Harvey Cushing's Operative Treatment of Metastatic Breast Cancer to the Central Nervous System in the Early 1900s

The early 1900s were a time ready for exploration into the treatment of metastasis of breast cancer to the central nervous system (CNS). In fact, the state of surgical treatment of breast cancer itself was undergoing a significant transition. The first half of the 1800s had been characterized by a firm belief that removal of the breast in cases of carcinoma was a foolish endeavor.¹ By the late 1800s, this notion began to change as William Halsted, MD, a respected surgeon at The Johns Hopkins Hospital, published research suggesting that more radical surgical procedures may improve survival.¹ Throughout the late 1800s and early 1900s, Halsted continued his work in this area with Harvey Cushing, MD, as his resident.²

The ability of breast cancer to metastasize to the CNS, and elsewhere throughout the body, was a well-established concept by the early 1900s. William Osler, MD, described several instances of metastasis, as did Halsted.^{3 - 5} Data from necropsies revealed that in rare cases, metastatic spread of carcinoma into the spinal cord was possible as well.⁶ Although some physicians believed that the symptoms of breast cancer metastases to the brain could be confused with cerebral embolisms,^{4 ,7} in general, symptoms were well recognized and were known to be similar to those of primary tumors (including headache, drowsiness, vertigo, and vomiting as well as localizing symptoms particular to specific regions of the brain).⁸ In fact, several theories for the cause of breast cancer metastasis had begun to emerge. Many of these focused on the role of lymphatics; however, dissemination through the bloodstream was believed to be particularly relevant in the case of cerebral metastasis.^{3 ,9}

Despite awareness of neurologic metastases, accurate estimates of its occurrence largely were lacking.¹⁰ Furthermore, in cases in which these metastases were recognized, surgeons of the era adhered to a general rule not to operate.⁸ In 1906, Schlesinger wrote, “Brain tumors can be primary or secondary. Surgical treatment concerns itself of course with the primary growths only.”^{8 (p3)} In actuality, the field of neurologic surgery itself was in the early years, and textbooks of the era reflected this fact. Henry Butlin's tome, On the Operative Surgery of Malignant Disease,¹¹ noted that merely 12 years before the turn of the 20th century, intracranial operations were purely experimental. Mumford's 1910 textbook entitled The Practice of Surgery stated that the

treatment of brain tumors has only recently begun to emerge from a position of almost hopeless chaos, and to-day even many competent general surgeons are skeptical of any practical benefit from operations.^{12 (p657)}

In 1906, Cushing, while still at The Johns Hopkins Hospital, was credited with emphasizing the value of decompression in the relief of increased intracranial pressure and for creating improvements in technique.¹³ We report herein on Cushing's first attempts at treatment of breast cancer metastases to the CNS.

The Johns Hopkins Hospital surgical records from 1896 to 1912 were reviewed. The 4 cases in which Cushing recorded the diagnosis as a metastasis of breast cancer to the CNS were selected for further study.

In early August 1905, a 40-year-old woman presented with gradual onset of a “pins and needles” sensation in her legs and difficulty moving her lower extremities. By hospital admission, her loss of sensation was practically complete to approximately 3 in (7.62 cm) below the nipple line, and she had flaccid paralysis of her legs and abdomen. In addition, the patient required bladder catheterization. Her medical history included a complete operation by Halsted for breast adenocarcinoma with metastasis to the axilla 1 year earlier.

Three days after the woman's current admission, Cushing performed a laminectomy for metastatic carcinoma with complete paraplegia. He wrote in his operative note (Figure 1):

Grahic Jump Location

Figure 1. Copy of Harvey Cushing, MD's, original postoperative illustration in August 1905. Labeling from top to bottom reads “Sketch Showing Situation of Growth (in red). Tumor split by dural incision. Collar of Tumor at TIII Vertebra.” Reproduced with permission from the Alan Mason Chesney Medical Archives of The Johns Hopkins Medical Institutions.

It was found that the third thoracic spine [was] loose, evidently a pathological fracture having taken place. Four or five of the upper thoracic spines were cut away and malignent [sic] disease was encountered . . . [the growth] seemed to entirely surround the cord. The dura was first entered . . . so that the cord was freed from the compression . . . A small strip of dura and enclosing tumor was removed for future pathological study. The growth was found not to be closely adherent to the dura, but could be easily separated from it by a blunt instrument. As much as possible of the tumor was removed, and the visible areas of necrotic tissue in the neighborhood were easily gotten away.

In the days after the operation, the patient's general condition was reported as good, but she had no return of movement or sensation. Before being discharged from the hospital in mid-September, she began to experience chills, fevers, sweats, and some attacks of nausea and vomiting. A note by Cushing placed in the patient's file on October 5 states:

Pt. gradually emaciated and died in a somewhat comatose condition at her home. The autopsy was performed by Dr Mayer, who found the abdominal organs riddled with metastases. Liver, intestines, spleen, kidneys, etc, were full of metastatic new growths.

In October 1909, a 49-year-old woman presented with a 3-month history of right-sided weakness and an attack of severe headache and temporary left hemianopsia the previous month. Since her attack, she experienced worsening headaches and nausea. The patient described instances of falling, inability to write, and awkward experiences of dropping objects. Physical examination revealed bilateral choked discs, complete loss of proprioception, and diminished motor strength and sensation on the right side. Her surgical history included an oophorectomy and left breast amputation in 1907. Cushing performed an extirpation of the tumor from the superior left parietal lobe. He recorded the following (Figure 2):

Grahic Jump Location

Figure 2. Copy of Harvey Cushing's original postoperative sketch illustrating the location of the lesion removed. The Illustration is entitled “Operator's 3/4 view of left hemisphere.” Labeling from right to left reads “Longitudinal sulcus, Central fissure, Right hemisphere.” Reproduced with permission from the Alan Mason Chesney Medical Archives of The Johns Hopkins Medical Institutions.

On reflecting the dura there came into view a grayish red, fairly definitely outlined tumor . . . about the size of a five-cent piece . . . By slow dissection . . . it was possible to tilt the growth forward and to expose it well down to a depth of 2½ or 3 cm. The tumor was then fully outlined by careful dissection and was scooped out without any loss of blood.

After the surgery, the patient showed dramatic improvement and regained muscle and sensory function on her right side. She was discharged after 1 month in the hospital in good health apart from some visual disturbances that had developed during postoperative recovery. In light of the patient's remarkable recovery, Cushing apparently consulted several physicians to be sure of her cancer diagnosis. The file contains several letters, including a reply from Maurice Richardson, MD:

My dear Dr. Cushing, . . . 
I saw Mrs. [reference removed] . . . On examination I found the left nipple had ulcerated away, and a mass could be felt in the breast and in the axilla. I removed the breast on June 26, 1907 . . . You can, I think, put it down in the records of this case that there is no question that the disease of the breast was cancer.
Yours very sincerely, M. H. Richardson.

The file also includes a reply from Frank Mallory, MD, a Harvard pathologist to whom Cushing sent a section of the tumor. Mallory returned the specimen embedded in paraffin with a note in which he stated that he was unsure of the diagnosis but suspected “an epithelial or neuroglia tumor.”

Some of the patient's symptoms returned in early December, less than 3 weeks after hospital discharge. She was readmitted to the hospital with headache, nausea, and visual problems. During this hospitalization, the patient developed relentless vomiting, sunk into a coma, and died within 6 days. Cushing performed the autopsy of the brain. His 4-page autopsy note describes multiple metastases that he encountered and provides theories on how each contributed to the patient's symptoms. On his autopsy note, Cushing labeled the diagnosis as “carcinoma metastases” and listed “metastatic carcinoma (mammary)” as the diagnosis on the first page of the patient's hospital record.

In April 1911, a 45-year-old woman presented with gradual onset of general numbness and attacks of vertigo in the previous 2 months. She was noted to have a great increase in intracranial tension but no definite localizing sign of tumor. She had a severe headache and a 5-week history of nausea and occasional attacks of vomiting. Her medical history was significant for a left breast amputation 18 months earlier for carcinoma. Her right breast also was amputated 6 months earlier due to a “cyst.” Physical examination revealed a marked choked disc in both eyes. The patient was drowsy, apathetic, and inattentive.

Cushing performed a right subtemporal decompression. His short operative note read:

A fairly generous opening was made without complications. There was no especial bleeding. The dura was freely incised disclosing a tense brain. There was some escape of fluid. The temporal muscles were exceedingly thin but were finally closed fairly satisfactorily.

Immediately after the operation, the patient had “no especial discomforts.” Her headaches and vomiting returned 10 days postoperatively, although these events were less severe than before the surgery. She was discharged after 3 more days in the hospital and subsequently died approximately 2 weeks later. An autopsy was permitted with sole examination of the brain after the body was embalmed. No gross abnormalities were revealed, but on sectioning, multiple cerebral and cerebellar carcinomatous metastases were noted to be in “almost every position” of the brain (Figure 3). After microscopic examination, the diagnosis was recorded as metastatic growth from breast cancer.

Grahic Jump Location

Figure 3. View of a brain section taken at autopsy and stored in the microfilm records of The Johns Hopkins Hospital. Arrowheads highlight the multiple metastases.

In February 1912, a 39-year-old woman presented with a recent decline in consciousness and a history of paroxysmal headaches with confusion of increasing severity that began on Christmas day. Her medical history was significant for a severe blow to her head before Christmas and an incomplete operation for breast cancer in January 1910 for a lump she had noticed in July of the previous year. After the initial operation, the patient underwent 2 rounds of x-ray treatment and another surgery when “other glands were noticed.” Cushing performed a subtemporal decompression. In his operative note he wrote:

Though the x-ray showed a point of tenderness . . . corresponding with the area at which the patient had received the trauma . . . it seemed best to attempt to relieve the immediate pressure symptoms by a decompression rather than to attempt to explore the possible lesion. There was no difficulty in the procedure and no special bleeding. A very tense dura was exposed and fortunately the brain was exceedingly wet and an abundance of fluid was evacuated reducing tension to normal.

Initially, the patient's condition improved; however, within 5 days, she experienced increasing pain, nausea, and vomiting. On postoperative day 11, Cushing performed a lumbar puncture, which immediately relieved some of the patient's symptoms. Two days later, when the symptoms returned, he performed a second lumbar puncture, but this did not improve the patient's condition; she developed worsening pain in her right leg and hip. In his notes, Cushing recorded “possibly a spinal lesion, inequality of reflexes.” After 2 additional days passed, Cushing performed an operation entitled “right extensive craniotomy for presumed metastatic tumor (patient's condition desperate).” He recorded:

The operation was nevertheless undertaken with full realization of the possibilities. Lumbar puncture was performed and about 40 cc of fluid withdrawn . . . A bone flap was then . . . turned back, disclosing the dura only moderately tense, due to the lumbar puncture. The dura was opened, no growth being disclosed . . . The patient took the anesthetic admirably, and seemed to me after its administration, as far as her appearance and general condition were concerned, better at the end than for three or four days before.

Within 4 days, however, symptoms returned. Another lumber puncture was performed, but the patient gradually deteriorated and died 1 week after her second operation.

We report herein on 4 of Cushing's cases while at The Johns Hopkins Hospital, where the patients presented with a history of breast carcinoma and symptoms suggestive of a tumor in the CNS. In each of these cases, metastasis was suspected. In contrast to the standard practice of his contemporaries, however, Cushing chose to attempt surgical treatment. Each patient ultimately reported some degree of temporary improvement in her symptoms, although these symptoms returned within days, and all the patients died within 2 months of their surgical procedures.

In the 3 cerebral metastasis cases, Cushing's treatments focused on the relief of symptoms caused by increased intracranial pressure, a common effect of cerebral metastasis.¹⁴ In the case of patient 4, for example, Cushing performed 4 lumbar punctures and 2 operations to reduce pressure. For a spinal metastasis, Cushing performed a laminectomy and intradural exploration. Throughout this work, Cushing clearly identified the condition of the patient as critical; he even included the words “patient's condition desperate” in the title of one of his operations.

Cushing was hampered in his ability to localize masses for resection and to establish whether the cancer consisted of a single solitary lesion or multiple lesions due to lack of modern imaging modalities, such as magnetic resonance imaging or even computed tomography, which are readily used today to assist in diagnosis.¹⁵ Although Cushing intended to remove the lesion in the case of patient 4, he could not visualize the lesion during surgery. In fact, the only patients in whom he was able to remove the tumor were patients 1 and 2. Furthermore, successful removal hinged on the presence of a solitary lesion, and surgeons of Cushing's era erroneously predicted this to be the case twice as often as they should have.¹⁶

The hasty demise of these 4 patients of Cushing likely tempered his desire to take on such cases; conversely, their temporary relief of symptoms increased his willingness to do so when faced with such a patient. Patient 2, for example, experienced dramatic relief and improvement immediately after her surgery, only to be followed by a heart-wrenching decline approximately 2 months later. After leaving The Johns Hopkins Hospital, Cushing continued to operate on patients with metastatic breast carcinoma, although he did not take on many of these patients. In 1932, Cushing acknowledged that the 3.2% of his cases at Brigham Hospital that were brain metastases were not representative of the true incidence of these lesions “for we refrain when possible from accepting patients with obvious intracranial metastases since so little can be done for them by surgical procedures.”^{10 (p105)} Nevertheless, Cushing and others reported a few cases in which patients survived more than a few months after the removal of a metastatic lesion.¹⁰ A 1926 study published in The Annals of Surgery showed that surgery on patients with cerebral metastases did not increase their survival.^{10 ,17} Regarding these findings, Cushing commented:

Operations nevertheless may not infrequently afford a vast degree of symptomatic relief for which patients and their relatives are most grateful. Hence, when the unfortunate victims of these disorders once come to be admitted to the hospital wards, it is difficult to refuse their appeals to give them at least the chance of temporary palliation of symptoms which a decompression may afford.^{10 (p105)}

Note that this opinion stood in stark contrast to beliefs expressed by some individuals of the era. In 1926, for example, Shelden advised that surgical procedures for cerebral metastases never be undertaken.¹⁸

The reluctance of Cushing and others to take on these patients did provide a degree of difficulty in arriving at good statistics regarding the frequency of breast cancer metastases because various hospitals took on different percentages of these cases. Nevertheless, owing to their willingness to examine these patients throughout the 1900s, Cushing and other surgeons began to realize and put forth the knowledge that the breast is a common source of cerebral metastasis.^{10 ,19 - 23} As evidence mounted that decreased intracranial pressure afforded some symptomatic relief, further studies appeared on the ability of corticosteroids to also counter cerebral swelling and decrease symptoms in instances of cerebral metastases, for example, French and Galicich's use of dexamethasone or Kofman's study on prednisone.¹⁶

Ironically, with advancements in medicine, CNS metastases are becoming a greater problem today due to the increasing longevity of breast and other cancer survivors.²⁴ Surgery has become a cornerstone of treatment for many patients; present-day analysis has demonstrated that aggressive surgical management of spinal metastases from breast cancer can preserve or improve neurologic functioning and relieve pain with acceptably low complication rates.²⁵ Additional data suggest that resections of brain metastasis remain safe up to the ninth decade of life.²⁶ Furthermore, the evolution of technology has allowed for the emergence of additional therapies, such as whole-brain radiotherapy, stereotactic radiosurgery, and external beam irradiation, which were not imaginable during Cushing's time.^{27 - 29} As numerous therapeutic options emerge, the review and evaluation of the ongoing research on these treatments have become critical.³⁰ The American Association of Neurological Surgeons, the Congress of Neurological Surgeons, and the American Association of Neurological Surgeons/Congress of Neurological Surgeons Joint Tumor Section recently funded an initiative aimed at developing and managing evidence-based clinical practice parameter guidelines for brain metastases.^{30 - 31} For single brain metastases, the evidence supports surgical resection combined with postoperative whole-brain radiotherapy in patients with good performance status and limited extracranial disease.²⁹ Nevertheless, patients with metastatic breast cancer still are without a cure for their condition.³² Our goals of treatment remain the same as those of Cushing when he first attempted surgery on his patients: our therapies aim for palliation and preservation of neurologic functioning.¹⁵