- Email: [email protected]
Oxygen saturation monitoring using resonance Raman spectroscopy
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Q11 17 18 19 Q1 20 21 22 23 24 25 Q2 26 27 28 29 30 Q3 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Oxygen saturation monitoring using resonance Raman spectroscopy Ivo Torres Filho, MD, PhD,a,b,c,* Nguyen M. Nguyen,a Rizwan Jivani,a James Terner, PhD,d Padraic Romfh,e Daryoosh Vakhshoori, PhD,e and Kevin R. Ward, MDf a
Department of Anesthesiology, Virginia Commonwealth University and the Virginia Commonwealth University Reanimation Engineering Science Center, Richmond, Virginia b Department of Emergency Medicine, Virginia Commonwealth University and the Virginia Commonwealth University Reanimation Engineering Science Center, Richmond, Virginia c Damage Control Resuscitation, U.S. Army Institute of Surgical Research, Fort Sam Houston, Texas d Department of Chemistry, Virginia Commonwealth University and the Virginia Commonwealth University Reanimation Engineering Science Center, Richmond, Virginia e Pendar Medical, Cambridge, Massachusetts f Department of Emergency Medicine and the Michigan Center for Integrative Research in Critical Care, University of Michigan
article info
abstract
Article history:
Background: The knowledge of hemoglobin oxygen saturation (SO2) and tissue oxygenation is
Received 28 September 2015
critical to identify the presence of shock and therapeutic options. The resonance vibrational
Received in revised form
enhancement of hemoglobin allows measurement of oxy- and deoxy species of hemoglobin
2 December 2015
and resonance Raman spectroscopy (RRS-StO2) has been successfully used to measure
Accepted 4 December 2015
aggregate microvascular oxygenation. We tested the hypothesis that noninvasive oxygen
Available online xxx
saturation measured by RRS-StO2 could serve as surrogate of systemic central venous SO2.
Keywords:
muscle, and liver were compared with measurements of central venous SO2 using tradi-
Oxygen saturation
tional multi-wavelength oximetry. Various oxygenation levels were obtained using a
Raman spectroscopy
stepwise hemorrhage while over 100 paired blood samples and Raman-based measure-
Methods: In anesthetized rats, measurements of RRS-StO2 made in oral mucosa, skin,
Shock
ments were performed. The relationships between RRS-StO2 and clinically important
Hemorrhage
systemic blood parameters were also evaluated. RRS-StO2 measurements were made in 3-
Tissue oxygenation
mm diameter tissue areas using a microvascular oximeter and a handheld probe. Results: Significant correlations were found between venous SO2 and RRS-StO2 measurements made in the oral mucosa (r ¼ 0.913, P < 0.001), skin (r ¼ 0.499, P < 0.01), and liver (r ¼ 0.611, P < 0.05). The mean difference between sublingual RRS-StO2 and blood sample SO2 values was 5.4 1.6%. Sublingual RRS-StO2 also correlated with lactate (r ¼ 0.909, P < 0.01), potassium (r ¼ 0.757, P < 0.01), and pH (r ¼ 0.703, P < 0.05). Conclusions: Raman-based oxygen saturation is a promising technique for the noninvasive evaluation of oxygenation in skin, thin tissues, and solid organs. Under certain conditions, sublingual RRS-StO2 measurements correlate with central venous SO2. ª 2015 Elsevier Inc. All rights reserved.
* Corresponding author. Damage Control Resuscitation, U.S. Army Institute of Surgical Research, 3698 Chambers Pass Fort, Sam Houston, TX 78234. Tel.: þ210 460-9744; fax: þ210 539-6244. E-mail address: [email protected] (I.T. Filho). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2015.12.001
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
2
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
1.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
Introduction
The noninvasive determination of tissue oxygenation is critical to identify the presence of shock and the suitability of therapeutic measures aimed at its resolution. The vibrational bands of hemeproteins provide medically relevant information such as hemoglobin oxygen saturation (SO2). Under proper excitation, the resonance vibrational enhancement of hemoglobin allows simultaneous identification and measurement of the proportion of oxy- and deoxy species of hemoglobin in a concentration-dependent manner with a single excitation wavelength. A chemical fingerprint of the species can be obtained with little or no interference from other compounds [1e3]. Because the profile of the microvasculature and blood volume in tissue is heavily weighted toward a venous composite (80%), aggregate measures of tissue oxygenation are representative of the post-extraction compartment of tissues. In global shock states such as hemorrhage, tissue oxygenation from certain tissues may compare well to central measures such as central venous or mixed venous hemoglobin oxygen saturation. In previous studies from our group [1e3], resonance Raman spectroscopy (RRS-StO2) has been successfully used to measure aggregate microvascular tissue oxygen saturation in the skin of normotensive rats using two laser excitations (406.7 and 532 nm). Encouraging results were also found in large animal models of hemorrhagic shock [4]. In this study, we tested the hypothesis that noninvasive tissue oxygen saturation measured by RRS-StO2 could serve as surrogate of systemic central venous SO2. For this purpose, we compared noninvasive RRS-StO2 measurements made in several different tissues (oral mucosa, skin, muscle, and liver) with measurements of central venous SO2 using traditional multi-wavelength CO-oximetry. To provide a wide range of oxygenation levels, a protocol of stepwise hemorrhage was conducted while paired blood samples and Raman measurements were performed. We also evaluated quantitatively the relationship between RRS-StO2 and clinically important systemic blood parameters such as lactate and potassium.
2.
Materials and methods
2.1.
Experimental animals
37 C using a thermostatically controlled heating blanket (Harvard Apparatus, Holliston, MA). A tracheostomy was performed to maintain a clear airway. The right jugular vein was cannulated with PE-90 tubing advanced to the entrance of the right atrium, used for the collection of blood samples and for injection of Euthasol at the end of the data acquisition process. The right femoral vein was cannulated with PE-50 tubing for infusion (0.17e0.24 mg/kg1h1) of alfaxalone/alfadolone acetate (Saffan; Schering-Plough Animal Health, Welwyn Garden City, England) anesthesia by a microprocessor-controlled infusion/withdrawal syringe pump (model PHD 2000; Harvard Apparatus, Holliston, MA) once the experimental procedure began. The right femoral artery was cannulated with phosphate-buffered saline (PBS)efilled PE-50 tubing connected to a pressure transducer to continuously measure arterial blood pressure (AP), for the controlled-hemorrhage process and for collection of blood samples. The AP was monitored using a pressure transducer Q4 connected to a computer for continuous data acquisition (DA100C, MP150, Acknowledge 3.9.0, Biopac Systems). All lines were flushed with heparinized PBS (10 i.u. mL1 heparin) after each blood collection to inhibit clot formation. The vein and artery on the left side were not cannulated, and this thigh was used for Raman monitoring.
2.3. Resonance Raman spectroscopyebased O2 saturation measurements A microvascular oximeter (Pendar Medical, Cambridge, MA) was used to measure tissue oxygenation using RRS-StO2. The device has a laser (405 nm, 4 mW) coupled to a complex plastic fiber optic cable and a handheld probe that illuminates a tissue area (approximately 3 mm in diameter, less than 1 mm deepS). The oxyhemoglobin and deoxyhemoglobin molecules can be excited by the laser light into distinct vibrational states resulting in a differential wavelength shift of the scattered light. The charge-coupled deviceebased spectrometer of the oximeter captures the spectrum of the scattered light every second, with distinct sharp peaks linearly proportional to the relative concentrations of oxyhemoglobin and deoxyhemoglobin. The device contains dedicated software to calculate the RRS-StO2 using the spectral peaks as the relative ratio of concentration of oxygenated hemoglobin to that of total hemoglobin. (Fig. 1)
2.4. This study was approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University. We used 13 SpragueeDawley rats (Harlan, Indianapolis, IN) weighing 241e501 g. The rats were kept on a commercial diet with free access to drinking water, in a facility with constant temperature (21 Ce23 C) and humidity (w50%) that rotated through 12-h light/dark cycles.
2.2.
Surgical preparation
The animals were anesthetized with 2% isoflurane and kept under constant anesthesia during cannulations (1.5 L/min balance 21% O2). The core temperature was maintained at
Animal groups and testing sites
Two groups of animals were studied. In one group of animals, the gracilis muscle, the sublingual surface, and the surface of the paw were prepared for examination, as described in the following section. In the second group of animals, the surface of the liver was prepared. The RRS-StO2 readings were taken from each site by gently placing the Raman probe directly over each tissue area. To prepare the “muscle” site, the left thigh was shaved and void of hair using a commercial hair removal cream. The skin layer was cut and folded back to expose the gracilis muscle that was immediately covered with Saran wrap to maintain the exposed tissue void of contact with atmosphere. PBS was administered (0.5 mL, i.v.) as a hydration
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
3
Fig. 1 e Composite picture showing the display of the device used for estimating tissue oxygen saturation in the oral mucosa based on Raman spectroscopy (StO2) during four levels of oxygenation (AeD). The intensities of the strongest resonance Raman hemoglobin peaks, indicated by arrows (deoxyhemoglobin, Ideoxy and oxyhemoglobin, Ioxy), vary as oxygenation level changes (as a consequence of hemorrhage). These intensities are used by the device to calculate StO2 as indicated by the formula on the left, and the result appears on each digital panel. Also shown is the formula used in the calculation of Raman shift where linc and lscatt are the incident and Raman scattered light wavelengths (in cm), respectively. fluid after exposure of the muscle. The site named “oral” was prepared by manually opening the mouth of the rat to access the sublingual surface. The surface of the front paw was also used and the site named “skin”. A small portion of the liver was exposed by first removing the skin hair from the abdominal region and pulling this skin portion back. A cut was made at the linea alba just below the diaphragm, and the abdominal muscles were folded back to expose a segment of the liver. Any small bleeding areas on the muscles were quickly cauterized, and Saran wrap was placed over the entire exposed area. The Saran wrap used was produced in 1999 and has been previously tested as a good barrier from contamination from room oxygen. The animals were then allowed time to stabilize (10e15 min).
2.5.
Biochemistry and baseline measurements
Blood samples from the femoral artery (0.1 mL each) were collected into heparinized glass capillary tubes (Clinitubes, Radiometer, Copenhagen, Denmark) and immediately replaced by an equal volume of heparinized PBS. Total hemoglobin concentration, hemoglobin O2 saturation, methemoglobin, and O2 content were measured with a multi-wavelength CO oximeter (OSM3, Radiometer). Blood pH, glucose, potassium, lactate, partial pressures of oxygen and carbon dioxide were measured with a blood gas analyzer (ABL 700; Radiometer, Copenhagen, Denmark). Immediately after RRS-StO2 measurements, venous blood samples were drawn from the superior vena cava into capillary tubes. About 5 min later, a second set of baseline measurements was made.
2.6.
Hemorrhage
The target hemorrhage volume was 45% of total blood volume (total blood volume was calculated as 6% of total body mass) and was achieved in three steps of 15% per blood withdrawal. Hemorrhage was performed manually and drawn from the right femoral artery line using a clean 5-mL plastic syringe. Each step was followed by a 15-min period for AP and heart rate to stabilize. After stabilization, RRS-StO2 readings in each of the three sites and blood samplings were made. The process was repeated until the target hemorrhage volume was reached.
2.7.
Data analysis
The first step in the estimation of the RRS-StO2 was the calculation of the peak ratio of the highest intensities of the Raman bands for oxygenated (Ioxy) and deoxygenated (Ideoxy) hemoglobin at 1360 cm1 and 1375 cm1, respectively (Fig. 1). The RRS-StO2 was then estimated (in %) using the peak ratio and adjustment coefficients previously obtained in calibration experiments. The spectra were analyzed with customized software specifically developed to streamline processing of Raman signals and estimate RRS-StO2 online. As shown in Figure 1, 20e80 spectra were averaged to increase precision of the RRS-StO2 determinations that were updated on the integrated monitor each second.
2.8.
Statistical analysis
Descriptive statistics are expressed as means standard error of mean. The coefficient of variation (calculated as mean
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
4
391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455
Q10
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
Table e Results of oxygen saturation measurements made in four different sites using Raman spectroscopy and in venous blood samples from 10 animals during baseline and after hemorrhage. Venous SO2 (%)
Mean SEM 42.4 3.7 n 27 Range 9.3e75.4 CV (%) 45.1 RRS-StO2 versus venous SO2
Resonance Raman spectroscopy O2 saturation (%) Oral
Skin
Muscle
Liver
47.9 2.9 27 18.0e77.0 31.5 r ¼ 0.913***
21.8 2.0 27 5.0e48.6 48.0 r ¼ 0.499**
63.7 4.2 27 6.0e88.0 88.0 r ¼ 0.249
22.1 3.4 12 15.3e32.2 27.5 r ¼ 0.611*
CV ¼ coefficient of variation; n ¼ number of measurements; SEM ¼ standard error of mean. The correlation coefficients (r) refer to least-squares linear regressions between Raman-estimated oxygen saturations (RRS-StO2) in each site and CO-oximeter measured saturations using venous blood samples (last row). All correlation coefficients were statistically significant (***P < 0.001, **P < 0.01, *P < 0.05), except for “muscle” (P ¼ 0.210).
divided by the standard deviation times 100) was used as an index of variability. Pearson correlation was used to quantify the degree to which any two variables are related. The statistical tests were performed using commercial computer software (OriginLab Origin 8 and Microsoft Excel 2010), and power analysis was performed for all correlations (Sigmaplot for Windows, v. 12.0). All P values correspond to two-tailed tests with significance set at 0.05.
3.
Results
Anesthetized rats were studied before and during hemorrhagic hypotension. Under these conditions, progressive reductions in arterial pressure, local blood flow, and oxygenation were observed. During baseline, all animals showed similar mean arterial pressure and heart rate, averaging 120.3 6.7 mm Hg and 415 10 min1, respectively, which were lowered to 59.8 3.4 mm Hg and 393 15 min1 after hemorrhage. A microvascular oximeter was used to make tissue oxygen saturation measurements based on the principles of RRS-StO2 using dedicated software (Fig. 1). Immediately before and after RRS-StO2 measurements, blood was sampled and analyzed using the CO oximeter. Table summarizes the results of SO2 measurements made in all animals during baseline and after hemorrhagic hypotension. The Raman probe was manually and sequentially positioned on the surface of four tissues. A power analysis was performed for all correlations. The sample size and correlation coefficient averaged 23 and 0.6, respectively, whereas a power of 0.8 was obtained for an alpha of 0.05. Measurements of RRS-StO2 in the sublingual and liver surfaces showed significant correlation with venous SO2, and these tissues showed the smallest variability of RRS-StO2 measurements. On average, RRS-StO2 values estimated from Raman spectra in the liver were lower than SO2 values using blood samples by 11.0 3.4%. Raman measurements in the surface of the exposed gracilis muscle showed the largest variability of RRSStO2 and no significant correlation with venous SO2. The correlation between RRS-StO2 in the front paw (skin tissue) and venous SO2 was weaker but still statistically significant. The strong correlation between RRS-StO2 values estimated from resonance Raman spectra in the sublingual surface and SO2 values measured from venous blood samples using a CO
oximeter is presented in Figure 2. The mean difference between RRS-StO2 values estimated from Raman spectra in this tissue and SO2 values using blood samples was 5.4 1.6%. The Raman-based O2 saturation values in the sublingual surface also correlated with various parameters measured in the blood collected from the vena cava (Fig. 3): RRS-StO2 was negatively correlated with blood lactate (r ¼ 0.909, P < 0.01) and potassium (r ¼ 0.757, P < 0.01) and positively correlated with pO2 (r ¼ 0.910, P < 0.01) and pH (r ¼ 0.703, P < 0.05). Although the blood hemoglobin levels fell from a baseline of 14.2 1.1 g/dL to 12.1 1.0 g/dL during hemorrhage, the fraction of methemoglobin remained below 1%.
4.
Discussion
Gas exchange takes place in vessels that contain most of the circulating blood: arterioles, venules, and capillaries [5]. The
Fig. 2 e Values of oxygen saturation estimated from Raman spectroscopy (StO2) and measured from venous blood samples using an oximeter (SO2). The illustrated data are from probes positioned on the oral mucosa (sublingual surface) of seven anesthetized rats. Each point represents a single saturation determination (Raman and oximeter). The least-squares regression was calculated considering measurements during baseline and after hemorrhage (P < 0.001).
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585
5
Fig. 3 e Values of oxygen saturation estimated from Raman spectroscopy (StO2) and levels of different biochemical parameters from venous blood samples. A statistically significant correlation was found in all cases. Data were obtained from seven rats and include measurements during baseline and after hemorrhage.
Raman signal used for estimating RRS-StO2 is a combination of arteriolar, capillary, and venular SO2. The contribution will be proportional to the volumes of these vessels. Because most of the blood volume resides in capillaries and venules [5], the RRS-StO2 mainly reflects the capillary and venular compartments. We tested the hypothesis that RRS-StO2 could serve as surrogate of systemic venous SO2. Significant correlations were found between venous SO2 and RRS-StO2 measurements made in the oral mucosa, skin, and liver. In a swine model of hemorrhagic shock, we have recently shown that RRS-StO2 measurements in the oral mucosa tracked simultaneous changes in central venous SO2 [4]. Previous pilot studies (using large lasers and nonportable collection systems) have also indicated that mucosal RRS-StO2 could track clinically relevant biochemical parameters such as lactate [6]. In the present study, we confirm and expand those preliminary findings while providing quantitative data on the relationship between RRS-StO2 and these parameters. Although different regulatory mechanisms may partially explain dissimilarities in microhemodynamics and macrohemodynamics, microvascular perfusion is also affected by local factors such as leukocyte adhesion, endothelial dysfunction, and rheological alterations. Nevertheless, because it can be imaged using minimally invasive methods [7,8], the sublingual area has been investigated in experimental animals and humans. Its adequacy as a noninvasive site to study systemic changes has been demonstrated by various investigators [9,10]. Although RRS-StO2 measurements have not been reported, decreases in microvascular flow indexes in sublingual microcirculation have been documented after hemorrhagic hypotension in pigs [11] and sheep [12]. The reduction in functional vascular density of the sublingual mucosa matched the
decreases in oxygen delivery in a model of progressive hemorrhage using human volunteers [13]. In addition, the noninvasive assessment of mucosal microvascular perfusion and oxygenation may help to identify patients with potential need for fluid therapy in hemorrhage [14] and sepsis [15,16]. In this context, new methods of high specificity such as Raman spectroscopy, that provide RRS-StO2 estimations using simple algorithms (Fig. 1) and backed by solid preclinical studies [1e4,6] may offer ample potential clinical importance. By using low-power deep violet laser excitation in a mucosal surface, the interfering resonance Raman signals of myoglobin from oral musculature were avoided due to only minimal penetration in this wavelength region. Moreover, low-intensity excitation minimized sample degradation in a tissue not covered by the stratum corneum. The presence of a relatively thick epidermis did not prevent good RRS-StO2 measurements because a significant correlation between RRSStO2 from skin (paw) and venous SO2 was found (Table). Although low-coherence spectroscopy has been used to estimate skin SO2 [17], those determinations were only preliminary, and no attempts were made to correlate skin RRS-StO2 with venous SO2. A significant correlation between RRS-StO2 from liver and venous SO2 was found (Table). These are the first data of RRSStO2 on liver. The fact that RRS-StO2 in the liver were lower than central venous SO2 is not surprising as oxygenation changes in the splanchnic bed may change earlier and be more severe than is reflected in the aggregate measure of central venous SO2 [18]. Moreover, the liver is mostly supplied with venous blood from the portal vein. Near-infrared spectroscopy (NIRS) has been used in the rat to study the effect of hypoxia on liver oxygenation and energy metabolism [19]. In
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650
6
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
that study, the relationship between tissue and venous SO2 651 was not investigated although in pigs, a good correlation was 652 found between tissue oxygenation measured by NIRS and PO2 653 in the hepatic vein [20]. In children, NIRS was used trans654 cutaneously (optode placed over the palpable liver), and the 655 656 measurements did not correlate with SO2 measured from 657 blood samples taken from the inferior vena cava [21]. The 658 authors pointed several reasons for the lack of correlation 659 including the fact that the NIRS values were mainly deter660 mined by non-hepatic measures such as arterial saturation 661 and vascular resistances. Although widely available, NIRS re662 quires care in interpretation due to issues related to the 663 biophysics of measurement and lack of a standard to validate 664 the measurement. Like pulse oximetry, NIRS is based on the 665 principle of relating light absorption to oxyhemoglobin and 666 deoxyhemoglobin concentrations. For example, because op667 668 tical path length and light scattering are not constant in adults 669 or children [22,23], it has been difficult to establish the exact 670 error in the measurement because NIRS lacks a standard to 671 compare it against. 672 We found no correlation between RRS-StO2 from muscle 673 and central venous SO2. One possible explanation for this 674 finding is that optical detection can be obscured by competing 675 background signals, such as those from myoglobin in muscle 676 tissue or other absorbing chromophores [24e26]. As with the Q5 677 respective NIR absorptions, we have previously observed that 678 679 resonance Raman signals of myoglobin are similar to those of 680 hemoglobin [3]. Therefore, these may have interfered with the 681 Q6 estimations of RRS-StO2 from muscle. The fact that myoglobin 682 and hemoglobin have significantly different P50 values, it is 683 not surprising that signals contaminated with myoglobin may 684 not reflect central venous SO2. The fact that the measure685 ments cannot distinguish myoglobin and hemoglobin does 686 not present a major limitation as a highly significant correla687 tion was found between RRS-StO2 and a clinically relevant site 688 (sublingual mucosa). Moreover, using 405-nm excitation, 689 Raman measurements in the most accessible area in the body 690 691 (the skin) can be readily performed without any interfering 692 signals from myoglobin in deeper tissue. 693 694 4.1. Limitations and applications 695 696 Although we tried to track changes in RRS-StO2 and central 697 venous SO2 over a wide range of oxygenation levels, we did 698 not investigate levels reached near death or very high SO2 699 levels. Because animals were not resuscitated, the correla700 tions are unknown during resuscitation using different fluids. 701 702 Only one species (rats) was used although the results are in 703 complete agreement with data from pigs using the same 704 methodology [4]. A low-power continuous laser beam was 705 used to avoid harmful effects associated with intense pulsed 706 laser beams because laser safety limits for the application of 707 ultrashort pulses on human tissues are more stringent than 708 those with continuous wave lasers [27]. In addition to 709 providing specific quantitative data on the relationship be710 tween SO2 and RRS-StO2, this study shows the suitability of a 711 device based on Raman spectroscopy for “point and shoot” 712 oxygenation measurements in thin tissues and in solid or713 714 gans, therefore valuable in multiple clinical settings. This is 715 especially important as a recent study showed that an
abnormal sublingual microvascular flow index was independently associated with an increased risk of hospital death in critically ill patients of 36 ICUs worldwide [28]. Also, evaluations of sublingual microcirculation by physicians and nurses were highly sensitive and specific for sublingual microcirculatory abnormalities [29]. A direct comparison between RRS- Q7 StO2 and the sublingual microcirculation using methods such as sidestream dark field imaging [7], was outside the scope of the present work because SDF currently does not measure O2 saturation. Nevertheless, it is likely that RRS-StO2 measurements will correlate with parameters derived from SDF images.
5.
Conclusion
In summary, RRS-StO2 is a promising technique for the noninvasive evaluation of oxygenation in various tissues such as sublingual mucosa, skin, and liver. The RRS-StO2 measurements in the sublingual mucosa correlate with central venous SO2 and other relevant biochemical parameters such as lactate, potassium, and pH. Additional testing will be required to test its ability to track changes in tissue oxygenation during resuscitation.
Acknowledgment The authors would like to thank Bruce Spiess, Brian Berger, and M. Hakam Tiba for the help during various phases of the study. The opinions or assertions contained herein are the private views of the author and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense. Authors’ contributions: I.T.F. and K.R.W. designed the study. N.M.N. and R.J. performed the research. J.T., P.R., and D.V. contributed to critical aspects on Raman spectroscopy. I.T.F, N.M.N., and R.J. contributed to data analysis. I.T.F. drafted the article. All the authors critically reviewed the article. I.T.F., J.T., and K.R.W. have intellectual property on the use of resonance Raman spectroscopy for tissue oxygenation monitoring assigned to Virginia Commonwealth University. P.R. and D.V. are officers in Pendar Medical, which have licensed the Raman spectroscopy technology from Virginia Commonwealth University.
Disclosures The authors reported no proprietary or commercial interest in Q8 any product mentioned or concept discussed in this article.
references
[1] Torres Filho IP, Terner J, Pittman RN, Proffitt E, Ward KR. Measurement of hemoglobin oxygen saturation using Raman microspectroscopy and 532-nm excitation. J Appl Physiol (1985) 2008;104:1809.
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
781 782 783 784 785 786 Q9 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
[2] Torres Filho IP, Terner J, Pittman RN, Somera LG 3rd, Ward KR. Hemoglobin oxygen saturation measurements using resonance Raman intravital microscopy. Am J Physiol Heart Circ Physiol 2005;289:H488. [3] Ward KR, Barbee RW, Reynolds PS, Filho IP, Tiba MH, et al. Oxygenation monitoring of tissue vasculature by resonance Raman spectroscopy. Anal Chem 2007;79:1514. [4] Tiba MH, Draucker GT, Barbee RW, Terner J, Filho IT, et al. Tissue oxygenation monitoring using resonance Raman spectroscopy during hemorrhage. J Trauma Acute Care Surg 2014;76:402. [5] Wiedeman MP. Dimensions of blood vessels from distributing artery to collecting vein. Circ Res 1963;12:375. [6] Ward KR, Torres Filho I, Barbee RW, Torres L, Tiba MH, et al. Resonance Raman spectroscopy: a new technology for tissue oxygenation monitoring. Crit Care Med 2006;34:792. [7] Goedhart PT, Khalilzada M, Bezemer R, Merza J, Ince C. Sidestream dark field (SDF) imaging: a novel stroboscopic LED ring-based imaging modality for clinical assessment of the microcirculation. Opt Express 2007;15:15101. [8] Groner W, Winkelman JW, Harris AG, Ince C, Bouma GJ, et al. Orthogonal polarization spectral imaging: a new method for study of the microcirculation. Nat Med 1999;5:1209. [9] De Backer D, Dubois MJ. Assessment of the microcirculatory flow in patients in the intensive care unit. Curr Opin Crit Care 2001;7:200. [10] Verdant CL, De Backer D, Bruhn A, Clausi CM, Su F, et al. Evaluation of sublingual and gut mucosal microcirculation in sepsis: a quantitative analysis. Crit Care Med 2009;37:2875. [11] Maier S, Holz-Holzl C, Pajk W, Ulmer H, Hengl C, et al. Microcirculatory parameters after isotonic and hypertonic colloidal fluid resuscitation in acute hemorrhagic shock. J Trauma 2009;66:337. [12] Dubin A, Pozo MO, Ferrara G, Murias G, Martins E, et al. Systemic and microcirculatory responses to progressive hemorrhage. Intensive Care Med 2009;35:556. [13] Ward KR, Tiba MH, Ryan KL, Filho IP, Rickards CA, et al. Oxygen transport characterization of a human model of progressive hemorrhage. Resuscitation 2010;81:987. [14] Pranskunas A, Koopmans M, Koetsier PM, Pilvinis V, Boerma EC. Microcirculatory blood flow as a tool to select ICU patients eligible for fluid therapy. Intensive Care Med 2013; 39:612. [15] Boerma EC, Mathura KR, van der Voort PH, Spronk PE, Ince C. Quantifying bedside-derived imaging of microcirculatory abnormalities in septic patients: a prospective validation study. Crit Care 2005;9:R601. [16] Ospina-Tascon G, Neves AP, Occhipinti G, Donadello K, Buchele G, et al. Effects of fluids on microvascular perfusion in patients with severe sepsis. Intensive Care Med 2010;36:949. [17] Bosschaart N, Faber DJ, van Leeuwen TG, Aalders MC. In vivo low-coherence spectroscopic measurements of local
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
7
hemoglobin absorption spectra in human skin. J Biomed Opt 2011;16:100504. Ba ZF, Wang P, Koo DJ, Cioffi WG, Bland KI, et al. Alterations in tissue oxygen consumption and extraction after trauma and hemorrhagic shock. Crit Care Med 2000; 28:2837. Seifalian AM, El-Desoky H, Delpy DT, Davidson BR. Effect of graded hypoxia on the rat hepatic tissue oxygenation and energy metabolism monitored by near-infrared and 31P nuclear magnetic resonance spectroscopy. FASEB J 2001;15: 2642. El-Desoky AE, Jiao LR, Havlik R, Habib N, Davidson BR, et al. Measurement of hepatic tissue hypoxia using near infrared spectroscopy: comparison with hepatic vein oxygen partial pressure. Eur Surg Res 2000;32:207. Weiss M, Schulz G, Fasnacht M, Balmer C, Fischer JE, et al. Transcutaneously measured near-infrared spectroscopic liver tissue oxygenation does not correlate with hepatic venous oxygenation in children. Can J Anaesth 2002;49:824. Benaron DA, Kurth CD, Steven JM, DelivoriaPapadopoulos M, Chance B. Transcranial optical path length in infants by near-infrared phase-shift spectroscopy. J Clin Monit 1995;11:109. Duncan A, Meek JH, Clemence M, Elwell CE, Fallon P, et al. Measurement of cranial optical path length as a function of age using phase resolved near infrared spectroscopy. Pediatr Res 1996;39:889. Beilman GJ, Myers D, Cerra FB, Lazaron V, Dahms RA, et al. Near-infrared and nuclear magnetic resonance spectroscopic assessment of tissue energetics in an isolated, perfused canine hind limb model of dysoxia. Shock 2001;15: 392. Chance B, Wang NG, Maris M, Nioka S, Sevick E. Quantitation of tissue optical characteristics and hemoglobin desaturation by time- and frequency-resolved multiwavelength spectrophotometry. Adv Exp Med Biol 1992;317: 297. Nighswander-Rempel SP, Kupriyanov VV, Shaw RA. Relative contributions of hemoglobin and myoglobin to near-infrared spectroscopic images of cardiac tissue. Appl Spectrosc 2005; 59:190. Matousek P, Draper ER, Goodship AE, Clark IP, Ronayne KL, et al. Noninvasive Raman spectroscopy of human tissue in vivo. Appl Spectrosc 2006;60:758. Vellinga NA, Boerma EC, Koopmans M, Donati A, Dubin A, et al. International study on microcirculatory shock occurrence in acutely ill patients. Crit Care Med 2015;43:48. Lima A, Lopez A, van Genderen ME, Hurtado FJ, Angulo M, et al. Interrater reliability and diagnostic performance of subjective evaluation of sublingual microcirculation images by physicians and nurses: a multicenter observational study. Shock 2015;44:239.
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Q11 17 18 19 Q1 20 21 22 23 24 25 Q2 26 27 28 29 30 Q3 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.JournalofSurgicalResearch.com
Oxygen saturation monitoring using resonance Raman spectroscopy Ivo Torres Filho, MD, PhD,a,b,c,* Nguyen M. Nguyen,a Rizwan Jivani,a James Terner, PhD,d Padraic Romfh,e Daryoosh Vakhshoori, PhD,e and Kevin R. Ward, MDf a
Department of Anesthesiology, Virginia Commonwealth University and the Virginia Commonwealth University Reanimation Engineering Science Center, Richmond, Virginia b Department of Emergency Medicine, Virginia Commonwealth University and the Virginia Commonwealth University Reanimation Engineering Science Center, Richmond, Virginia c Damage Control Resuscitation, U.S. Army Institute of Surgical Research, Fort Sam Houston, Texas d Department of Chemistry, Virginia Commonwealth University and the Virginia Commonwealth University Reanimation Engineering Science Center, Richmond, Virginia e Pendar Medical, Cambridge, Massachusetts f Department of Emergency Medicine and the Michigan Center for Integrative Research in Critical Care, University of Michigan
article info
abstract
Article history:
Background: The knowledge of hemoglobin oxygen saturation (SO2) and tissue oxygenation is
Received 28 September 2015
critical to identify the presence of shock and therapeutic options. The resonance vibrational
Received in revised form
enhancement of hemoglobin allows measurement of oxy- and deoxy species of hemoglobin
2 December 2015
and resonance Raman spectroscopy (RRS-StO2) has been successfully used to measure
Accepted 4 December 2015
aggregate microvascular oxygenation. We tested the hypothesis that noninvasive oxygen
Available online xxx
saturation measured by RRS-StO2 could serve as surrogate of systemic central venous SO2.
Keywords:
muscle, and liver were compared with measurements of central venous SO2 using tradi-
Oxygen saturation
tional multi-wavelength oximetry. Various oxygenation levels were obtained using a
Raman spectroscopy
stepwise hemorrhage while over 100 paired blood samples and Raman-based measure-
Methods: In anesthetized rats, measurements of RRS-StO2 made in oral mucosa, skin,
Shock
ments were performed. The relationships between RRS-StO2 and clinically important
Hemorrhage
systemic blood parameters were also evaluated. RRS-StO2 measurements were made in 3-
Tissue oxygenation
mm diameter tissue areas using a microvascular oximeter and a handheld probe. Results: Significant correlations were found between venous SO2 and RRS-StO2 measurements made in the oral mucosa (r ¼ 0.913, P < 0.001), skin (r ¼ 0.499, P < 0.01), and liver (r ¼ 0.611, P < 0.05). The mean difference between sublingual RRS-StO2 and blood sample SO2 values was 5.4 1.6%. Sublingual RRS-StO2 also correlated with lactate (r ¼ 0.909, P < 0.01), potassium (r ¼ 0.757, P < 0.01), and pH (r ¼ 0.703, P < 0.05). Conclusions: Raman-based oxygen saturation is a promising technique for the noninvasive evaluation of oxygenation in skin, thin tissues, and solid organs. Under certain conditions, sublingual RRS-StO2 measurements correlate with central venous SO2. ª 2015 Elsevier Inc. All rights reserved.
* Corresponding author. Damage Control Resuscitation, U.S. Army Institute of Surgical Research, 3698 Chambers Pass Fort, Sam Houston, TX 78234. Tel.: þ210 460-9744; fax: þ210 539-6244. E-mail address: [email protected] (I.T. Filho). 0022-4804/$ e see front matter ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jss.2015.12.001
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
2
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
1.
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
Introduction
The noninvasive determination of tissue oxygenation is critical to identify the presence of shock and the suitability of therapeutic measures aimed at its resolution. The vibrational bands of hemeproteins provide medically relevant information such as hemoglobin oxygen saturation (SO2). Under proper excitation, the resonance vibrational enhancement of hemoglobin allows simultaneous identification and measurement of the proportion of oxy- and deoxy species of hemoglobin in a concentration-dependent manner with a single excitation wavelength. A chemical fingerprint of the species can be obtained with little or no interference from other compounds [1e3]. Because the profile of the microvasculature and blood volume in tissue is heavily weighted toward a venous composite (80%), aggregate measures of tissue oxygenation are representative of the post-extraction compartment of tissues. In global shock states such as hemorrhage, tissue oxygenation from certain tissues may compare well to central measures such as central venous or mixed venous hemoglobin oxygen saturation. In previous studies from our group [1e3], resonance Raman spectroscopy (RRS-StO2) has been successfully used to measure aggregate microvascular tissue oxygen saturation in the skin of normotensive rats using two laser excitations (406.7 and 532 nm). Encouraging results were also found in large animal models of hemorrhagic shock [4]. In this study, we tested the hypothesis that noninvasive tissue oxygen saturation measured by RRS-StO2 could serve as surrogate of systemic central venous SO2. For this purpose, we compared noninvasive RRS-StO2 measurements made in several different tissues (oral mucosa, skin, muscle, and liver) with measurements of central venous SO2 using traditional multi-wavelength CO-oximetry. To provide a wide range of oxygenation levels, a protocol of stepwise hemorrhage was conducted while paired blood samples and Raman measurements were performed. We also evaluated quantitatively the relationship between RRS-StO2 and clinically important systemic blood parameters such as lactate and potassium.
2.
Materials and methods
2.1.
Experimental animals
37 C using a thermostatically controlled heating blanket (Harvard Apparatus, Holliston, MA). A tracheostomy was performed to maintain a clear airway. The right jugular vein was cannulated with PE-90 tubing advanced to the entrance of the right atrium, used for the collection of blood samples and for injection of Euthasol at the end of the data acquisition process. The right femoral vein was cannulated with PE-50 tubing for infusion (0.17e0.24 mg/kg1h1) of alfaxalone/alfadolone acetate (Saffan; Schering-Plough Animal Health, Welwyn Garden City, England) anesthesia by a microprocessor-controlled infusion/withdrawal syringe pump (model PHD 2000; Harvard Apparatus, Holliston, MA) once the experimental procedure began. The right femoral artery was cannulated with phosphate-buffered saline (PBS)efilled PE-50 tubing connected to a pressure transducer to continuously measure arterial blood pressure (AP), for the controlled-hemorrhage process and for collection of blood samples. The AP was monitored using a pressure transducer Q4 connected to a computer for continuous data acquisition (DA100C, MP150, Acknowledge 3.9.0, Biopac Systems). All lines were flushed with heparinized PBS (10 i.u. mL1 heparin) after each blood collection to inhibit clot formation. The vein and artery on the left side were not cannulated, and this thigh was used for Raman monitoring.
2.3. Resonance Raman spectroscopyebased O2 saturation measurements A microvascular oximeter (Pendar Medical, Cambridge, MA) was used to measure tissue oxygenation using RRS-StO2. The device has a laser (405 nm, 4 mW) coupled to a complex plastic fiber optic cable and a handheld probe that illuminates a tissue area (approximately 3 mm in diameter, less than 1 mm deepS). The oxyhemoglobin and deoxyhemoglobin molecules can be excited by the laser light into distinct vibrational states resulting in a differential wavelength shift of the scattered light. The charge-coupled deviceebased spectrometer of the oximeter captures the spectrum of the scattered light every second, with distinct sharp peaks linearly proportional to the relative concentrations of oxyhemoglobin and deoxyhemoglobin. The device contains dedicated software to calculate the RRS-StO2 using the spectral peaks as the relative ratio of concentration of oxygenated hemoglobin to that of total hemoglobin. (Fig. 1)
2.4. This study was approved by the Institutional Animal Care and Use Committee of Virginia Commonwealth University. We used 13 SpragueeDawley rats (Harlan, Indianapolis, IN) weighing 241e501 g. The rats were kept on a commercial diet with free access to drinking water, in a facility with constant temperature (21 Ce23 C) and humidity (w50%) that rotated through 12-h light/dark cycles.
2.2.
Surgical preparation
The animals were anesthetized with 2% isoflurane and kept under constant anesthesia during cannulations (1.5 L/min balance 21% O2). The core temperature was maintained at
Animal groups and testing sites
Two groups of animals were studied. In one group of animals, the gracilis muscle, the sublingual surface, and the surface of the paw were prepared for examination, as described in the following section. In the second group of animals, the surface of the liver was prepared. The RRS-StO2 readings were taken from each site by gently placing the Raman probe directly over each tissue area. To prepare the “muscle” site, the left thigh was shaved and void of hair using a commercial hair removal cream. The skin layer was cut and folded back to expose the gracilis muscle that was immediately covered with Saran wrap to maintain the exposed tissue void of contact with atmosphere. PBS was administered (0.5 mL, i.v.) as a hydration
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325
3
Fig. 1 e Composite picture showing the display of the device used for estimating tissue oxygen saturation in the oral mucosa based on Raman spectroscopy (StO2) during four levels of oxygenation (AeD). The intensities of the strongest resonance Raman hemoglobin peaks, indicated by arrows (deoxyhemoglobin, Ideoxy and oxyhemoglobin, Ioxy), vary as oxygenation level changes (as a consequence of hemorrhage). These intensities are used by the device to calculate StO2 as indicated by the formula on the left, and the result appears on each digital panel. Also shown is the formula used in the calculation of Raman shift where linc and lscatt are the incident and Raman scattered light wavelengths (in cm), respectively. fluid after exposure of the muscle. The site named “oral” was prepared by manually opening the mouth of the rat to access the sublingual surface. The surface of the front paw was also used and the site named “skin”. A small portion of the liver was exposed by first removing the skin hair from the abdominal region and pulling this skin portion back. A cut was made at the linea alba just below the diaphragm, and the abdominal muscles were folded back to expose a segment of the liver. Any small bleeding areas on the muscles were quickly cauterized, and Saran wrap was placed over the entire exposed area. The Saran wrap used was produced in 1999 and has been previously tested as a good barrier from contamination from room oxygen. The animals were then allowed time to stabilize (10e15 min).
2.5.
Biochemistry and baseline measurements
Blood samples from the femoral artery (0.1 mL each) were collected into heparinized glass capillary tubes (Clinitubes, Radiometer, Copenhagen, Denmark) and immediately replaced by an equal volume of heparinized PBS. Total hemoglobin concentration, hemoglobin O2 saturation, methemoglobin, and O2 content were measured with a multi-wavelength CO oximeter (OSM3, Radiometer). Blood pH, glucose, potassium, lactate, partial pressures of oxygen and carbon dioxide were measured with a blood gas analyzer (ABL 700; Radiometer, Copenhagen, Denmark). Immediately after RRS-StO2 measurements, venous blood samples were drawn from the superior vena cava into capillary tubes. About 5 min later, a second set of baseline measurements was made.
2.6.
Hemorrhage
The target hemorrhage volume was 45% of total blood volume (total blood volume was calculated as 6% of total body mass) and was achieved in three steps of 15% per blood withdrawal. Hemorrhage was performed manually and drawn from the right femoral artery line using a clean 5-mL plastic syringe. Each step was followed by a 15-min period for AP and heart rate to stabilize. After stabilization, RRS-StO2 readings in each of the three sites and blood samplings were made. The process was repeated until the target hemorrhage volume was reached.
2.7.
Data analysis
The first step in the estimation of the RRS-StO2 was the calculation of the peak ratio of the highest intensities of the Raman bands for oxygenated (Ioxy) and deoxygenated (Ideoxy) hemoglobin at 1360 cm1 and 1375 cm1, respectively (Fig. 1). The RRS-StO2 was then estimated (in %) using the peak ratio and adjustment coefficients previously obtained in calibration experiments. The spectra were analyzed with customized software specifically developed to streamline processing of Raman signals and estimate RRS-StO2 online. As shown in Figure 1, 20e80 spectra were averaged to increase precision of the RRS-StO2 determinations that were updated on the integrated monitor each second.
2.8.
Statistical analysis
Descriptive statistics are expressed as means standard error of mean. The coefficient of variation (calculated as mean
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
4
391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455
Q10
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
Table e Results of oxygen saturation measurements made in four different sites using Raman spectroscopy and in venous blood samples from 10 animals during baseline and after hemorrhage. Venous SO2 (%)
Mean SEM 42.4 3.7 n 27 Range 9.3e75.4 CV (%) 45.1 RRS-StO2 versus venous SO2
Resonance Raman spectroscopy O2 saturation (%) Oral
Skin
Muscle
Liver
47.9 2.9 27 18.0e77.0 31.5 r ¼ 0.913***
21.8 2.0 27 5.0e48.6 48.0 r ¼ 0.499**
63.7 4.2 27 6.0e88.0 88.0 r ¼ 0.249
22.1 3.4 12 15.3e32.2 27.5 r ¼ 0.611*
CV ¼ coefficient of variation; n ¼ number of measurements; SEM ¼ standard error of mean. The correlation coefficients (r) refer to least-squares linear regressions between Raman-estimated oxygen saturations (RRS-StO2) in each site and CO-oximeter measured saturations using venous blood samples (last row). All correlation coefficients were statistically significant (***P < 0.001, **P < 0.01, *P < 0.05), except for “muscle” (P ¼ 0.210).
divided by the standard deviation times 100) was used as an index of variability. Pearson correlation was used to quantify the degree to which any two variables are related. The statistical tests were performed using commercial computer software (OriginLab Origin 8 and Microsoft Excel 2010), and power analysis was performed for all correlations (Sigmaplot for Windows, v. 12.0). All P values correspond to two-tailed tests with significance set at 0.05.
3.
Results
Anesthetized rats were studied before and during hemorrhagic hypotension. Under these conditions, progressive reductions in arterial pressure, local blood flow, and oxygenation were observed. During baseline, all animals showed similar mean arterial pressure and heart rate, averaging 120.3 6.7 mm Hg and 415 10 min1, respectively, which were lowered to 59.8 3.4 mm Hg and 393 15 min1 after hemorrhage. A microvascular oximeter was used to make tissue oxygen saturation measurements based on the principles of RRS-StO2 using dedicated software (Fig. 1). Immediately before and after RRS-StO2 measurements, blood was sampled and analyzed using the CO oximeter. Table summarizes the results of SO2 measurements made in all animals during baseline and after hemorrhagic hypotension. The Raman probe was manually and sequentially positioned on the surface of four tissues. A power analysis was performed for all correlations. The sample size and correlation coefficient averaged 23 and 0.6, respectively, whereas a power of 0.8 was obtained for an alpha of 0.05. Measurements of RRS-StO2 in the sublingual and liver surfaces showed significant correlation with venous SO2, and these tissues showed the smallest variability of RRS-StO2 measurements. On average, RRS-StO2 values estimated from Raman spectra in the liver were lower than SO2 values using blood samples by 11.0 3.4%. Raman measurements in the surface of the exposed gracilis muscle showed the largest variability of RRSStO2 and no significant correlation with venous SO2. The correlation between RRS-StO2 in the front paw (skin tissue) and venous SO2 was weaker but still statistically significant. The strong correlation between RRS-StO2 values estimated from resonance Raman spectra in the sublingual surface and SO2 values measured from venous blood samples using a CO
oximeter is presented in Figure 2. The mean difference between RRS-StO2 values estimated from Raman spectra in this tissue and SO2 values using blood samples was 5.4 1.6%. The Raman-based O2 saturation values in the sublingual surface also correlated with various parameters measured in the blood collected from the vena cava (Fig. 3): RRS-StO2 was negatively correlated with blood lactate (r ¼ 0.909, P < 0.01) and potassium (r ¼ 0.757, P < 0.01) and positively correlated with pO2 (r ¼ 0.910, P < 0.01) and pH (r ¼ 0.703, P < 0.05). Although the blood hemoglobin levels fell from a baseline of 14.2 1.1 g/dL to 12.1 1.0 g/dL during hemorrhage, the fraction of methemoglobin remained below 1%.
4.
Discussion
Gas exchange takes place in vessels that contain most of the circulating blood: arterioles, venules, and capillaries [5]. The
Fig. 2 e Values of oxygen saturation estimated from Raman spectroscopy (StO2) and measured from venous blood samples using an oximeter (SO2). The illustrated data are from probes positioned on the oral mucosa (sublingual surface) of seven anesthetized rats. Each point represents a single saturation determination (Raman and oximeter). The least-squares regression was calculated considering measurements during baseline and after hemorrhage (P < 0.001).
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585
5
Fig. 3 e Values of oxygen saturation estimated from Raman spectroscopy (StO2) and levels of different biochemical parameters from venous blood samples. A statistically significant correlation was found in all cases. Data were obtained from seven rats and include measurements during baseline and after hemorrhage.
Raman signal used for estimating RRS-StO2 is a combination of arteriolar, capillary, and venular SO2. The contribution will be proportional to the volumes of these vessels. Because most of the blood volume resides in capillaries and venules [5], the RRS-StO2 mainly reflects the capillary and venular compartments. We tested the hypothesis that RRS-StO2 could serve as surrogate of systemic venous SO2. Significant correlations were found between venous SO2 and RRS-StO2 measurements made in the oral mucosa, skin, and liver. In a swine model of hemorrhagic shock, we have recently shown that RRS-StO2 measurements in the oral mucosa tracked simultaneous changes in central venous SO2 [4]. Previous pilot studies (using large lasers and nonportable collection systems) have also indicated that mucosal RRS-StO2 could track clinically relevant biochemical parameters such as lactate [6]. In the present study, we confirm and expand those preliminary findings while providing quantitative data on the relationship between RRS-StO2 and these parameters. Although different regulatory mechanisms may partially explain dissimilarities in microhemodynamics and macrohemodynamics, microvascular perfusion is also affected by local factors such as leukocyte adhesion, endothelial dysfunction, and rheological alterations. Nevertheless, because it can be imaged using minimally invasive methods [7,8], the sublingual area has been investigated in experimental animals and humans. Its adequacy as a noninvasive site to study systemic changes has been demonstrated by various investigators [9,10]. Although RRS-StO2 measurements have not been reported, decreases in microvascular flow indexes in sublingual microcirculation have been documented after hemorrhagic hypotension in pigs [11] and sheep [12]. The reduction in functional vascular density of the sublingual mucosa matched the
decreases in oxygen delivery in a model of progressive hemorrhage using human volunteers [13]. In addition, the noninvasive assessment of mucosal microvascular perfusion and oxygenation may help to identify patients with potential need for fluid therapy in hemorrhage [14] and sepsis [15,16]. In this context, new methods of high specificity such as Raman spectroscopy, that provide RRS-StO2 estimations using simple algorithms (Fig. 1) and backed by solid preclinical studies [1e4,6] may offer ample potential clinical importance. By using low-power deep violet laser excitation in a mucosal surface, the interfering resonance Raman signals of myoglobin from oral musculature were avoided due to only minimal penetration in this wavelength region. Moreover, low-intensity excitation minimized sample degradation in a tissue not covered by the stratum corneum. The presence of a relatively thick epidermis did not prevent good RRS-StO2 measurements because a significant correlation between RRSStO2 from skin (paw) and venous SO2 was found (Table). Although low-coherence spectroscopy has been used to estimate skin SO2 [17], those determinations were only preliminary, and no attempts were made to correlate skin RRS-StO2 with venous SO2. A significant correlation between RRS-StO2 from liver and venous SO2 was found (Table). These are the first data of RRSStO2 on liver. The fact that RRS-StO2 in the liver were lower than central venous SO2 is not surprising as oxygenation changes in the splanchnic bed may change earlier and be more severe than is reflected in the aggregate measure of central venous SO2 [18]. Moreover, the liver is mostly supplied with venous blood from the portal vein. Near-infrared spectroscopy (NIRS) has been used in the rat to study the effect of hypoxia on liver oxygenation and energy metabolism [19]. In
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650
6
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
that study, the relationship between tissue and venous SO2 651 was not investigated although in pigs, a good correlation was 652 found between tissue oxygenation measured by NIRS and PO2 653 in the hepatic vein [20]. In children, NIRS was used trans654 cutaneously (optode placed over the palpable liver), and the 655 656 measurements did not correlate with SO2 measured from 657 blood samples taken from the inferior vena cava [21]. The 658 authors pointed several reasons for the lack of correlation 659 including the fact that the NIRS values were mainly deter660 mined by non-hepatic measures such as arterial saturation 661 and vascular resistances. Although widely available, NIRS re662 quires care in interpretation due to issues related to the 663 biophysics of measurement and lack of a standard to validate 664 the measurement. Like pulse oximetry, NIRS is based on the 665 principle of relating light absorption to oxyhemoglobin and 666 deoxyhemoglobin concentrations. For example, because op667 668 tical path length and light scattering are not constant in adults 669 or children [22,23], it has been difficult to establish the exact 670 error in the measurement because NIRS lacks a standard to 671 compare it against. 672 We found no correlation between RRS-StO2 from muscle 673 and central venous SO2. One possible explanation for this 674 finding is that optical detection can be obscured by competing 675 background signals, such as those from myoglobin in muscle 676 tissue or other absorbing chromophores [24e26]. As with the Q5 677 respective NIR absorptions, we have previously observed that 678 679 resonance Raman signals of myoglobin are similar to those of 680 hemoglobin [3]. Therefore, these may have interfered with the 681 Q6 estimations of RRS-StO2 from muscle. The fact that myoglobin 682 and hemoglobin have significantly different P50 values, it is 683 not surprising that signals contaminated with myoglobin may 684 not reflect central venous SO2. The fact that the measure685 ments cannot distinguish myoglobin and hemoglobin does 686 not present a major limitation as a highly significant correla687 tion was found between RRS-StO2 and a clinically relevant site 688 (sublingual mucosa). Moreover, using 405-nm excitation, 689 Raman measurements in the most accessible area in the body 690 691 (the skin) can be readily performed without any interfering 692 signals from myoglobin in deeper tissue. 693 694 4.1. Limitations and applications 695 696 Although we tried to track changes in RRS-StO2 and central 697 venous SO2 over a wide range of oxygenation levels, we did 698 not investigate levels reached near death or very high SO2 699 levels. Because animals were not resuscitated, the correla700 tions are unknown during resuscitation using different fluids. 701 702 Only one species (rats) was used although the results are in 703 complete agreement with data from pigs using the same 704 methodology [4]. A low-power continuous laser beam was 705 used to avoid harmful effects associated with intense pulsed 706 laser beams because laser safety limits for the application of 707 ultrashort pulses on human tissues are more stringent than 708 those with continuous wave lasers [27]. In addition to 709 providing specific quantitative data on the relationship be710 tween SO2 and RRS-StO2, this study shows the suitability of a 711 device based on Raman spectroscopy for “point and shoot” 712 oxygenation measurements in thin tissues and in solid or713 714 gans, therefore valuable in multiple clinical settings. This is 715 especially important as a recent study showed that an
abnormal sublingual microvascular flow index was independently associated with an increased risk of hospital death in critically ill patients of 36 ICUs worldwide [28]. Also, evaluations of sublingual microcirculation by physicians and nurses were highly sensitive and specific for sublingual microcirculatory abnormalities [29]. A direct comparison between RRS- Q7 StO2 and the sublingual microcirculation using methods such as sidestream dark field imaging [7], was outside the scope of the present work because SDF currently does not measure O2 saturation. Nevertheless, it is likely that RRS-StO2 measurements will correlate with parameters derived from SDF images.
5.
Conclusion
In summary, RRS-StO2 is a promising technique for the noninvasive evaluation of oxygenation in various tissues such as sublingual mucosa, skin, and liver. The RRS-StO2 measurements in the sublingual mucosa correlate with central venous SO2 and other relevant biochemical parameters such as lactate, potassium, and pH. Additional testing will be required to test its ability to track changes in tissue oxygenation during resuscitation.
Acknowledgment The authors would like to thank Bruce Spiess, Brian Berger, and M. Hakam Tiba for the help during various phases of the study. The opinions or assertions contained herein are the private views of the author and are not to be construed as official or as reflecting the views of the Department of the Army or the Department of Defense. Authors’ contributions: I.T.F. and K.R.W. designed the study. N.M.N. and R.J. performed the research. J.T., P.R., and D.V. contributed to critical aspects on Raman spectroscopy. I.T.F, N.M.N., and R.J. contributed to data analysis. I.T.F. drafted the article. All the authors critically reviewed the article. I.T.F., J.T., and K.R.W. have intellectual property on the use of resonance Raman spectroscopy for tissue oxygenation monitoring assigned to Virginia Commonwealth University. P.R. and D.V. are officers in Pendar Medical, which have licensed the Raman spectroscopy technology from Virginia Commonwealth University.
Disclosures The authors reported no proprietary or commercial interest in Q8 any product mentioned or concept discussed in this article.
references
[1] Torres Filho IP, Terner J, Pittman RN, Proffitt E, Ward KR. Measurement of hemoglobin oxygen saturation using Raman microspectroscopy and 532-nm excitation. J Appl Physiol (1985) 2008;104:1809.
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780
j o u r n a l o f s u r g i c a l r e s e a r c h x x x ( 2 0 1 5 ) 1 e7
781 782 783 784 785 786 Q9 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
[2] Torres Filho IP, Terner J, Pittman RN, Somera LG 3rd, Ward KR. Hemoglobin oxygen saturation measurements using resonance Raman intravital microscopy. Am J Physiol Heart Circ Physiol 2005;289:H488. [3] Ward KR, Barbee RW, Reynolds PS, Filho IP, Tiba MH, et al. Oxygenation monitoring of tissue vasculature by resonance Raman spectroscopy. Anal Chem 2007;79:1514. [4] Tiba MH, Draucker GT, Barbee RW, Terner J, Filho IT, et al. Tissue oxygenation monitoring using resonance Raman spectroscopy during hemorrhage. J Trauma Acute Care Surg 2014;76:402. [5] Wiedeman MP. Dimensions of blood vessels from distributing artery to collecting vein. Circ Res 1963;12:375. [6] Ward KR, Torres Filho I, Barbee RW, Torres L, Tiba MH, et al. Resonance Raman spectroscopy: a new technology for tissue oxygenation monitoring. Crit Care Med 2006;34:792. [7] Goedhart PT, Khalilzada M, Bezemer R, Merza J, Ince C. Sidestream dark field (SDF) imaging: a novel stroboscopic LED ring-based imaging modality for clinical assessment of the microcirculation. Opt Express 2007;15:15101. [8] Groner W, Winkelman JW, Harris AG, Ince C, Bouma GJ, et al. Orthogonal polarization spectral imaging: a new method for study of the microcirculation. Nat Med 1999;5:1209. [9] De Backer D, Dubois MJ. Assessment of the microcirculatory flow in patients in the intensive care unit. Curr Opin Crit Care 2001;7:200. [10] Verdant CL, De Backer D, Bruhn A, Clausi CM, Su F, et al. Evaluation of sublingual and gut mucosal microcirculation in sepsis: a quantitative analysis. Crit Care Med 2009;37:2875. [11] Maier S, Holz-Holzl C, Pajk W, Ulmer H, Hengl C, et al. Microcirculatory parameters after isotonic and hypertonic colloidal fluid resuscitation in acute hemorrhagic shock. J Trauma 2009;66:337. [12] Dubin A, Pozo MO, Ferrara G, Murias G, Martins E, et al. Systemic and microcirculatory responses to progressive hemorrhage. Intensive Care Med 2009;35:556. [13] Ward KR, Tiba MH, Ryan KL, Filho IP, Rickards CA, et al. Oxygen transport characterization of a human model of progressive hemorrhage. Resuscitation 2010;81:987. [14] Pranskunas A, Koopmans M, Koetsier PM, Pilvinis V, Boerma EC. Microcirculatory blood flow as a tool to select ICU patients eligible for fluid therapy. Intensive Care Med 2013; 39:612. [15] Boerma EC, Mathura KR, van der Voort PH, Spronk PE, Ince C. Quantifying bedside-derived imaging of microcirculatory abnormalities in septic patients: a prospective validation study. Crit Care 2005;9:R601. [16] Ospina-Tascon G, Neves AP, Occhipinti G, Donadello K, Buchele G, et al. Effects of fluids on microvascular perfusion in patients with severe sepsis. Intensive Care Med 2010;36:949. [17] Bosschaart N, Faber DJ, van Leeuwen TG, Aalders MC. In vivo low-coherence spectroscopic measurements of local
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
7
hemoglobin absorption spectra in human skin. J Biomed Opt 2011;16:100504. Ba ZF, Wang P, Koo DJ, Cioffi WG, Bland KI, et al. Alterations in tissue oxygen consumption and extraction after trauma and hemorrhagic shock. Crit Care Med 2000; 28:2837. Seifalian AM, El-Desoky H, Delpy DT, Davidson BR. Effect of graded hypoxia on the rat hepatic tissue oxygenation and energy metabolism monitored by near-infrared and 31P nuclear magnetic resonance spectroscopy. FASEB J 2001;15: 2642. El-Desoky AE, Jiao LR, Havlik R, Habib N, Davidson BR, et al. Measurement of hepatic tissue hypoxia using near infrared spectroscopy: comparison with hepatic vein oxygen partial pressure. Eur Surg Res 2000;32:207. Weiss M, Schulz G, Fasnacht M, Balmer C, Fischer JE, et al. Transcutaneously measured near-infrared spectroscopic liver tissue oxygenation does not correlate with hepatic venous oxygenation in children. Can J Anaesth 2002;49:824. Benaron DA, Kurth CD, Steven JM, DelivoriaPapadopoulos M, Chance B. Transcranial optical path length in infants by near-infrared phase-shift spectroscopy. J Clin Monit 1995;11:109. Duncan A, Meek JH, Clemence M, Elwell CE, Fallon P, et al. Measurement of cranial optical path length as a function of age using phase resolved near infrared spectroscopy. Pediatr Res 1996;39:889. Beilman GJ, Myers D, Cerra FB, Lazaron V, Dahms RA, et al. Near-infrared and nuclear magnetic resonance spectroscopic assessment of tissue energetics in an isolated, perfused canine hind limb model of dysoxia. Shock 2001;15: 392. Chance B, Wang NG, Maris M, Nioka S, Sevick E. Quantitation of tissue optical characteristics and hemoglobin desaturation by time- and frequency-resolved multiwavelength spectrophotometry. Adv Exp Med Biol 1992;317: 297. Nighswander-Rempel SP, Kupriyanov VV, Shaw RA. Relative contributions of hemoglobin and myoglobin to near-infrared spectroscopic images of cardiac tissue. Appl Spectrosc 2005; 59:190. Matousek P, Draper ER, Goodship AE, Clark IP, Ronayne KL, et al. Noninvasive Raman spectroscopy of human tissue in vivo. Appl Spectrosc 2006;60:758. Vellinga NA, Boerma EC, Koopmans M, Donati A, Dubin A, et al. International study on microcirculatory shock occurrence in acutely ill patients. Crit Care Med 2015;43:48. Lima A, Lopez A, van Genderen ME, Hurtado FJ, Angulo M, et al. Interrater reliability and diagnostic performance of subjective evaluation of sublingual microcirculation images by physicians and nurses: a multicenter observational study. Shock 2015;44:239.
5.4.0 DTD YJSRE13603_proof 24 December 2015 12:47 pm ce
833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884