Near-infrared spectroscopy of the lower limb in patients undergoing endovascular aneurysm repair: A systematic review

Near-infrared spectroscopy of the

lower limb in patients undergoing

endovascular aneurysm repair: A

systematic review

Version 2

Ilya Zorikhin-Nilsson

Version 2

Degree Project, 15 ECTS

c

Ilya Zorikhin-Nilsson, 2015

Examiner: Torbj¨orn Nor´en MD, PhD, Assoc. Prof.

Supervisor: Bengt Hammas MD, PhD

¨ OREBRO UNIVERSITY SE-701 82 ¨Orebro, Sweden www.oru.se Phone: +46-(0)19-30 30 00

undergoing endovascular aneurysm repair: A systematic

review

Introduction: Prior research shows that near-infrared spectroscopy is a reliable means of monitoring lower limb perfusion. Furthermore, endovascular aortic re-pair leads to ischemia that may cause irreversible damage to the limb.

Objective: To survey clinical trials that compare interventions with and with-out near-infrared spectroscopic monitoring of the leg during endovascular aortic repair surgery, and to determine if there is a difference in clinical outcome. Method: A systematic review using the PubMed search engine was conducted. Twenty eight articles were found, of which 15 met the initial inclusion criteria. Four remained after filtering, but none was deemed adequate for final inclusion. Result: The observed apparent lack of proper clinical trials rendered it impos-sible to determine whether the clinical outcome could be improved or not. Conclusion: Finally, it was concluded that additional research, of the clinical sort, is required to determine possible intervention benefits or drawbacks.

Keywords: review, near-infrared, spectroscopy, lower limb, endovascular, aneurysm, repair, outcome

Abstract i

1 Introduction 1

1.1 Aim . . . 1

1.2 Related work . . . 2

1.3 Endovascular aortic repair . . . 2

1.4 Ischemic cell injury . . . 3

1.5 Hemoglobin and oxygen transport . . . 3

1.6 Brief history of oximetry . . . 5

1.7 Introduction to the physics of optical oximetry . . . 7

1.7.1 Transmittance oximetry . . . 7 1.7.2 Reflectance oximetry . . . 10 2 Method 12 2.1 Search terms . . . 12 2.2 Inclusion criteria . . . 13 2.3 Exclusion Criteria . . . 13 2.3.1 Strict . . . 13 2.3.2 Soft . . . 13 2.4 Ethical considerations . . . 13 3 Results 14 3.1 Search results . . . 14 3.2 Findings . . . 15 4 Discussion 16 5 Conclusion 17 ii

1

Introduction

Oxygen is a vital requirement of human metabolism. Oximetry, the science of measuring its saturation level in blood (SO2), has many applications

within the field of medicine. It can, for instance, be used to determine the function of the heart and lungs or to supervise oxygenation of the brain during surgical procedures[1][2].

According to Harrison[3] studies have shown that hypoxia substantially reduces the anti-bacterial properties of neutrophils, that it hampers the dif-ferentiation of fibroblasts and the synthesis of collagen in healing wounds. Furthermore, due to the greater prevalence of ischemic hypoxia in some lower extremity wounds he suggests that it may be beneficial to monitor wound oxygenation in order to evaluate healing viability and to prevent infection.

In situations with risk of damage to the lower limbs, e.g. due to hypoxia due to prolonged tissue hypoperfusion, which for example may occur dur-ing endovascular aortic repair (EVAR) procedures, we suspect that non-invasive oximetric instruments based on near-infrared spectroscopy (NIRS) may provide an opportunity to introduce timely countermeasures through which necrosis and subsequent amputation can be avoided. The above will be the main hypothesis that we aim to explore in this work.

1.1

Aim

The PICO-method[4] is commonly used to design answerable questions in evidence-based medicine and was utilized to outline the goals of this study further.

P – Population : Patients having undergone EVAR surgery. I – Intervention : Intraoperative NIRS monitoring of the leg. C – Comparison : Without intraoperative NIRS monitoring. O – Outcome : Amputation.

To summarize, the objective of this systematic review is to survey interven-tions with and without NIRS monitoring of the leg, during EVAR-surgery specifically, and to determine whether there is a difference in outcome con-cerning rate of amputation in the time frame of one week, or not.

1.2

Related work

In a pilot study, Boezeman et al.[5] arrive at the conclusion that NIRS is a reliable means of monitoring lower limb perfusion during surgery of the central aorta. Taillefer et al.[6] have systematically reviewed the clin-ical efficacy of cerebral NIRS in adult heart surgery and found it to be a promising technology, but concluded that more research is required. A similar survey was made by Zacharias et al.[7] who support the notion that additional clinical trials are needed.

Furthermore, while close to the subject, but to a greater extent touching the aftermath of amputation, a systematic review on “The use of transcuta-neous oximetry to predict healing complications of lower limb amputations” has been made by Arsenault et al.[8], which shows that transcutaneous oxygen saturation measurements are able to predict healing outcomes after lower limb amputations. However, it was unable to determine if the addi-tional information provided by the instruments could be of clinical benefit. Bongard et al.[9] conclude that the technology when used pedally can be utilized to “predict an amputation within a few months” in patients with severe arterial occlusive disease. In a systemic review, Vardi et al.[10] have assessed the technology “as a method for the diagnosis and evaluation of peripheral vascular disease” where they found that it can be of assistance to the diagnosis and evaluation of patients affected by the condition at hand.

Finally, in a recent review Mesquida et al.[11] have summarized “the exist-ing evidence” on the utility of NIRS in the context of critically ill patients. They found that “the lack of randomized controlled trials” is an obstacle, but that the technology might be useful in settings “where cardiovascu-lar performance needs to be challenged”, for example at the removal of mechanical ventilation, as during veno-venous extracorporeal membrane oxygenation (VV-ECMO).

1.3

Endovascular aortic repair

Endovascular aortic repair (EVAR) is a method by which a stent is intro-duced through a femoral (or radial or brachial) artery to support the aortic wall e.g. in patients with abdominal aortic aneurysms. The stenting proce-dure effectively causes an obstruction of the arteries which may bring a risk of ischemic injury to the leg at the time of (and after) surgery, according

to the findings of Jonsson et al.[12] and Maleux et al.[13]. Invented in 1986 by Volodos et al., the method can be said to be relatively modern.[14][15]

1.4

Ischemic cell injury

By definition ischemia is a state of tissue hypoperfusion, i.e. where local blood supply is insufficient to cover local metabolic needs. It is important to distinguish it from hypoxia, as non-ischemic hypoxic states allow for the transport of nutrients and continued metabolism through anaerobic glycolysis, depending on tissue. Normal aerobic cellular respiration reduces oxygen to release energy to reform ATP from ADP. In turn, ATP is required to fuel the plasma membrane Na+/K+ pump, the failure of which leads to an osmotic pull that swells the cell and that damages e.g. the cytoskeleton and the plasma membrane. Cell destruction occurs from necrosis and to some extent from apoptosis. The process can be mitigated by inducing hypothermia which reduces the metabolic requirements, while suppressing chemical processes that lead to inflammation as well as the creation of free radicals. Additional injury may occur during reperfusion, possibly partly due to free radicals being formed through the reoxygenating process and the reduced capability of tissue to cope with it, due to existing damage.[16]

1.5

Hemoglobin and oxygen transport

During equilibrium between a mixture of ideal gases and a liquid, a result of thermodynamics, Henry’s Law[17], states that the concentrations of each dissolved gas in the liquid is proportional to their corresponding partial pressures. Concretely for oxygen that dissolves in blood, at the alveolar interface, we have

[O₂]_dis= k_O2· P_O2, (1) where k_O2 is the solubility constant and P_O2 the partial pressure of oxygen in the alveolus. At standard temperature and pressure (STP), [O2]dis= 3

(ml O2/l of arterial blood). With a cardiac output of 5 l/min, and

instan-taneous equilibration, we get 15 ml O2/min, which accounts for about 6%

of the required amount, typically estimated at 250 ml O2/min.[18]

Luckily, evolution has arrived at a solution for a more efficient way of trans-porting oxygen. By putting forth a protein that both has a high (variable) affinity for oxygen and a large carrying capacity an additional 200 ml O₂/l blood can be delivered. That protein is hemoglobin (Hb).

As with many other proteins in our body, hemoglobins come in many dif-ferent types and shapes. The most common adult version of hemoglobin, HbA, contains four monomers, each containing two main units. The first, the heme, is a heterocyclic organic ring, a so called porphyrin, that carries a Fe2+ ion in its center. Each iron ion has the capacity of binding one oxygen molecule. The second unit, the globin, is a polypeptide that either consists of 141 or 146 amino acids, designated α or β respectively. It forms helices that envelop the heme to modulate its affinity to oxygen, so that it is less prone to bind oxygen in conditions of low local PO2 (T-state), which

is useful in areas of high oxygen demand. In ventilated alveoli, on the other hand, affinity has to be high and the molecule enters a so called R-state.[18]

Additionally, the iron-porphyrine compound of the heme consists of many conjugated bonds, where electrons are shared across the molecule. The larger the molecule the less electromagnetic energy is required to excite one electron by absorbing a photon. Red is at the edge of the visible spec-trum of light and corresponds to low energy. If the heme carries an oxygen molecule, the typical red color can be recognized, while a more purple tone is associated with deoxygenated heme. This property is used in oximetry. Absorption coefficient spectra of hemoglobin are shown in figure 1.[18]

Another type of hemoglobin is methemoglobin (metHb), that carries an oxidized, ferric (Fe3+), iron ion, that is unable to bind oxygen. MetHb accounts for about 1.5% of the total Hb amount in non-pathologic condi-tions. Yet another type is myoglobin (Mb), which is available in muscle. It is capable of carrying only one O2, but with a higher affinity, and thus

passes on oxygen from normal hemoglobin to tissue that gets oxygenated faster. The prenatal hemoglobins (the embryonic Gower 1, Gower 2, Port-land and the fetal HbF) are characterized by different combinations of α and β chains. Fetal hemoglobin shares the property of Mb of having a high affinity to oxygen, with the principle being to transfer oxygen from the HbA of the mother to the fetus. HbF rarely exceeds 1 − 2% of the total Hb in the adult. Other so called minor-component hemoglobins exist as well and may become of greater significance during episodes of glycosylation, for example in diabetes mellitus patients, where they are useful as long term blood glucose level markers. One of them, the HbA2 accounts for 2.5%

Figure 1: Absorption coefficient spectra of hemoglobin species at a concen-tration of 150 g/l and a molar mass of 64.5 kg/mol (based on data compiled by Prahl[19]).

hemoglobin HbS, associated with hemolytic anemia, but that famously re-duces the complications of malaria.[18]

It is important to keep the existence of the different hemoglobin types in mind as they have different properties and potentially may contribute to errors in oximetric readings by interfering with the normal absorption spectra of hemoglobin.[20]

Oxygen saturation The oxygen saturation SO2 can be defined as “the

ratio of concentrations of oxyhemoglobin to total hemoglobin”[21]

SO₂= [HbO2] [HbO2] + [Hb]

, (2)

where HbO2and Hb correspond to oxy- and deoxyhemoglobin respectively.

1.6

Brief history of oximetry

Before the advent of oximeters clinicians observed the degree of cyanosis and the rate of chest movement so as to obtain rough estimates of oxygena-tion and recovery time. In the mid 19th century when N2O was introduced

part of normal clinical practice. However, as was later discovered, less than half of the physician staff could successfully detect cyanosis unless SO2< 80% this way. The ocular method was too inaccurate.[22]

The first oximeter of practical clinical use was invented by D. van Slyke in 1922. He used a chemical setup in which a sample of blood was added to a chamber of a flask. Opsonin was then used to detach oxygen from hemoglobin. A vacuum was applied to drive the solution to another flask where the oxygen separated at a partial pressure that could be measured by a manometer. In the next step another sample of blood was saturated with oxygen and the procedure was repeated. By reading the manome-ter values the concentration of oxyhemoglobin in the first sample could be calculated[23]. But, it was a slow procedure and sampling a patient for blood interferes with his or her respiratory function.[24]

As an alternative to arterial blood, also starting in the 1920ies, oxyhe-moglobin concentrations in blood samples taken from capillaries in the fin-gers or in the ear lobes were shown to differ no more than 5% from arterial readings in 90% of the cases. It was faster, but invasive still.[24]

In 1900, Vierordt had observed that by applying a tourniquet to a finger, he could modulate the intensity of light transmitted through its pertain-ing tissue[22]. Later, Kramer and Sarre discovered that whole blood light transmission could be approximated by the Beer-Lambert law.[24]

In his 1942 article “The Oximeter, an Instrument for Measuring Contin-uously the Oxygen Saturation of Arterial Blood in Man” G. A. Millikan[21] described a device consisting of a miniature lamp bulb, two color filters and a couple of selenium photo cells that weighed only 30 grams and that could be mounted on the ear. He thus coined the term “oximeter”. The first filter produced light with wavelength components that were equally absorbed by oxy- and deoxyhemoglobin, while the second filter produced a color that was absorbed differently in each of the hemoglobin types. Thanks to this setup, arterial samples that otherwise could invoke anxiety and pain in the patient could be avoided for the first time. The Millikan oximeter operated according to the principles described in section 1.7, and had an average of 3% difference to the arterial blood gas sample values that were used as reference.

Brinkman et al. developed the first light reflection oximeters in 1949. A sensor was attached to the foreheads of patients, like a cyclop[22]. Further on, in 1958 a flow oxy(hemo)meter was developed by Sochivko et al., which allowed simultaneous monitoring of oxyhemoglobin concentrations in both arterial and venous blood.[24]

In the early 1970ies, T. Aoyagi studied transmissions made through the auricle and discovered a pulsatile component in the signal which cancelled out if he subtracted the IR-signal from the red one. But, he did find that it would not cancel out during hypoxemia induced by breath-holding, and decided to use that fact to measure saturation. Based on the assump-tion that blood only contained either reduced or oxygenated hemoglobin Aoyagi achieved sufficient accuracy with his device. Pulse oximetry was invented.[24]

A few years later, in 1977, J¨obsis discovered that infrared light absorp-tion by human tissue was negligible and that it could be used for brain monitoring.[24][25]

Cost has been a disadvantage for the adoption of tissue spectrophotometry. Instruments earlier carried price tags in the order of hundreds of thousands SEK, but new technologies have been able to reduce costs of microspec-trophotometers and made SO₂ measurements more accessible, according to Harrison[3].

1.7

Introduction to the physics of optical oximetry

In general, there are two types of optical oximetry. The first being trans-mittance oximetry, which relies on measuring light that has passed through tissue, while the second type is characterized by measuring reflected light. Manual, ocular methods, can perhaps be said to belong to the second group. 1.7.1 Transmittance oximetry

Transmittance T through a material is defined as the fraction of the inten-sities of the outgoing Iout and incoming light Iin. The Beer-Lambert law,

equation 3, states that the transmittance T relates exponentially to prod-uct of the path length of the light beam l (m) within the substance and the absorption coefficient µ (m−1) which is a material (and concentration) specific property.

Figure 2: As light passes through a medium, absorption occurs.

T = Iout I_in = e

−µl ₍₃₎

The absorbance A, on the other hand, is defined as the natural logarithm of the aforementioned intensities, thus the Beer-Lambert law can be expressed as follows:

A= − logIout Iin

= − log T = µl (4)

which holds for a uniform material with negligible scattering, and which re-lates the absorption coefficient spectrum, figure 1, to the absorbed amount.

As perhaps can be seen from that figure absorption is relatively low in the near infrared spectrum (above 700 nm) compared to visible light (390 to 700 nm), which means that red and infrared light will manage to pass deeper into or even through tissue. If you have watched a bright light from behind your eyelids or perhaps through your finger, you will have noticed a particular red tone that reveals that most other colors have been filtered out by the tissue. In addition, we also observe that there is a major dif-ference in absorption between oxy- and deoxyhemoglobin starting at about 600 nm (red) and ending at about 800 nm (infra-red).

For N simultaneous absorbers, of significance, in a solution, as in the case with our two oxy- and deoxyhemoglobin species, we get contributions from each in the following way

µ (λ ) =

N

∑

i=1

µi(λ ). (5)

convert the absorption coefficients into something that singles out their in-herent respective substance concentration contributions. By formally con-verting the expression into one involving the molar absorption coefficient ε = ε (λ ) instead, we get µ (λ ) = log 10 · N

∑

With N = 2, in our scenario, we simply arrive at

µ10(λ ) = c1ε1+ c2ε2= [Hb]εHb+ [HbO2]εHbO2. (7)

But, unfortunately we have two unknowns and only one equation. To get past this problem, we discover the need to make measurements at two differ-ent wavelengths. We introduce the absorbance ratio ρ = A_λ1/A_λ2= µ10(λ1)·l

µ10(λ2)·l,

which we can measure, and obtain an expression for oxygen saturation

SO2=

ε (λ1)Hb· ρ − ε(λ2)Hb

(ε(λ1)Hb− ε(λ1)HbO2) · ρ + ε(λ2)HbO2− ε(λ2)Hb

, (8)

by using the definition, equation 2. Now, this is quite messy, however, if we look at the expression in parenthesis in the denominator, we see that it might be possible to pick a λ1 so that the parenthesis vanishes

alto-gether. To do this we pick a filter that generates a wavelength where both species of hemoglobin show the same molar absorption coefficient (obtain-able from figure 1, i.e. where the curves intersect, a so called “crossover wavelength”)[21]

SO₂= ε (λ1) · ρ − ε(λ2)Hb ε (λ2)HbO2− ε(λ2)Hb

. (9)

A further benefit of this choice of wavelength is that A_λ1= ε(λ1) · (c1+ c2) · l

will be proportional to the total hemoglobin concentration c = c1+ c2.

Ad-ditionally, this linear expression for SO₂ shows that we should select λ2 in

a way so that the numerator does not equal zero, i.e. a wavelength where both hemoglobin species have distinct optical properties. The oxygen satu-ration can now be acquired optically by simply measuring ρ and computing the result, which concludes the discussion of the basic operating principles of transmission oximeters.

Pulse oximetry The pulse oximeter specifically measures the arterial oxygen saturation, by making a series of temporal measurements using at least two diodes firing in sequence.

Absorption values conceptually similar to the ones shown in figure 3 are obtained as a current where the AC-component reflects the varying arte-rial absorption. For each wavelength the AC component is divided by the respective DC component, which can be related to SO₂ as follows from 1.7, for example.

Figure 3: Overview of the component contributions to the total absorption over time in pulse oximetry.

1.7.2 Reflectance oximetry

In measurements involving tissue that cannot be shone through, trans-mission oximetry is of little use. For this reason a technique involving reflectance oximetry is applied.

Figure 4: Concept of scattering and absorption in skin.

As light passes through tissue it scatters upon encountering larger struc-tures like cells or tissue layer changes in general, while a fraction gets ab-sorbed by chromophoric molecules like hemoglobin. A part of the emitted light manages to travel back to the sensor, as illustrated in figure 4.

Scattering leads to the fact that mean photon travel times (or distances) through tissue increase. Studies have quantified so called “differential path-length factors” (DPF) that reflect this change for different kinds of tissue. For the average adult head the value was found to be 6.3 whereas the value for the adult leg was measured to be 5.51, which indicates that leg tissue scatters less light at the particular wavelengths studied compared to brain tissue.[26]

This complicates the use of the unmodified Beer-Lambert law, especially as different wavelengths have different scattered path lengths, and it can be modified in numerous ways to account for different factors of importance like the amount of light lost to scatter G[27]

A₁₀= − log₁₀Iout I_in = N

∑

Furthermore, by using multiple receivers it is possible to place the transmit-ter in such a way so that light arriving at a proximal point can be assumed to have travelled mostly through surface tissue while light entering a distal sensor can be assumed to have passed through deeper levels of interest. This method is called “spatially resolved spectroscopy” (SRS). Computer models of pathways support the notion that photons that travel deeper down inside tissue to a larger extent end up more distally to the source. The values computed for the peripheral layers can thus be used to correct the distal readings to get the saturation value for the area of interest.[26][28]

Phase shift techniques, like phase modulated spectroscopy, can be used as well[29]. A laser light of a known frequency beamed into a target returns light that will be slightly out of phase due to the energy loss which is pro-portional to path length. By measuring it, absorption and scattering can be quantified to compute changes in cerebral HbO2 and Hb concentrations.

Near infrared spectroscopy NIRS A typical NIRS oximeter consists of a light source, a combined emitter/receiver, a processor and a display[11]. Laser photodiodes emit light at specific wavelengths, while photomultipliers or photodiodes may be used at the receiving end. The processor then performs the computations required to convert changes in attentuation into hemoglobin concentration.[26]

2

Method

As a basis for the systematic search the PubMed search engine was used.

2.1

Search terms

The following paragraphs describe how the search term was built.

To begin with, we are interested in the application of a specific technol-ogy or measurement method. The search term “[(”Spectroscopy, Near-Infrared”[Mesh]) OR (INVOS) OR (NIR*) OR (”Blood Gas Monitoring, Transcutaneous”[Mesh]) OR (”Oxygen Consumption/analysis”[Mesh])]” should cover the transcutaneous monitoring instruments of interest, and perhaps more. It is compulsory to our search.

Now, we need to add the compulsory measurement locale: “AND [(leg) OR (lower limb) OR (”Lower Extremity”[Mesh])]”.

The main procedure which is EVAR should also be included as a must. For purposes of completeness the search includes all vascular surgical pro-cedures in general:

“AND [ (”Endovascular Procedures”[Mesh]) OR (”Vascular Surgical Proce-dures”[Mesh]) OR (”Aneurysm”[Mesh]) OR (”aneurysm repair”) OR (”En-dovascular aneurysm repair”) OR (*EVAR)].

Finally, the outcome must be described:

“AND [(”Amputation”[Mesh]) OR (”Treatment Outcome”[Mesh]) OR (”Vas-cular Surgical Procedures/adverse effects”[Mesh]) OR (”Endovas(”Vas-cular Pro-cedures/adverse effects”[Mesh])]”.

In total we get “( (”Spectroscopy, Near-Infrared”[Mesh]) OR (INVOS) OR (NIR*) OR (”Blood Gas Monitoring, Transcutaneous”[Mesh]) OR (”Oxygen Consumption/analysis”[Mesh]) ) AND ( (leg) OR (lower limb) OR (”Lower Extremity”[Mesh]) ) AND ( (”Endovascular Procedures”[Mesh]) OR (”Vas-cular Surgical Procedures”[Mesh]) OR (”Aneurysm”[Mesh]) OR (”aneurysm repair”) OR (”Endovascular aneurysm repair”) OR (*EVAR) ) AND ( (”Amputation”[Mesh]) OR (”Treatment Outcome”[Mesh]) OR (”Vascular Surgical Procedures/adverse effects”[Mesh]) OR (”Endovascular Procedures/ adverse effects”[Mesh]) )”.

2.2

Inclusion criteria

Inclusion criteria were set as follows. Research of relevance was deemed to be either clinical trials, comparative studies or validation studies. For sim-plicity, as the search was performed in English and in a US-based database, the required language was set to English.

2.3

Exclusion Criteria

Two types of separate exclusion criteria were set up to allow for the option of viewing EVAR as a subset of all vascular surgical procedures that may lead to ischemia of the lower limbs, which should be covered by the search. 2.3.1 Strict

At least one separately studied patient group of the study must clearly have undergone an EVAR procedure. In general this indirectly adds a date interval of interest, ranging from the time of invention, i.e. the first experimental trials in 1986, until today.

2.3.2 Soft

If the assumption is added that EVAR is a subset of all vascular surgical procedures that may cause ischemia of the lower limbs, and that the leg is agnostic to any intervention that causes ischemia through halted tissue perfusion past a. femoralis, general performance of NIRS in the area may be evaluated, and in turn, indirectly EVAR.

2.4

Ethical considerations

Considering that this systematic review covers published and ethically ap-proved studies, readily available in PubMed, it is deemed that no new, or additional, ethical approval is required.

3

Results

The database search was successfully run. The results, and the processing of articles, are documented in the sections that follow.

3.1

Search results

In May 2015, a search without the beforementioned inclusion criteria gen-erated 28 results of which 15 met the inclusion criteria. One article was written in Polish, but featured an English abstract. Since all search results potentially may include studies of relevance in their reference lists, and since the amount found was relatively small, no study was excluded at this point.

Of the 28, 17 were excluded after reading their abstracts. Articles that were excluded focused on angioplasty in diabetics (n = 6), surgery (n = 5), bone marrow treatments (n = 2), ischemic preconditioning (n = 1), hypogastric revascularization (n = 1), or investigated ABI (n = 1) or acid-base changes (n = 1). Eleven were left for full text evaluation, of which four met the original inclusion criterion. As a result, the seven that did not meet the requirements were set to be scanned for references as a verification of the search. Although of general interest, no new references of relevance were found. The article filtering process is illustrated in figure 5.

After having applied the strict criterion, no articles were left past a full text review, as none of them studied EVAR specifically.

Filtering the inclusion-passed articles through the soft criterion resulted in zero articles. Two articles evaluated NIRS detection sensitivities rather than treatment outcomes in patients. The other two applied the technology to manage critical (chronic) limb ischemia outside of surgery.

3.2

Findings

4

Discussion

Considering that near-infrared spectroscopy and endovascular aneurysm re-pair are relatively novel technologies, clinical EVAR has only been used, at a scale, at least locally, for less than ten years[30], the initial expectation was that few clear-cut clinical trials would be available as a basis for any far-reaching conclusions to be drawn. But, the fact that no study at all combined the two, despite the existence of trials that point at the possible complications of EVAR and the possible prospects of oximetry in manage-ment of ischemia, was unexpected. This study was not able to provide any results on the matter other than to heavily emphasize the absence of, and perhaps the need for, additional clinical trials. Although a respectable amount of filtered studies applied NIRS to the lower limbs, they mainly evaluated if readings could be indicative to the status of the limbs. The researchers did not follow up patients to see if their readings could bring actual clinical benefit to treatment or prognosis, which would have been of interest.

Perhaps more fundamentally, the clincal validity of oximetry in general, does not seem to have been fully established, as research by Taillefer[6], Zacharias[7], Arsenault[8] and Mesquida[11] indicates. According to Owen-Reece et al.[26] Schwartz et al. found the cerebral saturation in six deceased subjects to be greater than the lowest values for healthy adults. However, many of these studies were made some time ago, e.g. the paper by Schwartz was published in 1997, and may as such not reflect the current state of the field. Studies that would validate the clinical use of oximetry pre-suppose that a range of reported results from various applications exist, and that they are up to date with clinical practice.

For reasons of perspective, a giant intervention review on perioperative pulse oximetry, including data from 22 992 patients, performed by Peder-sen et al.[31], updated in 2014, found that pulse oximetry indeed reduces incidence of hypoxaemia by 1.5 to three times, but found no evidence that it “affects the outcome of anaesthesia for patients”. However, they simulta-neously state that blinding to participant allocation could not be done as staff had to be able to respond to instrument readings, which likely is an important ethical factor to consider. They argue that it could be possible that the additional data provided might not be important enough in com-parison to the experience and skill of the operator, but that the question

remains unanswered.

During the process of this study, the thought occured that a more gen-eral approach to the application of oximetry in ischemic limbs could be more fruitful. As the reasons for ischemia vary, especially when comparing micro- and major branch occlusions, generalizations from one to the other are dangerous and perhaps not valid. In this study the decision was made to focus on ischemia arising from EVAR (i.e. an acute major branch occlu-sion), other studies would have to investigate the feasibility of oximetry in, for example, chronic ischemia, separately.

When thinking about the search terms, the required ANDs were deemed to be in line with the minimum requirements of the subject studied. They come at a price, however, every AND limits the amount of found articles, which on the other hand requires a precise amount of ORs to expand the search to at least fully contain the area of interest with a minimum of re-dundancy.

Another limitation lies in the process of manual filtering of articles, which includes a measure of subjectivity, and which ideally should be performed by multiple researchers independently, or in an automated fashion. Addi-tionally, the selected search engine PubMed, in itself, may possibly include biases toward English language publications.

A natural future development, warranted by the inconclusive nature of this study, would be to launch a pilot study into the clinical outcome of the use of near-infrared oximetry during endovascular aortic repair procedures.

5

Conclusion

The aim of this study could not be reached, likely due to a lack of clinical trials on the subject at hand, and no recommendations on the use of near-infrared spectroscopy could be made.

References

[1] Schachner T, Bonaros N, Bonatti J, Kolbitsch C. Near infrared spec-troscopy for controlling the quality of distal leg perfusion in remote access cardiopulmonary bypass. European Journal of Cardio-thoracic Surgery. 2008;34(6):1253–1254.

[2] Murkin JM, Adams SJ, Novick RJ, Quantz M, Bainbridge D, Iglesias I, et al. Monitoring brain oxygen saturation during coronary bypass surgery: A randomized, prospective study. Anesthesia and Analgesia. 2007;104(1):51–58.

[3] Harrison DK. Optical Measurement of Tissue Oxygen Saturation. Int J Low Extrem Wounds. 2002;1(3):191–201.

[4] Huang X, Lin J, Demner-Fushman D. Evaluation of PICO as a knowl-edge representation for clinical questions. AMIA Annual Symposium proceedings / AMIA Symposium AMIA Symposium. 2006;p. 359–363. [5] Boezeman RPE, Kelder JC, Waanders FGJ, de Vries JPPM. Contin-uous surveillance of lower limb perfusion during aortic surgery with near-infrared spectroscopy: a pilot study. Vascular and endovascular surgery. 2011;45(5):407–411.

[6] Taillefer MC, Denault AY. Cerebral near-infrared spectroscopy in adult heart surgery: systematic review of its clinical efficacy. Canadian journal of anaesthesia = Journal canadien d’anesth´esie. 2005;52(1):79– 87.

[7] Zacharias DG, Lilly K, Shaw CL, Pirundini P, Rizzo RJ, Body SC, et al. Survey of the clinical assessment and utility of near-infrared cerebral oximetry in cardiac surgery. Journal of Cardiothoracic and Vascular Anesthesia. 2014;28(2):308–316.

[8] Arsenault KA, Al-Otaibi A, Devereaux PJ, Thorlund K, Tittley JG, Whitlock RP. The use of transcutaneous oximetry to predict heal-ing complications of lower limb amputations: A systematic review and meta-analysis. European Journal of Vascular and Endovascular Surgery. 2012;43(3):329–336.

[9] Bongard O, Kr¨ahenb¨uhl B. Predicting amputation in severe ischaemia. The journal of bone and joint surgery. 1988;70-B(3):465–467.

[10] Vardi M, Nini a. Near-infrared Spectroscopy for Evaluation of Periph-eral Vascular Disease. A Systematic Review of Literature. European Journal of Vascular and Endovascular Surgery. 2008;35(1):68–74. [11] Mesquida J, Gruartmoner G, Espinal C. Skeletal muscle oxygen

satu-ration (StO2) measured by near-infrared spectroscopy in the critically ill patients. BioMed Research International. 2013;2013.

[12] Jonsson T, Larzon T, Jansson K, Arfvidsson B, Norgren L. Limb ischemia after EVAR: an effect of the obstructing introducer? Journal of endovascular therapy : an official journal of the International Society of Endovascular Specialists. 2008;15(6):695–701.

[13] Maleux G, D P, Koolen M, Heye S. Complications after Endovascular Aneurysm Repair. RadioGraphics. 2009;1(212):3–9.

[14] Volodos NL. The First Steps in Endovascular Aortic Repair : How It All Began. Journal of Endovascular Therapy. 2013;20(Suppl):I3–I23. [15] Bergqvist D. Val av spr˚ak och forum – avg¨orande f¨or efterm¨alet. L¨

akar-tidningen. 2010;107(38):2259–2261.

[16] Kumar V, Abbas AK, Fausto N, Aster JC. Robbins and Cotran Patho-logic Basis of Disease. 8th ed. Philadelphia: W.B. Saunders; 2010. [17] Schroeder DV. An Introduction to Thermal Physics. San Francisco:

Addison Wesley Longman; 1999.

[18] Boron WF, Boulpaep EL. Medical physiology: a cellular and molecular approach. 2nd ed. Philadelphia: Elsevier Saunders; 2012.

[19] Prahl S, Gratzer WB, Kollias N. Raw Data: Optical Absorption of Hemoglobin;. [Online] [Accessed on 2015-05-14]. Available from: http: //omlc.org/spectra/hemoglobin/summary.html.

[20] Verhovsek M, Henderson MPa, Cox G, Luo HY, Steinberg MH, Chui DHK. Unexpectedly low pulse oximetry measurements associated with variant hemoglobins: A systematic review. American Journal of Hema-tology. 2011;86(8):722–725.

[21] Millikan GA. The oximeter, an instrument for measuring continuously the oxygen saturation of arterial blood in man. Review of Scientific Instruments. 1942;13(10):434–444.

[22] Severinghaus JW. Monitoring oxygenation. Journal of Clinical Moni-toring and Computing. 2011;25(3):155–161.

[23] Van Slyke DD, Neill JM. The determination of gases in blood and other solutions by vacuum extraction and manometric measurement. I. Journal of Biological Chemistry. 1924;61(2):523–73.

[24] Zislin BD, Chistyakov AV. The history of oximetry. Biomedical Engi-neering. 2006;40(1):53–56.

[25] J¨obsis FF. Discovery of the near-infrared window into the body and the early development of near-infrared spectroscopy. Journal of Biomedical Optics. 1999;4(4):392–396.

[26] Owen-Reece H, Smith M, Elwell CE, Goldstone JC. Near infrared spectroscopy. British journal of anaesthesia. 1999;82(3):418–426. [27] Madsen PL, Secher NH. Near-infrared oximetry of the brain. Progress

in Neurobiology. 1999;58(6):541–560.

[28] McCormick PW, Melville S, Goetting MG, Balakrishnan G. Regional Cerebrovascular Oxygen Saturation Measured by Optical Spectroscopy in Humans. Stroke. 1991;22:596–602.

[29] Chance B, Cope M, Gratton E, Ramanujam N, Tromberg B. Phase measurement of light absorption and scatter in human tissue. Review of Scientific Instruments. 1998;69(10):3457–3481.

[30] Larzon T. EVAR/TEVAR;. [Online] [Accessed on 2015-05-25]. Avail-able from: http://www.usorebro.se/sv/Behandlingar/Hjarta/ Aorta/EVARTEVA/.

[31] Pedersen T, Nicholson A, Hovhannisyan K, Am Ml, Af S, Sr L. Pulse oximetry for perioperative monitoring. Cochrane Database of System-atic Reviews. 2014;(3).