Primary bone tumor arises from the tissues or cells present within bone.Because many types of cells are present within the medullary space of bone and adjacent to the bone surface,a number of histologic tumor types are possible,including tumors derived from osteoblasts,cartilage cells,fat,fibrous tissue,vascular elements,and hematopoietic and neural tissues. These tumor types include benign and malignant, the latter all referred to as sarcoma.

The most common of primary malignant bone sarcoma is osteosarcoma, which makes up approximately 35.1%of all bone tumors; the second is chondrosarcoma, then is fibrosarcoma.The common benign bone lesions are giant cell tumor,enchondroma,and so on.

Metastatic lesions in bone are very common.In man,the most common metastatic tumors found in bone are from the lung and prostate,whereas in women the most common is from breast.

Overview of treatment of bone tumors
Treatment for primary bone cancers can include surgery, chemotherapy, or radiation therapy, alone or in combination. The type of treatment used depends on the type, location, and size of the tumor, as well as the patient's age and general health.

Surgery is often the primary treatment for bone cancer. When operating to remove bone tumors, surgeons remove some of the surrounding bone and muscle to be sure that they are removing as much cancerous tissue as possible. If the operation is on an arm or leg, the limb should be tried to preserve and its functionality to maintain.
Chemotherapy is often used prior to or following surgery, to either shrink the tumor before surgery or manage and control the tumor after surgery. Such therapy -known as adjuvant therapy -increases the chances for a complete cure by destroying microscopic accumulations of cancer cells before they have an opportunity to grow larger.
Radiation therapy is sometimes given together with surgery, to destroy tumors or to reduce the size of the tumor. Radiation therapy may also be used to kill remaining cancer cells after surgery, or treat tumors that cannot be surgically removed -- sometimes in combination with chemotherapy. In patients with cancer that has spread to the bones, radiation therapy may also be used to relieve symptoms, including pain. External-Beam Radiation Therapy and brachytherapy with radioactive seeds are commonly used types of radiation in bone tumor.

Treatment of benign bone lesions depend upon tumor types. Generally, the following modalities may be selected: (1) curettage, (2) curettage and cytotoxic agents such as phenol, zinc chloride, alcohol, and H2O2, (3) curettage and a physical adjuvant (polymethylmethacrylate and cryosurgery, (4) primary resection, (5) radiation therapy, and (6) embolization, which is practiced in unresectable tumors.
History of cryosurgery for bone tumors
In 1966, Gage et al20 published their initial findings on the biologic effect of cryotherapy on bone. These authors produced bone necrosis in laboratory animals by circulating liquid nitrogen around the femurs and observed subsequent bone regeneration from the periosteum and endosteum. In 1968 Ralph C. Marcove introduced cryosurgery into orthopedic oncology for the treatment of primary and metastatic bone tumors by repetitive freezing. He was awarded the first prize in scientific research at the 162nd annual convention of the Medical Society of the State of New York (1). Marcove and Miller38 used cryotherapy in the treatment of metastatic carcinoma of the proximal humerus in 1969. They used cryosurgery for treatment of various benign and metastatic bone tumors, as well.36,37,39,40,42 Marcove et al41,43 described the use of cryosurgery in the treatment of primary bone sarcomas. During the 1970s, Marcove et al42 pioneered the development of cryotherapy in the treatment of giant cell tumor of bone and described the effectiveness of a direct pour method in freezing the walls of a curetted cavity.
Since then more orthopedic surgeons dealing with skeletal tumors have adopted the technique, and the clinical results and experimental data of cryosurgery with specific reference to the skeletal system have been published regularly.

Indication of cryosurgery
In general the following orthopedic bone tumors are suitable for cryosurgery:

• simple bone cyst
• aneurysmal bone cyst
• giant cell tumor
• eosinophilic granuloma
• enchondroma and chondrosarcoma grade 1
• fibrous dysplasia
• miscellaneous

In specific circumstances the technique can be of value for the treatment of malignant lesions of bone. These circumstances are:

• Marginal resection of the tumor is, due to its location, not possible or induces unacceptable morbidity, such as in vertebral (chordoma) and pelvic lesions.
• Marginal or wide resection is not indicated but beneficial for local control of the tumor, as with bone metastases.

Cryosurgery procedure for Bone Tumors

Cryoprobe design

The basic design of cryoprobes suitable for orthopedic oncological purposes can be divided into open and closed systems (8). Since today liquid nitrogen is most commonly used, the focus is specifically on this particular cryogen.

Open systems: liquid nitrogen is sprayed directly on tissue and has been shown to be an effective, if not the most effective, way of cooling. The liquid nitrogen cools the tissue by boiling, which occurs as heat is extracted from the tissue surface. When the liquid is forced out of the nozzle, the sudden drop of pressure causes partial vaporization, a phenomenon called “flashing”. Liquid nitrogen drops in the spray will start boiling immediately when they come into contact with a higher surface temperature. The initial vaporized gas can form a vapor layer between the liquid nitrogen and the target tissue. This vapor layer or “film” acts as an insulator and prevents further cooling of the surface.

Closed systems:It employ two different principles for creating low temperatures at the end of a probe. The first principle applied in cooling a cryoprobe is to allow liquid nitrogen to boil at the end of the tip of the probe; by boiling it extracts latent heat from its surroundings, cooling it at the same time. As long as liquid nitrogen is passed through the tip fast enough to maintain boiling, the temperature at the tip of the probe will remain at the boiling point of liquid nitrogen: -196oC. The other principle used to cool the end of a cryprobe in closed systems utilizes the Joule-Thomson effect. If a pressurized gas is allowed to expand to a much lower pressure, its temperature will drop. The magnitude of change in temperature depends on the change in pressure and the physical characteristics of the substance used. The most commonly used gas in this kind of probe is nitrous oxide.

The operational control of cryoprobes in general is carried out by simply stopping the flow of the cryogen. The decision is mainly based on experience. In addition, monitoring devices are available.

Basic technique

The basic technique of cryosurgery requires an effective use or exploitation of biological mechanisms leading to cell death under the influence of low temperature. It has been shown that the most effective way to achieve this is rapid freezing and slow thawing, done in repetitive cycles. To establish these suitable cryobiological circumstances an efficient technique is needed in situ to extract heat from the tissue.

The requirements for rapid freezing are:

• A cryogen suitable for the lesion, not only providing potential cooling power, but also in sufficient quantities.
• Furthermore it must be safe (nontoxic and non-flammable), easy to store and to transport, and preferably inexpensive. For orthopedic purposes only liquid nitrogen meets all these criteria, used either in closed probe systems or as a spray.
• The contact area between the lesion and the cryogenic device should be as large as possible, facilitating rapid transport of heat.
• The surface of a lesion in orthopedic oncology is always irregular and frequently large. These surface properties nearly always necessitate the use of a liquid nitrogen spray instead of a probe (Fig. 14.1). In orthopedic oncology closed cryoprobes are used in very small bones like those of the finger (Fig. 14.2) and in “tricky” locations like the cervical spine (Fig. 14.3)

The principle for slow thawing requires only patience on the part of the surgeon. Ideally, frozen tissue should be allowed to completely thaw without assistance. While rapid freezing can be achieved in less than 30 seconds, spontaneous thawing in the orthopedic setting can take up to ten minutes or more. The thawing rate is most influenced by nearby heat sources such as blood vessels.
It is necessary to repeat the freeze and thaw cycles several times, because living tissue is capable of resisting thermal injury and it is technically difficult to achieve optimal conditions for cell death in all areas of many lesions. To compensate, repetition of freeze/thaw cycles is a practical solution and provides safety especially at the periphery of the lesion. After the first cycle, thermal conductivity in the tissue is increased, and the specific heat capacity and vascularity are decreased. This precondition make the next cycles more effective by virtue of faster cooling and slower thawing rates(9-12).

Surgical Technique

Standard orthopedic oncologic exposures are used. When the exact location of a lesion or the proximity of a growth plate is unclear, an image intensifier is used. To avoid inadvertent freezing of the skin, wide retraction is mandatory. The use of extremity tourniquets may be ill advised, because normal circulation decreases the risk of freezing nearby neurovascular bundles and skin. Normally it is not necessary to dissect the neurovascular bundle away from the lesion to be frozen.

A sufficient oval window is made in the cortex using a drill or saw. The tumor is resected as thoroughly as possible using a curette (Fig. 14.4B). Care is taken to avoid pushing tumor into the uninvolved medullary canal.

To monitor the cryosurgical process and local extent of the freeze, thermocouples are positioned in and around the lesion.

Cryosurgery is then performed (Fig.14.4C). Three cycles of cryosurgery are performed using a machine producing a liquid nitrogen spray. This spray is directed into the lesion in every direction, until the whole cavity is wetted and becomes frosted. The duration of each freeze is based on the temperature readings and visual observation. Intralesional temperatures of at least -50oC are aimed for and are necessary to induce tissue necrosis (13, 14). Warming up to 20oC takes place by spontaneous thawing.

The defect remaining after curettage and cryosurgery is filled with autograft or allograft bone chips. The cortical window is frozen as well and is replaced or used as a source of bone graft (Fig. 14.4D).
A careful soft tissue reconstruction is very important, restoring the periosteum, if possible, and covering the bone with muscle. A wound drain is always used and it should only be removed when drainage falls below 30ml per 24 hours. Perioperative antibiotics are routinely used. If the strength of weightbearing bones, in particular the diaphysis of the femur, is compromised by the lesion and cryosurgery, prophylactic internal fixation is advised, using a plate and screws. An intramedullary enforcement is ill advised, because it has risk of contaminating the entire intramedullary compartment with tumor cells. Partial weightbearing is usually necessary until three months after the operation.

Monitoring

A cryosurgical procedure with a satisfactory result can be achieved based only the physician’s clinical judgement and experience. However, there are several reasons to use precise monitoring devices.

• To ensure that lethal temperatures are reached. The temperature of frozen tissue cannot be determined by its appearance, as frosted tissue looks the same at any freezing temperature.
• An accurate measurement of the depth of the freezing is important not only for obtaining adequate margins, especially if the lesion has a malignant nature, but also to avoid an unwarranted extension of the freezing and potential morbidity.
• Local temperature measurement enables a more precise determination of the ending of the thawing phase.
• Not only local monitoring is important, but systemic monitoring is also recommended, since cryosurgery has been associated with potentially lethal circulatory and pulmonary complications.
• Monitoring and preferably recording of the cryosurgical procedure enables the physician to evaluate the procedure itself, for instance the freezing rate, and better interpret follow-up results with respect to the procedure.

Local Monitoring: To supplement clinical judgement during cryosurgery, a range of monitoring devices and techniques has been developed.Thermocouples are by far the only suitable device for intraoperative (real-time) temperature monitoring during cryosurgery for bone tumors. It consist of a copper/copper-nickel alloy and are mounted in the tip of a 50-mm-long and 0.8-mm-diameter injection-like needle. Their accuracy is better than 0.1 oC with a response time of 0.3 sec. Temperature data acquisition is done using a digital multi-meter equipped with a thermocouple scanner card. Graphic real-time visualization of the course of the temperatures is accomplished by connecting the multi-meter to a personal computer running appropriate software. Next to performing real-time display, this program stores all temperature data for later analysis (25). The locations of the thermocouples should be chosen from a strategic point of view as illustrated in Fig. 14.3.

Correct placement of the thermocouples and verification of their positions with radiographs is sometimes necessary, especially when structures have to be protected from a freeze injury, for instance joints and growthplates (Figs. 14.6 and 14.7).

Using the temperature measuring system with real-time graphic visualization, it has been shown that in a cryosurgical system using a liquid nitrogen spray, intralesional temperatures of -150oC are achieved within seconds (freezing-rate>10oC/s-1). Furthermore, the maintenance of a temperature of -50oC for 40 sec, and the tissue in the frozen state for 3min, followed by spontaneous slow thawing and repetition of freeze/thaw cycles, are all of importance to maximize tissue destruction (26).

After spraying has been stopped and thawing is allowed, the extralesional thermocouples show a further lowering of temperature, representing the retraction of tissue heat, which is used for the thawing of the more central parts of the frozen lesion. The non-linear increase of the temperature in frozen tissue and its subzero plateau phase is explained by additional energy needed for the transition of ice into water, which halts the temperature increase temporarily. In Figs. 14.8 and 14.9 all these phenomena are shown next to almost identical patterns of the freeze/thaw cycles, indicating that cryosurgery utilizing a spray is a reproducible method.

Systemic monitoring: Whenever a gas is introduced into a body cavity there is always the hazard of intravascular introduction of gas bubbles, especially when pressure is allowed to develop. Gas emboli in the vascular circulation can cause serious hemodynamic complications (27-30). Therefore, in addition to routine systemic monitoring of the patient, end-tidal gas analysis is indicated using a mass spectrometer measuring inspired and end-tidal O2, CO2, N2O, N2 tensions and anesthetic vapor concentration. Using real-time recording of the gas analysis breath by breath makes detection of any exhaled N2 possible, which is associated with venous nitrogen gas embolism. In this way one may be able to take appropriate action in time to prevent serious hemodynamic complications.

Monitoring the freeze/thaw cycles during a cryosurgical procedure with temperature recordings in and outside the lesion in orthopedic oncology is of importance and very helpful in facilitating an effective, reproducible cryosurgical procedure and in controlling the extent of the freeze avoiding local complications. Systemic monitoring is of paramount significance for the safety of the patient.

Clinical results

Cryosurgery utilizing liquid nitrogen is practiced in orthopedic oncology for the treatment of primary benign and malignant bone tumors as well as for secondary metastases to bone. In benign and low-grade malignant stage IA skeletal tumors, it is used as an adjuvant treatment to intralesional resection (curettage) to extend the surgical margin of resection. By this method the procedure can be considered to be marginal according to oncologic principles (2). The advantage of this kind of treatment, as compared to local resection, is that as much as possible of the supportive function of bone is preserved and that reconstructive surgery can be limited.

In high grade sarcomas, cryosurgery has been used as the primary treatment with variable results (3). Cryosurgery used in the treatment of bony metastases has to be considered as palliative and in this respect helpful in local control of the malignancy (4). Theoretically, adjuvant therapy may consist of systemic chemotherapy, radiotherapy and physical adjuvants like phenol, hypertonic saline merthiolate, polymethylmethacrylate (PMMA) cement applied locally, and cryosurgery.

It seems that cryosurgery is specially amenable to treatment for the following bone tumors:

Simple bone cyst
Simple bone cyst is a tumor of bone of unknown origin. It tends to occur in the metaphyses of long bones, particularly of the humerus and femur. Although histologically completely benign, it

frequently weakens the integrity of bone resulting in pathological fracture, which is often the presenting feature. The treatment options for simple bone cysts include observation, injection and surgical curettage. Until Scaglietti et al. introduced the technique of steroid injections (31,32), the most prevalent treatment method for simple bone cyst has been curettage followed by bone grafting, with recurrent rates that vary from 12 to 48% (33-42). Adjuvant therapy after curettage is indicated to destroy residual tumor cells and lower tumor recurrence.

A retrospective study in children with unicameral bone cysts treated with curettage, cryosurgery and bone grafting showed that 5 of 42 (12%) treated patients suffered a local recurrence with a mean clinical follow-up of 24.5 months (43). In another study of 13 patients 2 local recurrences were reported (44).

Aneurysmal bone cyst
Aneurysmal bone cyst is a rare benign tumor-like lesion of bone of unknown origin. Lack of understanding about its origin and growth makes treatment empirical. The most common treatment has been curettage with bone grafting which has a substantial rate of recurrence (33,46-49). Lower recurrence rates can be achieved by marginal or wide resection but are accompanied by loss of bone and the need for reconstruction (33,50,51).

A review of the literature on the results of treatment modalities, including cryosurgery, is listed in Table 14.1.
The recurrence rates for irradiation with or without curettage are similar, 11.8% and 13.8%, respectively. Curettage with or without bone grafting is accompanied by a high recurrence rate of 30.8%. Cryosurgery as an adjuvant after curettage has a recurrence rate of 12.8%. In two series describing the use of cryosurgery as adjuvant treatment a local recurrence rate of 4% is reported in one. The other series reported a recurrence rate of 18%, which, after additional treatments with the same technique, decreased to 4% (48). It may be concluded that cryosurgery for aneurismal bone cysts has comparable results to marginal resection in terms of control of the tumor (52). However, after marginal resection extensive reconstructive surgery is needed, with associated morbidity.

Giant cell tumor
Giant cell tumor is considered to be a benign lesion, however during its clinical course it shows very aggressive features with the potential for destruction of bone and joints and soft tissue intrusion. Furthermore, 3% of giant cell tumors are primarily malignant or will undergo malignant transformation and will metastasize either after radiation therapy or after several local recurrences. The anatomic locations in which giant cell tumors commonly occur are the femoral condyles, tibial plateau, proximal humerus and distal radius.

Expendable bones with giant cell tumor (proximal fibula, ribs) should be resected. However,recurrence after curettage is the largest problem. Hutter et al24 reported that recurrence rates in giant cell tumors treated by curettage alone were higher than those in tumors treated by resection or curettage in combination with physical adjuvants. In a large series of 648 patients with giant cell tumor treated by curettage,an average local recurrence rate was high as 40.8% (265 patients). Therefore, contained lesions should be treated with curettage followed by local adjuvant treatment (cryosurgery, cytotoxic agents etc.) (Fig.14.12).

The treatment results published in the literature on curettage followed by cryosurgery of giant cell tumor are summarized in Table 14.2.
Martin et al reported that between 1983 and 1993, 102 patients with giant cell tumor of bone were treated at three institutions. Sixteen patients (15.9%) presented with already having had local recurrence. All patients were treated with thorough curettage of the tumor, burr drilling of the tumor inner walls, and cryotherapy by direct pour technique using liquid nitrogen. The average followup was 6.5 years (range, 4-15 years). The rate of local recurrence in the 86 patients treated primarily with cryosurgery was 2.3% (two patients), and the overall recurrence rate was 7.9% (eight patients). Six of these patients were cured by cryosurgery and two underwent resection. Overall, 100 of 102 patients were cured with cryosurgery. Overall function was good to excellent in 94 patients (92.2%), moderate in seven patients (6.9%), and poor in one patient (0.9%).

In the case of a joint destroyed by a giant cell tumor, commonly with an intra-articular pathological fracture, marginal resection and reconstruction are advised. However, preservation of the joint is possible in selected cases using curettage, cryosurgery, bone graft, cement and osteosyntheses (Fig. 14.13).

Eosinophilic granuloma of bone

Eosinophilic granuloma (EG) of bone is part of a spectrum of diseases known as Langerhans cell granulomatosis, all characterized by proliferation of specific histiocytes, designated as Langerhans type, in normal tissue. These histocyte proliferations, called eosinophilic granuloma, can occur unifocally or multifocally and be part of systemic illness known as Hand-Schuller-Christian disease and Letterer-Siwe disease (66, 67).

The etiology is unknown and therefore its treatment is empirical. Eosinophilic granuloma of bone was most commonly treated with chemotherapy, surgery and irradiation (68, 69). Currently intralesional instillation of steroids is the first choice of treatment, with good results (67, 68, 70-72). However, some lesions fail to respond to steroid injection therapy or are unsuitable for injection therapy due to their size, location, loss of bony containment and/or soft tissue intrusion. Especially in case of cortical destruction with or without an impending pathological fracture and possible neurological damage in spinal cases (Fig. 14.3), a primary surgical treatment seems feasible. Schreudet et al. described six patients with eosinophilic granuloma of bone with these lesion characteristics and treated them with curettage, cryosurgery and bone grafting. No recurrences were reported (73).

Enchondroma and chondrosarcoma grade 1

Much controversy exists about the methods of evaluating, staging and final treatment of enchondroma, transient chondroid tumor and low-grade chondrosarcoma, because their accurate differentiation is hampered by their radiographic and histological similarity and clinical behavior. The continuous spectrum of biological behavior of these entities is unpredictable. Histologic features of enchondroma, chondroid tumor and chondrosarcoma may coexist in the same lesion. It is established that a benign tumor such as enchondroma can undergo malignant transformation into secondary chondrosarcoma (75, 78, 79), especially in patients with multiple enchondromas as in Ollier disease and Maffuci syndromc (80).

Because of low risk recurrence of enchondroma, transient chondroid tumor and low-grade chondrosarcoma in extremities, limited surgery with or without adjuvant therapy is advocated (81-83).

In view of the above, one of the advantages of cryosurgery is critical: since the dosage of this local adjuvant treatment can be well controlled, the extent of the cryosurgical margin can be adjusted (Fig. 14.8, 14.16 and 14.17).

In a series of 21 patients with enchondroma and 15 with either transient chondroid tumor or chondrosarcoma grade 1, all located in the extremities, and with a follow-up of more than two years, there was one local recurrence of an enchondroma located in a finger (84).

Fibrous dysplasia

Fibrous dysplasia is a benign tumor-like lesion of immature fibrous connective tissue and poorly formed immature trabecular bone (85). Fibrous dysplasia can compromise the structural integrity of affected bones leading to recurrent fractures and skeletal deformities. The clinical picture of fibrous dysplasia is diverse. Its manifestation can be monostotic, polyostotic, or polyostotic in combination with skin pigmentation and dysfunction of the endocrine system (McCune-Albright syndrome).

Circumscribed lesions may remain asymptomatic for many years during adult age and need no further treatment. In most cases fibrous dysplasia presents with pain or a pathological fracture, usually noted between 5 and 20 years of age.

When operative treatment is indicated in circumscribed monostotic fibrous dysplasia, a single procedure of curettage, cryosurgery and bone grafting has shown satisfactory oncological results and functional outcome (86).

Extended lesions may result in bony deficiency or deformities necessitating additional internal fixation, massive allografts or corrective osteotomics. The beneficial effects of additional cryosurgery in extended lesions and particularly in McCune-Albright syndrome is unclear (86).

Miscellaneous
Chondroblastoma, chondromyxoid fibroma, intramedullary hemangioma and schwannoma are all very rare benign bone tumors suitable for curettage and local adjuvant therapy with cryosurgery. No data presenting series of patients with the results of cryosurgery with respect to these bone tumors are available in the literature.

Complications

As any new surgical therapies, cryosurgery for bone tumors will be accompanied by complications. In general, complications will diminish with improvement of the technique itself.

Wound infections
Intralesional resection (curettage) of an intramedullary tumor will leave behind a cavity with a lot of dead space. Cryosurgery results in an additional amount of tissue necrosis. Furthermore, most surgeons fill this defect with a “dead” homologous bone graft and sometimes an osteosynthesis is added as well. All these factors are strong mediators for developing a bacterial infection.Delayed wound healing, persistent wound drainage with or without positive cultures, superficial and deep infection probably occurs in about 4% of the skeletal tumors treated with cryosurgery (43, 44, 48, 60, 62, 64, 87).
For prevention of the infection after a cryosurgical procedure the following elements are of importance:

• The use of perioperative broad-spectrum anti-biotics until 24 hours after the operation, comparable to regimens in use for prosthetic replacements.
• Adequate drainage of wound fluids. Clinical observation has taught that there seems to be some kind of reactive increased blood flow in the area of the cryosurgery. When this blood is allowed to form a hemathoma it may become infected.
• Adequate wound exposure with retraction of skin, and protection with gauze is necessary to avoid accidental freezing of the skin.
• Would closure with sufficient soft tissue coverage.

Venous gas embolism
During cryosurgery liquid nitrogen is either sprayed or poured into the bony cavity and since its boiling point is -196°C, nitrogen gas bubbles are rapidly produced at room temperature. In general, whenever a gas is introduced into a body cavity there is the hazard of intravascular introduction of gas bubbles especially when pressure is allowed to develop. Gas emboli in the vascular circulation can cause serious hemodynamic complications (27, 28).

Dwyer et al. reported a pressured incident of venous gas embolism during a cryosurgical procedure in which a dramatically increased end-tidal nitrogen tension was noted, but without any hemodynamic complications (88).

In the literature, one fatal case due to venous gas embolism during cryosurgery has been described. The occurrence was explained by the blocking of the exit of gaseous nitrogen from the bone by intentional digital occlusion of the opening in the bone cortex (89).

De Vries conducted experiments in rats and rabbits to evaluate the problem of bone marrow embolism during cryosurgery. It was concluded that the intravasation of bone marrow was caused by increased intramedullary pressure, and embolization of bone marrow was encountered but not on a large scale. Most of the bone marrow intravasations remained local in the extraosseous veins (90).

Schreuder et al. reported on two patients who showed signs of impairment of pulmonary circulation during cryosurgical procedures, as indicated by a sudden significant drop in end-tidal CO2 and corresponding changes in blood pressure and heart rate. They suggested that these features represent venous gas embolism because of their rapid development at the same time as the instillation of the liquid nitrogen and the fact that the symptoms disappeared rather quickly after the cryosurgery was ended (Fig. 14.19) (30). Solid particle embolism by marrow or fat is less likely, because this of embolism is provoked by mechanical elevation of the intramedullary pressure as in intramedullary nailing and introduction of a prosthesis (91, 92).

In a clinical experiment Schreuder et al. used a mass spectrometer to perform end-tidal gas analysis in all patients who were treated by cryosurgery. The mass spectrometer measured inhaled and end-tidal O2, CO2, N2O, N2 tensions and anesthetic vapor concentration breath by breath. In 15 cases analyzed, they did not detect any exhaled N2 during cryosurgery. Also, the measured O2, CO2, N2O tensions and anesthetic vapour concentration were completely normal (30).

The mechanism of N2 embolism is unclear. When during cryosurgery the surface of the cavity is becoming extremely cold, the additional sprayed liquid nitrogen will not be able to vaporize. Instead, the bone marrow develops properties comparable to a sponge and sucks the liquid nitrogen into small marrow spaces. This liquid nitrogen may get trapped in these marrow spaces. When thawing or a rise in the temperature is allowed, the trapped liquid nitrogen will boil and vaporize, building extremely high pressures. It may be possible that under these circumstances liquid nitrogen or gaseous nitrogen is pushed into the venous circulation. That N2 dissolves in blood first is highly unlikely because of its very low Oswald solubility coefficient (C = 0.015 at 37oC). The risk is increased when the site of the tumor is located in a richly perfused area such as the metaphysic of the long bones where the major nutrient arterial and venous blood supply enters the medullar cavity. Unfortunately, the metaphysic is the location of preference for many bony tumors suitable for cryosurgical treatment.

When using cryosurgery, one should never block the entrance to the bony cavity.

Fractures
Postoperative fracture is the most common and serious complication associated with cryosurgery.25,42 In published series of 15 or more causes of bone tumors treated with cryosurgery, the postoperative fracture rate is 10% (43, 44, 48, 60-62, 64, 87, 93).
Fracture is an inherent risk after reconstruction of any large bone defect, and especially after cryosurgery near a weightbearing joint. After cryosurgery, bone necrosis and disruption of osteoid extend the period through which reossification occurs and delay bone healing.34 Vigorous freezing increases the likelihood of cure at the cost of higher rate of pathologic fractures, whereas inadequate freezing of bone surrounding the tumor may predispose to local recurrence. Marcove et al42 made only a minimal attempt to reconstruct these defects and reported a 25% fracture rate that is similar to the fracture rate of the current series when internal fixation was not used. The fractures they reported occurred before the use of polymethylmethacrylate combined with internal fixation.

The effect of cryosurgery on the strength of bone was tested by Fisher et al. The mandibles of rats had a reduction in strength of approximately 30% eight weeks after cryosurgery. The gradual loss of strength in these bones paralleled observed radiographic osteolysis. At four months the mandibles had regained strength accompani9ed by clear radiographic evidence of sclerosis (94). Although not investigated in this experiment, the gradual loss and return of strength in cryosurgically treated bone also parallels histologic evidence of bone resorption, repair and remodeling (95, 96).

Postoperative fracture of the remodeling bone subjected to cryosurgery is not only very distressing for the patient, but also compromise the orthopedic oncologic status; first there was intra-compartmental disease, because of the fracture it has now potentially changed to an extra-compartmental disease.

It seems that fractures are most likely to occur four to eight weeks after the cryosurgical treatment, but they can occur even 8 months after cryosurgery (Figs. 14.20-14.22). Diaphyseal lesions are most prone to fracture (Fig. 20). Therefore, prophylactic internal fixation is advised, but is mandatory when the femur is involved (Fig. 14.17). An intramedullary enforcement is ill advised, because it carries the risk of contaminating the entire intramedullary compartment with tumor cells. Titanium alloys are preferred because these implants create little interference on MRI, making tumor follow-up less difficult. Partial weightbearing is usually necessary until three months after the operation.

Experience and improvements in technique have reduced the fracture rate to an acceptable level. Since prophylactic osteosynthesis is used, sometimes in combination with cement and or auto/allo bone graft, fractures are no longer seen (60).

Damage to epiphysis
Benign bone tumors, especially simple bone cysts and aneurismal bone cysts tend to occur in patients of immature skeletal age. Furthermore these tumors commonly develop in the metaphysic, often adjacent or very close to the epiphysis. Damage to the epiphysis either by the tumor itself or by the use of cryosurgery is very well possible and may result in arrest or disturbance of normal growth.

Malawer and Dunham reviewed 25 pediatric patients with aggressive benign tumors all treated by cryosurgery. They saw 2 patients with damage to the epiphysis necessitating epiphyseodesis of the contralateral side (64). During surgery no attempt was made to prevent the epiphysis from freezing, control of tumor was their first priority.

Schreuder and Conrad reported on 42 treated simple bone cysts of which 11 were located in the proximal metaphysic adjacent or close to the epiphysis. During surgery care was taken not to damage the adjacent physis by curettage and if the physis was exposed to the cyst, it was separated from freezing by several layers of surgical gel foam. No growth disturbances were seen, but two local recurrences were encountered (43).

Figure 14.23 demonstrates the long term follow-up of the patient in Fig 14.6. Damage to an epiphysis is noted, even as the result of not very low temperatures.

Whether an epiphysis is damaged by the bone tumor or the treatment will not always become clear and in many cases may be the result of both. To minimize the risk of damage, protection of an exposed epiphysis by gel foam seems feasible, but on the other hand it lowers the effectiveness of the cryosurgery, which may result in a local recurrence of the bone tumor, which potentially will definitively damage the epiphysis.

Degenerative osteoarthritis
Some bone tumors occur almost always extremely close to the major joints, like giant cell tumor and chondroblastoma. Damage to the articular surface either by the tumor itself (intra-articular fracture) or treatment (cryosurgery) may be anticipated. Malawar et al. demonstrated in an experiment using dogs, that cryosurgery can produce bone necrosis 7 to 12 millimetres away from the surface of the cavity being treated, in contrast to the minimal zone of necrosis produced by the heat of polymerization of polymethyl-methacrylate. They found that cryosurgery had no effect on articular cartilage (97)

Aboulafia et al. described a technique for treatment of large subchondral tumors around the knee which extended to within two millimeters of the articular surface. Curettage, cryosurgery and composite reconstruction with bone graft, bone cement and osteosynthesis were used as an alternative to primary joint-sacrificing resection. Of nine tumors (six giant cell tumors, one chondroblastoma, one chondrosarcoma and one fibrosatcoma) there was one recurrence, retreated in the same fashion. All 9 patients had an excellent functional outcome. Only two patients had mild degenerative cartilage changes (65).

Among 120 cases all treated with cryosurgery and a follow-up of more than one year, Schreuder et al. saw 3 cases of secondary osteoarthritis, all giant cell tumors of which two were situated in the distal radius (Fig. 14.21) and one in the proximal humerus (Fig. 14.12.3) (44). Malawer et al. reported two of 48 patients with giant cell tumor around the knee joint developing radiographic and clinical evidence of degenerative changes (60).

It seems that articular cartilage can resist low temperatures to some degree. In practice local control of the tumor (especially in case of intra-articular pathological fracture) has priority, dealing with osteoarthritis seems in those circumstances to be of concern later (98).

Damage to nerves
Nerve palsy is a complication of cryosurgery which was recognized very early in the introduction of cryosurgery for bone tumors. Marcove saw in 128 patients, all treated for various types of bone tumors, nine. (mostly transient) nerve palsies (48, 62, 87, 99). Schreuder et al. saw in 165 cryosurgical procedures seven nerve palsies; one peroneal nerve failed to regain its function. Not the cryosurgery but rather surgical traction was very likely the cause of this persistent palsy. Three patients with sacral lesions (two giant cell tumors and one chordoma) suffered permanent, only partially recovered loss of function of sacral roots. Two palsies of the radial nerve and one of the peroneal nerve were shown to be completely reversible.

The radial nerve seems especially at risk due to its proximity to the humeral bone (44).

If nerves are frozen their function is only temporarily impaired. Most neuropraxias resulting from freezing will resolve in 6 weeks to 6 months. Very likely, regenerating nerve fibers can grow down the nerve sheaths since they are left intact. Furthermore, the vital nerve cell nucleus is located at some remove in the dorsal root ganglion.

Discussion

Cryosurgery is sometimes used in addition to surgery in some patients with bone cancer. After a bone tumor is removed, liquid nitrogen is used to freeze the tumor cavity to subzero temperatures, killing microscopic tumor cells and decreasing the chance of tumor recurrence.
Cryosurgery for treatment of bone tumors has the advantages of preserving limb and joint function, excellent functional outcome, and low recurrence rate when compared with other joint preservation procedures. For these reasons, it is recommended as an adjuvant to curettage for bone tumors such as most giant cell tumors of bone.
As compared to other adjuvant modalities cryosurgery has some advantages. It is an active local adjuvant with no systemic side-effects, non-toxic to the patient, extremely powerful, and the dosage can be controlled, so that the procedure can be customized to the specific type of bone tumor and location.
Conclusion
Modern cryosurgery instruments, which are perfectly equipped for successful treatment, has been used for treatment of a lot of bone tumors with higher successful rate. In this, cryosurgery is comparable to every other advancement in osteo-oncology. The goal of continuing research is to improve current techniques,that is necessary, in particular to respond to demands for safety and expanded indications.