.0 x 4.6 cm [14]. The total numbers of placentomes increase from early pregnancy to mid pregnancy with a tendency to decrease towards the end of gestation [14].

The buffalo has an epitheliochorial placentation [16]. Numerous binucleate trophoblastic cells are present in the single layered trophoblastic villous cover [16]. The frequency of binucleate cells increases throughout gestation [17]. The buffalo placentome has simple slightly conical villi branching less than in cattle, thus indicating different and less complex feto-maternal interdigitation in buffaloes [18].

Figure 1. The buffalo placenta. (Courtesy Dr G. N. Purohit, Department of Veterinary Gynecology and Obstetrics, College of Veterinary and Animal Sciences, Rajasthan University of Veterinary and Animal Sciences, Bikaner Rajasthan, India.)

The umbilical cord in the buffalo is composed of two thick arteries and two thin veins [19]. A pair of artery and vein diverges at right angles to traverse as placental vessels in the allanto-chorion [19]. The umbilical vessels have exceptional vascular structure and the diameter and thickness of the umbilical artery is more than the vein up to 95 days of gestation, but subsequently there is no difference between their sizes [19]. The placenta has a large vascularized allantoic sac. The amnion is avascular and covered by a single layered epithelium [20]. Amniotic plaques are found from 79 – 220 days of gestation in the amnion of buffaloes [21]. The amniotic plaques are made of polygonal or oval cells with granular cytoplasm, rich in glycogen and free of mucin and lipid [21]. The weight of the fetal membranes and the size of the allanto-chorion, amniotic vesicle and the umbilical cord show a progressive increase during gestation [13]. The volume of the allantoic fluid increases continually during gestation, whereas the amniotic fluid increases only up to the eighth month, after which the fluid volume declines [13]. The morphological aspects of placental haematomes of water buffalo placenta have been described [22] and they are considered areas for placental transfer of iron in buffaloes [22]. The buffalo placenta produces some amount of progesterone [17]. Pregnancy associated glycoprotein molecules have been isolated from water buffalo placenta [23] and sera of pregnant buffaloes [24].

1.2. Incidence

Among domestic ruminants, retained fetal membranes are a reproductive abnormality unique to the cow and water buffalo [25]. The placenta is normally expelled between 30 min to 8 h after fetal delivery. The normal time of expulsion of the placenta recorded by Kunbhar [26] was 3-8 h. Other investigators [27-29] reported 4.56-4.93 h as the normal time for expulsion of fetal membranes in buffaloes.

The earliest reports on retained fetal membranes in the buffalo appear to date back to the 1940s [30,31]. The incidence of retention of placenta in buffaloes ranged from 10-15% [26]. Roy and Luktuke [32] also reported the same incidence in buffaloes. Slightly higher incidences (16-18%) were also reported [1,33,34,35]. The incidence of retention of placenta increases with parity, twins and premature births [36-38] and varies from country to country, year to year and from herd to herd [33].

1.3. Etiology

The detachment of the fetal membranes from the buffalo’s uterus occurs by mechanisms that are still not well understood. The contractions of the uterus that occur at calving obviously result in changes in the size and shape of the placentomes which may lead to a loosening of the fetal cotyledon; more important, probably, is the marked reduction in blood flow to the uterus after the expulsion of the calf. There appears to be general agreement among authors that placental retention is deemed to have occurred if the fetal membranes remained attached more than 8-12 h after the birth of the calf [26,28,29,35]. Retention of fetal membranes is perceived by veterinarians and buffalo owners as a potentially more serious affection than the same condition in cattle [39]. Cotyledon proteolysis and decreasing adhesiveness of the cotyledon-caruncle interface fluids seem to be the key factors in the release of the fetal membranes. Collagenase activity of cotyledon villi during delivery is increased in healthy buffaloes and decreased in buffaloes with retained fetal membranes (Azawi, unpublished data). The cellular sources of collagenase and proteolytic enzymes for placental release in the buffaloes are unknown. In laboratory animals and humans, myometrial cells, fibroblasts, and leukocytes have been identified as sources of collagenase in the uterus. Lack of uterine motility is not considered as a reason for primary retention, because uterine motility is normal or above normal in buffaloes with primary retention of fetal membranes [40]. Serotonin has been proposed as a signal to begin the massive collagen degradation that occurs in the postpartum uterus [41]. It has been suggested that the role of serotonin during delivery is to stop blood circulation between the placenta and the fetus and to trigger uterine proteolysis. Blood concentrations of serotonin are extremely high (54μM) in the fetal calf during pregnancy and dramatically decrease (13μM) at parturition to near adult cow concentrations. Coincident with decreased blood serotonin concentrations is decreased serotonin concentration in fetal membranes (cotyledon) during placental detachment [41]. Thus, a physiologic pattern of serotonin concentration related to placental detachment during parturition exists in blood and placental tissues.

Several important effects of serotonin have been recognized: it has a proliferative effect on numerous cell types, including cultured bovine placentome cells, which may favor placental attachment [41]; it inhibits secretion of proteolytic enzymes in bovine placentome cells, which may prevent placental detachment during pregnancy [42]; it affects cortisol secretion, which may trigger parturition [46]; it is a powerful narcotic in newborn calves and probably in fetal calves, which may keep the fetus asleep during pregnancy [41].

These four effects of serotonin can support a pivotal role for serotonin in placental detachment, activation of the fetal-adrenal axis, and fetal awakening after prolonged gestational narcosis.

The following observations suggest that activation of biochemical mechanisms for the expulsion of placenta may occur before delivery:

1. Placentomes collected from the uterus at term (≥270 days) are relatively smaller than those collected at Day 240;
2. Dexamethasone administered before delivery, instead of immediately after delivery, can induce placental retention;
3. Injection of relaxin (collagenase inducer) after delivery inhibits the retained fetal membrane-inducing property of dexamethasone injection;
4. Softening (collagen breakdown) of the cervix occurs before delivery; and
5. Plasma collagenase activity increases at term, but before labor. It is suspected that collagenase is activated before delivery, about 9 hours before the release of the placenta in the buffalo (Azawi, unpublished data).

It is proposed that the biochemical disturbance leading to RFM may be triggered either before or during delivery. The collagenolytic activity of cotyledon villi is decreased in cows with RFM, and persistence of type III collagen is observed in cows with RFM. Activity of matrix metalloproteinase-9 (MMP-9) in cows with retained placenta is lower than in cows without retained placenta, and specific active forms of MMP-2 are absent in retained placenta. The differences in enzyme activity between cows with and those without retained placenta may affect the hydrolysis of collagen and subsequent release of the fetal membranes [41]. These observations suggest that a deficiency of collagenase may be involved in the hydrolysis of type III collagen. RFM may also be due, in some cases, to the presence of an anticollagenase system in the placenta, since intraplacentome injections of collagenase are unable to hydrolyze collagen in 15% of cows with RFM. Collagenases are calcium-dependent enzymes; however, the decreased level of serum calcium found in cows with RFM does not preclude collagenase activity. Several factors have been related to failure of cotyledon-caruncle detachment. Hormone imbalances existing before delivery are effective in inducing RFM. Progesterone, more than estrogen, inhibits uterine collagenases and slows uterine involution. Clinical studies on buffaloes have shown higher progesterone in buffaloes with retained fetal membranes [43-45] compared to buffaloes that had normal expulsion of the fetal membranes. Dexamethasone increases synthesis and utilization of progesterone by cotyledon tissues in the cow. These changes may contribute to blocking postpartum expression of cotyledon collagenases. Moreover, it has been found that glucocorticoids down-regulate collagenases. Because of inconsistent reports, the mechanism for dexamethasone and prostaglandin F2 alpha (PGF2α) to induce delivery with a high incidence of REM remains unclear. It has been proposed that increases in prepartal PGF2α metabolites and cortisol may constitute an indicator of RFM in cows. In one study, bovine production of PGF2α by cotyledons predominated when membranes were released, whereas prostaglandin E2 (PGE2) production predominated if fetal membranes were retained. In the rabbit, however, PGE2 and PGF2α increased collagenase activity. The fact that previous retention is a significant risk factor suggests that RFM may be due in part to a random expression of a malfunctioning gene that regulates uterine involution, including cotyledon proteolysis. Gene expression of proteolytic enzymes can be modulated by steroidal hormones; however, the prepartal hormonal signal to block cotyledon proteolysis and cause RFM has not been identified. The role of leukocytes and chemotaxis in the etiology of RFM has been widely discussed. Deficient neutrophil phagocytic activity, decreased migration, and decreased superoxide anion production have been proposed as factors in the pathogenesis of RFM in cattle. In fact, circulating neutrophils from cows with RFM produced less superoxide anion than did neutrophils from control cows. Positive chemotaxis resulted in a RFM incidence of 2.6%, and negative chemotaxis, 35.6%. Moreover, leukocytes are a mobile source of collagenases and may be involved in uterine regression and release of placenta. In one study, severe oxidative stress with obvious increase in blood malondialdehyde and nitric oxide and decrease in catalase, superoxide dismutase and ascorbic acid were observed in Egyptian buffaloes with retained placenta [1].

A shorter intrauterine life of the fetus has been put forward as one predisposing cause of placental retention in buffaloes [46] along with advanced involution of the placentomes, edema or maceration of the chorionic villi, hyperemia of the placentomes, placentitis and uterine inertia [47]. Abortions caused by Brucella, although less frequent in the buffalo [48,49], could also cause retained fetal membranes in the buffalo [48,49].

Clinical studies in buffaloes with retained fetal membranes have shown lower levels of circulating glucose, calcium, phosphorous, protein, copper and altered levels of serum cholesterol [50-52]. Similarly deficiencies of vitamin A, D, and E and selenium increase the risks of placental fetal retention in buffaloes [46,53-55]. Clinical evidence of retained fetal membranes is usually grossly visible by shreds or tissues hanging out of the vulvar lips (Fig. 2 and Fig. 3).

Figure 2. Buffaloes with retained fetal membranes hanging out of the vulvar lips. (Courtesy Dr G. N. Purohit, Department of Veterinary Gynecology and Obstetrics, College of Veterinary and Animal Sciences, Rajasthan University of Veterinary and Animal Sciences, Bikaner Rajasthan, India.)

Figure 3. Buffaloes with retained fetal membranes hanging out of the vulvar lips. (Courtesy Dr G. N. Purohit, Department of Veterinary Gynecology and Obstetrics, College of Veterinary and Animal Sciences, Rajasthan University of Veterinary and Animal Sciences, Bikaner Rajasthan, India.)

1.4. Pathophysiology

Prepartum inhibition of cotyledon proteolysis may occur prior to and during parturition in cows [56] and little is known about these changes in the buffalo although they are considered to be similarly operative. The resultant inhibition of liquefaction of cotyledon-caruncle adhesive fluids delays the separation of fetal structures from maternal structures [56]. The increased demands of late pregnancy, parturition and lactation probably increase the production of reactive oxygen species and the oxidative stress probably prolongs the dissociation of fetal cotyledon and maternal caruncles. Studies on Egyptian buffaloes have shown high amounts of reactive oxygen markers in buffaloes that had retained fetal membranes [1]. Lack of uterine contractions might result in failure of expulsion of the partially or completely separated fetal membranes. When the fetal membranes are retained, the fetal membrane mass loses its blood supply, but continues to grow for one or several days. The lack of blood circulation to retained membranes and their foul odor suggest that these membranes are necrotic or nearly necrotic tissues. When membranes are kept in controlled laboratory conditions, however, they are capable of active utilization of oxygen and glucose, and more than 30% of the cells can exclude vital dye for 3 or more days. When membranes are kept at 1° to 2°C, they stay metabolically active for 8 weeks or longer. Fetal membranes have an outstanding potential to "survive" without a live fetus. These properties suggest that RFM may respond to ischemia, anoxia, and bacteria by releasing biochemicals that cause inflammation, thus predisposing the cow to metritis. The retention of fetal membranes can be a substantial risk factor for toxic puerperal metritis in the buffalo. The retained fetal membranes release inflammatory biochemicals resulting in immunosuppression, increased vascular permeability, endometrial damage and decreased chemotaxis. Inflammatory biochemicals have systemic effects. Bacterial colonization may follow with resultant massive uterine infection (metritis). The systemic reaction results in clinical symptoms of depression, restlessness, arched back, straining, dysuria, diarrhea, reduced milk production, loss of appetite and high temperature in affected buffaloes [4,39,40]. Decomposition and fragmentation of placental tissue and chorionic villi occurs during retained fetal membranes in buffaloes with concomitant hyperplasia in the chorionic epithelial cells of the villi [5,6]. The presence of uterine infection delays the resumption of postpartum ovarian cyclicity and conception [57]. Thus presence of retained fetal membranes in the uterus for prolonged periods leads to the development of uterine infection [4,58-60], poor milk production (Fig. 4) and poor subsequent reproduction [9].

Figure 4. Pathophysiology of placental retention. (Courtesy Dr. G. N. Purohit, Department of Veterinary Gynecology and Obstetrics, College of Veterinary and Animal Sciences, Rajasthan University of Veterinary and Animal Sciences, Bikaner Rajasthan, India.)

1.5. Therapy

The treatment objectives for retained fetal membranes are to cause early detachment of the membranes in order to reduce the occurrence of metritis, decrease milk losses, improve reproductive efficiency, and decrease veterinary expenses. Untreated buffaloes are more often affected by endometritis and evidence repeat breeding compared to treated buffaloes. Manual removal had been a common practice in the past. Manual removal of retained membranes should be avoided at all costs because uterine infections are more frequent and more severe after this form of intervention than they are in untreated buffaloes. Oxytocin 40-60 IU intramuscularly if used immediately after calving, reduces the rate of RFM [39]. If the placenta is not expelled after the administration of oxytocin, the treatment still has a beneficial effect if manual removal of the placenta is done, because to some extent, oxytocin loosens the attachment between cotyledons and caruncles [32]. The use of oxytocin is of questionable value 24 hours after calving because by this time, the response to oxytocin becomes poor [56]. Unfortunately, in general in buffalo practice, the veterinarian is not consulted until after 24 h of the retention of fetal membranes because until then the farmer has hoped for a spontaneous expulsion. The layman's practice of tying a weight on the placenta should be discouraged because the weight causes the buffalo to strain and results in premature and incomplete breaking of placenta, leaving a part of it still in the uterus. This weight may also cause invagination of the uterine horn and prolapse of the uterus may occur. Manual removal of retained fetal membranes is contraindicated if the buffalo has a fever because uterine damage increases the risk of septicemia and perimetritis. Manual removal of fetal membranes has been reported for buffaloes [7,52] although dangers of uterine trauma cannot be excluded even with careful manual removal. When confronted with manual removal of fetal membranes, the clinician should first wash the perineum thoroughly and assess whether the placental separation has occurred or not and manual removal should be delayed for 12-24 h in the absence of proper loosening. Sufficient intrauterine and parenteral antibiotics [26,52] should be administered to promote uterine contractions and prevent uterine infection subsequent to manual removal of placenta. Ethnoveterinary practices have been used by buffalo farmers for hastening the expulsion of bubaline placenta [11], however, their efficacy continues to be poor with little scientific evidence. The concentrations of prostaglandin F metabolites are known to be lower in buffaloes retaining their fetal membranes compared to buffaloes that had uneventful parturitions [61,62]. Thus injections of prostaglandins to buffaloes with retained fetal membranes could be useful.

A new approach for the treatment of retained fetal membranes is the injection of collagenase into the umbilical arteries in cattle [56] but has not yet been used in buffaloes because such an approach can be carried out in animals immediately after calving and it is not usual for buffalo owners to consult a clinician until after 24 h of retention of fetal membranes. This approach may be an improvement to traditional treatments because it is specifically directed at correcting of the lack of cotyledon proteolysis. Bacterial collagenase from Clostridium histolyticum is used because it can degrade several types of collagen. It is commercially available, affordable, and it does not cause residual blood clotting in placenta. In this procedure, the umbilical cord is located and is identified by two firm arteries and two veins (pencil diameter) that slip off the fingers when palpated. Once the cord is located, a second hand is introduced into the vagina and the cord is retracted by alternating hands in the vagina. Once the umbilical cord is in the vulva, the arteries are clamped with Kelly forceps. Collagenase solution (200,000 U, plus 40 mg calcium chloride and 40 mg sodium bicarbonate dissolved in 1L saline) is injected rapidly. The injection of collagenase solution can be accomplished easily by using a hand pressurized 1000 ml saline bag attached to an intravenous administration set. The needle tip of the set is directly inserted into the artery (and ligated) without using a catheter. To ensure perfusion of the entire placenta, a volume of 1L is injected. Oxytetracycline (100 mg total dose, which is approximately 30 mg/kg fetal membranes) for intravenous injection can be added to 1 L of collagenase solution if an antibiotic is desired; however, final pH of the solution should be adjusted to approximately 7.5. Five hundred mL of collagenase solution are injected into each artery or 1000 ml into one artery if only one is available. Thirty-six hours later, the retained membranes are easily extracted by gentle traction if they have not been expelled spontaneously. Intrauterine and systemic antibiotics do not hasten detachment of retained membranes. On the contrary, inactivation of collagenase by tetracycline in some tissues has been reported. The future treatment for retained fetal membranes may be based on the development of an inexpensive injectable substance capable of triggering the activation of cotyledon proteases (collagenases).

Approaches for reducing the incidence of retained fetal membranes in buffaloes probably include supplementation of vitamin A, D, E and selenium in deficient areas and selection of animals for the reduced incidence, besides minimizing stress at parturition.

2. Postpartum Uterine Infections

Postpartum metritis and endometritis are the most important disorders in buffaloes [40], causing heavy economic losses due to prolonged days open and prolonged intercalving intervals, resulting in involuntary culling [62,63]. Uterine function is often compromised in buffaloes by bacterial contamination of the uterine lumen after parturition, insemination and wallowing; pathogenic bacteria frequently persist, causing genital diseases, a key cause of infertility [64]. The major problems faced by buffalo breeders and farmers include poor reproductive efficiency and prolonged intercalving intervals [64-68]. This can be attributed to factors such as harsh environments [69], lack of year round feed supply and minimal managerial input [70], in the majority of farming systems under which buffaloes are raised [68]. In every survey of the factors causing endometritis, metritis and toxic puerperal metritis, dystocia and retained fetal membranes were identified as of major importance in buffaloes [1,39,63]. The presence of pathogenic bacteria in the uterus causes inflammation, histological lesions of the endometrium, delays in uterine involution and perturbs embryo survival [71,72]. In addition, uterine bacterial infection, bacterial products or the associated inflammation, suppress pituitary LH secretion and perturb postpartum ovarian follicular growth and function, which disrupt ovulation in buffaloes and cattle [9,73]. The incidence rate of uterine infection in buffaloes was much higher than in cattle [74-76]. The annual incidences of uterine infections in postpartum dairy buffaloes were 20 to 75% [76,77] and in dairy cattle the proportions were 10 to 50% [78]. Postpartum metritis is one of the most important disorders in buffaloes [79-85]. Toxic puerperal metritis (i.e. acute septic metritis) is characterized by increased rectal temperature, depression, anorexia and fetid watery vulvar discharge [39,40]. Toxic puerperal metritis can be a severe problem, and is considered a uterine infection that is life threatening [39,40,86-88].

2.1. Definition and Incidence

Metritis and endometritis denote inflammation of the uterus. Metritis involves the endometrium, the underling glandular tissues and the muscular layer [60,78]. Endometritis involves only the endometrium which includes the superficial (luminal) epithelium, the underlying stratus compactum (stromal cells and gland necks) and the stratum spongiosum (gland bodies and stroma) [89], and without systemic signs [71]. These diseases share common etiological factors, predispose to one another and, largely, share common treatment [40]. The current knowledge and available literature regarding metritis and endometritis in buffaloes is very limited and most of the studies on uterine infection were done in cattle.

2.2. Classification of Uterine Infections

Several systems have been described in attempts to classify and define uterine infection. Uterine infections are generally classified according to clinical signs and degree of severity, which is in agreement to definitions used by theriogenologists [90]. However, frequently the definition or characterization of the various manifestations of uterine disease either lack precision, or definitions vary among research groups, and/or were not validated as to their effect on reproductive performance, making the assessment of the effects of treatment difficult. Often the term endometritis incorrectly includes metritis and endometritis or is determined solely based on transrectal palpation of an enlarged uterus [90]. During the 15th International Congress on Animal Reproduction [91], it was suggested that the research field would be aided by clear definitions of uterine disease that researchers could adopt. Sheldon [92] provided a clear clinical definition of uterine diseases. Adopting similar criterion in buffaloes’ toxic puerperal metritis was defined as an acute systemic illness due to infection of the uterus with bacteria, usually within 10 days after parturition [39]. The following clinical signs characterize toxic puerperal metritis in buffaloes: a fetid red-brown watery uterine discharge and usually, pyrexia, reduced milk yield, dullness, inappetence or anorexia, elevated heart rate and apparent dehydration may also be present [39]. The term metritis is used for animals that are not systemically ill, but have an abnormally enlarged uterus and a purulent uterine discharge detectable in the vagina [71]. Clinical endometritis is characterized by the presence of purulent (>50% pus) or mucopurulent (approximately 50% pus, 50% mucus) discharge detectable in the vagina and/or appearing at the vulva (Fig. 5) after 26 days postpartum [92]. A new technique for the diagnosis of endometritis has been used recently in bovine gynecology such as uterine cytology, mainly to detect subclinical endometritis in clinically healthy cows [93]. The proportion of polymorphonuclear neutrophils (PMN) in the total number of endometrial cells is indicative for subclinical endometritis [94]. Different threshold values for the proportion of PMN have been suggested, varying from 5 to 18% [94,95]. Reports on the use of endometrial cytology for the diagnosis of clinical endometritis in buffaloes are limited, however, one recent study described endometrial cytology as the most reliable method of diagnosing endometritis in buffaloes [4]. Subclinical endometritis can be defined as endometrial inflammation of the uterus usually determined by cytology in the absence of purulent material in the vagina. A cow with subclinical endometritis is defined by >18% in uterine cytology samples. Recently Dubuc [95] defined postpartum endometritis by its negative effect on subsequent reproductive performance, cytological and clinical diagnostic criteria were taken together to determine the optimal definition of endometritis. They also suggested that clinical endometritis terminology may not be appropriate and that purulent vaginal discharge may be more descriptive. Cows may be classified according to their uterine health status as purulent vaginal discharge only, cytological endometritis only, or both purulent vaginal discharge and cytological endometritis.

Figure 5. Buffalo with endometritis showing mucoid vaginal discharge. (Photo Courtesy Dr G. N. Purohit, Department of Veterinary Gynecology and Obstetrics, College of Veterinary and Animal Sciences, Rajasthan University of Veterinary and Animal Sciences, Bikaner Rajasthan, India.)

2.3. Pathogenesis

Following calving, the uterus of buffaloes becomes contaminated with bacteria [39,40]. Some of these bacteria are harmful and others are not [39]. When harmful bacteria are present; the uterus may become infected [96,97]. One should differentiate between uterine contamination and uterine infection. The uterus of postpartum buffaloes is usually contaminated with a range of bacteria, but this is not consistently associated with clinical disease [97]. Infection implies adherence of pathogenic organisms to the mucosa, colonization or penetration of the epithelium, and/or release of bacterial toxins that lead to establishment of uterine disease [39]. The development of uterine disease depends on the immune response of the buffalo, as well as the species and number (load or challenge) of bacteria [39,40,58]. When the numbers of pathogenic bacteria in the uterus of postpartum buffaloes are sufficiently large they overcome the uterine defense mechanisms and cause life threatening infection [39,58]. The postpartum uterus has a disrupted surface epithelium in contact with fluid and tissue debris that can support bacterial growth [58]. The outcome of uterine contamination depends on the number and virulence of the organisms present [39,40,58], as well as the condition of the uterus and its inherent defense mechanism [98]. A mild to severe endometritis occurs in 90% of postpartum buffaloes during the second through fourth weeks postpartum [40]. Resolution of the inflammation in cattle occurs with time, firstly being restored in the normal cow by 40 to 50 days postpartum [10]. No information is available on the resolution of postpartum endometritis in buffaloes after normal parturition. The interval from calving to clinically completed involution of the uterus in buffaloes varied widely with a minimum of 25 days [93] and a maximum of 74 days [99,100]. No study is available on the spontaneous clinical resolution of postpartum endometritis in buffaloes. In cattle, approximately three quarters of cows with postpartum endometritis had spontaneous clinical resolution [99]. The central question is why buffalo cows have a persistent infection after the postpartum period without spontaneous clinical resolution of the postpartum endometritis leading to prolonged days open and prolonged intercalving intervals. This could be due to the prolonged interval from calving to clinically completed involution of the uterus in dairy buffaloes [100] and the period of postpartum anestrus, as anestrus is usually longer in buffalo than in cattle [48,101]. Further studies in postpartum buffaloes concerning the release of acute phase proteins after parturition are needed to understand the impairment of spontaneous clinical resolution of postpartum endometritis in buffaloes. These proteins are likely to play an important role during the inflammatory process of the uterus, since they help to promote tissue repair following the inflammatory process, hydroxyproline or prostaglandin (PG) F2a metabolites enhance neutrophil chemotaxis and the ability of neutrophils to ingest bacteria, and plasminogen activators which are specific serine proteases, convert plasminogen to plasmin. A variety of species of bacteria, both gram-positive and gram-negative aerobes and anaerobes, can be isolated from the early postpartum bubaline uterus [39,40]. Most of these are environmental contaminants. Buffaloes with certain periparturient problems have a reduced ability to control uterine infections. Excess stretching of the uterus, as with hydrops allantois, traumatization of genital tissues during dystocia or obstetric manipulation, predispose buffaloes to postpartum metritis [64]. Metabolic disorders, some traditional practices by farmers and herdsmen such as insertion of the hand or implements in the vagina of a buffalo to stimulate milk letdown, as well as unhygienic conditions under which animals are allowed to calve, can diminish uterine tonus. In addition, some farmers suture the buffalo's vulva to prevent uterine prolapse immediately after parturition [40]. Lochia is then retained beyond the normal period, providing a medium for bacterial multiplication [55]. Phagocytosis by uterine leukocytes is reduced in buffaloes with dystocia, retained fetal membranes and metritis [64]. If the uterus is severely debilitated, any variety of contaminating organisms can cause a toxic puerperal metritis [39,40,64]. In less severe cases, endometritis occurs and may become persistent impairing fertility [10,39,40,64].

2.4. Bacterial Causes of Uterine Infections

The most common cause of uterine infection is the pathogenic microorganisms affecting productivity and fertility of buffaloes [4,58,59,62,64]. Pathogenic organisms isolated from an infected uterus are found generally in livestock environments and are capable of infecting other tissues and organs [39]. Thus, uterine infections are classified as non-specific infections [102]. They are called non-specific infection because the initial colonizing bacterium is not known and the specific bacteria causing the signs of infection are not identified [78]. Numerous bacteria in a variety of combinations have been isolated from an infected bubaline uterus. Arcanobacterium pyogenes and E. coli are usually associated with uterine infection in buffaloes and cattle [58,59,62,64]. The composition of the uterine flora changes somewhat at each recontamination, no specific combination of organisms is associated consistently with postpartum infections [103]. Nevertheless, Arcanobacterium pyogenes either alone or in combination with other bacteria such as the anaerobic Fusobacterium necrophorum and Bacteroides spp [39,40] is often associated with uterine infections [39,62,64]. The potential of intra-uterine oxygen reductase declines in the presence of infection [104] and aerobic bacteria decline too, thereby creating an anaerobic environment. This drop in intrauterine oxygen reductase potential may be associated with either microorganism metabolism or increased oxygen consumption by polymorphonuclear inflammatory cells. Of the anaerobic bacteria cultured from cases of uterine infection, Fusobacterium necrophorum and Bacteroides spp. have been identified [62,64]. When A. pyogenes was isolated from uterine fluids, buffaloes developed severe endometritis and usually were infertile at first service [75]. Azawi [39] suggested that organisms other than A. pyogenes and gram-negative anaerobes such as F. necrophorum, as well as, E. coli, Streptococcus spp., Staphylococcus spp., and Pseudomonas spp. are responsible for toxic puerperal metritis. The growth of anaerobic bacteria may enhance the establishment of A. pyogenes and lead to the development of severe uterine infections [39,40]. Indeed, F. necrophorum produce leukotoxin [105,106] while Bacteroides produce substances that prevent bacterial phagocytosis and A. pyogenes produce a growth factor for F. necrophorum [40]. Bacteroides and Fusobacterium species are prevalent in the indigenous flora on all mucosal surfaces. Tissue necrosis and poor blood supply lower the oxidation-reduction potential, thus favoring the growth of anaerobes [107,108]. In addition, F. necrophorum is frequently a secondary invader and mixed infection with A. pyogenes is not uncommon [10]. In addition F. necrophorum produces a variety of extra-cellular products including hemolysin, hemagglutinin, adhesions, platelet aggregation factor, proteases and DNase. The significance of these products relative to virulence is not clear [109]. Azawi [39] suggested that the earlier appearance of E. coli in the uterus affects the phenotype and function of polymorphonuclear cells, and this might support the co-infection by A. pyogenes at a later time.

2.5. Uterine Defense Mechanism

Anatomical and functional barriers provide effective defense against reproductive tract invasion by environmental organisms as well as nonspecific and specific immune responses [39]. Dhaliwal [106] reported that uterine defense mechanisms against contaminating microorganisms were maintained in several ways: anatomically, by the simple or pseudostratified columnar epithelium covering the endometrium; chemically by mucus secretions from the endometrial glands; immunologically, through the action of polymorphonuclear inflammatory cells and humoral antibodies, but the degree of interaction is not clear. Disruptions of these mechanisms allow opportunist pathogens, mostly microorganisms found in the posterior gastro-intestinal tract and around the perineal area [58], to colonize the endometrium and cause an endometritis [58,59,96,97]. A degree of bacterial contamination of the uterus usually occurs during, or immediately after parturition [1,13]. Bacterial contamination of the uterus may also occur during coitus or insemination [39,40,106]. Also in buffaloes, bacterial contamination of the vagina and other external reproductive organs might occur during wallowing [39,40]. Whether or not a persistent infection of the uterus becomes established, depends upon the level of contamination, the animal's uterine defense mechanisms and the presence of substrates (such as devitalized tissue) for the growth of bacteria [64].

Under normal circumstances, there are several mechanisms, which prevent opportunist pathogens from colonizing the genital tract. The major anatomical barriers between the contaminated external environment and the relatively sterile environment of the uterus, include the vulva, the vestibule (guarded by a muscular sphincter), and the cervix. It should be noted that, although the vulva may appear of little consequence as a barrier, it is, in fact, remarkably efficient at preventing fecal contamination of the tubular genitalia [106-108] as in cattle, while in buffaloes the larger soft loosely arranged vulvar tissue might reduce the efficacy of this barrier [40]. In cattle and buffaloes, the cervix is a formidable barrier composed of a series of mucosal lined collagenous rings [40,106]. In addition, the cervical-vaginal mucus (especially the scant, tenacious mucus of the luteal phase) can function as a physical barrier for organisms that would otherwise ascend through the reproductive tract [107]. The circular and longitudinal layers of the uterine musculature provide physical propulsion of particular material, including microbes.

Epithelial cells are the first to make contact with potential pathogens that enter the uterus [108,109]. Epithelial and stromal cells interactions are critically important for endometrial function, with stromal cells affecting epithelial cells through both the release of soluble factors and turnover of extracellular matrix [108]. Conversely, epithelial cells affect stromal cell function through the release of soluble factors and cell-to-cell contact. Pierro [110] suggested that PGE2 regulate epithelial cell proliferation and could be mediated indirectly by uterine stroma.

Estradiol and progesterone have both opposing and complementary effects on the female genital tract with estradiol stimulating epithelization (especially of the vaginal lining and endometrial gland), and vascularization of the endometrium [111]. Progesterone aids in endometrial gland differentiation and enhances uterine gland secretions, reduces cervical mucus production, prevents uterine contractility [58], and acts as a counter influence to estradiol in immune protective responses of the reproductive tract [107]. Cattle are resistant to uterine infections when progesterone concentrations are basal and they are susceptible when progesterone concentrations are increased [78]. For example, spontaneous uterine infection in cattle does not usually develop until after formation of the first postpartum corpus luteum, although bacterial contamination can be sufficient to induce the onset of puerperal metritis very soon after calving when progesterone concentrations are basal [78,108]. Postpartum cows that received intrauterine infusions of Arcanobacterium pyogenes and E. coli when progesterone concentrations were basal did not develop uterine infections, whereas all cows developed uterine infections when the bacteria were infused after the onset of luteal function and progesterone concentrations had begun to increase [11,23,57]. In addition, none of the animals, that received intrauterine infusions of Arcanobacterium pyogenes and E. coli during the estrus phase developed a uterine infection, however, all those that received Arcanobacterium pyogenes and E. coli infusions during the luteal phase of the estrous cycle, developed uterine infections [40,58,108]. The previous examples clearly support the idea that progesterone converts the uterus from an organ that is resistant to one that is susceptible to infection. In cycling buffalo cows, the uterus is usually under the influence of progesterone during the greater part of the cycle. That is, the non-pregnant uterus is in the luteal phase (under the influence of progesterone) for about 14 to 15 days of its 21-day cycle (i.e., from about day 3 to 17 after estrus and ovulation) [101]. The uterus is under the most significant influence of estradiol, with no progesterone to counter its effects, for about 1 day (immediately preceding standing estrus). It has been reported that Murrah buffaloes have higher overall plasma estradiol concentrations compared to swamp buffaloes. The estradiol values at estrus are 31±1.70 pg/ml [112-114], compared to the lower values of 12.9 pg/ml in swamp buffaloes [57,113]. The high estradiol concentrations that occur at estrus and parturition cause changes in the number and proportions of circulating white blood cells, with a relative neutrophilia and a "shift to the left" [10]. Moreover, at estrus, the blood supply to the uterus is increased under the influence of estradiol, whilst at parturition there is a massive blood supply to the gravid uterus. This increased blood supply, coupled with the migration of white cells from the circulation to the uterine lumen, enables vigorous and active phagocytosis of bacteria to occur [108]. Estradiol also causes an increase in the quantity and nature of vaginal mucus, which also plays an important role in the defense of the uterus against bacteria by providing a protective physical barrier and by flushing and diluting the bacterial contaminants [104,107]. The immune functions of the uterus are known to be upregulated when estrogens are increased [106]. It is difficult to determine whether increased estrogens during follicular phase induced the upregulation or whether upregulation was due to the removal of the suppressive effects of progesterone [106,107]. Wira [109] demonstrated that changes in ovarian estrogens and progesterone regulate uterine immune function. The effect of estrogens and progesterone may seem antagonistic at first, but the two hormones seem to orchestrate uterine immune function in favor of the animal. Indeed, uterine immune function is up-regulated at estrus when there are many opportunities for the introduction of pathogens and down-regulated during the luteal phase when the uterus is capable of supporting a conceptus. This down-regulation during the luteal phase seems to allow the uterus to tolerate a fetal allograft [10,39]. The most critical factor in uterine defense against infection is the rapid, physical clearance of inflammatory debris from the uterus after insemination or calving [39,40]. Buffaloes compared to cattle have difficulty in clearing this debris from the uterine cavity because they have a lower estradiol secretion than in cattle during the estrous phase, decreasing uterine drainage [115,116].

2.6. Therapy

The treatment of uterine infections including; toxic puerperal metritis, endometritis, and metritis in buffaloes, should be directed towards saving the life and/or salvaging the affected buffalo and improving fertility [10,58,87]. As such, fluid replacements, antibiotics, antipyretics, liver tonics and antihistaminics are suggested along with drugs that favor uterine clearance. Several studies indicate that the presence of aerobic and anaerobic bacteria in the uterus contributes to reducing fertility in buffaloes with uterine infection [39,40]. Therefore, ideally, therapy for uterine infection should eliminate pathogens from the uterus, and should result in a short as possible withdrawal period for milk and meat. Success in the treatment of uterine infections depends on, evacuation of uterine fluids, susceptibility of the infectious agents to the drug used, concentration and number of times the drug is used, and the exposure of the entire endometrium to the drug. Evacuation of the uterus contributes to the success of further antibiotic therapy. Evacuation can be done by repeated palpations of the uterus by the veterinarian and/or the use of hormones to expel the fluid or hasten the onset of estrus [78]. Estrus is usually the best way of stimulating uterine contractions and expelling the fluids [90]. When fluids are expelled, the effectiveness of antibiotics in clearing the remaining infection is improved. Draining uterine fluids from buffaloes with toxic puerperal metritis is a common, but apparently ill-advised procedure. Manipulation of a friable infected uterine wall may exacerbate the problem [30,40].

2.6.1. Antibiotics

The following criteria are important when choosing antibiotics for the treatment of uterine infections:

1. The antibiotic should be active against the main uterine pathogens and should maintain its activity in the environment of the uterus. The isolation of anaerobic bacteria from the postpartum uterus has resulted in recognition of this site as an anaerobic environment [117]. Therefore, antibiotics that are ineffective under anaerobic conditions, such as the aminoglycosides, are not recommended for the treatment of the postpartum uterus [105,118]. The uterine lochia consists of organic fluids and debris and contains a variety of Gram-positive and Gram-negative aerobic and anaerobic bacteria [73]. Consequently, a broad-spectrum antibiotic that is active in the presence of organic debris is indicated [119]. This eliminates the sulfonamides, which are ineffective in the presence of tissue breakdown products [119-121]. Intrauterine administration of penicillin would also not be recommended owing to the likelihood of penicillinase production by some of the bacterial species present [78].
2. The antibiotic should be present in sufficient concentration at the site of infection (i.e., the sub-endothelium). This depends on the properties of both the antibiotics and the vehicle [121,122]. The pharmacokinetic properties of the antibiotic preparation should allow the rapid distribution of the antibiotic throughout the uterine cavity, and good penetration of the antibiotic into the endometrium [122].
3. The preparation should not inhibit the normal defense mechanisms and should be well tolerated and not induce irritation in the endometrium [123]. Antibiotic therapy cannot sterilize the uterus nor prevent the continual recontamination that occurs during the early postpartum weeks [124]. Successful resolution of a uterine infection requires effective uterine defense mechanisms [125]. Buffaloes with an abnormal puerperium, including buffaloes with dystocia, retained placenta, or metritis have decreased activity of uterine phagocytes [10]. In addition, most antiseptics and many antibiotics have been shown to depress phagocytosis for several days after intrauterine administration [119]. This effect is most pronounced after intrauterine infusion owing to the high concentrations achieved in the uterine lumen. The use of Lugol’s iodine solution destroyed phagocytic activity of the uterine leukocytes for several days after intrauterine application [119]. Intrauterine infusions may sometimes be credited with successful antimicrobial treatments of endometritis when the beneficial effect was actually an irritation of the endometrium [126,127]. In cyclic buffaloes, this may induce prostaglandin release, luteolysis and removal of the inhibitory effects of progesterone on the uterine defense mechanism [126]. This might be considered justification for use of irritating intrauterine products, but this should be discouraged [120]. Fertility of rats infused with Lugol’s iodine was depressed for a long time after histological resolution of the necrotizing endometritis [128]. In addition, with infusions of irritating drugs, there is a possibility of oviductal irritation and its consequences [90].
   Oxytetracycline is a broad spectrum antibiotic and is indicated for the treatment and control of infections caused by or associated with oxytetracycline sensitive, rapidly growing bacteria [87]. Its antibacterial efficacy against many infections caused by Gram-positive and Gram-negative bacteria are well-documented [120]. The antibiotic may also be used by the intrauterine route [64,87,120]. Intrauterine administration represents a useful therapy, especially in the treatment and prophylaxis of postpartum endometritis in the buffalo [64,120,127]. The direct intrauterine administration of oxytetracycline produces immediate therapeutic concentration in the caruncles and endometrium of both healthy and affected animal [137], and because of its relatively low absorption into the bloodstream [134], the therapeutic action is largely confined to the uterine lumen and endometrium. Tetracycline’s are known to be active under anaerobic conditions and are partly inactivated by the purulent material, cell debris at pH levels found in the affected uteri [10]. Girardi [129] observed a remarkable variability in the absorption of the antibiotic, after intrauterine administration, in relation to the vehicle used, and emphasized that the oxytetracycline passage through the uterine mucosa is strongly influenced by the carrier. However, intrauterine therapy is the preferred treatment for endometritis, but there are indications for the use of systemic route of administration. Higher antibiotic concentrations throughout the genital tract are achieved with systemic administration than with intrauterine therapy [120].
   Systemic treatment is best if antibiotics are subjected to degradation by conditions in the uterine lumen. A buffalo with toxic puerperal metritis has an invasion of microorganisms into deeper layers of the uterine wall [39]. This would imply that effective concentrations of the antibiotic would be desired in all regions of the genital tract. Bretzlaff [120] found no significant differences in plasma to genital tract tissue ratio of oxytetracycline when administered systemically at early postpartum. Intrauterine infusions in early postpartum buffaloes with toxic puerperal metritis were associated with decreased drug absorption. These were reflected in reduced plasma concentrations and, therefore, lower tissue concentrations of oxytetracycline in sub-endometrial tissues [120]. Recent studies recommended the use of antibiotics administered systemically in buffaloes with toxic puerperal metritis [42,43]. Such an approach is especially useful during the first 10 days postpartum since during this time there is plenty of tissue debris and fluid in the uterus which renders ineffective most of the antibiotics administered by the intrauterine route. A useful aid to clear uterine contents could be the intrauterine administration of normal saline (which dissociate the debris) followed by IM administration of uterine motility enhancers such as prostaglandins.

2.6.2. Hormones

The effective use of hormones in uterine infection requires knowledge of both normal reproductive endocrinology and the therapeutic characteristics of available hormonal preparations. Administration of prostaglandin F2α at postpartum shortens the postpartum interval in buffaloes [39], and may influence days to first estrus [130], by enhancing lochia evacuation. On day 32 after calving, PGF2α treatment decreased postpartum intervals [131]. The use of PGF2α has clearly been shown to be the treatment of choice of buffaloes with endometritis [39]. The decrease in progesterone and increase in estrogen concentrations associated with luteolysis and follicular growth result in maximal resistance of the uterus to bacterial infection. PGF2α has the least harmful effects and milk does not have to be discarded [39,40]. Estradiol has been recommended by some to stimulate myometrium contractions, phagocytosis and mucus production [10]. Others, [58,64,71] concluded that administration of estradiol did not have beneficial effects on metritis and reproductive performance and rather reduced milk production in some lactating buffaloes. Sheldon [102] recommended the use of estradiol benzoate by intrauterine route for the treatment of postpartum infection. The most important side effect of using estradiol in the treatment of metritis is the propulsion of uterine inflammatory exudates into the oviduct [90]. In addition, the use of estradiol has been banned in some countries owing to certain higher residues [131]. Thus the administration of estradiol in lactating buffaloes is not suggested. The use of oxytocin for therapy of uterine infections in buffalo is not justified probably because oxytocin acts only on a uterus primed with estrogen which declines shortly after parturition in this species [132-133]. In some studies the postpartum single or regular oxytocin administration enhanced the uterine involution and reduced the calving to service interval [134-135] whereas another study failed to observe such benefits [136].