The Era of Intracavitary Radium Therapy

Prostate cancer was rarely diagnosed a century ago, but prostatitis, benign hyperplasia, and even tuberculosis of the prostate were treated by X-irradiation (15,16). Successful treatment of prostate cancer by X-rays was first reported in France in 1904 (17). Treatment of prostatic disease with radium was first reported in Paris, at a meeting of the Assoçiation Francaise d’Urologie in October 1909 (18). Ernst-Louis Desnos treated hypertrophy with a series of urethral and rectal applications (19). Henri Minet treated cancers of the prostate, bladder, and ureter with a silver tube containing 10 mg of radium, applied through a urethral catheter or a suprapubic cystotomy (20). Urologist Octave Pasteau and radium therapist Paul-Marie Degrais also began treating prostate cancer with intracavitary radium in 1909, but their first reports did not appear for several years (21). Pasteau’s rationale for preferring brachytherapy to prostatectomy was that “in cancer of the prostate the curative treatment by operation is in truth illusory; it is dangerous, and gives the most temporary results,” whereas these tumors are “particularly susceptible to the influence of radium” (22). They had used a silver capsule, containing 10 to 50 mg of radium sulfate, placed near the tip of a 17Fr coudé urinary catheter (Figure 1.3). Five treatment sessions, each lasting 2 to 3 hours, were delivered over 2 weeks. The series could be repeated periodically (annual maintenance treatments were prescribed for patients who had enjoyed a complete response). Desnos, Minet, and Degrais (who coauthored the first comprehensive radium therapy text in 1909 [23]) understood the need to filter caustic beta particles and soft gamma rays with a radiodense capsule, and that bremsstrahlung radiation (arising from the capsule) should be filtered by less dense material (rubber tubing).

Figure 1.3

Pasteau and Degrais’s radium-bearing urethral catheter. The catheter was slowly advanced until urine began to drip out, at which point it was withdrawn until the dripping stopped. In this way, the radium was properly positioned in the prostatic urethra.

Source: From Ref. (21). Public domain, from Pasteau O, Degrais. De l’emploi du radium dans le traitement des cancers de la prostate (The use of radium in the treatment of prostate cancer). J Urologie Med Chirur. 1913;4:341–366.

Prostate brachytherapy was performed in Vienna in 1909 (24). Rudolf Paschkis and Wilhelm Tittinger reported the case of a 32-year-old man treated with radium at the Rothschild Hospital. The patient had been admitted for urinary retention, and digital examination suggested locally advanced, unresectable cancer of the prostate. A catheter could not be passed, so cystotomy was performed, exposing a nodular, ulcerating tumor infiltrating the bladder neck. The tumor was treated with a capsule containing 4.7 mg of radium bromide applied through the bladder fistula. Treatments lasted 20 minutes and were repeated at 2 week intervals. After 10 months of treatment, the tumor had vanished and the patient was voiding through his urethra. The case was the first to have pathologically confirmed malignancy prior to treatment, and complete clinical response following it (25).

Although Hugh Hampton Young introduced his radical prostatectomy procedure for cancer in 1904 (26), he rarely performed it, as it was uncommon for patients to be diagnosed with organ-confined disease (27). Young had attended the International Congress of Medicine in London in 1913, where he heard Pasteau and Degrais present their experience with radium therapy. He acquired 102 mg of radium and developed his own system of delivering treatment through the rectum, urethra, and bladder, as well as by applying external radium placques (essentially “crossfiring” the tumor). A single application site was treated in a daily “seance” (treatment session) lasting 1 to 2 hours. Treatment sites were alternated and carefully mapped (Figure 1.4), so that no mucosal segment was irradiated twice; in this way, urethritis, cystitis, and proctitis were avoided (28). A typical course of treatment delivered 3,000 to 4,000 mg h of radium therapy. Results were gratifying; Young reported “amazing resorption of extensive carcinomatous involvement of prostate and seminal vesicles” resulting in the “disappearance of pain and obstruction . . . which is indeed remarkable” (29). He treated 500 patients with radium therapy between 1915 and 1927 (30), and his textbook of urology devoted many more pages to radium therapy than to radical prostatectomy (28).

Figure 1.4

Record of the dates and locations of rectal applications. A similar record was kept of urethral and bladder neck applications.

Source: From Ref. (28). Young HH. Treatment of carcinoma of the prostate. In: Young HH, Davis DM, eds. Young’s Practice of Urology: Based on a Study of 12,500 Cases. Vol 1. 1st ed. Philadelphia, PA: WB Saunders; 1926:644–671. Used with permission.

The Era of Interstitial Radon

James Douglas, a Canadian-American mining engineer and executive, became interested in radium after losing a daughter to breast cancer. He was appalled that she had to travel to Europe to be treated with radium that had been mined in the United States. He joined with surgeon Howard Kelly (America’s leading gynecologist, one of Johns Hopkins Medical School’s “Big Four”) in lobbying Congress to nationalize American radium-bearing lands. When Congress declined to do so, Kelly and Douglas entered into a collaborative effort with the United States Bureau of Mines. They established the National Radium Institute in 1913, with Kelly and Douglas providing the capital and the Bureau supplying the mining and processing expertise. The institute leased 16 carnotite claims in Colorado’s Paradox Valley for 3 years. The ore was transported, by burro and rail, to their processing plant in Denver. Operations ceased in 1917, after 8.5 g of radium was refined. One-half gram was donated to government hospitals, and the remaining radium was divided between Kelly and Douglas. Douglas donated his 4 g to New York’s Memorial Hospital, with the stipulation that the hospital become dedicated to the treatment of cancer (31).^j

Radium’s specific activity (ratio of activity to mass) is low, due to its long half-life (1,600 years). In practical terms, it takes at least a week to deliver a curative dose with radium needles. This would be particularly awkward for the treatment of prostate cancer, as the sources would be left in an open suprapubic or perineal wound for an extended period (32). The solution to this problem lies in radon, radium’s first daughter product (Table 1.1). As most of the therapeutic gamma rays exiting a radium tube were actually emitted by daughter product “radium C” (²¹⁴Bi), radium (²²⁶Ra) itself was unnecessary for “radium therapy.” Treatment with radium C would be challenging, due to its 20 minute half-life, but radon (known as radium emanation until 1923) could serve as a reservoir for radium C. Radon has a very high specific activity, owing to its short (3.8 day) half-life; despite being a gas, 1 Ci of radon has a volume of less than 1 mm³. Because of its high specific activity, an “emanation” needle could be much thinner than a radium needle. Consequently, radium salts were kept in an aqueous solution, and the emitted radon gas was harvested for therapeutic applications. Unfortunately, the collected gas was mostly composed of water vapor, hydrogen and oxygen (from electrolysis of the water), helium (from alpha particles), and chlorine (from the hydrochloric acid used to keep the radium ions in solution). Harvard biophysicist William Duane had spent 7 years as a research associate of the Curies, much of that time focusing on the purification of radon. On his return to the United States, he built a radium emanation plant at Boston’s Collis P. Huntington Hospital, which he replicated at Memorial Hospital (33,34). Memorial’s entire 4 g of radium was kept in solution (Figure 1.5), and the purified radon was encapsulated in short lengths of glass capillary tubes, 0.3 mm in diameter (Figure 1.6), which were inserted into hypodermic needles. The radon-bearing needles were used for temporary implantation (the needles’ steel filtered most beta particles and soft gamma rays).

Figure 1.5

Memorial’s emanation plant. All of Memorial’s radium was kept in solution in the safe (bottom right). The emitted radon was captured and purified.

Source: From Ref. (33). Failla G. The physics of radium. In: Clark JG, Norris CC, eds. Radium in Gynecology. Philadelphia, PA: JB Lippincott Co; 1927:63. Used with permission.

Figure 1.6

Capillary glass radon tubes, inserted into “serum” needles for temporary implantation.

Beginning in 1915, Memorial’s urologist, Benjamin Barringer, used these needles for outpatient treatment of prostate cancer (35). With the patient in the lithotomy position, Barringer anesthetized the perineum prior to implanting a 15 cm needle, under the guidance of a finger in the rectum, into a lateral prostate lobe. The needle, bearing 50 to 100 mCi of radon in its distal 3 cm, was left in place for 4 to 6 hours before being retracted and inserted into the other lateral lobe. The seminal vesicles were often treated through a transrectal puncture. Treatments were repeated as necessary, at intervals of several months (36). Barringer reported highly favorable tumor responses.

With abundant quantities of short half-life radon, it became appealing to perform permanent implants. At first, “bare” glass tubes were implanted into tumors, but this practice resulted in painful sloughing of necrotic tissue. Memorial’s physicist, Gioacchino Failla, recognized the offender to be unfiltered caustic beta particles. He remedied the problem by encasing the radon in a 0.3 mm thick envelope of gold (Figure 1.7) that filtered out 99% of beta particles while allowing more than 80% of therapeutic gamma rays to pass (37). Barringer implanted up to 20 seeds, each containing 1.5 to 2.0 mCi of radon, into the prostate, typically delivering 4,000 mCi h of treatment (38). Barringer’s techniques were adopted at other institutions (39,40), and a “gold” radon seed industry was established that persisted in the United States for decades (41).^k

Figure 1.7

A gold-encased radon seed. Note the resemblance of the dimensions to modern seeds.

Introduction of Man-Made Radionuclides

The large majority of prostate cancer patients undergoing radium or radon brachytherapy developed recurrence (38). This is not surprising, as most men diagnosed in that era had advanced disease that could not be cured by any means. The use of prostate brachytherapy waned after Charles Huggins (1901–1997) discovered that prostate cancer responds to androgen deprivation (1941) (42), but interest revived on recognition that castration was only temporarily effective.

Congress passed the Atomic Energy Act after World War II, establishing the Atomic Energy Commission. The Oak Ridge Laboratories were transferred to civilian control and directed to provide radioisotopes for peaceful purposes, including medical applications. One of the first radionuclides made available was radiogold (Au-198); its short half-life (2.7 days) is comparable to that of radon, but is safer to handle because it does not generate megavoltage photons and has no radioactive daughter products. Microparticles of forming a colloid for instillation into pleural or peritoneal cavities (to suppress malignant effusions and ascites) (43), or injected into lymphomatous masses and solid tumors (44). The first radiogold prostate implant, at the University of Iowa in 1951, was unplanned (45). The prostate of an 80-year-old man with hormone-refractory Stage C disease was surgically exposed for radon seed implantation, but the seeds were not available. Colloidal gold was at hand, and was infiltrated into the prostate. The treatment was without apparent toxicity, the bulky tumor resolved, and follow-up biopsy was negative (46). The urologist, Rubin Flocks, began infiltrating colloidal gold into the prostate and seminal vesicles of men with Stage C disease, through suprapubic and perineal approaches. An enthusiastic report on 20 cases was published in the Journal of Urology the following year (47).

There were compelling reasons to consider colloidal gold as a suitable radionuclide for prostate brachytherapy. It is a beta emitter that deposits 90% of its energy within millimeters (it was assumed that the fascia investing the prostate and seminal vesicles would limit the colloid’s migration). There was evidence that gold microparticles would be phagocytosed by macrophages, which, on circulating to draining lymph nodes, would irradiate D1 metastases (48). In reality, treatment did not work as expected. Dense tumor nodules resisted infiltration; the colloid had to be injected under pressure, resulting in spattering that contaminated drapes, scrubs, and shoes (Figure 1.8). Radiation exposure to personnel was so high that surgical teams were rotated to avoid accumulation of prohibitive doses (49). Much of the injected material leaked out of the prostate, pooling in the pararectal gutters, causing severe rectal injury (50). Some of the gold microparticles entered the circulation, and autoradiographs demonstrated hepatic accumulation. Although radiogold did percolate through regional lymphatics, it did not penetrate tumor-congested nodes (50). Flocks devised several maneuvers to overcome these difficulties: Grossly abnormal lymph nodes were resected; hyaluronidase and epinephrine were mixed into the colloid to improve distribution and reduce vascular uptake; and small volumes of highly concentrated suspension were used to reduce leakage from the gland (49). It became apparent that the procedure was only effective for the smallest tumors (50), and 80% of posttreatment biopsies were positive (51). Flocks eventually resorted to perineal prostatectomy, using radiogold as adjuvant therapy (infiltrating the colloid into periprostatic fascia and vascular pedicles) (46). He defended the procedure, claiming better local control (95%) for Stage C disease, compared to published prostatectomy series (70%–80%) (52). Toxicity, however, was significant: delayed healing in 80% and “persistent urethro-cutaneous fistula” in 2% (53). Use of colloidal gold continued at the University of Iowa Hospitals until its manufacture ceased in 1977; thereupon, radiogold grains were substituted.

Figure 1.8

Device used at the University of Iowa to inject colloidal gold under pressure. It was heavily shielded, to reduce the operator’s exposure.

Source: From Ref. (49). Flocks RH, Culp DA. Radiation Therapy of Early Prostate Cancer. Springfield, IL: Charles C. Thomas; 1960. Public domain.

Urologist C. Eugene Carlton (1930−) initiated a prostate brachytherapy program at Baylor Hospital in 1965. He chose to implant Au-198 grains (rather than colloid) because of ease of handling and more accurate placement. The procedure began with lymph node dissection, followed by incision of the endopelvic fascia and mobilization of the prostate, allowing implantation under direct visualization (54). Initially, a single gold grain was implanted in the tumor nodule; eventually, the procedure entailed implantation of 6 to 10 grains, distributed within the gland (55) (Figure 1.9). Cobalt (later, linac) teletherapy began 2 to 3 weeks later; the radio-opaque grains served as fiducial markers, identifying the prostate. Although the procedure was designed for Stage C disease, early results were so promising (58% negative biopsies) that it was also used to treat patients with organ-confined disease (54). Toxicity consisted mostly of thrombophlebitis and temporary extremity or genital edema (secondary to the node dissection). The incidence of proctitis was 16%, but less than 1% of patients required colostomy (56). Impotence was reported to develop in only 5% of men who were potent prior to treatment. The procedure became so popular that many private urologists at affiliated hospitals participated. But the Baylor program had serious flaws. The radiogold grains were delivered to Baylor once weekly, and their activity at time of implantation varied widely (between 2 and 9 mCi, depending on the day of implantation). Dosimetry was crude; dose was estimated by assuming that the entire implant activity was deposited at the geometrical center of the prostate, and the delivered dose was defined as the isodose that subtended a diameter equivalent to that of the gland (56). It is difficult to encompass the gland with so few sources, even if they were well placed. Years later, formal dosimetric evaluation demonstrated that these implants typically delivered less than a third of the prescribed dose (57). It is not surprising that, with longer follow-up, treatment outcomes were disappointing (58,59).

Figure 1.9

A Baylor implant. It is difficult to achieve adequate coverage with only six sources, even with a high energy radionuclide.

Source: From Ref. (55). Hudgins PT. Irradiation of prostatic cancer combined with abdominal exploration. In: Fletcher GH, ed. Textbook of Radiotherapy, 2nd ed. Philadelphia, PA: Lea & Febiger; 1975:768–772. Used with permission.

Ulrich K. Henschke (1914–1980) came to New York to head Memorial’s brachytherapy service in 1955. He had spent the previous 3 years at Ohio State University, where he collaborated with William Myers in the introduction of Au-198 and Ir-192 into clinical practice (60,61). In New York, however, most of his permanent implants used radon, because the daily seed requirement was unpredictable (he was called to the operating room whenever a surgically exposed tumor was found to be unresectable) (62) and large quantities of radon seeds were produced by Memorial’s radon plant. In 1963, health physicist Donald Lawrence sought funding from Memorial for production of an I-125-impregnated suture (63). Henschke provided encouragement and modest financial support, but advised that the radionuclide be encapsulated in a seed.^l Within months, Lawrence sent him iodine seeds for animal studies, and Henschke began performing human I-125 implants for lung cancer in 1965 (64). Henschke’s protégé, Basil Hilaris (1928−), assumed leadership of the brachytherapy service upon Henschke’s departure in 1967. Brachytherapy had been used at Memorial as salvage therapy for locally recurrent prostate cancer (following failed radiation or hormonal therapy) since 1956 (65), and Hilaris proposed I-125 prostate brachytherapy as primary treatment to Memorial’s chief of urology, Willet Whitmore, Jr. (1917–1995). Whitmore was receptive; he had attempted aggressive surgical resections early in his career, but by 1963 (66), acknowledged the futility of radical prostatectomy to control locally advanced disease. He concluded that intervention for prostate cancer should consider quality of life, and “need not necessarily involve an effort at cancer cure” (67).

Memorial’s I-125 implant procedure began with the patient in a modified lithotomy position (68). A Foley catheter was inserted and an O’Connor drape (with an appendage allowing insertion of a finger into the rectum) was placed. A midline or paramedian incision extended from the umbilicus to the pubis. External, hypogastric, and obturator nodes were dissected. Fat was removed from the anterior surface of the prostate, but the puboprostatic ligaments were left intact. The endopelvic fascia was incised, mobilizing the lateral margins of the gland, but the prostate was not dissected from the rectum (69). The radiation therapist then inserted empty 15 cm long 16-gauge steel needles into the gland, spaced approximately 1 cm apart. The needles were slowly advanced until sensed by a finger in the rectum (Figure 1.10). Gland dimensions, measured intraoperatively, were used to calculate prostate volume, which determined total implant activity. The number of seeds needed for the implant was derived by dividing the calculated total implant activity by the activity of the available seeds (the ideal seed strength was eventually determined to be 0.5 mCi). Memorial physicist Lowell Anderson developed nomograms to rapidly calculate seed requirements and spacing (70). An applicator was developed to implant the seeds (71) (Figure 1.11). Bleeding could be heavy (median blood loss was 1 L), and almost half the patients required transfusion (72). The Foley catheter was removed 1 to 3 days postoperatively, and the patient was discharged a week later. Postoperative irradiation was delivered to patients found to have lymphatic metastases or bulky tumors (73). Operative mortality was rare (0.5%). The most distressing complications (venous thrombosis, pulmonary embolism, lymphocele, lymphedema) were attributed to lymph node dissection. Impotence or incontinence occurred in fewer than 10% of cases (72).

Figure 1.10

Retropubic implantation of needles toward a finger in the rectum.

Source: From Ref. (67). Whitmore WF Jr. Proceedings: the natural history of prostatic cancer. Cancer. 1973;32(5):1104–1112. © 1975 Memorial Sloan-Kettering Cancer Center. Used with permission.

Figure 1.11

Seed implantation using an early applicator designed by Felix Mick (when he was employed by Memorial Hospital).

Source: From Ref. (71). Hilaris BS. A Manual for Brachytherapy. 2nd ed. New York, NY: Memorial Hospital; 1970:65. © 1970 Memorial Sloan-Kettering Cancer Center. Used with permission.

A computer program was used to calculate dose distribution from postimplant radiographs (see the “Computer Dosimetry” section; Figure 1.12). Without accurate delineation of the target volume, however, the adequacy of an implant was difficult to determine. The dose covered by a volume equivalent to that of the prostate (calculated from intraoperative measurements) was deemed the “matched peripheral dose” (mPD; Figure 1.13) (74). This metric was misleading, as there was no indication that the target and the treated volumes coincided (75) (published radiographs suggest that they often did not [68]), and intraoperative measurement was later found to underestimate prostate volume (76). Disease control was monitored by digital examination, acid and alkaline phosphatase levels, and bone scans. Local control (as determined by palpation) was 80% at 5 years if the mPD exceeded 10,000 rads (100 Gy) (74). Of the 40% of patients found to have nodal metastases, fewer than half survived 5 years, and were found not to have benefited from nodal dissection or irradiation; thereafter, nodal dissection was eliminated from the procedure (73).

Figure 1.12

Computer-generated dose distribution from an open retropubic implant.

Source: From Ref. (68). Hilaris BS, Whitmore WF, Batata M, et al. Cancer of the prostate. In: Hilaris BS, ed. Handbook of Interstitial Brachytherapy. Acton, MA: Publishing Science Group; 1975:219–234. © 1975 Memorial Sloan-Kettering Cancer Center. Used with permission.

Figure 1.14

Determination of matched peripheral dose (mPD) from a computer-generated dose–volume histogram. The mPD was the dose delivered to a volume of tissue equivalent to the intraoperatively calculated prostate volume.

Source: From Ref. (68). Hilaris BS, Whitmore WF, Batata M, et al. Cancer of the prostate. In: Hilaris BS, ed. Handbook of Interstitial Brachytherapy. Acton, MA: Publishing Science Group; 1975:219–234. © 1975 Memorial Sloan-Kettering Cancer Center. Used with permission.

More than a thousand patients were implanted with iodine seeds at Memorial Hospital between 1970 and 1986. It was appreciated that quality implants controlled very early disease, but few patients had presented with early disease, and few implants delivered the prescription dose. Disease-free survival curves never plateaued (77), and reports of disappointing long-term control rates (78,79) led to abandonment of the procedure.

The Incorporation of Sonography

Physicists involved in the discovery of radium also uncovered the principles underlying sonography. The piezoelectric effect (the property of certain crystals to develop an electric potential when mechanically stressed) was described in 1880 by Marie Curie’s future husband (Pierre, 1859–1906) and brother-in-law (Jacques, 1856–1941). The following year, Marie Curie’s thesis adviser (Jonas Gabriel Lippmann, 1845–1921) predicted that a change in electric potential would alter a crystal’s dimensions (91). These phenomena underlie the function of the ultrasound transducer: The generation and detection of sound waves. The first practical application of sonography was a device to detect German U-boats (sonar), patented in 1916 by Pierre Curie’s doctoral student (Paul Langevin, 1872−1946^m). Sonography was later used by industry as a nondestructive method for detecting metal flaws and fatigue (replacing X-rays and gamma rays for that purpose) (92). Ultrasound was applied by physiatrists in the 1930s to therapeutically heat subsurface tissues (93). Diagnostic applications were developed in the late 1940s; initial attempts measured the transmission of ultrasound waves through tissue (hyperphonography) (94), but detection of reflected waves was investigated by 1950 (95). Sonography for detection of cancer was described in 1957 (96).

An inventive and mechanically inclined Danish surgical resident, Hans Henrik Holm (1931−), became interested in sonography during a radiology rotation. He visited physicist Carl Hellmuth Hertzⁿ (1920–1990) in Lund, Sweden. Hertz had explored medical applications of sonography with cardiologist Inge Edler (“father of echocardiography,” 1911–2001) and neurosurgeon Lars Leksell (“father of radiosurgery,” 1907–1986). Holm was awarded a state grant to obtain an ultrasound unit, and duplicated Lund’s multidisciplinary methodology by collaborating with a cadre of young physicians, as well as with the Welding Institute (a state technology laboratory) to adapt and develop ultrasound apparatus for clinical use.^o Equipment was designed to be mobile, so that bedside procedures could be performed. The group developed techniques for interventional sonography in the 1970s, including percutaneous biopsy, drainage, pericardiocentesis, amniotic fluid sampling, and percutaneous nephrostomy.

In 1974, the Welding Institute introduced a probe with transducers for transurethral and transrectal imaging (97,98). A “fixing sledge” (stepper unit) that retracted the probe at 5 mm intervals facilitated planimetric volume determinations (98). A metal template mounted on the probe shaft directed prostate and seminal vesicle biopsy (99). By 1980, Holm was using ultrasound guidance to implant I-125 seeds (separated by chromic suture spacers) into liver metastases and pancreas tumors (100,101). The prescription dose was 160 Gy, and most patients also underwent adjuvant teletherapy. By 1982, he was implanting I-125 seeds into cancerous prostates, under the direction of axial imaging from a rectal probe mounted on a sledge-stepper (102). Preplanning and implantation were performed with the patient in the lithotomy position. A modified Memorial Hospital nomogram determined the implant activity needed to deliver 160 Gy, based on Henschke’s system of dimension averaging (103). A 3 cm thick acrylic template was attached to the probe shaft (Figure 1.17). After immobilizing the gland with an empty needle passed through the template, needles preloaded with seeds and spacers were inserted. Needles that were to be advanced most deeply (in the central gland) were placed first. After proper needle position was confirmed by transverse sonographic imaging, the seeds were deposited by stabilizing the stylet while the needle was retracted. The ultrasound probe was then retracted 5 mm, and the next deepest set of needles was placed; in this fashion, concentric circles of needles were inserted and their seeds deposited. Postimplant dosimetry was performed on orthogonal radiographs the following day (Figure 1.18).

Figure 1.17

Acrylic template mounted on B&K 1850 probe. Needle hole numbers coordinated with the number of the needle to pass through it.

Source: From Ref. (105). Holm HH. The history of interstitial brachytherapy of prostatic cancer. Semin Surg Oncol. 1997;13(6):431–437.

Figure 1.18

Postimplant radiograph of a Holm implant.

Source: From Ref. (116). Torp-Pedersen S, Holm HH, Littrup PJ. Transperineal I-125 seed implantation in prostate cancer guided by transrectal ultrasound. In: Lee F, McLeary RD, eds. The Use of Transrectal Ultrasound in the Diagnosis and Management of Prostate Cancer. New York, NY: Alan R Liss; 1987:151.

A 1989 paper reported that 33 patients had undergone implantation followed by teletherapy (40–47.4 Gy in 20 fractions) with as little as a 2 week interval between brachytherapy and teletherapy (106). Of 25 patients undergoing postimplant biopsy and/or transurethral resection, 12 had pathological evidence of persistent disease. Forty-five percent of patients had suffered “severe” late complications (hemorrhagic proctitis, anal ulceration, rectovesical fistula, or “severe persisting radiation cystitis”). The combination of disappointing disease control and high morbidity led to abandonment of the program in 1987 (107).

By then, several centers in Europe (108–111) and the United States had adopted Holm’s technique. Stefan Loening, at the University of Iowa, visited Holm in 1984 and began performing ultrasound and fluoroscopically guided transperineal implants in October (112). His technique differed from Holm’s in that he used a Mick applicator to implant Au-198 grains under axial and sagittal ultrasound imaging (113). One hundred seventy-nine patients were implanted in Iowa within 7 years (114). Patients with bulky tumors were treated with a combination of brachytherapy and teletherapy (115). Response was monitored by digital examination, prostate shrinkage on serial sonography, and biopsy. Roughly half of the 12 month biopsies were positive, but some became negative at 24 or 36 months (115). Toxicity was modest.^p

After visiting Holm in 1984, Seattle urologist Haakon Ragde (1927−) proposed the institution of an ultrasound-directed brachytherapy program to radiation oncologist John Blasko (1943−). Blasko had reservations; the recently introduced prostate specific antigen (PSA) test had revealed that most patients who had undergone retropubic implantation had developed recurrence. It was unclear whether failure was due to poor seed distribution or the inability of I-125 to eradicate prostate cancer. Other concerns were regarding dose and sequencing. Blasko believed that a 160 Gy implant (without teletherapy) should be adequate to control low-grade tumors. If brachytherapy was to be combined with teletherapy (for more aggressive or bulkier disease), the implant dose should be reduced and teletherapy should be delivered first (to avoid concurrent dosing). With these alterations in place, Ragde and Blasko performed their first implant in November of 1985.

They used Brüel and Kjær (B&K) equipment (Figure 1.19). Implantation was preceded by a volume study, acquiring axial images separated by 0.5 cm. Their target volume was several millimeters wider than the prostate. The treatment plan consisted of placing seeds 1.0 cm apart (the holes on their original template were separated by 1.0 cm) throughout the target volume (placing some seeds in extracapsular locations). The total implant activity was determined by nomogram (initially, Holm’s nomogram [117]; later a modification of the Memorial nomogram [118]), and individual seed strength calculated by dividing total implant activity by the number of seeds in the plan. Computer dosimetry checked the adequacy of the preplan. Eighty to hundred seeds, of 0.30 to 0.40 mCi, were implanted. The seeds (separated by chromic suture spacers) were preloaded into 18 gauge needles.

Figure 1.19

Seattle setup, circa 1987. Note the B&K 1850 probe, used by Holm, Loening, and the Seattle group. The shaft was so thin that it had to be propped up to prevent sagging. The stepper and template were also B&K equipment, but the stepper immobilizer was custom fabricated by a retired Boeing machinist.

Source: Courtesy of John Blasko.

The procedure was performed under spinal anesthesia, with the patient in the lithotomy position. The ultrasound probe was positioned to recapitulate the volume study images. After stabilizing the gland with two empty needles, the base of the prostate was viewed on axial imaging, and the central needles (which would implant seeds at the base) were inserted first. After the central needles’ seeds were discharged, the probe was retracted 1.0 cm, and a second cohort of needles was placed. In this manner, all needles were inserted and discharged. Pubic arch interference was overcome either by freehand needle angling, or by drilling holes through the bone (118)! Postimplant dosimetry was performed on orthogonal films taken 2 weeks after implantation.

Their first patient tolerated the procedure well, but it would be several months before they would perform a second implant; thereafter, 273 men were implanted within 4 years (119). Ragde recruited patients to ultrasound and PSA screening clinics by advertising on air and in printed media.^q The procedure had become so popular that Peter Grimm (1952−) and Timothy Mate (1949−) were soon incorporated into the program. The technique and equipment evolved. Spacing between template holes was reduced to 0.5 mm. Implants were reserved for prostate volumes of less than 60 mL, and pubic arch drilling was abandoned. The order of needle insertion was changed (anterior, not central, needles were implanted first). Seed activity varied, eventually settling at 0.36 mCi. Mate was concerned about the quality of the implants, and had physicist Douglas Schumacher write a computer program to calculate dose distribution from postimplant CT scans (120). CT-based postprocedure dosimetry revealed implant deficiencies, prompting modifications in technique (such as eliminating periurethral seeds) that enhanced dose distribution.

The Seattle group reported favorable treatment outcomes at meetings and in publications (121,122). Visits from interested urologists and radiation oncologists became common; to relieve the congestion of operating room visitors, the group instituted monthly training sessions. The equipment evolved; initially, they had to cut their own chromic spacers and have their needles sharpened after every 10 cases. Their success soon attracted industry attention, and a symbiotic relationship developed. Disposable brachytherapy needles (with echogenic tips), precut spacers, stiffened Vicryl seed strands, and palladium seeds were introduced. Industry helped introduce the procedure to physicians and the public, and lobbied for physician reimbursement codes.

The initial Seattle technique was identical to the procedure performed by Holm in Denmark; why were the outcomes different? The Seattle patients had earlier disease (with more favorable prognosis) because they had been diagnosed as a result of screening. Blasko’s modified dosing and sequencing reduced the intensity of therapy, resulting in less morbidity.

The ultrasound-guided transperineal procedure was rapidly accepted in the United States, accounting for a growing percentage of patients treated for prostate cancer. Some radiation oncologists made it the focus of their practice, and introduced innovations. Urologist Nelson Stone and radiation oncologist Richard Stock correlated implant dose (123,124) and adjuvant therapy (125,126) with disease control (127,128) and urinary (129,130), rectal (131,132), and sexual toxicity (133,134). Their publications have helped establish guidelines for dosing and normal tissue constraints. They adopted intraoperative treatment planning and used a computer to monitor dose distribution as seeds were deposited, allowing real-time implant modification. Radiation oncologist Frank Critz was one of the few brachytherapists to have enjoyed success in both the retropubic (135) and transperineal (136) eras. Critz adopted ultrasound-guided implantation after taking the Seattle Prostate Institute course (1992); at its peak, his program had implanted more than 1,000 men annually. Critz was an advocate of stringent PSA criteria for disease control, adopting the urologist’s postprostatectomy standard of achieving and maintaining a PSA level less than or equal to 0.2 ng/mL (137). His excellent, durable outcomes, reported in urology journals (138,139), legitimized prostate brachytherapy as a highly effective alternative to surgery.

Afterloading, Remote Afterloading, and High Dose Rate Brachytherapy

Although delayed loading of radium into previously implanted applicators had been performed as early as the first decade of the 20th century (140,141), afterloading was not seriously pursued as a radiation safety measure until the 1950s (142,143). The implantation of inert applicators facilitates deliberate, unhurried procedures (especially important for trainees), and eliminated exposure to the brachytherapist, operating-theater personnel, recovery room nurses, and radiology technicians (as well as people in the hallways through which the implanted patients passed in transit to their hospital room). It allowed dosimetric determination of optimal source distribution prior to loading. In 1953, Ulrich Henschke, Arthur James, and William Myers (at Ohio State University) described temporary interstitial brachytherapy by afterloading Au-198 seeds into previously implanted nylon tubes (144). Henschke later introduced the use of Ir-192 seeds for this purpose (145,146).^r

Afterloading nylon tubing with Ir-192 became an integral part of the “Paris System” (147). Court and Chassagne described the Gustave-Roussy low dose rate (LDR) prostate technique in 1977 (148). Following lymph node and prostate dissection (and suprapubic resection of adenomatous hypertrophy), the prostate was grasped between the thumb and index finger of the left hand. Steel needles were advanced through the perineum and into the prostate by the right hand. The needles were replaced by plastic loops, through which lead wires were threaded for orthogonal films. Two or 3 days later (following dosimetric calculations), Ir-192 wires replaced the lead in the loops. Sixty to 70 Gy was delivered over 6 days. A slightly less invasive technique was described by Miller in 1979; beginning in 1972, 16 patients underwent Ir-192 LDR temporary prostate implantation at the Duke University Medical Center (149). The implant was preceded by small-field prostate irradiation delivering 20 Gy in 10 fractions using anterior and posterior megavoltage beams. The implant began with a lower abdominal incision to gain access to the pelvis; neither lymph node dissection nor mobilization of the prostate was performed. Closed-end needles were passed through the perineum under the guidance of a finger in the rectum and a hand in the pelvis, directing the tips 1.5 to 2.0 cm superior to the prostate base. The three sides of a “triangular volume implant” were defined by the ischia and rectum. Following confirmation of needle parallelism by orthogonal radiographs, the needles were fixed by anchoring buttons sewn to the perineum. Six centimeter lengths of Ir-192 wire were afterloaded into the needles, delivering 45 to 50 Gy in 4 to 5 days. There were several unique features of this program: Although it was an “open” technique, neither lymph node dissection nor prostate mobilization was performed. The patient was supine with legs abducted. Extracapsular tumor extension was encompassed by needles inserted through the obturator foramina. Small bowel was protected from irradiation by suturing the bladder dome to the posterior surface of the pubic arch, and by maintaining partial bladder filling by means of a Y-tube.

Nisar Syed (1949−) was a surgeon prior to training in radiation oncology at Manchester’s Christie Hospital and with Henschke at Howard University. He came to the University of Southern California in 1974, where his colleagues included gynecological oncologist Philip Di Saia and physicist David Neblett. Syed and Neblett developed site-specific acrylic templates to fix transperineal needles during treatment (151). Disappointed with the results of retropubic prostate implants (more than half of his biopsied patients were found to have persistent disease), he began using a template-guided transperineal technique to temporarily implant the prostate with Ir-192 (152). After lymph node dissection and exposure of the prostate, needles were advanced to the bladder neck under guidance of a hand in the pelvis. The templates contained concentric circles of holes; initially, each hole held a needle containing a ribbon with seven seeds of 0.4 mg Ra eq strength. After several patients developed serious complications, however, the needles were differentially loaded, with central and pararectal ribbons containing half-strength seeds (153). The implants delivered 30 Gy to the prostate over 40 hours; an additional 40 Gy was delivered by linear accelerator, beginning 10 to 14 days after the implant was removed. He treated 200 patients with this technique between 1977 and 1985; of 74 patients biopsied from 4 to 24 months after treatment, only 16% had evidence of persistent disease (153).

Although manual afterloading reduced exposure to the brachytherapist and operating theater personnel, it did not address exposure to the physician loading or unloading the implant, the staff member preparing or restoring the sources, or the nursing staff caring for the implanted patient. To eliminate all exposure to personnel, remote afterloading was introduced in 1962 at Stockholm’s Radiumhemmet (154). The single channel unit, based on the source transport system of a teleradium unit (155), remotely delivered a Cs-137 source on a flexible cable into a hollow applicator. This unit, and others that soon followed (Cervitron, Curietron), essentially reproduced standard Manchester gynecological distribution and dose rates. The GammaMed, an afterloading unit with a high activity Ir-192 source deployed through a single channel, was introduced in 1964 for stereotactic treatment of brain tumors. The same year, Henschke introduced the concept of an ingenious afterloading device that could simulate an infinite variety of source loadings by cycling a single high-activity source (Figure 1.20) (156). He suggested treating patients in minutes (rather than days), declaring, “On the basis of our limited experience with such short treatment times in the last three years, we feel that they may be used with impunity if the total dose is divided into more fractions” (157).

Figure 1.20

Memorial Hospital’s proposed afterloader. The source’s oscillation was programmed by a tie bar (that determined the length of the cycling path) and a cam (that dictated dwell times). An afterloader was not put into use at Memorial until 1967.

Source: From Ref. (156). Henschke UK, Hilaris BS, Mahan GD. Remote afterloading for intracavitary radiation therapy. Prog Clin Cancer. 1965;10:127–136.

Few brachytherapists (or radiobiologists) shared Henschke’s confidence that fractionated high dose rate (HDR) treatments would be safe. When the Cathetron, a HDR remote afterloading unit (using capsules of Co-60 that delivered a dose more than 100 times more rapidly than standard radium sources) was introduced at London’s Charing Cross Hospital, clinicians approached its use carefully. They performed animal experiments to assess the relative potency of large fractions (158). Toxicity data from hypofractionated teletherapy regimens were considered (159). Physicist William Liversage applied his cell survival equation (a predecessor of the linear–quadratic equation) to calculate appropriate HDR fractionation schemes (160). After the applicator was placed, but prior to delivery of therapy, very low activity sources were inserted into the applicator and the dose delivered to the rectum was measured at 1 cm intervals (161). The first patients to be treated were women who were to undergo hysterectomy for endometrial or cervical cancer, because they would be treated to a lower dose, and the resected irradiated uterus would undergo pathological examination (159). When compared to patients treated with traditional radium loading, the uteri of HDR patients showed more frequent complete responses, and the patients had fewer complications (162).

Originally, HDR units utilized Co-60 sources. Ir-192, however, has a greater specific activity, and its adoption for HDR allowed fabrication of 1 mm source capsules, amenable to interstitial application. After visiting Holm, urologist Hagen Bertermann proposed the institution of an ultrasound-directed prostate implantation program at the Christian-Albrechts-University Hospital in Kiel, Germany. His radiotherapy colleagues agreed to collaborate, but preferred to use their HDR equipment rather than I-125 seeds. Beginning in August of 1985, they treated prostate cancer with HDR brachytherapy and external radiation (163). The prostate was targeted with 40 Gy teletherapy in 2 Gy fractions from a linear accelerator (164). Interposed with teletherapy were two fractions of HDR brachytherapy, each delivering 15 Gy to the prostate peripheral zone. The implants were preplanned; needle insertion, treatment, and needle removal were all performed in a single operation in the brachytherapy suite, while the patient was anesthetized. Initially, only two to four needles were inserted. PSA-based recurrence-free survival was 69% at 10 years (a very respectable result, considering that a third of their patients had pretreatment PSA greater than 20) (165). Grade 3 rectal and urinary complications occurred in 4% and 2% of patients, respectively (166).

Timothy Mate already had experience with HDR for gynecological malignancies and had performed ultrasound-guided prostate seed implantation when he instituted a prostate HDR program at Seattle’s Swedish Hospital (Figure 1.21) (167). Flexible cystoscopy was added to the procedure to position needle tips under bladder mucosa (168). Based on his gynecological experience, he prescribed 12 to 16 Gy, delivered in four fractions spread over 40 hours. Teletherapy, upto 50 Gy, was begun 2 weeks after brachytherapy. Acute urinary toxicity was low (compared with his experience with permanent implantation), but 8% of patients developed urinary stricture at 5 years. Recurrences were uncommon (8%) if pretreatment PSA was less than 20. He treated 104 patients between 1989 and 1995; Andy Grove, CEO of Intel, was among them. Prostate brachytherapy gained important publicity when Grove’s enthusiastic report became the cover story of Fortune Magazine (169).

Figure 1.21

Timothy Mate and an afterloaded prostate implant.

Source: Courtesy of Timothy Mate.

Ultrasound-directed HDR brachytherapy was initiated at William Beaumont Hospital in 1991 (170). The procedure was performed on outpatients, in the brachytherapy suite, under spinal anesthesia. A computer generated the treatment plan (including needle position and source dwell times) intraoperatively. Dose distribution was recalculated during implantation to account for actual needle position. Initially, patients with intermediate- or high-risk disease were treated with three HDR applications, interdigitated with teletherapy (46 Gy). Between 1991 and 1995, the HDR fraction size rose from 5.5 to 6.5 Gy.; thereafter, the number of HDR applications was reduced to 2, and fraction size gradually rose from 8.25 Gy to 11.5 Gy. It became apparent that disease control strongly correlated with HDR fraction size (171,172), resulting in a recalculation of the α/β ratio for prostate cancer (from 10 to less than 2) (173). The Beaumont Group and the California Endocurietherapy Cancer Center have published promising results using HDR monotherapy for favorable- and intermediate-risk disease (174,175).