Inflammatory Mitochondria

Happy New Year everyone! Last post for the year before many of us scatter apart in the upcoming one.


Motivation: Every lecture on sepsis starts with a Venn diagram showing an overlap between SIRS (systemic inflammatory response syndrome) and general infection with the intersection defined as sepsis.  I get the sepsis part, but I have always wondered about other non-infectious causes of SIRS like pancreatitis or trauma.  Since the injured tissue is our own and subject to constant immunologic surveillance, what exactly is so inflammatory about it when injured? Recently, a group in Boston examined this issue and proposed a remarkable hypothesis to explain the phenomenon.


Paper: Circulating Mitochondrial DAMPs Cause Inflammatory Responses to Injury. Zhang, Q. et. al.  Nature (2010) 464: 104-107.  http://www.nature.com/nature/journal/v464/n7285/full/nature08780.html

Hypothesis: The authors observed that when innate immunity responds to infection, specific signals expressed on invading micro-organisms called pathogen-associated molecular patterns (PAMPs) are recognized.  Some of these specific signals like N-formylated peptides are also expressed by mitochondria, which are thought to have a bacterial origin.  The authors postulated that when human tissue is injured, release of the mitochondrial products is inflammatory.

Method (clinical): Plasma was collected from 15 patients presenting with acute accidental trauma.  The age range was 17-71.  Patients did not either have significant medical co-morbidities or have major open or intestinal injuries. For controls, plasma was collected from healthy volunteers aged 26-61.  Most did not have chronic illnesses except two controls with Type II diabetes.

Results: I will just highlight some of the key results from the paper.

When trauma patients were compared to healthy controls, the concentration of mitochondrial DNA (mtDNA) in trauma plasma was 2.7 ug/mL (sem 0.94) compared to barely detectable in control plasma.  Bacterial products (tested by measuring ribosomal subunit 16S RNA) was absent from all samples.  To demonstrate the inflammatory and activating nature of mitochondrial products, human neutrophils exposed to mitochondrial products showed signs of activation by producing cytokines like IL-8 and expressing proteases such as matrix-metalloprotease 8 (MMP-8).  Other downstream activation cascade markers were also increased. In culture, human neutrophils became activated and migrated towards regions with mitochondrial products.  In the paper, the authors also identify specific receptors in neutrophils that are likely involved in sensing mitochondrial products.

To test biological significance, rats were intravenously given mitochondrial degradation products equivalent to a 5% liver injury.  Remarkably, as in sepsis ARDS, rats showed marked oxidative lung injury, increased pulmonary permeability, accumulation of IL-6, and PMN infiltration into airways.  Rat livers also demonstrated increased PMN accumulation.

Discussion: The paper has clinical relevance because it showed (1) that after trauma, the concentration of mitochondrial products is increased in plasma and (2) that mitochondrial products are inflammatory and can lead to injury at least in animal models.  While tissue injury may release a number of products that are inflammatory, this paper pins down a novel pathway that at the very least contributes to the process and provides concrete mechanistic basis of why tissue injury is inflammatory.  A direct consequence of describing these interactions could be development of inhibitors that could pharmacologically decrease the PMN activation process and decrease the severe damage that tissue injury like pancreatitis produces.

The Two Hour Rule

Motivation: On the residency interview trail, I was put on a sample rotating ward team at a hospital, and one of the patients examined by the team had a significant pressure ulcer.  After examining the patient, the team talked about the importance of frequent repositioning to prevent pressure ulcers.  The standard of care is repositioning every two hours.  After listening to this case, I wondered how the figure of two hours entered clinical practice.  If I sit in the same position for even an hour, I feel pretty uncomfortable.  Turns out that the two hour guideline is partly expert opinion.  There have been some trials with inconsistent results.  Are there studies to measure effectiveness?

Paper: Frequent manual repositioning and incidence of pressure ulcers among bed-bound elderly hip fracture patients. Rich, S. et. al. Wound Repair and Regeneration (2010): 1-9.

Methods: The study examined bed-bound elderly patients 65 or older who underwent surgery for hip fracture.  During the first five days of hospitalization, study nurses assessed the frequency of repositioning through chart review.  Study nurses also determined the presence of stage 2+ pressure ulcers (at least partial thickness dermal loss or presence of blister) at baseline (within five days of hospitalization) with follow-up on alternating days for 21 days.  The study was conducted in seven hospitals in Maryland and two in Pennsylvania.

Results:

A total of 269 patients entered the study.  Overall, only 53% of patients were repositioned at least every two hours.  There was wide inter-hospital variability with repositioning rates varying from 23-77% among hospitals.  Patients most likely to be repositioned were those with pressure ulcer at time of admission.

Effect of repositioning: Pressure ulcers developed in 12% of patients repositioned frequently compared to 10% of patients repositioned less frequently.  The difference in incidence rate  was statistically not-significant (Incident Rate Ratio (IRR): 1.22 with 95% CI of 0.65-2.30).  Even after adjusting for covariates such as BMI, different support surfaces, peripheral vascular disease, nutritional status, disease severity, comorbidities, and other variables, the incident rate ratio (IRR) was not significantly, IRR: 1.12 (0.52, 2.42).

Conclusion:

In this study, frequent repositioning was not associated with reduced incidence of pressure ulcers.  In this study, there are a couple of important caveats.  The first is that the repositioning data was taken in a cross-sectional fashion in the first five days while pressure-ulcer incidence was followed for 21 days.  This technique does not capture what repositioning the patients received in the subsequent 16 days after the visit (although presumably hospitals which have repositioning in the first few days would continue the practice).  Also, the frequency of repositioning was determined from chart recordings by nurses.  The recordings may not accurately reflect the actual frequency of repositioning and may over-represent the frequency of repositioning.  Another significant finding is that the standard of care repositioning is only followed for 53% of patients despite nationwide emphasis on this issue. 

Overall, the data suggest that even with repositioning every two hours, the strategy may not be significant. Adjusting for pressure support techniques did not make a difference. Other factors such as exact position of reposition and more novel pressure support techniques may turn out to be important. While this study does not conclusively show that repositioning every two hours is ineffective, the study certainly highlights the need for a large scale randomized trial to examine the issue robustly.

Gray Hair

Motivation: For the past two days, I was caught in the blizzard at Minneapolis and spent a lot of idle time reading newspapers at the airport.  One of the faces that caught my attention was that of the WikiLeaks founder Julian Assange - his hair is just so white.  Even looking around the airport, I saw that hair graying proceeds at remarkably different rates.  We all associate hair graying with ageing, but is it also linked with diseases of ageing like vascular disease?  Turns out that this issue has also been studied.

Paper: "Gray hair, baldness, and wrinkles in relation to myocardial infarction: The Copenhagen City Heart Study", Schnohr, P. et. al. American Heart Journal, 1995 (130): 1003-1010.  http://www.ahjonline.com/article/0002-8703(95)90201-5/abstract?source=aemf

Method: The Cophenhagen City Heart Study is a prospective study of 20,000 adult men and women.  In the study, 7,163women and 5,837 men were physically examined between 1976-1978.  Five years later, the subjects were again examined (91% follow-up rate).  Incidence of death and MI were ascertained from questionnaires and from hospital records.  During physical exam, gray hair, baldness, wrinkles, ear lobe crease, and arcus senilis were noted.  Follow-up time for patients was on average 12 years.

Results:
The paper provides excellent statistics of the prevalence of gray hair, baldness, wrinkles, etc. divided by gender and age groups (really interesting), but I will skip to the meat of the matter regarding correlation.  Overall, in the observed time period, women suffered from 226 MI and men suffered from 524 MI.

Gray hair: After adjusting for known risk factors including age, smoking, BMI, cholesterol, etc., a significant correlation existed for men but not for women.  For men, the relative risk was 1.4 (CI: 0.9-2.0) for moderately gray hair and 1.9 (1.2 to 2.8) for completely gray hair.  For women, the relative risk was 1.1 (0.7-1.6) for moderately gray hair and 1.4 (0.9-2.0) for completely gray hair.  Note that in the subgroup, about 25% of women dyed their hair and had to be excluded.  Among women, dyeing hair did not increase incidence of MI, RR - 1.3 (0.6-2.7).

Baldness: Frontoparietal region baldness was associated with incidence of MI in men but not in women.  In men, the relative risk was 1.6 (1.1-2.3) compared to men with no bald triangle.  No significant association was found in crown-top baldness.

Wrinkles: Facial wrinkles over all age groups were not significantly associated in either sex with MI.  For subjects younger than 55,  in men but not women, severe facial wrinkles were associated with MI, RR for men 1.6 (1.1-2.3).

Conclusion: At least in men, after adjusting for conventional known risk factors of MI, completely gray hair was associated with increased risk of MI.  While the same association did not hold true for women, the case is a bit murkier.  First of all, women, as expected, experienced fewer MI than men.  Also, a large fraction of women (~25%) used hair dyes and had to be excluded, which biased the analysis.  So, graying hair may still be a risk factor or may not be a risk factor for women.  Other interesting results were that frontoparietal baldness and severe facial wrinkling in subjects younger than 55 were both associated with MI in men rather than women.  These results suggest that craniofacial ageing may result from different processes in men and women.

Overall, this was one of the largest prospective studies examining the issue of clinical ageing signs and MI.  But, as with any prospective studies, the presence of confounders cannot be ruled out.

The Great Heart Failure Mystery

This post is best viewed from the blog website: http://siriasis.blogspot.com/


Motivation: One of the "bread and butter" diagnoses in inpatient medicine is systolic heart failure exacerbation - we all see it, we all treat it, but do we understand it?  I first wondered about this question last year during my basic medicine clerkship.  The etiology of systolic heart failure  is usually answered with some notion of overstretching the heart.  For me, the difficulty was that it is easy to see how overstretching a single muscle fiber can lead to weakness, but do we really overstretch an entire organ - especially an already dilated organ (like the enlarged heart in dilated cardiomyopathy)?

The traditional teaching and general consensus for decompensated systolic heart failure is that with volume overload, pre-load exceeds the capacity of the Frank-Starling compensation and then an inverse relation holds with increased filling resulting in decreased cardiac output.  The presumed basis is that there may be lack of productive overlap of muscle fibers with increased filling.  But, is there solid empirical data for this explanation?

The answer is shockingly no.  I have even asked cardiologists to confirm this assessment.  Of course, lack of data does not imply that the traditional teaching is false.  So, are there physiological studies about the response of the dilated heart to filling?

Paper: Evidence of the Frank-Starling Mechanism in the Failing Human Heart.  Holubarsch, C., et. al. Circulation (1996); 94: 683-689. http://circ.ahajournals.org/cgi/content/full/94/4/683

Method: Five human hearts with end-stage dilated cardiomyopathy were excised from patients receiving heart transplants.  The average ejection fraction of these hearts was 17%.  These hearts were kept alive artificially (really cool!)  using oxygenated blood.  Pressure and volume relationships were directly measured.  The hearts were paced using epicardial leads.  For controls, two healthy hearts were used which were originally destined for transplantation but could not be transplanted due to technical reasons.

Results:
For this experiment, I think that pictures tell more than any description.  So, I will paste some of the key graphs and point out the pertinent results:

In the graph on the left, the x-axis describes the volume of the left ventricle.  In Figure A, as the volume increases, the pressure inside the ventricle increases.




In B, we see that as the volume increases the diastolic pressure increases and the generated systolic pressure increases as well (the Frank Starling mechanism).  For perspective, the hearts here are really being filled up - the average diastolic volume is about 120 mL.





C.  Observe in this figure that as the LV is being filled up, the net generated pressure (which is simply systolic pressure minus diastolic pressure) decreases despite an intact Frank-Starling mechanism (the diastolic pressure increases at a faster rate than the systolic pressure at high volumes).

In the figure on the right, to investigate the state of Frank-Starling effect on muscle strips rather than whole organ, individual muscle strips were isolated from healthy hearts (top curve) and from failing hearts (bottom curve).  The total tension generated was measured in response to different muscle lengths.  As seen, the healthy heart was able to generate a lot more force in response to increased stretch.  But, while the response of the failing heart was more moderate, there is no evidence of "falling-off" the Starling curve.






Conclusion: This paper does not conclusively answer the physiologic basis of dilated cardiomyopathy but sheds some light on couple of areas.  (1) Even in failing hearts with weak contractility, the Frank-Starling effect is likely preserved even when stretched out by volume overload.  (2) Besides preload, cardiac contractility is controlled by two other factors: heart rate (increased contractility at increased heart rates) and neurohumoral responses (such as to catecholamines).  It may be possible that increased volume somehow upsets the responses to the latter two mechanisms rather than disrupting the Frank-Startling mechanism.  Anyway, there is a story yet to be told ....