Technology-Guided Therapy: A Research Program within a Hospital

Figure 1. Virtual liver remnant (dark orange) projected at (a) 34% of total liver volume before portal vein embolization, and (b) 43% of total liver volume after portal vein embolization.

by Amber L. Simpson, PhD

Technology is increasingly being used to help clinicians make more informed treatment decisions in the presence of imperfect data. My research lab strives to translate technology to clinical use by working directly with surgeons, radiologists, and oncologists. We develop novel computing strategies for image-guided surgery, provide virtual surgical planning, and define prognostic imaging markers. Our hope is to provide enabling technologies for precision oncology.

Virtual Organ Volumetry for Surgical Planning

Full recovery from major hepatic resection requires a healthy, well-perfused liver remnant that is capable of regenerating to its pre-resection volume. Several studies have shown that the percentage of functional liver parenchyma remaining after major hepatic resection is one of the few reliable predictors of postoperative hepatic dysfunction and morbidity [1,2]. Having demonstrated that liver volumetry based on preoperative scans can be performed accurately [3], we then developed patient-specific virtual models that provide the surgeon with three-dimensional measurements of parenchyma, tumor, and delicate vasculature in relation to resection lines. Today, we routinely perform detailed virtual planning for individual patients undergoing major liver resection.

As an example, the image in Figure 1a is from virtual surgical planning software for a patient with a potentially small liver remnant (dark orange), and includes the remnant volume projection at 34 percent of the original liver volume. To stimulate growth of the liver remnant prior to resection, the patient underwent portal vein embolization. Figure 1b shows the same patient with increased remnant liver volume (dark orange) at 6 weeks after embolization, including the remnant volume projection at 43 percent of original liver volume. Because growth rate after portal vein embolization is a good predictor of hepatic resection outcome [4], this patient proceeded to resection.

Figure 1. Virtual liver remnant (dark orange) projected at (a) 34% of total liver volume before portal vein embolization, and (b) 43% of total liver volume after portal vein embolization.
Figure 1. Virtual liver remnant (dark orange) projected at (a) 34% of total liver volume before portal vein embolization, and (b) 43% of total liver volume after portal vein embolization.

Image-Guided Liver Surgery

Image-guided surgery systems provide a virtual roadmap, much like a global positioning system, through a patient’s anatomy. The liver is soft, compressible, and undergoes displacement when detached from its ligaments and positioned for resection. Consequently, the positioned liver no longer corresponds exactly in shape to the liver visible on preoperative cross-sectional imaging. This discrepancy makes it difficult to locate small, deeply placed tumors, especially after chemotherapy when the liver parenchyma has become fatty or fibrotic. Surgical navigation systems are routinely used to more easily localize these tumors. Figure 2 shows the image displayed by one such guidance system when the tumor is successfully located on CT, after the tumor was not found using standard ultrasound alone.

Figure 2. Tumor successfully located using CT-based navigation combined with ultrasound: (a) green probe indicates tumor on the image-guidance display of CT, and (b) yellow arrow indicates the tumor displayed in ultrasound. Initially, the tumor was not able to be localized using standard ultrasound alone.
Figure 2. Tumor successfully located using CT-based navigation combined with ultrasound: (a) green probe indicates tumor on the image-guidance display of CT, and (b) yellow arrow indicates the tumor displayed in ultrasound. Initially, the tumor was not able to be localized using standard ultrasound alone.

The challenge still is to provide surgeons with real-time, continuous visualization of internal structures (including tumors and vasculature) relative to their surgical objectives, thus guiding intervention while avoiding injury. The solution in development at MSK uses stereo-vision technology (Figure 3), where dual video cameras capture organ shape in much the same way that our eyes perceive depth. Continuous imaging technology represents a new paradigm of surgical guidance that has the potential to further improve surgical accuracy.

Figure 3. Stereo-vision system for continuous surgical guidance, in development at MSK.
Figure 3. Stereo-vision system for continuous surgical guidance, in development at MSK.

Prognostic Imaging Markers

Radiologists assess subtle changes in contrast enhancement patterns on images to determine diagnosis and treatment response. These heterogeneity changes can be measured with a quantitative tool called texture analysis. Figure 4 shows a standard diagnostic abdominal CT scan with a region of interest (ROI) magnified. Changes in pixel values across the magnified region can be appreciated. Texture analysis describes the spatial variation of pixel intensity in a ROI where the ROI could be any biologically relevant structure such as parenchyma, tumor, or node.

Figure 4. Variation in pixel intensity visible within a magnified region of interest of a standard diagnostic CT scan.
Figure 4. Variation in pixel intensity visible within a magnified region of interest of a standard diagnostic CT scan.

With support from Cycle for Survival, we are evaluating tumor texture as a preoperative prognostic marker in pancreatic ductal adenocarcinoma (PDAC). PDAC is one of the most lethal cancers worldwide with a 5-year overall survival rate of 6 percent [5]. Complete surgical resection, achievable in 10 percent to 15 percent of patients, is the only curative treatment [6]. Thus, determining preoperative prognostic factors is crucial for these patients. At MSK, we have demonstrated that extracting texture from pre-treatment scans can provide patient prognosis stratification. The receiver operating characteristic (ROC) curve in Figure 5 demonstrates the power of texture variables to predict survival with 83% accuracy (using cross validation). This marker was independent of all other clinical variables. Although this proof of principle needs further validation across a sufficiently powered study, it nonetheless shows promise for the evaluation of PDAC patients.

From virtual surgical planning, to image-guided surgery, to prognostic imaging markers, we continue to develop technologies with an eye toward the ultimate goal of precision therapy for hepatic and pancreatic tumors.

Figure 5. ROC curves obtained with different sets of texture features, extracted from tumor region using cross-validation method
Figure 5. ROC curves obtained with different sets of texture features, extracted from tumor region using cross-validation method

References

  1. Shoup M, Gonen M, D’Angelica M, Jarnagin WR, DeMatteo RP, Schwartz LH, Tuorto S, Blumgart LH, Fong Y. Volumetric analysis predicts hepatic dysfunction in patients undergoing major liver resection. J Gastrointest Surg. 2003;7(3):325–330. 
  2. Schindl MJ, Redhead DN, Fearon KCH, Garden OJ, Wigmore SJ. The value of residual liver volume as a predictor of hepatic dysfunction and infection after major liver resection. Gut. 2005;54(2):289–296.
  3. Simpson AL, Geller DA, Hemming AW, Jarnagin WR, Clements LW, D’Angelica MI, Dumpuri P, Gonen M, Zendejas I, Miga MI, Stefansic JD. Liver planning software accurately predicts postoperative liver volume and measures early regeneration. J Am Coll Surg. 2014;219(2):199–207.
  4. Leung U, Simpson AL, Araujo RLC, Gönen M, McAuliffe C, Miga MI, Parada EP, Allen PJ, D’Angelica MI, Kingham TP, DeMatteo RP, Fong Y, Jarnagin WR. Remnant Growth Rate after Portal Vein Embolization Is a Good Early Predictor of Post-Hepatectomy Liver Failure. J Am Coll Surg. 2014 Oct;219(4):620–630. PMID: 25158914
  5. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin. Jan;62(1):10–29. PMID: 22237781
  6. Alexakis N, Halloran C, Raraty M, Ghaneh P, Sutton R, Neoptolemos JP. Current standards of surgery for pancreatic cancer. Br J Surg. 2004;91(11):1410–1427. PMID: 15499648

Amber L. Simpson, PhD, Assistant Attending Computational Biologist, Hepatopancreatobiliary Service, Department of Surgery, Memorial Sloan Kettering Cancer Center

Be the first to comment

Leave a Reply

Your email address will not be published.


*