Uterine Health in Dairy Cows

1. Uterine Health in Dairy
Cows
Stephen LeBlanc
University of Guelph, Canada

2. Risk factors for uterine disease
 Species of bacteria
 Virulence factors
 Strain
 Level of contamination
 Dry matter intake
 Energy and lipid
metabolic health
 Stressors & hormonal
changes
 Hypocalcemia
Bacteria
Immune response
Regulation of inflammation
Pathological
inflammation
Involution
Pathological
inflammation

3. The uterus is not
sterile in virgin
heifers or
pregnancy
• 10 virgin heifers in
estrus and 5 pregnant
cows (3rd trimester;
slaughter samples Moore
et al JDS 2017)
• 43 pregnant cows –
slaughter samples.
Bacteria identified in
91% Karstrup et al Therio 2017

4. Uterine microbiome
• Culture-independent
techniques give a
different picture than
routine bacteriology
• Cows with metritis
have lower diversity
of uterine microbes
than healthy cows
Bicalho et al JDS 2017
Jeon et al PLoSONE 2016

5. Uterine E coli
Some strains of E. coli have an adhesion factor
that make them uterine adapted pathogens
1. Silva et al 2009 – 72 isolates from 12 healthy cows and 18 with
metritis – no associations of phylogenetic group or specific
virulence factors
• Strain associations with herd
2. Sheldon et al 2010 – 114 isolates from 64 cows from 1 herd
sampled at weeks 1 – 4
• Association of 1 group with metritis
• Strains found in cows with metritis were more adherent and invasive to
epithelium and LPS induced greater inflammatory response in culture
3. Bicalho et al 2010
• 611 E coli isolates from 374 cows in week 1 in 4 herds, screened for 32
virulence genes; subset of 117 cows screened for PVD in week 4
• 6 virulence genes associated with reproductive tract disease
• Adherence factor (fimH) was the single strongest predictor of metritis (OR
= 6)

6. Vaccination against reproductive tract
disease
• Total of 371 heifers from 1 herd vaccinated at ~ 50 and 20 d
before calving with vaccinate containing whole killed E. coli, F.
necrophorum and T. Pyogenes ±FimH, leukotoxin, and pyolysin
• SC vaccination with bacteria ± proteins
• decreased metritis incidence
• 12 to ~4% (Dx only at 6 DIM)
• 28 to 16% (daily Dx by farm staff)
• Decreased time to pregnancy
• 120 to 94 median DIM
• No effect on clinical endometritis (lavage visual at 35 DIM; 9%) or on
prevalence of E. coli or F. necro at 2 or 6 DIM, or T. pyo ay 35 DIM
Machado et al 2014

7. What impacts does
this have on this cow?
Can we improve her
outcomes?

8. Giuliodori et al., 2013
Metritis – mild or severe affects milk
production

9. Giuliodori et al., 2013 Dubuc et al., 2011
Only severe metritis, but both PVD and
endometritis impair reproduction

10. Uterine microbiome
• Clinical cure is associated
with greater diversity
• Failure to cure is associated
with increased abundance of
“core uterine pathogens”
• Bacteroides
• Fusobacterium
• Porphyromonas
• [Not E. coli] Jeon et al VetMicro 2018
Cows with metritis
NB n = 3 per group
Cured Not cured

11. Metritis treatment
• Case definition = T > 39.5 + fetid discharge
• 1) n = 406 cows (8 large herds) randomized to control, 1.1 or
2.2. mg/kg ceftiofur for 5 d Chenault et al , 2004
• 2) n = 982 cows (15 herds) randomized to 2 doses CCFA (3 d
apart) or saline control McLaughlin et al JDS 2012
• Clinical cure (no fever; no fetid discharge) at day 14:
•Ceftiofur: 77% Saline: 62%
•CCFA: 74% Saline: 55%

12. Metritis treatment
• n = 528 cases of fetid discharge (n=312) or metritis (fetid + T ≥ 39.5 C;
n=216) from 1 large dairy in Florida; 268 matched healthy cows in
comparison group
• Blocked by metritis type, randomly assigned to 11 mg/kg ampicillin or 2.2
mg/kg ceftiofur q 24 h x 5 d
• PreSynch 50 and 64 DIM, bred on estrus after 2nd PGF or 5-d Ovsynch with
TAI at 84 DIM
• Among cows with fetid discharge only, 53% developed ≥ 1 day of fever
while on antibiotics
Lima et al JDS 2014

13. Lima et al JDS 2014
• Absence of fever +
fetid discharge at day
12 not different ~67%
• No interactions of
treatment with metritis
type, parity, dystocia
Metritis treatment
Clinical cure

14. Lima et al JDS 2014
• Lower prevalence of PVD
at 32 d with ampicillin
than ceftiofur
• No difference in
endometritis at day 39
• Cows with metritis still
had much higher
prevalence
Metritis treatment
PVD and
endometritis

15. Outcome Ampicillin Ceftiofur Healthy
comparison
P
Cyclic 64 DIM 75% 76% 75% 0.96
Pregnant 1st AI
30 d 29% 29% 32% 0.87
60 d 28% 28% 31% 0.91
Loss 6% 6% 11% 0.52
Lima et al JDS 2014
These outcomes not different between
fetid discharge and metritis
Metritis treatment – Reproductive
performance

16. • Diagnostic criteria inconsistent and not well
validated (Sannmann et al 2012)
• Systematic review (Haimerl & Heuwieser 2014) and meta-
analysis (Haimerl et al 2017)
• 23 studies; narrow set of research questions but overall
high quality of evidence
• Cows with  2 of : T > 39.5; fetid discharge;
dullness/off-feed can rationally be treated with
systemic antibiotics
• Ceftiofur, ampicillin, or penicillin likely to be
similarly efficacious (Smith et al 1998; Drillich et al 2001; Chenault et
al 2004; Lima et al 2014; review by Reppert VCNA 2015)
Metritis treatment - Summary

17. NSAID vs antimicrobial for initial
treatment of metritis
• 558 cases of metritis (T ≥ 39.5 and fetid
discharge); 6 farms in Germany
• Randomized to ketoprofen or ceftiofur
• If T ≥ 39.5 on day 4-7:
• KET → 3 d of ceftiofur
• CEF → + 2 d of ceftiofur
Pohl et al JDS 2016

18. NSAID vs antimicrobial for initial
treatment of metritis
• KET 3.4 times greater odds of extended treatment
• Similar outcomes with 1.8 vs. 3.6 doses of
ceftiofur per case
Pohl et al JDS 2016

19. •Expected outcomes:
• Absence of fever in 80 – 100% of cows 2-6 days
after treatment
• Absence of fetid discharge 33 – 77%
• No difference in reproductive performance for
treated cases of metritis (Dubuc et al 2011; Lima et al
2014)
•More field research is needed to refine
treatment selection criteria (i.e. discharge ±
temperature), and assess antimicrobial
resistance, animal welfare, and economics
Metritis treatment - Summary

20. Cows with post-partum metritis
had lower DMI during the post-
and pre-partum periods
0
4
8
12
16
20
-13 -10 -7 -4 -1 2 5 8 11 14 17 20
Healthy
Mildly Metritic
Severely Metritic
DMI(kg)
Day from Calving
Clinical signs
of infection
Huzzey et al. 2007

21. Neutrophil myeloperoxidase (iodination)
activity – a measure of killing activity
Cows with
subclinical
endometritis at
28 DIM n = 43
Cows with metritis
< 14 DIM n = 18
Cows with normal
uterine health n = 22

22. Increased Haptoglobin Precedes
Reproductive Disease
Huzzey et al 2009
Hp ≥ 1 g/L at 3 DIM OR = 6.7 for metritis Se = 50% Sp
= 87%
n=12
n=23
n=32
Hp > 0.8 g/L in week
+1 increased odds of
metritis, PVD, and
endometritis (OR =
1.6 – 2.2) (Dubuc et al
2010)

23. Inflammation
• Acute/classical inflammation
• Injury → heat, pain, swelling, redness
• Infection → fever, decreased feed intake
• Disease causes inflammation
• Metabolic/chronic inflammation
• No clinical signs
• Oxidative stress, obesesity, fat mobilization (?) → Slight
elevations in pro-inflammatory mediators and acute phase
proteins; changes in signalling
• Contributes to insulin resistance
• Inflammation contributes to disease
Adapted from Barry Bradford

24. Associations with reproductive
tract inflammatory disease
20 1-1 ≥ 3
↓ DMI
↑ NEFA
↑ BHB
↓ PMN oxidative burst
↑ IL-6 ↑ IL-6
↑ IL-8
↑ Hp and SAA
↑ IL-10
± IL-1
± TNFα± TNFα
↑ DEFB5
↑TAP
↑PTGS
↑IFN

25. Concepts of inflammatory response
1 42 3 5
UterineInflammation
weeks
Healthy
Endometritis –
excessive
inflammation
Endometritis –
inadequate response

26. Reproductive tract infection and
inflammation
Endo-
metritis
PVD
Each affects
15 – 20%
of cows

27. • Endométrite cytologique (≥ 21 JEL)
• Cytobrosse
Uterine Cytology

28. Subclinical endometritis: endometrial
cytology
R. Lefebvre, Université de Montréal

29. Endometritis – common and
important
Study n DIM Dx Method
& PMN cut-
point
Prevalence Median days to pregnancy
Un-
affected
Endo-
metritis
Kasimanickam
2004
228
(2 herds)
20-33 Brush >18% 35% 112 141
Gilbert 2005 141
(5 herds)
40-60 Flush
> 5%
53% 118 206
Barlund 2008 221
(8 herds)
28-41 Brush >8% 12% 121 145
Galvao 2009a 202
(1 herd)
48-54 Flush 5% 29% 112 126
Galvao 2009b 406
(1 herd)
32-38 Flush 7% 38% 121 151
Dubuc 2011 2072
(6 herds)
32-38 Brush 6% 13% cyto
7% both
132 148
195
Deguillaume
2012
168
(3 herds)
21-35 Brush >6% 45% 125 144

30. Impact of Reproductive Disease on Time
to Pregnancy
SurvivalDistributionFunction
0.00
0.25
0.50
0.75
1.00
darpreg
0 50 100 150 200 250 300
Proportionofcowsnotpregnant
DIM
Unaffected
ENDOMETRITIS only
PVD only
BOTH
Median days open
-24d--
----36d----
--------61d--------
Dubuc et al JDS 2011

31. PVD ≠ endometritis
• Poor agreement (42% of cows with PVD had
concurrent endometritis; Dubuc et al 2010a)
• Different risk factors (Dubuc et al 2010b)
• Additive effects on reproduction and pregnancy loss
(Dubuc et al 2010; Lima et al 2013)
• Cervicitis is a co- or independent factor in many cases
PVD BOTH Endometritis

32. Bromfield et al., 2015 Ribeiro et al., 2016
Mechanisms of uterine disease effects
on fertility

33. Mechanisms of harms
from reproductive
disease
• Exposure to LPS inhibits the
development and competence of
oocytes (Magata and Shimizu 2017).
Plasma LPS not different between
cows with high or low pathogen load
(Williams et al 2007) or ovulatory vs. non-
ovulatory first dominant follicles,
follicular LPS concentrations were
greater in non-ovulatory group
(Cheong et al., 2017)
• Some found no association of
postpartum uterine disease with
anovulation (Ribeiro and Carvalho, 2017);
endometritis increased anovulation
(19 vs. 13%; Dubuc et al., 2012)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0
5
10
15
20
25
30
35
40
0 0.001 0.01 0.1 1 10
ICM:TEratio
%blastocyts
LPS concentration in media (ug/mL)
IVF, % cleaved to blastocyst Day 8 ICM:TE
(Magata and Shimizu 2017)
(Cheong et al., 2017)

34. Mechanisms of harms from
reproductive disease
• Cows with early postpartum uterine infection with bacteria that
are understood to be pathogenic has smaller CL and circulating
progesterone concentrations were lower (Williams et al., 2007)
◼ High  Low growth density of uterine pathogens

35. Mechanisms of harms from
reproductive disease
• Metritis or inflammatory disease in
early lactation had fewer cleaved
and live embryos at 5 or 6 d after
AI, lesser pregnancy at 45 d after AI,
and greater pregnancy losses
(summarized in Ribeiro and Carvalho, 2017)
• Fewer, lower quality embryos in
cows with endometritis (Gilbert review
RFD 2012)
•  ICM:trophectoderm in embryos
exposed to LPS (Magata & Shimizu 2017)
or to inflammation (Hill & Gilbert, 2008) –
likely less viable
Calvingto1stAI,%

36. • PVD is associated with
bacterial infection,
especially Trueperella
pyogenes
• Many/most cows with
endometritis do not have
concurrent bacterial
infection (by conventional
microbiology) (McDougall
et al 2011; Madoz et al
2014)
Brodzki et al RDA 2014
Neutrophils with ingested bacteria
Inflammation
vs. infection

37. Days in milk
0 100 200 300 400 500
Proportionofcowsnotpregnant
(Survivalfunctionestimate)
0.0
0.2
0.4
0.6
0.8
1.0
Untreated cases
Cephapirin IU
Prostaglandin F2
LeBlanc et al, 2002 JDS 84:2237
Treatment of PVD and time to pregnancy

38. Denis-Robichaud & Dubuc, 2015
2259 cows Examined at 35 ± 7 DIM by Metricheck and cytobrush
76% 24%57% 19% 25%Prevalence
Treatment of PVD or endometritis

39. Randomized controlled trial of PGF
-7 7 14 21 28 35 42 5649 630 300
Disease recording
Exam for
PVD & Endo AI,
pregnancy
diagnosis
NEFA,
BCS
NEFA, BHB,
Haptoglobin
P4 P4 P4 P4
BCS
PGF
All cows
2072 cows in 6 herds
in Ontario, Canada and New York, USA
CCFA
High risk
n=1048

40. PGF Therapy - Clinical Cure
Week 5 status n “Cure” at week 8 P
PGF twice Control
PVD +
(ignoring CYTO)
323 72% 58% .01
• No effect of PGF on cure of endometritis
•Among cows ‘clean’ at week 5 (n=1432), no
effect of PGF on new PVD or endometritis (6 –
8 % incidence)
• No overall effect of PGF on pregnancy rate
Dubuc, LeBlanc et al unpublished

41. Prevalence of CL
Cows with PVD (n = 332)
Variable Week postpartum
3 5 7 9
Cows with serum
P4 > 1 ng/ml (%) 23 43 63 65
Cumulative
proportion (%)
with  1 sample
with P4 > 1 ng/ml
- 49 70 79
CL status at PGF
None 31%
Once 32%
Both 37%
Dubuc, LeBlanc et al unpublished

42. n = 318
Cows with PVD at week 5
Accounting for parity, dystocia/RP/twins, BCS at
calving, herd
2 shots PGF: HR = 1.2 {0.95 – 1.6}
P = 0.07
No PGF*CL interaction
PGF treatment of PVD

43. Clinical Trial in NZ
756 cows with PVD > 14 DIM
26% with palpable CL
Outcome (%) IU Cephapirin to all
CL-dependent
treatment
CL: PGF
No CL: cephapirin
Model P
value
Clinical cure 14 d 84 81 .66
Submission 21 d 86 88 .98
Pregnant 1st service 42 37 .60
42 d In Calf 64 62 .64
Final In Calf (12 wk) 86 87 .46
McDougall et al 2013

44. Anti-inflammatory therapy for
endometritis
• Treatment (Priest et al 2012)
• 213 cows with > 14% PMN at 14 DIM randomized to carprofen 3X
q 3 d starting at 21-25 DIM
• No effect on PMN% at 42 DIM (92-96% < 14%)
• Carprofen decreased time to pregnancy among PMN >14% at 14
DIM
• No effect on production, metabolites, or % anovular at 45 DIM
• Prevention (Meier et al 2014)
• 639 pastured cows in 2 herds in NZ
• Carprofen 1,3,5 or 19, 21, 23 DIM
• No effect on
• PVD or PMN% at 13-24 or 30-49 DIM
• Production (c/f Farney et al 2013 salicylate effects)
• Metabolites (small  in albumin, BHBA ~ 35 DIM)

45. Reproductive health monitoring
3 14 2821 42
Metricheck
Metritis –
daily
screening
????