Cryoablation and Resection Influences Patient Survival for Soft Tissue Sarcomas

Impact on Survivorship and Local Recurrence

Abstract

Cryosurgical ablation has been used successfully for treating various carcinomas and bone tumors; however, few studies report the use of cryosurgery followed by tumor resection for the treatment of soft tissue sarcomas. We evaluated local recurrence, progression of disease, histologic necrosis from the cryoablation, complications, patient survival, and functional outcomes using this approach. We retrospectively reviewed 38 patients with no prior treatment for their neoplasm who underwent cryosurgery followed by wide excision of soft tissue sarcomas. Three patients developed local recurrence. Sixteen patients had evidence of more than 95% tumor necrosis, and 11 of these had 100% histologic necrosis. We observed a difference in overall survival and disease-free survival based on the adequacy of freezing process. Patients with more than 95% necrosis had a survival rate of 94% at 2 years and 86% at 5 years, while those with less than 95% necrosis survived 53% at 2 years and 34% at 5 years. Complications included transient neuropraxia (13%), superficial wound infections (8%), and seroma (21%). Cryosurgical ablation appears a safe and effective method of devitalizing tumor cells of soft tissue sarcomas.

Level of Evidence: Level IV, therapeutic study. See the Guidelines for Authors for a complete description of levels of evidence.

Soft tissue sarcomas have been treated with various modalities including surgery, radiation, chemotherapy, or a combination of these treatments. Achieving local tumor control while preserving limb function is an important goal of treatment. The use of chemotherapy as a surgical adjuvant has had equivocal results in the treatment of soft tissue sarcomas.^6,7,18 The combination of surgery and radiation therapy has been generally effective in treating high-grade sarcomas, but results have not been particularly impressive in the treatment of low-grade tumors.²

Toxicity and serious complications are often associated with the use of these adjuvant modalities leading to poor functional outcomes. Radiation therapy for soft tissue sarcomas of the extremities causes substantial morbidity, including epidermolysis, fibrosis, stiffness, delayed wound healing, wound infection, radionecrosis of bone, degenerative arthritis, peripheral neuropathy, and the later development of radiation-induced sarcoma.^21,23,35,45 The morbidity of systemic chemotherapy in the treatment of cancer is well-recognized, and the efficacy of this modality in the treatment of soft tissue sarcomas is controversial.^6,7,18 Additionally, these adjuvants are often contraindicated in certain patients, including the elderly, those with multiple medical conditions, those who have previously been treated with the maximum allowable dose of radiation, and those with tumors unresponsive to these modalities. In the past, such patients had no option other than to undergo surgical resection without prior treatment of their tumors. Although wide resection alone should theoretically remove all tumor, neoadjuvant therapy has the benefit of sterilizing the adjacent soft tissues, shrinking the tumor, and destroying microscopic satellite lesions that may be left behind, thereby facilitating resection. Along these lines, cryoablation may be an appropriate adjunct to make limb salvage safer by decreasing the amount of histologically viable tumor.

Cryoablation is a recognized approach to the treatment of various malignancies including those of the liver,^38,39 prostate,^3,17,19 head and neck,⁴³ eye,^10,22 and skin.^24,44 The range of application of cryosurgical techniques to the treatment of cancer is widely diversified and steadily increasing in scope. This technique was first shown to be effective in devitalizing tissue in the 1850s when iced saline solutions were used to treat breast and uterine carcinomas.¹² In 1907, liquid nitrogen was used to freeze skin carcinomas, and in the 1950s a multitude of carcinomas were treated with this method.¹³ The concept of a cryoprobe was first introduced in 1963 for use in neurosurgical procedures.¹⁴

Cryosurgery is most useful in easily accessible areas of the body, but recent advances in technology have made it possible to treat tumors in areas previously thought relatively inaccessible. Cryoprobes can provide a controllable and predictable area of necrosis and their use is the preferred method of treatment in many situations.^{11,13,14,30,36,41} Dramatic advances in imaging technology, especially the development of computer-enhanced real-time intraoperative ultrasonography have allowed for the safe and precise monitoring of deep-seated lesions that before could not be visualized adequately. Cryosurgical ablation has been extensively used in the treatment of various neoplasms of bone.^{2,20,25-29,42} There are no published data suggesting cryosensitivity of certain cell types, and so we could assume any tumor can be devitalized with adequate freezing. To date, however, soft tissue sarcomas have not been treated extensively with this modality, although these tumors are present in relatively accessible locations.

We wanted to determine whether cryoablation followed by tumor resection for the treatment of soft tissue sarcomas produces histologic necrosis comparable to that of other neoadjuvant treatments and whether it influences patient survival, local recurrence, and functional outcome.

MATERIALS AND METHODS

We retrospectively searched our orthopaedic oncology database and identified 133 patients with soft tissue sarcomas treated from August 1997 to April 2003. We included patients with a histologically confirmed diagnosis of nonmetastatic soft tissue sarcoma, no prior treatment with neoadjuvant chemotherapy or radiation therapy prior to surgical resection, and the use of cryoablation as an intraoperative adjuvant. Patients who were treated with neoadjuvant radiation and/or chemotherapy in addition to cryoablation were excluded from this study because the amount of resulting necrosis could not be attributed to the cryosurgical procedure alone. Thirty-eight patients met the criteria. We reviewed their charts for information pertaining to demographics, location of the lesion, radiographic evaluation, surgical management, histology, complications, and functional outcome. A power analysis indicated 652 patients would be required to demonstrate a statistically significant increase in recurrence-free survival and overall survival as a result of cryoablation. The study included 20 male and 18 female patients with an average age of 59 years (range, 28-92 years). The minimum followup was 6 months (average, 37 months; range, 6-98 months). All living patients had a minimum of 24 months of followup. Patients who died of disease before reaching 24 months of followup were also included in the study.

Locations of the tumors were the thigh (13), leg (4), forearm (3), chest wall (3), buttocks (3), pelvis (3), shoulder (1), arm (1), and hand (1). We used the surgical staging system adopted by the American Joint Committee on Cancer⁹ to classify the tumors: Stage IA (4), Stage IB (4), Stage IIA (2), Stage IIB (19), and Stage IIIB (9). Diagnoses included malignant fibrous histiocytoma (16), myxoid liposarcoma (6), extraskeletal chondrosarcoma (5), leiomyosarcoma (3), synovial cell sarcoma (3), malignant Schwannoma (2), epithelioid sarcoma (1), and fibrosarcoma (1).

All patients had been referred to the senior author (LRM) for treatment. At presentation, a workup including detailed history and physical examination, plain radiographs, magnetic resonance imaging (MRI), and computerized tomography (CT) scan of the chest were obtained according to a standard protocol. After appropriate staging, all patients underwent incisional biopsy to confirm the diagnosis of a soft tissue sarcoma.

We performed surgical resection through a longitudinal incision made over the sarcoma. Appropriate skin flaps or fascial cutaneous flaps were raised and a cuff of normal tissue was left around the lesion. Neurovascular structures not in direct continuity with the pseudocapsule of the tumor were mobilized and retracted. After taking care to adequately protect the surrounding soft tissues, neurovascular structures, and skin, cryosurgical ablation commenced. Ultrasonography of the lesion was performed using 5-MHz or 7.5-MHz intraoperative transducers (3535 Ultrasound Scanning System, B-K Medical Systems, Wilmington, MA) to assist with strategic placement of the cryoprobes. Probes 3 mm, 8 mm, or 10 mm in diameter were inserted sequentially into the lesion at the discretion of the surgeon, depending on which cryosystem was used. The initial five cases were performed using the Cryotech LCS 3000® system (Candela Laser Corporation, Wayland, MA), and the ensuing seven cases were performed with the CMS AccuProbe® 450 cryosurgical system (Cryomedical Sciences, Inc, Rockville, MD). Both systems circulate gaseous nitrogen through the cryoprobes to effect freezing of tissue. Since 2001 we have been using the CRYOcare® system (Endocare Inc, Irvine, CA), which uses argon gas and therefore has the advantage of freezing faster and colder than liquid nitrogen to achieve temperatures of up to −186°C. For all cryoablative procedures, the probes were placed to ensure a complete freeze of the tumor as well as a minimum 10-mm cuff of normal tissue. Intralesional temperatures of −160°C were maintained for 15 minutes with constant monitoring using thermocouples (Fig 1). The freeze cycle was also monitored with ultrasonography by visualizing the freezing interface as an advancing hyperechoic hemispheric rim with complete acoustic shadowing (Fig 2). The ice ball was extended 10 mm beyond the tumor margin into normal tissue in an attempt to freeze satellite lesions that may have been present in the reactive zone of the tumor. Once the ice ball reached the appropriate size, the tumor was allowed to slowly thaw by circulating helium gas through the probes and maintaining a temperature of 0°C for 5 minutes. The cycle was then repeated for a total of two cycles of fast-freeze/slow-thaw. Warm distilled water was then used to aid with thawing of the ice ball that encompassed the tumor, as well as to enhance the tumoricidal effect. At the end of the second thawing cycle, the probes were removed and the holes left by the cryoprobes were packed with Gelfoam® (Upjohn Co, Kalamazoo, MI). Wide resection of the tumor was then performed often necessitating the use of osteotomes to resect the frozen tissue. The cryoablation was considered successful if an ice ball encompassing the entirety of the tumor and a wide margin was visualized and successfully performed on all 38 patients. Overall, the entire cryoablative process added approximately 60 minutes to the procedure. No procedure required intraoperative termination due to technical reasons or complications. The wound was closed in the usual manner and a suction drain inserted.

Fig 1: Seven 8-mm cryoprobes inserted into the sarcoma around the ellipsed biopsy tract are seen at the center of the image. Two smaller 1.2-mm temperature transducers inserted at the periphery of the lesion monitor temperature at the margins of the ice ball. The surrounding skin is retracted and protected with use of laparotomy pads to prevent necrosis as a result of freezing.

Fig 2: An ultrasound image of two cryoprobes labeled proximal and distal to a sarcoma of the thigh is shown. Between the cryoprobes, the advancing ice ball can be visualized as a hyperechoic mass.

All patients received postoperative antibiotics until the drain was removed, and they were discharged from the hospital when medically stable. Patients were monitored for complications related to their surgery.

Gross specimens were assessed and measured for tumor dimensions at the time of resection. The average tumor volume was 848 cm³ (range, 12-6800 cm³). Tissues were processed according to previously described methods^37,40 with the goal of obtaining representative samples from all areas of the specimen in order to find viable residual tumor. Each tumor was initially inked and cut along the long axis resulting in slab sections of opposing hemispheres. The slab sections were cut both parallel and perpendicular into small cubes from end-to-end and the location of the segments labeled in order to map the specimen. Histologic analysis was performed from each mapped segment as well as from random samples to obtain representative sections from all areas of the tumor. An average of 30 histologic sections from each tumor was made. Permanent histologic specimens were reexamined retrospectively by two pathologists (ANF, AC) specializing in musculoskeletal neoplasms, who agreed with the reported findings. Microscopically, tumors were classified as either low- or high-grade on the basis of established criteria, including the degree of nuclear pleomorphism, degree of differentiation, and number of mitoses per high-power field. Multiple morphologic parameters were reviewed, including the presence or absence of tumor, anatomic distribution of residual tumor when present, qualitative and quantitative analysis of viable tumor when present, and a review of resection margin status. A clear margin was defined as a minimum of 10 mm between the surgical margin and evidence of tumor. Quantitative analysis was generated by reviewing and gathering data in a slide-by-slide fashion, totaling and then averaging the values for extent of necrosis as well as residual tumor. Extent of necrosis was determined relative to the percentage of residual viable tumor.

Patients were followed postoperatively at regular intervals every 3 months for the first 2 years, every 6 months between 2 years and 5 years after treatment and yearly thereafter. Each followup included physical examination, MRI of the extremity, CT scan of the chest, and specific questions about symptoms and complications. Functional evaluation was performed by administering the Musculoskeletal Tumor Society (MSTS) survey for the functional outcome of surgical management of musculoskeletal tumors⁸ at the latest followup. This survey equally weights pain, function, emotional acceptance, ambulatory supports, walking ability, and gait to obtain an overall numeric rating, as well as a subjective rating of patient satisfaction.

Statistical analysis was performed using the Kaplan-Meier method to generate survivorship curves with corresponding 95% confidence intervals for overall patient survival and disease-free survival using the GraphPad Prism® software (GraphPad Software, Inc, San Diego, CA). Time zero was defined as the date of surgical resection. The endpoint of evaluation for disease-free survival was defined as the time to development of either local recurrence or metastatic disease. For overall survival the end-point was confirmed tumor-related mortality. All other deaths unrelated to the neoplasm were censored. The log-rank test and Cox proportional-hazards regression model were used to analyze the association between percent necrosis and recurrence or overall survival. A p value of less than 0.05 was considered significant.

RESULTS

Histologic evaluation of permanent specimens revealed 16 of the 38 patients (42%) had evidence of near-complete (≥95%) tumor necrosis, and 11 of these (29%) had complete (100%) histologic necrosis (Fig 3). However, the majority of specimens revealed less than 95% necrosis. In 58% of cases, cryoablation was not effective in achieving complete devitalization of tumor tissue. The percent necrosis for these 22 cases was 90% (3), 80% (4), 70% (1), 50% (4), 40% (3), 30% (1), 20% (2), 10% (2), and 5% (2). In all cases the viable tissue seen in the resected tumor was only at the periphery or rim, indicating that there was inadequate freezing and the ice ball did not encompass the entire tumor. Six of these patients had relatively large tumors with a volume greater than 1000 cm³ (mean, 1894 cm³). In these cases only a 1- to 2-cm rim of viable tumor remained at the periphery. All tumors were resected with a clear margin of 10 mm between the surgical margin and evidence of tumor. With the numbers available, we found no difference in percent tumor necrosis for those tumors less than 1000 cm³ and more than 1000 cm³, indicating there should be no contraindication to treating large tumors with cryoablative techniques. We observed no bias toward a certain grade of tumor demonstrating either an entirely favorable or unfavorable amount of tumor necrosis (Table 1).

TABLE 1: Pathologic Necrosis Based on American Joint Committee on Cancer (AJCC) Grade

Fig 3A: B. (A) A histologic biopsy specimen of patient with malignant fibrous histiocytoma before treatment shows malignant histiocyte-like cells growing in a storiform pattern with nuclear pleomorphism and cellular atypia. (B) A postresection specimen of the same patient exhibits an excellent response to treatment with more than 95% histologic necrosis after cryoablation (Stain, hematoxylin & eosin; original magnification, ×100).

We detected no difference in the recurrence rate in those patients with complete freezing of the tumor indicated by evidence of more than 95% necrosis from the rate in patients with less than 95% necrosis. The recurrence-free survival was 87% at both 2 and 5 years for patients with more than 95% necrosis, and 87% and 72% respectively for those with less than 95% necrosis based on Kaplan-Meier survivorship analysis (Fig 4). Local tumor recurrence occurred in three patients with an average time from surgical resection to the development of local recurrence of 16 months (range, 5-37 months). All tumors were malignant fibrous histiocytomas and all demonstrated less than 95% necrosis. Tumor necrosis for these patients was an average of 50% (range, 10-50%) and average tumor volume was 3132 cm³ (range, 956.3-6300 cm³). All patients were treated with reexcision and adjuvant cryoablation and are currently without evidence of disease.

Fig 4: Kaplan-Meier survivorship curves with 95% confidence intervals show no difference in recurrence-free survivorship based on percent necrosis.

Overall patient survival was less (p = 0.006) in patients who had tumors that were not adequately frozen: those patients who had less than 95% necrosis had poorer survival at 2 and 5 years that those with greater than 95% necrosis (Fig 5).

Fig 5: Kaplan-Meier survivorship curves with 95% confidence intervals show that patients with adequate freezing indicated by more than 95% histologic necrosis have a greater (p = 0.006) overall patient survival than those patients who had less than optimal tumor freezing. The 2-year and 5-year survivorships are 94% and 86% respectively for patients with more than 95% necrosis. For patients with less than 95% necrosis, survival is 53% at 2 years and 34% at 5 years.

We observed a decrease (p = 0.021) in disease-free survival in those patients with less than complete freezing of their tumors. Disease-free survival for patients with more than 95% necrosis was 85% at 2 and 5 years respectively. For patients with less than 95% necrosis, disease-free survival decreased to 60% at 2 years and 50% at 5 years (Fig 6).

Fig 6: Kaplan-Meier survivorship curves with 95% confidence intervals show that patients with adequate freezing indicated by more than 95% histologic necrosis have a greater (p = 0.021) disease-free survival than those patients who had less than optimal tumor freezing. Patients with more than 95% necrosis had disease-free survival of 85% at 2 and 5 years respectively. Patients with less than 95% necrosis had 60% and 50% 2 and 5 year disease-free survival respectively.

The most common complication associated with the procedure was the development of serous wound drainage and seroma occurring in 8 patients (21%). This complication is seen frequently with cryoablative procedures due to both the resulting liquefaction of residual frozen tissue as well as the large volumetric defect created as a result of resection. The average tumor volume resected in the patients who developed seromas was 1369 cm³ (range, 16.5-6800 cm³). None of the patients with serous drainage or seromas became infected. The wound drainage subsided within 6 weeks (range, 2-6 weeks). In six patients, seromas resolved after aspiration performed between 1 and 4 weeks postoperatively. Two patients with large seromas that recurred after aspiration required surgical evacuation. No patients in this study required soft tissue coverage with muscle flaps.

Peripheral nerve palsies occurred in five patients (13%) who all had their respective nerves deliberately frozen as we believed mobilizing the nerves from their respective locations in close proximity to the tumor would potentially contaminate the surgical margin. There were three palsies of the sciatic nerve and one of the posterior tibial nerve, which occurred after resection of tumors of the posterior compartment of the thigh. The remaining patient developed palsies of both the radial and ulnar nerves after resection of a tumor from the arm. All five cases initially had complete motor and sensory neuropraxias but demonstrated complete recovery of function, two within 1 week of cryoablation, two within 3 months, and the remaining patient within 6 months.

Superficial wound infections that resolved after a course of oral antibiotics occurred in three patients. No cases of deep wound infection, deep venous thrombosis, pulmonary embolism, or skin slough occurred in this study.

For the patients surviving at the time of this study, the average MSTS functional outcome score was 26 (range, 22-30) indicating patients achieved good to excellent postoperative function corresponding to 87% of baseline function. The patients who scored lowest were those who had experienced nerve palsies and therefore had prolonged recovery time. All patients in this study were able to return to activities of daily living and occupation-related activities.

DISCUSSION

Cryosurgery is a method of treating neoplastic tissue in situ to achieve devitalization prior to resection.^13,14 It is now a recognized approach to the treatment of malignant tumors and the range in application of this technique is widely diversified and steadily increasing in scope. Malignancies in various parts of the body, including bone,^2,20,25,28 have been treated with cryosurgical techniques. Experience in treatment of soft tissue sarcomas with this modality has been limited.³¹

As with most studies evaluating the effectiveness of a treatment regimen for soft tissue sarcomas, the present study is limited by small patient numbers and limited followup. These are rare tumors and acquiring adequate numbers of patients is difficult. However, this study may be useful for evaluating the effectiveness of an alternative treatment for a group of patients who otherwise have few options for adjunctive therapies. Although soft tissue sarcomas often take a substantial amount of time to recur or metastasize, previous studies have reported most patients who develop local recurrence will do so within 2 years of surgery.^6,7,18,23,45 For this reason we limited the study to those living patients with a minimum 2-year followup resulting in a relatively small group of patients. However, we also included those patients who died of disease before they reached 2 years of followup in order not to eliminate those patients who died early in the course of treatment.

The achievement of local control is essential in the successful treatment of soft tissue sarcomas.^21,23,45 Although wide resection alone should theoretically remove all tumor, neoadjuvant therapy has the benefit of sterilizing the adjacent soft tissues and destroying microscopic satellite lesions that may be left behind, thereby facilitating resection. Cryoablation may be an appropriate adjunct to make limb salvage safer for patients who cannot undergo chemotherapy or radiation by decreasing the amount of histologically viable tumor present during resection. In addition, extending the zone of cryoablation past the grossly apparent tumor may provide for a safer and greater margin by destroying microscopic satellite lesions that would otherwise have been left behind as viable tumor and ultimately lead to recurrence. In theory, this should decrease the local recurrence rate and may ultimately improve overall survival.

Successful cryoablation requires that the same volume of tissue must be frozen as would have been excised if the tumor was treated with conventional wide resection alone. The amount of tissue devitalized cryosurgically must extend past the reactive zone of a tumor and into surrounding normal tissue so that all satellite lesions are destroyed.^13,14,31 In this study, percent of tumor necrosis was used as a indicator of extent and adequacy of the tumor freezing process. Cryoablation was effective in producing more than 95% tumor necrosis in 42% of patients. The necrosis seen in these specimens was virtually complete and easily identifiable by histology. In the treatment of osteogenic sarcoma, necrosis of greater than 95% has been shown to correlate with a complete response to chemotherapy as well as improved prognosis.^1,46 For this reason we used 95% as the cut-off amount in the current study to avoid interpreting spontaneous necrosis of the tumor as a response to cryoablation.

The difference in overall survival and disease-free survival we found based on amount of histologic necrosis emphasizes the importance of complete freezing of the tumor. The fact that only 42% of tumors had either complete or near complete necrosis highlights the technical difficulty of the cryoablation procedure. Of the patients with less than 95% necrosis five were treated with the Cryotech® LCS 3000 and seven with the CMS Accu-Probe® 450 system. Both commercially available models of probes were designed for treatment of liver, prostate, and uterine tumors and can only accommodate a maximum of five probes simultaneously. When freezing large bulky tumors five probes are sometimes inadequate. In such cases, the tumor must be frozen in sections, which may not be ideal. The CRYOcare® system we presently use can accommodate up to eight probes simultaneously and thus a greater volume of tumor can be frozen at one time. In this study all cases without favorable histologic responses to cryoablation can be explained by inadequate freezing. The resected tissue specimens showed pockets of viable tumor or interspersed viable cells throughout the tumor margins. The cells at the center of the tumor that were adequately frozen became necrotic as expected, while at the periphery some viable cells remained due to the tumors not being completely frozen at their margins. Of the 14 patients with less than 50% necrosis, eight of these cases were performed early in our experience with this technique. The lack of complete necrosis can be attributed to several factors, including poor intraoperative imaging and the initial lack of use of thermocouples to monitor temperature. Early in our experience the propagating ice ball often could not be well visualized. We have not encountered this problem with the current use of more technologically advanced ultrasound transducers and have been able to more easily assess propagation of the ice ball to encompass the entirety of the tumor. Another problem we initially encountered was the inability to accurately assess the temperature of the tumor in all dimensions during propagation of the ice ball. We have since used thermocouples to monitor the temperature at the periphery of the ice ball and have been able to assure an adequate margin of devitalized tissue at the periphery of the tumor. We recommend the placement of multiple thermocouples at the expected margin of later resection to assure a temperature of at least −40°C for more than 1 minute.

The effects of freezing are nonselective; normal and malignant tissues are destroyed equally. An exception to this is the muscular wall of large arteries, which are relatively impervious to the effects of freezing.^4,5,15,16 Large vessels will show no evidence of thrombosis or wall damage after thawing.^4,16 For this reason, a tourniquet should not be used when performing a cryosurgical ablation of an extremity sarcoma. No patients in this study developed clinical evidence of thrombosis or vascular damage. The close proximity of major vessels to a sarcoma should therefore not be a contraindication to performing cryoablation. Care should still be taken to protect and retract vascular structures as far as possible from the tumor bed to prevent unnecessary freezing.

Peripheral nerves are another exception to the nonselective effects of freezing. Peripheral nerve palsies were documented after cryosurgical treatment of bone tumors.^28,29 Peripheral nerves that are frozen develop a neuropraxia and cease to function initially, but this resolves anywhere between 3 and 12 months after the initial insult.^28,29,32,33 Peripheral nerve palsies associated with cryosurgery appear transient, and there have been no reported cases of permanent nerve dysfunction.^28,29,32,33 Therapeutic cryosurgical ablation of intercostal nerve freezing has often been used to prevent postthoracotomy pain, and this surgically induced neuropraxia was also transient.³⁴ In the present study, all cases of transient neuropraxia completely resolved between 1 week and 6 months postoperatively.

We have previously demonstrated cryoablation is a safe and technically feasible treatment for soft tissue sarcomas of the extremity.³¹ Cryosurgical ablation should not be used in lieu of conventional treatment techniques, but may be an effective surgical adjuvant for certain patients. We currently limit our use of cryoablation in the treatment of soft tissue sarcomas to those patients who cannot or chose not to undergo other adjuvant treatments. Further randomized, controlled investigations of this technique are necessary to determine the effect on long-term patient survivorship. Our data highlight the importance of achieving complete freezing of the tumor and the technical challenges associated with the cryoablation process.

Acknowledgment

The authors thank Dr. Adrian Correa for his assistance with this study.