Laser assisted breast tumor excision under MRI guidance: a multi-variate study.

University of Minnesota Ph.D. dissertation. September 2009. Major: Mechanical Engineering. Advisor: Arthur G. Erdman. 1 computer file (PDF); xiii, 168 pages, appendices A-E. Ill. (some col.)Introduction: For the year 2008 it was predicted that 182,460 women would get diagnosed with breast cancer and that 40,480 would die due to this disease. These numbers have decreased over the past three decades and this is due to improvements in diagnostic and treatment techniques. This number can be further reduced by implementing MRI for screening, diagnosing and tumor removal. MRI shows promise to detect tumors invisible to other imaging techniques and provides sub-millimeter accuracy in detecting tumor margins. This means that if MRI is used as the surgeon’s eyes during a surgery to remove tumors then the number of cases where small tumor sections or margins are not removed should reduce dramatically. Background: To implement MRI for this treatment there is a gap in research that needs to be bridged. No investigations have been done to remove or treat tumors under real-time MRI guidance. The limitation is a viable tissue cutting modality. Lasers have excellent tissue cutting capabilities and most lasers are fiber deliverable, which are usually MRI compatible. If lasers are used in conjunction with a remote controlled probe position device, they can be used to facilitate breast tumor excision by separating the tumor from surrounding healthy tissue. If a tumor is cut including a thin healthy tissue margin, then it can be removed without any leakage of cancer cells into the circulation and lymphatic system. Thus a pathological complete tumor removal can be confirmed. Gap in research: Lasers have not been used extensively to cut breast tissue. Only one published work was found to cut breast tissue with lasers. This publication by Ansanelli in1986 in Lasers in Surgery and Medicine used CO2 laser to conduct mastectomies. The paper reported only post surgery recovery times. This publication, however, only helps reinforce that lasers can be used for cutting breast tissue in the MRI but CO2 lasers are not deliverable through MRI compatible fibers. Since breast is composed of a heterogeneous mix of fatty tissue, glandular tissue and fibrous tissue, laser ablation research to cut and drill these tissue types was searched. No publications were found to cut these breast tissue constituents. Several publications were found on characterizing the ablation mechanism for drilling holes in soft tissues. No publications were found to characterize ablation for cutting soft tissue. Objective: The objective of this thesis is to experimentally evaluate laser ablation as a modality to cut breast tissue to obtain some baseline data for ablation of this tissue type. The data are collected to obtain lasers and laser system settings where deep cuts can be achieved in ex vivo breast tissue and breast tissue analogue with minimum damage to the surrounding tissue (less than 5 mm recommended by surgeons). Materials and Methods: Experiments were carried out to study tissue cutting effects in ex vivo porcine liver tissue (used as a model for breast tissue in extensive studies) and ex vivo human breast tissue (for specific studies) using two lasers. An 808 nm continuous wave diode laser was chosen for its ability to cut tissue (demonstrated in several urology applications), its compact size and inexpensive manufacturing cost. Also a 1064 nm pulsed Nd:YAG laser was chosen because (a) it is absorbed well by water (peak at this wavelength in water’s absorption spectrum) and it is assumed that soft tissue hence breast tissue is composed of 70% water and (b) pulsed lasers are known to cause ablation with very little damage to the surrounding tissue. Light from the 808 nm laser was delivered onto the tissue sample surface via a 600 μm fiber optic cable placed in contact with the tissue. Non-contact mode of light delivery was used with 1064 nm laser; light was focused on the tissue surface using a lens. To simulate tissue cutting using scalpels, laser fiber in 808 nm laser experiments was moved over tissue and in 1064 nm laser experiments tissue was moved under a fixed beam-delivering head. The thickness of vaporized tissue and thermally damaged tissue were quantified using digital microscopy and histopathology analysis using haematoxylin and eosin (H&E) staining. Four experiments were carried out as a combination of the two lasers and tissue types. Experiment I: This experiment was carried out with liver tissue and 808 nm laser system in three parts. In the first part, a constant fiber velocity 0.05 inch/s was used and with each sample from one liver, the power was ramped up to identify the coagulation threshold (identified as whitening of tissue) and the ablation threshold (identified by the observation of a crater). Part two was a out with liver tissue and 808 nm laser system in three parts. In the first part, a constant fiber velocity 0.05 inch/s was used and with each sample from one liver, the power was ramped up to identify the coagulation threshold (identified as whitening of tissue) and the ablation threshold (identified by the observation of a crater). Part two was a randomized complete block design with two factors each at three levels, power (~15, 20, 25 W) and velocity (0.03, 0.05, 0.07 inch/s). Measurements of ablation width and depth and coagulation width and depth were made at two locations along the cut. In part three, each sample was cut twice at a fixed power (~20 W and velocity 0.05 inch/s). The first linear cut was 1 inch long and the second, ½ inch long. Measurements of ablation width and depth and coagulation width and depth were made in sections with one cut and two cuts. The objective was to find out if a change in tissue properties after the first cut affects ablation during the second cut. Results Experiment I: At 0.05 inch/s fiber velocity, coagulation occurred at 1.42 W and ablation at 2.94 W. Damage increased with increasing power and except for coagulation depth the other responses (measurements) showed significant variation with power. Ablation depth and width and coagulation depth and width showed significant increase with decreasing fiber velocity. Ablation depth did not show any significant variation, statistically, from one liver to another. The other 3 responses (ablation width, coagulation depth and width) were significant (p-value < 0.05). From part three of this experiment, it was determined that changes in tissue properties after one exposure do not affect further tissue cutting. The ablation depth after the second cut almost doubled without a significant increase in coagulation. For ablating liver tissue with this 808 nm system a power of 24.8 W and fiber velocity of 0.03 inch/s is most suitable for maximum cutting depth and thin coagulated tissue thickness. Experiment II: This experiment used liver tissue and 1064 nm laser. This experiment was also carried out in three parts. In the first part, coagulation and ablation threshold for drilling a hole were identified by increasing pulse duration (which increased pulse energy) while keeping frequency and number of pulses fixed at 15 Hz and 20 pulses. In the second part, ablation for drilling a hole in liver tissue was carried out in two sub-parts with two livers. In the first sub-part, pulse duration was varied at 3 levels, 4 ms, 7 ms and 10 ms while keeping pulse frequency and number of pulses per sample fixed at 15 Hz and 10 pulses. In the second sub-part, pulse frequency was varied at three levels 30, 45 and 60 Hz while keeping pulse duration and number of pulses fixed at 4 ms and 10 pulses respectively. The objective for this portion of the vi experiment was to characterize drilling with the system and provide a point of comparison of this experimental work with published work and also to contrast with ablation for cutting tissue (either moving the tissue or the beam). The third part of this experiment was carried out in three sub-parts for ablation for tissue cutting (tissue moved under laser beam) with two livers. In the first sub-part, pulse duration was varied (4, 7, 10 ms) at 15 Hz frequency and 0.03 inch/s tissue velocity. In the second sub-experiment frequency was varied (15 and 30 Hz) at 4 ms pulse duration and 0.01 inch/s tissue velocity. In the third sub-part one sample per liver was moved at 0.02 inch/s at 4 ms pulse duration and 15 Hz frequency. The objective of this part was to quantify the effects of ablation, determine the trends with maximum parameter variation, and to also compare ablation for drilling holes in tissue verses cutting tissue. Results Experiment II: During threshold experiment, coagulation in liver tissue was first noticed at 1.9 ms pulse duration and pulse energy of 126 J. Ablation was first noticed at 3 ms pulse duration at 275 J of energy per pulse. In part two of the experiment, with increasing pulse frequency, while drilling tissue, no ablation and only coagulation was seen. The hole depth after 20 pulses at 4 ms and 15 Hz was 140 μm and coagulation was 90 μm thick. Part three of this experiment, while cutting tissue, damage increased with increasing pulse duration, pulse energy, frequency and decreasing velocity. At 0.01 inch/s tissue velocity, 4 ms pulse duration and 15 Hz the ablation, the depth was 760 μm and coagulation was 200 μm thick. At the same settings the damage to drill tissue is very small indicating that while cutting tissue, changes in tissue property due to heat transfer enhances ablation damage. For ablation to cut liver tissue with this laser system the best combination of parameters is 4 ms pulse duration, 30 Hz frequency and 0.01 inch/s tissue velocity. Experiment III Experiment III & IV: These experiments were carried out with ex vivo human breast tissue and both laser systems. One parameter combination was chosen per laser and all samples were ablated in the cutting mode. The objective was to get the largest possible sample size with the small volume of available tissue for one parameter combination and use liver tissue data to correlate response for other parameter combinations. 4 tissue slides, ablated with the 808 nm laser at ~20 W and 0.03 inch/s fiber velocity were pathologically analyzed and 5 slides, ablated with 1064 nm laser at 4 ms pulse duration, 15 Hz frequency and 0.01 inch/s velocity, were pathologically analyzed. Results Experiment III & IV: The average ablation depth and width in breast tissue was 1.15 mm and 3.71 mm respectively with the 1064 nm laser. These were higher than the ablation depth and width caused by the 808 nm laser. With the 808 nm laser, ablation depth was 0.88 mm and width was 2.18 mm. Coagulation thickness was smaller with 808 nm laser (at 0.2 mm) than the 1064 nm laser (at 0.28 mm). In slides that had only fatty tissue the ablation extent with both lasers was high with no adjacent tissue coagulation. The 1064 nm laser is a better choice to make deep cuts by ablation in breast tissue. Conclusion: This research effort has been successful in providing the first baseline data for ablating ex vivo breast tissue with two laser systems and providing a reference data base of ex vivo liver tissue cutting under similar conditions with the two systems. This research proves the feasibility of using lasers for minimally invasive real-time MRI guided breast tumor removal surgery. The research data also helps to conclude that for this surgery, the 808 nm laser is a better choice than the 1064 nm system, even thought ablation depth achieved is slightly less. This is because the 808 nm laser is less expensive than the 1064 nm laser and is better suited for placement in the surgery suite as it does not need external water cooling like the 1064 nm laser. The 808 nm laser also has a smaller foot-print, it is portable and is compatible with cheaper fibers