The Role of Surgery Following Chemotherapy in Stage III Germ Cell Neoplasms

0022-5347 /83/1291-0039$02.00/0 Vol. 129, January

THE JOURNAL OF UROLOGY

Copyright© 1983 by The Williams & Wilkins Co.

Printed in U.S.A.

THE ROLE OF SURGERY FOLLOWING CHEMOTHERAPY IN STAGE III GERM CELL NEOPLASMS R. BRUCE BRACKEN, DOUGLAS E. JOHNSON, 0. HOWARD FRAZIER, CHRISTOPHER J. LOGOTHETIS, ANTONIO TRINDADE AND MELVIN L. SAMUELS* From the Departments of Urology, Thoracic Surgery and Medicine, University of Texas M. D. Anderson Hoepital and Tumor Institute, Houston, Texas

ABSTRACT

After curative chemotherapy 58 patients with metastatic germinal tumors and 2 with extragonadal germinal tumors underwent an operation either to confirm a clinical complete response or to remove a residual mass. Biomarker status was converted to negative in all patients. Retroperitoneal lymphadenectomy was done in 22 patients after a complete clinical response was achieved and in 23 after an apparent complete response was achieved except for a residual retroperitoneal mass. Thoracotomy was done for a residual mass in 13 patients, while 2 underwent bilateral thoracotomy. Scarring was found in 25 patients (42 per cent), teratoma in 14 (23 per cent) and active tumor in 17 (28 per cent). One patient with scarring and primary seminoma died of disseminated seminoma, while 1 with primary teratocarcinoma died of disseminated disease, for a false negative rate of 3 per cent. Four patients with a residual mass showed unexpected findings: 2 had granulomas in the lung, 1 had carcinoid of the lung and 1 had retroperitoneal neurofibroma. Thus, there was a 7 per cent incidence of a mass being other than scar, teratoma or residual tumor. There was 1 operative death and 10 deaths of tumor, for an 18 per cent death rate. Of the 17 patients with active tumor 9 responded to salvage chemotherapy with stable complete remission and a minimum followup of 18 months. In addition to the 60 patients the advanced disease in 10 was surgically debulked after failure of chemotherapy and this procedure was followed by salvage chemotherapy. All 10 patients died, with a median survival of 7 months. The lack of responsiveness to chemotherapy was not improved by incomplete removal of tumor. After the introduction of vinblastine and bleomycin in 1970, 1- 3 and the further addition of cis-platinum in 19744 •5 better methods were necessary to assess the completeness of response in patients with advanced testicular cancer. Initial reports of surgery following chemotherapy demonstrated that chemotherapy could eliminate cancer from retroperitoneal lymph nodes. 6- 8 When computerized tomography (CT) 9 and radioimmunoassay for a-fetoprotein (AFP) and human chorionic gonadotropin (HCG) became reliable 10 these noninvasive methods diminished the need to confirm surgically a drug-induced complete response. Presently, we have been able to limit surgery to patients who have a near complete response except for a much reduced mass, as determined by CT and sonography, that persists despite adequate chemotherapy. In addition, 10 patients underwent a cytoreductive operation in an attempt to control metastatic neoplasm that was poorly responsive to drug therapy. The results were poor uniformly. Because the indications for an operation following chemotherapy gradually have changed during the last decade 11 - 14 we herein review our surgical results.

surgically by retroperitoneal lymphadenectomy to confirm the ability of drug treatment to eliminate cancer from retroperitoneal lymph nodes. An additional 38 patients who had fulfilled all the criteria for a complete response except for having a residual mass were treated by an operation. Retroperitoneal lymphadenectomy was done in 23 patients for a residual retroperitoneal mass, and thoracotomy and wedge resection(s), segmental pulmonary resection or lobectomy (whichever was most appropriate) was done in 15 for residual pulmonary nodular lesions. In addition, 10 selected patients who had obviously active tumor were treated surgically when it was evident that unassisted chemotherapy would not result in cure. It was hoped that by debulking unresponsive tumor the refractory nature of chemotherapy would be altered. The masses were intrathoracic in 6 patients ~md intra-abdominal in the remaining 4. Patient age ranged from 15 to 51 years, with a median of 23 years. Disease presentation was classified according to the following criteria modified partly from that of Samuels and associates: 3 stage IHA-disease limited to supraclavicular nodes only; stage IIIB-1-~l elevated biomarkers and/or unilateral or bilateral gynecomastia with no demonstrable mass; stage IIIB-2-minimal pulmonary disease with up to 5 nodules in each lung field, the largest diameter of any single lesion being not >2 cm. and the total tumor volume being not >40 cm. 3 (based on each nodule being spherical and calculated by the formula for the volume of a sphere); stage IIIB-3-advanced pulmonary disease, including a mediastinal masi;,, neoplastic pleural effusion, superior vena caval syndrome or >5 nodules in each lung field; stage IIIB-4-advanced abdominal disease, including a palpable abdominal mass, ureteral deviation, ureteral obstruction with hydronephrosis and inferior vena caval deviation (minimal abdominal disease is stage II), and stage IIIB-5-visceral disease, except lung, including primarily the brain and liver, and invasion of the gastrointestinal tract, skin or bone.

MATERIALS AND METHODS

Between November 1, 1971 and December 31, 1979, 60 patients with metastatic germinal tumors, 58 of testicular origin and 2 of extrago;nadal origin, underwent an operation after curative chemotherapy. The population from which these patients were derived consisted of 318 patients with stage III disease treated duri;ng the same interval. Before 1977, when the reliability of the CT scan and radioimmunoassays for AFP and HCG was established at our hospital, the disease in 22 patients who had responded completely to chemotherapy was staged Accepted for publication April 23, 1982. * Requests for reprints: Department of Medicine, Room W-725, University of Texas M, D. Anderson Hospital and Tumor Institute, 6723 Bertner Ave., Houston, Texas 77030. 39

40

BRACKEN AND ASSOCIATES

Before referral 21 patients had undergone unsuccessful retroperitoneal lymphadenectomy, which was abandoned owing to massive retroperitoneal disease, 4 had had prior abdominal radiation therapy, and 2 had had prior attempted retroperitoneal lymphadenectomy and abdominal radiation therapy. Preoperatively, the patients received between 3 and 11 (median 5) courses of chemotherapy, which was given according to previously reported regimens. 1- 3 When a residual mass demonstrated stability to chemotherapy an operation was considered. All patients were evaluated with a complete history and physical examination, complete blood count, SMA-12, AFP and HCG determinations (after 1976 these were done by radioimmunoassay), and chest x-ray with pulmonary tomography, if necessary. Before 1976 patients underwent excretory urography, inferior venacavography, pedal lymphangiography and retroperitoneal sonography to evaluate the retroperitoneum. After 1976 CT was added in combination with ~1 of the aforementioned studies to examine the retroperitoneum and liver. An operation was performed when the results of these studies confirmed the clinical impression of either a complete response or a complete response except for a stable mass. We wish to emphasize that all patients had normal values for AFP and HCG preoperatively. In addition to the aforementioned study population 10 patients underwent a debulking operation in the hope that it would help control the active malignant process. Of these 10 patients 6 had ~1 markers elevated preoperatively. RESULTS

The results of retroperitoneal lymphadenectomy in the 22 patients with clinical complete response are shown in table 1. One must remember that the CT scan was not yet available to evaluate this group. In 14 patients only scar tissue remained and 13 have no evidence of disease 24 to 105 months (median 60 months) postoperatively. The remaining patient with stage III seminoma had diffuse retroperitoneal scarring, which could be biopsied but not removed totally. This patient died 4 months later of disseminated seminoma. Three patients with small

TABLE

areas of teratoma have no evidence of disease for up to 117 months postoperatively. The 5 patients with residual tumor were treated with aggressive postoperative chemotherapy: 3 were salvaged with an operation plus postoperative chemotherapy while 2 died of disseminated tumor. Thus, before the CT scan was used 5 of 22 patients were understaged after chemotherapy but only 2 died. When these results were reviewed we believed that 3 patients presented with masses large enough to be visualized by the CT scan but 2 had small masses (<2 cm.) that probably would have been missed. This would give a staging error of 9 per cent if all available procedures short of an operation were done. The results of retroperitoneal lymphadenectomy in the 23 patients with a residual retroperitoneal mass are shown in table 2. Of the 8 patients whose mass consisted of scar tissue 7 have no evidence of disease 30 to 41 months (median 33 months) postoperatively. Only 1 patient died of septic aortitis 2 months after multiple surgical procedures. This patient was free of tumor at postmortem examination. The 7 patients with teratoma are free of disease 18 to 74 months (median 30 months) postoperatively. One patient with seminoma is free of cancer 17 months after excision of a second primary tumor (retroperitoneal neurofibroma). Of the 7 patients with viable tumor present within the retroperitoneal mass 4 have died of cancer 7 to 30 months (median 12 months) postoperatively despite chemotherapy. The residual mass was unresectable in 2 of these patients. The remaining 3 patients with persistent viable tumor received further drug treatment and presently are free of disease and off all therapy 34 to 54 months postoperatively. The results achieved by excision of residual pulmonary lesions are illustrated in table 3. Of the patients 9 had solitary and 6 had multiple lesions. The largest lesion measured 4.5 cm. in diameter and the remaining lesions ranged between 0.5 and 2.5 cm. in diameter. Three patients died of disease. The remaining patients are free of disease and off all therapy 18 to 91 months (median 48 months) postoperatively. It is of interest that 1 patient with teratoma died of teratocarcinoma 28 months after thoracotomy.

1. Stage at presentation and primary histology compared to retroperitoneal findings in 22 patients with apparent complete response Total No. Scarring Teratoma Embryonal Ca Teratoca. Pts. Primary histology 1 16 _E 22

Seminoma Embryonal Ca Teratoca. Total No. (%)t

Stage at presentation 1 20 1

IIIB-2, mimmal lung disease with pos. lymphangiogram IIIB-4, advanced abdominal disease (No. died)t IIIA, supraclavicular node disease with pos. lymphangiogram

1* 11

2 14 (64)

13 (1) 1

1 2 3 (14)

3

l* 1*

3 (14)

2 (9)

3

1 2

2 (2)

* Died of tumor (14 per cent). Note active tumor was found in 23 per cent of the group.

t 9 patients had lung metastases. TABLE

2. Stage at presentation and primary histology compared to surgical findings in 23 patients with residual retroperitoneal mass Total No. Pts.

Scarring

Teratoma

Embryonal Ca

Teratoca.

Other

Primary histology Seminoma Embryonal Ca Teratoma Teratoca. Chorioca. Total No. (%)

4 9

3 3

2

It

7

1 23

1* 3 1 3

2

1

2

2

7 (30)

4 (17):j:

3

2

2 (2) 2 (2) 4

1 2 3

1 8 (35)

Stage at presentation IIIB-4, advanced abdominal disease (No. died) IIIB-4, advanced abdominal and lung diseaset Totals * Retroperitoneal neurofibroma, no malignancy found.

t Treatment death :j: Died of tumor.

·

10 (l)t _13__ 23

4

±

.I!.

8

7

1* 1

41

SURGERY IN STAGE III GERM CELL NEOPLASMS

In addition to the aforementioned population a cytoreductive operation was performed on 10 patients with advanced disease and either minimal or no response to chemotherapy (table 4). An operation did not appear to alter favorably the disease process in these patients and all died, with a median survival of 7 months postoperatively. Of these 10 patients 6 had elevated biomarkers preoperatively, compared to O of the 60 patients studied (p = 0.01). Postoperative complications occurred in 4 patients (6 per cent): 1 had a pneumothorax that persisted for 6 weeks after thoracotomy, 2 had arterial injuries (1 aortic and 1 common iliac) that defied primary repair and required bypass graft procedures to restore blood flow satisfactorily, and 1 died of septic aortitis discovered 1 month after the patient was discharged from the hospital, which proved refractory to 7 attempts at surgical correction. In the remaining patients convalescence was no worse than would be expected for patients who 3. Pulmonary findings compared to primary histology and presentation in 15 patients with residual pulmonary lesions

TABLE

Total No. Pts.

Scarring

Teratoma (No. died)

Embryonal Ca (No. died)

2 1

4 (1)

3

4

2t 2

3

0.70 0.60

TOTAL

FAIL

22

3

0

23 15 10

5

6

3

•

C

-~ 0

Q.

e 0..

0.50

10

0

0.40

Apparent Complete Remission Residual Abdominal Mass Residual Pulmonary Mass Surgical Debulking Non-failure

0.30 0.20

0.001

.!1 15

0.80

Other

3 (2)*

5 8

0.90

0.10 Teratoca.

Primary histology Embryonal Ca Teratoca. Chorioca. Seminoma Totals

Survival Following Surgery 1.00

2

3

0

12

24

36

48

60

72

84

96

108

120

Months

There is no significant difference in survival among apparent complete remissions, residual abdominal mass and residual pulmonary mass groups but all 3 are significantly better than surgical debulking group (p <0.001).

Stage at presentation IIIB-2, minimal lung disease IIIB-3, advanced lung disease!! Totals

6

1

1 (l)*

9

]_

3 (1)

~

15

3

4

3

3§

_!_ 2

3

Note that active tumor was found in 5 patients: 2 died of tumor and a third died of teratoma. * 1 patient died of endodermal sinus tumor. t A granuloma secondary to histoplasmosis was found in 1 patient and a carcinoid tumor was found in 1. t Granuloma of lung, etiology undetermined. § 2 patients died of granuloma and 1 of carcinoid of the lung. II Bilateral thoracotomy was done in 2 patients with advanced lung disease. Both patients are free of disease at 52 and 53 months, respectively. Teratoma was recovered in 1 patient and teratocarcinoma was recovered from both lungs in 1.

TABLE

4. Patients whose tumor was surgically debulked after poor

response to chemotherapy Pt. No.

Extent of Disease

Biomarker Elevated

Cell Type

Died Mos. Postop.

Retroperitoneal lymphadenectomy

~~

\ i

18

None

Teratoca. and endodermal sinus tumor Embryonal Ca

AFP

Embryonal Ca

7

AFP

Teratoca.

9

None

Embryonal Ca

7

None

Terato ca.

13

None

Teratoca.

9

HCG

8

AFP

9 nodules scar, 1 chorioca. Teratoca.

AFP

Terato ca.

6

1

Large retroperitoneal mass

AFP

2

Retroperitoneal mass and liver metastases Large retroperitoneal mass Retroperitoneal mass and liver metastases* Huge retroperitoneal mass Large abdominal mass

3 4

5 6

7

Thoracotomy 7 8

9 10

Large unresectable intrapulmonary mass Multiple pulmonary nodules 9X9X7cm. intrapulmonary mass Large lung mass requiring pneumonectomy

7

* Partial hepatectomy performed with retroperitoneal lymphadenectomy.

had not received prior chemotherapy. No intraoperative or postoperative pulmonary problems occurred despite the fact that patients received an average total dose of 750 mg. bleomycin preoperatively. Survival curves are given for the 4 groups (see figure). There is no significant difference among the 22 patients with a complete response, the 23 with a residual abdominal mass and the 15 with pulmonary disease. However, survival in each of these groups was significantly better than the group that underwent cytoreductive surgery (p <0.001). DISCUSSION

Of the 60 patients 49 (82 per cent) are free of disease after curative chemotherapy had induced a complete or near complete response except for a residual mass. No maintenance chemotherapy has been administered. Deaths were related to tumor in 10 patients and postoperative in 1. Of the 60 patients 17 (28 per cent) had active cancer and 9 (53 per cent) were salvaged. In addition, 2 patients with benign surgical findings (that is teratoma in 1 and scarring in 1) died of tumor. In 1 patient with a retroperitoneal mass and 2 with residual pulmonary lesions an operation alone probably was curative. In these individuals small areas of active disease were present deep within a mass of necrotic tissue and no salvage chemotherapy was given. In the remaining 6 patients surgery contributed to cure by excising cancer that remained after drug treatment and, more importantly, by indicating to the medical oncologist that additional chemotherapy was required. The finding of scar tissue or mature teratoma usually but not always solidifies the clinical impression of a complete response. However, 1 patient with seminoma and only scarring in the retroperitoneum died 4 months postoperatively, for a false negative scar rate of 1 in 25 (4 per cent). Also, 1 patient with teratoma in the lung died of teratocarcinoma, for a false negative failure rate of 1 in 14 (7 per cent). Thus, because these figures are low it is our policy to give no immediate further chemotherapy but to observe closely patients with teratoma and scarring via chest roentgenograms and biomarkers for an additional 2 years. Surgical debulking proved to be useless in 10 patients who responded poorly or not at all to initial chemotherapy. A

42

BRACKEN AND ASSOCIATES

minimum of 3 courses of chemotherapy (median 5) was administered preoperatively. Median survival postoperatively was 7 months. Therefore, we believe that a cytoreductive operation does not benefit patients with grossly active disease. Surgery after chemotherapy is difficult and potentially hazardous. However, apart from the need to bypass arterial injuries that do not lend themselves to standard primary repair, the consequences of an operation are no worse than in patients who had not had prior drug treatment. Goldiner and Schweizer reported on 5 patients with testicular disease in whom an acute respiratory distress syndrome developed 3 to 5 days postoperatively. 15 All 5 patients had received bleomycin 6 to 12 months preoperatively and the mean dose was stated to be 426 mg. Unfortunately, all 5 patients died despite ventilator support with positive end-expiratory pressure. We are at a loss to explain this complication. Our mean dose of bleomycin was 750 mg. and we have never experienced the adult respiratory distress syndrome postoperatively. We monitor bleomycin administration carefully with serial spirometry, blood gases, chest roentgenograms and gallium scan. Our incidence of bleomycin in the lungs is low (<2 per cent) and usually is diagnosed early, before the stage of florid symptomatology and chest roentgenographic findings. There were no cases of bleomycin in the lung in this group. We are optimistic that the CT scan will be reliable in the evaluation of the retroperitoneum. However, we must emphasize that the retroperitoneum is not normal in patients who have been treated successfully by chemotherapy for retroperitoneal malignant disease. In these patients the retroperitoneum contains large areas of necrosis and scar tissue that are the results of tumor eradication. However, the present sensitivity of CT scanning does not recognize these abnormalities unless a frank mass is present and, thus, an operation can be avoided when CT is normal. Since a third of the masses that persist in the retroperitoneum contain cancer and a third contain adult teratoma, finding a mass by CT after curative chemotherapy constitutes a reason for an operation. Presently, we believe that an operation probably should be avoided when curative chemotherapy produces a clinical complete response (negative biomarkers, lymphangiogram, sonogram, CT scan and tomograms of the lungs) and that it has no place when either tumor is active clinically and poorly responsive to chemotherapy or when biologic markers are elevated significantly. We recognize that the latter point is controversial, since some investigators may perform an operation with biomarker elevation. Donohue and associates recently have reported results in 26 patients with advanced disease who were treated initially with 4 courses of vinblastine, bleomycin and cis-platinum. 14 All patients had a persistent abdominal mass after chemotherapy. There were 9 patients (35 per cent) with active tumor found at operation: 3 remained in a continuous state free of disease and 6 had relapse. Of the patients with relapse 1 was salvaged with additional chemotherapy. However, the over-all salvage rate for the entire group was 80 per cent. These investigators emphasized that one cannot rely on biomarkers to indicate the presence of active disease after chemotherapy. Of their 9 patients with residual malignancy 5 were seropositive and 4 were seronegative. Vugrin and associates interposed an operation at an earlier time than most groups, that is after 2 or 3 induction courses of chemotherapy. 16 Their criteria for an operation are 1) sustained but incomplete regression to chemotherapy, 2) suspicion of a residual mass and 3) bulky disease at presentation, even if a complete remission was obtained by chemotherapy. Of their 48 patients 22 (46 per cent) showed active malignancy that could be removed completely in 11. Their protocol called for 2 additional induction courses of chemotherapy should malignancy be found. Their program is too early to evaluate as yet (median 13 months, range 5 to 33 months) but the high number of malignancies encountered suggests that additional preoperative induction courses are

needed. Of further interest was that 1 of 8 patients (12.5 per cent) with benign teratomas died of recurrent malignancy. The protocol of Hendry and associates again is another interesting variation on the theme. 17 Like Donohue and associates, 14 these investigators required a minimum of 4 courses of chemotherapy but if the disease was massive they increased the number to 6 and then operated to remove residual foci. If the disease was bulky but responded partially to 4 courses and then appeared stable, radiotherapy to the para-aortic nodes was introduced, followed by an operation for residual disease. Of 20 patients radiated 4 (20 per cent) had residual active malignancy compared to 5 of 21 (24 per cent) nonradiated patients. Thus, it appeared that radiotherapy added little to the over-all result. These investigators stressed the difficulty in producing salvage chemotherapy remissions in the malignant group, since only 2 of their patients responded favorably to further chemotherapy efforts. They also noted 3 failures among 31 patients (10 per cent) in the necrosis, fibrosis and teratoma groups. Preoperatively, they now prefer biomarkers to be normal and tumor masses to be stable after initial response. Responses from patients must be individualized on the basis of histology and volume of tumor at presentation, 18 and aggressive chemotherapy completed preoperatively is contemplated. It is clear that an operation is in no way a substitute for chemotherapy. Javadpour and associates recently have completed a randomized trial of chemotherapy versus surgical debulking followed by chemotherapy in patients with advanced disease. 19 There were no significant differences in response or survival between the 2 groups. In summary, an operation provides the oncologist with vital information on patients with a residual mass that will enable cessation of chemotherapy when cure probably has been achieved (scar or teratoma) or continuation of chemotherapy when active cancer remains. In our experience only a rare patient with scarring or teratoma will fail. A debulking operation in patients who fail chemotherapy does not improve survival or response to salvage chemotherapy. REFERENCES 1. Samuels, M. L., Holoye, P. Y. and Johnson, D. E.: Bleomycin combination chemotherapy for metastatic testicular carcinoma. Cancer Bull., 25: 53, 1973. 2. Samuels, M. L., Holoye, P. Y. and Johnson, D. E.: Bleomycin combination chemotherapy in the management of testicular neoplasia. Cancer, 36: 318, 1975. 3. Samuels, M. L., Lanzotti, V. J., Holoye, P. Y., Boyle, L. E., Smith, T. L. and Johnson, D. E.: Combination chemotherapy in germinal cell tumors. Cancer Treat. Rev., 3: 185, 1976. 4. Higby, D. J., Wallace, H. J., Jr., Albert, D. and Holland, J. F.: Diamminodichloroplatinum in the chemotherapy of testicular tumors. J. Urol., 112: 100, 1974. 5. Einhorn, L. and Donohue, J. P.: Chemotherapy for disseminated testicular cancer. Urol. Clin. N. Amer., 4: 407, 1977. 6. Donohue, J.P., Perez, J.M. and Einhorn, L. H.: Improved management of non-seminomatous testis tumors. J. Urol., 121: 425, 1979. 7. Comisarow, R.H. and Grabstald, H.: Re-exploration for retroperitoneal lymph node metastases from testis tumors. J. Urol., 115: 569, 1976. 8. Johnson, D. E., Bracken, R. B., Ayala, A. G. and Samuels, M. L.: Retroperitoneal lymphadenectomy as adjunctive therapy in selected cases of advanced testicular carcinoma. J. Urol., 116: 66, 1976. 9. Lee, J. K., McClennan, B. L., Stanley, R. J. and Sagel, S. S.: Computed tomography in the staging of testicular neoplasms. Radiology, 130: 387, 1979. 10. Scardino, P. T., Cox, H. D., Waldmann, T. A., McIntire, K. R., Mittemeyer, B. and Javadpour, N.: The value of serum tumor markers in the staging and prognosis of germ cell tumors of the testis. J. Urol., 118: 994, 1977. 11. Merrin, C. E. and Takita, H.: Cancer reductive surgery: report on the simultaneous excision of abdominal and thoracic metastases from widespread testicular tumors. Cancer, 42: 495, 1978.

s-:-JRGE.R':l Il"J ,3TAGE III GERJ½ CELL f,JEQPLAS.NIS

~:t h1errin, C.

13.

14.

15.

16.

17.

18.

19.

n,ac,uev. and T'akita, II .. J'vlu.ttirnodal treatn1ent of advanced ce:sucum1 tumor with radical reductive surgery and multisequentia! with cis platinum, bleomycin, vinblastine, vincristine and ac1tmorr,ycrn D. J. Urol., 120: 73, 1978. Garnick, M. B., Canellos, G. P., Richie, J. P. and Stark, J. J.: Sequential combination Pnon•m,,rs,rn, and for disseminated testicular cancer: au,rnu111u.u11u110µ1ca v1nblastine, and bleomycin remission-induction followed by cyclophosphamide and adriamycin. Cancer Treat. Rep., 63: 1681, 1979. Donohue, J. P., Einhorn, L. H. and Williams, S. D.: Cytoreductive surgery for metastatic testis cancer: considerations of timing and extent. J. Urol., 123: 876, 1980. Goldiner, P. L. and Schweizer, 0.: The hazards of anesthesia and surgery in bleomycin-treated patients. Semin. Oncol., 6: 121, 1979. Vugrin, D., Whitmore, W. F., Jr., Sogani, P. C., Bains, M., Herr, H. W. and Golbey, R. B.: Combined chemotherapy and surgery in treatment of advanced germ-cell tumors. Cancer, 47: 2228, 1981. Hendry, W. F., Goldstraw, P., Husband, J. E., Barrett, A., McElwain, T. J. and Peckham, M. J.: Elective delayed excision of bulky para-aortic lymph node metastases in advanced nonseminoma germ cell tumours of testis. Brit. J. Urol., 53: 648, 1981. Peckham, M. J., Barrett, A., McElwain, T. J., Hendry, W. F. and Ragahavan, D.: Non-seminoma germ cell tumours (malignant teratoma) of the testis. Results of treatment and an analysis of prognostic factors. Brit. J. Urol., 53: 162, 1981. Javadpour, N., Ozols, R. F., Barlock, A., Anderson, T. and Young, R. C.: A randomized trial of cytoreductive followed by chemotherapy versus chemotherapy alone in stage III (poor prognosis) testicular cancer. Abstract C-549. Proc. Amer. Ass. Cancer Res. Amer. Soc. Clin. Oncol., 22: 473, 1981. 1

EDITORIAL COMMENTS These authors emphasize the role of surgery in assessing the response and in supplementing chemotherapy for patients with advanced germ cell tumors. The 82 per cent apparent cure rate achieved in patients with advanced disease, having surgery after a complete or near complete response to chemotherapy, and with negative tumor markers is more or less analogous to experiences reported from Indiana Medical Center and Memorial Sloan-Kettering Cancer Center, and affirms: 1) the usefulness of surgery in recognizing clinically occult disease and in permitting pathologic assessment of residual masses, 2) the value of

assessn1ent ir,. further che1nothersalvage rate (53 per cent, of 17) in patients to have residual cancer at surgery, and the importance of complete resection and further'"'"""""''"'" and 4) the poor prognosis of patients who have surgical debulking in face of a poor response to initial chemotherapy. The optimal regimen and duration of chemotherapy before consideration of surgery, the natural history of unresected teratoma, the management of patients with persistently elevated markers or evidence of poorly responsive disease after initial chemotherapy and the indications for supplemental surgery in patients obtaining a clinical complete response with chemotherapy are some of the unsettled issues in this area. Willet F. Whitmore Department of Surgery Memorial Sloan-Kettering Cancer Center New York, New York This is an excellent paper that confirms several important points concerning surgical management following chemotherapy for patients with stage III testis cancer. The finding of a mass by CT after curative chemotherapy is reason in itself for surgery because a third of this group will have teratoma (with its continued growth and malignant transformation potential) and a third will have active cancer requiring further chemotherapy. Conversely, those patients achieving a solid complete remission (including negative CT scan) after curative chemotherapy need not have a retroperitoneal lymph node dissection since the subsequent relapse rate is less than 10 per cent and usually not in the abdomen. About a quarter to a third of post-chemotherapy partial remissions with an abdominal mass will have persistent malignancy in the specimen. If this specimen is completely resected salvage chemotherapy can rescue at least half of the patieHts. However, if the specimen is merely biopsied and not resected completely these patients do poorly, having already demonstrated relative resistance to chemotherapy. The remaining patients with excised scar tissue and teratoma rarely suffer relapse later with active disease and need no additional chemotherapy but close followup is still recommended. Surgical debulking is useless in nonresponsive tumor patients with evidence of active disease preoperatively. This group's historic contributions to the field are well exemplified by this fine work in this series of difficult cases. John P. Donohue Department of Indiana University Center Indianapolis, Indiana