Herpes zoster in patients with carcinoma of the lung

Herpes Zoster in Patients with Carcinoma of the Lung

RONALD

FELD,

M.D.,

F.R.C.P.(C),

F.A.C.P. WILLIAM GERRIT

K. EVANS, DeBOER,

M.D.

BSc.,

F.R.C.P.(C) MSc.,

Ph.D.

Toronto, Ontario, Canada

From the Department of Medicine, University of Toronto, and the Department of Biostatistics, The Ontario Cancer Institute (incorporating The Princess Margaret Hospital), Toronto, Ontario, Canada. This work was supported by a grant from the Ontario Cancer Treatment and Research Foundation (Project Number 360) and by a National Cancer Institute of Canada Clinical Trials Scholarship. Requests for reprints should be addressed to Dr. Ronald Feld, oepartment of Medicine, The Ontario Cancer Institute, 500 Sherbourrne Street, Toronto, Ontario, Canada M4X lK9. Mlanuscript accepted April 5, 1982.

Herpes zoster was observed in only four of 250 (1.6 percent) patients with small cell carcinoma of the lung, who were treated in a prospective, combined modality therapy trial. Induction chemotherapy in this study consisted of six courses of cyclophosphamide, doxorubicin, and vlncristine (CAV), followed by intrathoracic and cranial irradiation. Those with extensive disease also received single doses of upper half-body irradiation. Patients did not receive maintenance chemotherapy (CAV2 protocol). This contrasted with our previous study (CAV, protocol), which consisted of three courses of the same induction chemotherapy, the same intrathoracic irradiation, but with one year of oral maintenance chemotherapy. During the CAV, regimen, we observed that herpes zoster developed in 13 of 161(6.1 percent) patients in association with their therapy. A retrospective analysis of 6,576 patients with lung cancer revealed that herpes zoster developed in 56 (0.9 percent). This complication developed in 10 of 622 (1.6 percent) patients with small cell carcinoma of the lung, as compared to 46 of 5,954 (0.6 percent) patients with nonsmall ceil carcinoma of the lung. The rlsk of development of herpes zoster in the CAV, group was significantly greater than the historical group (p = 0.007) and was also greater than the CAV2 group (p = 0.031). However, there was no significant difference between the historical group and the CAVz group. Attempts to explain the differences in the rate of herpes zoster in our three studies and those in the literature suggest that the duration of therapy, the type of chemotherapy used, and the Improving survival rate may be important contributing factors to this complication in patients aggressively treated for small cell carcinoma of the lung. The literature and our own studies suggest that procarbazine is the most likely chemotherapeutic agent predisposing to this complication. Herpes zoster is seen most commonly in children and adults with hematologic malignancies, particularly Hodgkin’s disease [l-l 11. Patients undergoing renal transplantation also have an increased incidence of this infection [ 121. Although herpes zoster occurs in 4 to 25 percent of patients with Hodgkin’s disease [9], it has been reported as a complication in only 0.5 to 1.8 percent of patients with solid tumors [2,4] and only 0.2 percent of patients without malignancy

[21.

There is very little information on the frequency of herpes zoster in patients with lung cancer. Wright and Win& [2]noted that herpes zoster developed in six of 738 (0.8 percent) lung cancer patients, but in two patients herpes zoster developed before the diagnosis of malignancy. Schimpff et al. [4] noted that one of four patients with solid

December 1982

The American Journal of Medicine

Volume 73

795

HERPES ZOSTER IN LUNG CANCER PATIENTS-FELD

ET AL.

tumors in whom herpes zoster developed had carcinoma of the lung. Small cell carcinoma accounts for approximately 20 percent of all cases of lung cancer but has recently received much attention because of encouraging results of treatment with combination chemotherapy and radiotherapy [ 13-151. We have recently reported our experience with a group of 161 patients with small cell carcinoma of the lung treated with combination chemotherapy and radiotherapy [ 161. During this clinical trial, an unexpectedly high frequency of herpes zoster was observed [ 171. This observation led us to review retrospectively our prior experience with this infection in patients with all types of lung cancer and to evaluate it further in a more recent, prospective, combined modality therapy trial in small cell carcinoma of the lung [ 181. The results of these analyses are the subject of this report. PATIENTS AND METHODS

A retrospective review was undertaken to identify cases of herpes zoster in lung cancer patients seen at The Princess Margaret Hospital between 1962 and 1975. A computer printout was generated from the activity file of all patients with diagnoses of herpes zoster and lung cancer, and the lung cancers were further categorized by histologic cell type. We recognize that this methodology may underestimate the number of cases of herpes zoster, and the accuracy of the data is dependent both on the physician making the correct diagnosis or eliciting a history of herpes zoster and recording it in the chart. Nonetheless, this represents the only method of identifying the lung cancer population with herpes zoster prior to 1976. The data pertaining to patients with small cell carcinoma of the lung collected from the retrospective study were compared with that collected prospectively in two combined modality studies. During these prospective studies, all patients in whom herpes zoster developed after treatment were identified and documented in the following fashion. The diagnosis of herpes zoster was made on clinical grounds by the appearance of vesicular lesions in a sensory nerve root distribution. Viral studies were not routinely carried out. The herpes zoster was classified as localized (lesions within three dermatomes) or generalized (lesions disseminated over the skin or involving other organs). The data recorded on these patients included the site and extent of infection, temporal relationship to cancer therapy, and the relationship to radiotherapy fields. The outcome and sequelae of the infection were noted, as was the status of the malignancy at the time of the infection. Whine blood cell counts and differentials were recorded. In addition, the important characteristics of the patients relative to their malignancy were also noted. These included the extent of their disease, age, sex, performance status, response to therapy, and survival from the initiation of treatment. Most lung cancer patients in the retrospective group, including those with small cell carcinoma of the lung who were treated at our institution prior to 1976, were treated with ra-

796

December

1962

The American

Journal

of Medicine

diotherapy alone. A few received singlsagent chemotherapy with cyclophosphamide and a few received combination chemotherapy. During the prospective studies, patients were carefully staged using standard clinical methods including radioisotope scans of brain, bone, and liver, and bone marrow aspirates and biopsies. Patients were considered to have limited disease if disease was confined to a hemithorax and ipsilateral supraclavicular node. Extensive disease was defined as any spread beyond the limits defined for limited disease. Performance status was graded according to the Eastern Cooperative Oncology Group scale. In the first combined modality study (CAV,), the doses of the induction chemotherapeutic agents were as follows: cyclophosphamide 750 mg/m2, doxorubicin (Adriamycin’“) 50 mg/m*, and vincristine 2 mg. This induction regimen was repeated every three weeks for three cycles. In the second prospective study (CAVz), the induction was intensified: cyclophosphamide was administered in doses of 900 mg/m2, Adriamycin was given in doses of 45 mg/m*, and the vincristine dosage remained 2 mg. Also, the duration of induction therapy was extended from three to six cycles. Radiotherapy to the volume of pretreatment intrathoracic disease (2,500 rads in 10 treatments over two weeks) was given to all responding patients in the CAV, protocol. These patients were then treated with oral maintenance chemotherapy consisting of lomustine (CCNU) 50 mg/m* every six weeks, procarbazine 100 n-g/m* daily for 10 days every three weeks, and methotrexate 10 mg/m* twice weekly. This regimen was continued for one year or until relapse, when reinduction with the initial chemotherapy combination (cyclophosphamide, doxorubicin, and vincristine) was attempted. In the GAV2 protocol, radiation treatment to the primary site was the same but, in addition, all responding patients had prophylactic cranial irradiation (2,000 rads in five treatments) and patients with extensive disease had upper half-body irradiation. The single dose ranged from 300 to 600 rads. Maintenance chemotherapy was not used. If relapse occurred, reinduction was usually attempted with the combination therapy of cyclophosphamide, doxorubicin, and vincristine. Statistics. Survival curves were calculated by the actuarial or life table method [ 191. The cumulative probability of the development of herpes zoster was calculated as the complement of herpes zoster-free rate. The latter was calculated by the method of Kaplan and Meier [20]. Pair-wise comparisons of survival curves or zoster-free curves were tested for statistical significance with the Wilcoxon-Gehan test ]2ll. RESULTS The characteristics of the 161 patients in the CAV, protocol and the 250 patients in the CAVz protocol, as well as the 622 patients in the historical group, are shown in Table 1. Some of the data about the historical

group is not included because on these old charts the information on some of the patients was missing. There was a higher percentage of patients with limited disease and an increased proportion of women in

Volume 73

HERPES ZOSTER

TABLE I

Characteristics of Patients with Small Ceil Carcinoma of the Lung Historical Group (1962-1975)

Characteristic Total no. of patients No. of patients in whom herpes zoster developed Limited/extensive disease Male/female Age (yr) Median Range Previous treatment Performance status at start of treatment (ECOG scale) Grade 0 Grade 1 Grade 2 Grade 3

CAV, Group

CAVz Group

622 10 (1.6%)

161 13 (8.1%)

250 4 (1.6%)

... 4881134

69192 126133

119/131 171179

59 22-84 ?

59 34-8 1 0

60 32-81 0

26 115 15 5

18 195 28 9

... ...

... ...

ECCG = Eastern Cooperative

I

,

I

CAV, Grow 55of65(85%) 53 of 82 (65%) 108 of 147 (73%)

CAV, Group 83 of 105 (79%) 81 of 120 (66%) 164 of 225 (73%)

ET AL.

I

40

-

30

-

20

-

r

“\’ i

‘\’

?

\

‘1, k,

\

K:..

_...-, ‘: .___ ‘L. \ ‘,\

4-

‘\. ‘w

3.

‘\

‘\

Oncology Group.

Comparisonof Response Rates* in CAV, and CAV2 Groups:Response (CR and PR) after Three Courses of Cyciophosphamide, Doxorubicin, and Vincristine

Limited disease Extensive disease Overall

PATIENTS-FELO


2

TABLE ii

IN LUNG CANCER

‘-._

t

,y

4

3

2

0

.0171

Yeors

Figure 1. Actuarial survival curves for the historical group and the two prospective studies (CA V 1 and CA V 2). The number of patients still at risk at half-yearly intervals were as follows: 0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

622 161 250

271 114 179

117 55 70

61 24 27

33 14 9

19 11 1

14 6 0

9 5 0

9 3 0

The response rates only include the patients evaluable for response, although the inevaluable patients are included in other parts of the analysis. CR = complete response; PR = partial response.

Hist. CAVI CAV2

the CAV2 group as compared to the CAV, group. The age range was almost identical in all three groups. The distribution of patients by performance status was also similar in the CAVt and CAVs groups, with the majority of patients completely ambulatory and only mildly symptomatic. Using standard response criteria, a comparison of response rates in the two prospective studies, after three courses of cyciophosphamide, doxorubicin, and vincristine, revealed no significant differences. Table II shows that 81 percent of patients with limited disease and 66 percent of patients with extensive disease showed evidence of tumor regression (complete or partial remission). A detailed analysis of survival, with corrections for ail prognostic factors, using the iogrank method [22], revealed no significant differences between either group or any of the subgroups in the two

studies. Actuarial survival curves are shown in Figure 1 for the historical group and the two prospective studies. Among the 13 patients in the CAVt group in whom herpes zoster developed, we observed an increased proportion of patients with high Eastern Cooperative Oncology Group performance status, limited disease, and complete response-ail good prognostic factors for improved survival. As expected, the median survival time of this group was somewhat longer than that of the entire group (56 versus 46 weeks). This increased survival would of course result in increased time at risk for herpes zoster to develop in this subgroup of patients. Herpes zoster has developed in only four patients on the CAVs regimen, and therefore a detailed analysis of prognostic factors is not possible. Ten patients in the CAV, group had lesions within

l

December

Hist. = historical group.

1962

The American

Journal of Medicine

Volume 73

797

HERPES ZOSTER IN LUNG CANCER PATIENTS-FELD

r

I

r

1

8 I , I

r_______l

I

0

ET AL

I

----CAVI

- -- Pre-trial -CAV2

2

p=o.o31

p=o.700

Figure 2. Cumulative probability of the development of herpes zoster plotted versus time for the historical group (Pretrial) and the two prospective studies (CA V 1 and CA V 2). Number of patients at risk are similar to those in Figure 7.

3

Years

three dermatomes, while three patients had disseminated lesions in the skin only, with no major organ involvement. Initial vesicles were within the thoracic dermatomes in 11 patients and within the cervical and lumbar dermatomes in one patient each. Lesions were within the radiation field in seven patients. Six patients were in remission at the time that herpes zoster developed, while seven had relapsed. The infection developed in all patients from five to 25 months following the initiation of treatment. This infection developed in 11 patients while they were being treated with the oral chemotherapeutic combination of lomustine (CCNU), procarbazine, and methotrexate. Herpes zoster developed in the remaining two patients during reinduction with cyclophosphamide, doxorubicin, and vincristine, after relapse during oral chemotherapy. Although the number of patients in whom herpes zoster developed on the CAV;! regimen is small, three had localized lesions and one had disseminated lesions. The dermatomes involved included two thoracic and two cervical. One patient had lesions within the radiation field and two patients were in remission at the time that herpes zoster developed. Infection developed in the patients from two months to seven months following initiation of therapy. Herpes zoster developed in these patients either during induction therapy with cyclophosphamide, doxorubicin, and vincristine, or when no therapy was given. In the two studies, there were no consistent abnormalities of the total white blood cell count or lymphocyte

799

p=o.o07

December 1982

The American Journal of Medicine

count. Delayed hypersensitivity skin tests and phytohemagglutinin stimulation of lymphocytes were not carried out in either study. All patients recovered from herpes zoster. One patient from each study had postherpetic pain in the involved area. The retrospective analysis revealed that a total of 6,576 patients with lung cancer were seen at The Princess Margaret Hospital from 1962 to 1975. Herpes zoster developed in 58 (0.9 percent) of these patients. This complication developed in 10 of 622 patients (1.6 percent) with small cell carcinoma of the lung, as compared to 48 of 5,954 (0.8 percent) patients with non-small cell carcinoma of the lung. Therefore, the relative risk of this complication developing in patients with small cell carcinoma of the lung was twice that of patients with non-small cell carcinoma of the lung, despite the generally short survival of patients with small cell carcinoma of the lung between 1962 and 1975 (median survival time approximately 26 weeks). This increased risk did not reach a high level of statistical significance (p = 0.087, Fisher exact test, two-tailed value), but there appears to be a suggestion of an association between small cell histom and an increased propensity to herpes zoster infection. The actuarial survival curves for the historical group and the two prospective studies are shown in Figure 1. There was a significant difference in the survival rate in patients in the CAV, and CAV2 groups as compared to the historical group (p <0.0005). There was no difference between the survival rates in the CAV, and

Volume 73

HERPES ZOSTER IN LUNG CANCER PATIENTS-FELD

CA//* groups (p = 0.535). Actuarial zoster- free rates were calculated for the three groups. The cumulative probability of herpes zoster developing (l-zoster-free rate) at various times is shown in Figure 2. Using the Wilcoxon-Gehan test [21], the risk of herpes zoster developing in the CAV, group was significantly greater than in the historical group (p = 0.007) and also greater than in the CAV* group (p = 0.031); however, there was no significant difference between the historical group and the CAV2 group (p = 0.700). COMMENTS

Although we have previously reported a high rate of herpes zoster complicating aggressive treatment of patients with small cell carcinoma of the lung [ 171, this is a rarely mentioned complication in published studies concerned with the treatment of small cell carcinoma of the lung. To our knowledge, this complication has been described in only three other studies. Wittes et al. [23], while treating 17 patients with cyclophosphamide, vincristine, and procarbazine, observed two patients with herpes zoster; in both patients, localized infections developed in dermatomal distribution without dissemination. In one of these patients, postherpetic neuralgia developed. Sorensen et al. [24] treated 24 patients with small cell carcinoma of the lung with a combination of vinblastine, doxorubicin (Adriamycin), and procarbazine. They mentioned one patient who died during remission as a result of widespread herpes zoster. No other details were given. Recently, Huberman et al. [25] noted that in nine of 74 patients (12 percent) with small cell carcinoma of the lung, treated with aggressive chemotherapy, herpes zoster developed at some time during therapy. The chemotherapy used in their study consisted of a six-week, in-hospital induction course with cyclophosphamide, methotrexate, and lomustine (CCNU), followed by vincristine, doxorubicin, and procarbazine. Some patients were also treated with VP-16-213 and ifosfamide. In addition, patients were assigned at random to receive either a placebo or thymosin, 20 n-g/m* or 60 mg/m*, twice weekly for 12 doses [26]. Patients were not treated with thoracic or prophylactic cranial radiotherapy. The characteristics of the patients in Huberman’s series in whom herpes zoster developed are similar to those of our patients on the CAV, regimen. In both studies, herpes zoster occurred late in the course of treatment. Lesions developed in our patients from five to 24 months following initiation of treatment, while in those in Huberman’s study, lesions developed from 22 days to 18 months after initiation of chemotherapy, with a median time of 11 months. The most common site of initial dermatome involvement in both studies was the thoracic region. Both series had patients with disseminated herpes zoster, but all patients recovered from

December

ET AL.

this infection. In five of Huberman’s patienis, severe, postherpetic neuralgia developed, while this complication developed in only one of our patients in the CAV, group. The immunocompetence of the zoster patients was not well evaluated in either series. The mean white blood cell counts at the time of the development of herpes zoster were essentially normal in all but two patients in both studies. Five of nine patients in Huberman’s series had lymphopenia, while seven of our 13 patients had this abnormality. Huberman et al. [25] did perform skin tests with antigens for intermediate strength, purified protein derivative, mumps, Candida, histoplasmin, and streptokinase in 72 of their 74 patients, but these tests were not carried out in our study. Only one of eight patients with herpes zoster in their study failed to respond to any of the antigens employed, whereas 26 of the 64 patients without infection failed to react. All six patients with appropriate samples available had positive varicella-zoster complement fixation titers of 1 in 8 or greater, prior to chemotherapy. More elaborate tests of immunocompetence such as lymphocyte blastogenesis, proportion of B and T cells and T-cell subsets have not yet been studied. Both studies have identified that the patient population in whom herpes zoster is most likely to develop is one with good prognostic features of limited disease and favorable response to therapy. In approximately one half of the patients in both studies, herpes zoster developed while the patients were in complete remission. In addition, in both studies it is apparent that the risk of the development of herpes zoster increases with increasing survival. Our retrospective review suggests that, just as in the case of Hodgkin’s disease, the tumor type itself may predispose to the development of herpes zoster. Patients with small cell lung cancer appear to have an increased risk compared to patients with non-small cell carcinoma of the lung, despite the fact that the latter group had a better survival rate during this time period. A number of other factors predisposing to herpes zoster have been suggested by other investigators who have studied this infection in patients with Hodgkin’s disease and non-Hodgkin’s lymphoma [4,5,9]. These include predisposition in areas of disease, advanced (extensive) disease, predisposition in areas receiving radiotherapy, splenectomy, cutaneous anergy, and combination chemotherapy. The thoracic dermatomes were a common site of involvement by herpes. As the thorax is the primary site of the neoplasm in all cases and also an area treated by radiotherapy in all our cases, it was impossible to assess whether either of these factors predisposed to the development of zoster. However, in Huberman’s study, thoracic irradiation was

1982

The American

Journal of Medicine

Volume

73

799

HERPES ZOSTER IN LUNG CANCER PATIENTS-FELD

ET AL.

not used as part of primary therapy and yet the thoracic region was nonetheless a common site of zoster infection. Splenectomy was not a factor in any patient in either series. The most important predisposing factors may well be the intensity, duration, and type of systemic chemotherapy. The highest incidence of herpes zoster was seen in the CAV, protocol, which employed a short, intensive induction and a prolonged maintenance phase. The more intensive induction in the CAV2 protocol without maintenance chemotherapy was associated with the same incidence of herpes zoster as we had observed in the retrospective series treated almost exclusively with radiotherapy. In the CAV, protocol, maintenance chemotherapy consisted of lomustine (CCNU), methotrexate, and procarbazine. Among the three studies that previously described herpes zoster [23-251, the only drug in the maintenance chemotherapy that was present in all four trials was procarbazine. Although many other drugs overlapped during induction chemotherapy, as previously discussed, this did not apear to be the major time of risk for the development of herpes zoster. The relative infrequency of use of procarbazine in the treatment of small cell carcinoma of the lung might also explain the rarity of herpes zoster in previous reports. Huberman presented information on the dose/m2 of the various drugs that patients received, with and without herpes zoster. The only striking difference observed

was that patients in whom herpes zoster developed received more procarbazine (4,422 mg/m2 as compared to 2,764 mg/m2 in patients in whom herpes zoster did not develop). A statistical analysis was not carried out comparing these doses, nor was the possibility that this might be a major factor considered in the discussion [25]. The fact that procarbazine is included in the commonly used MOPP combination for treatment of Hodgkin’s disease-a disease in which chemotherapy is considered to be an important predisposing factor for herpes zoster [4,5,27]-lends further support to the notion that procarbazine itself may in some way predispose to herpes zoster. In conclusion, herpes zoster may complicate the treatment of patients with small cell carcinoma of the lung when combined modality therapy is used. Although improved survival as the result of more effective therapy may in part be responsible for the appearance of more cases of herpes zoster by increasing the period of risk, it appears that the duration of therapy and possibly the type of chemotherapy used may be important contributing factors to the increased incidence of herpes zoster seen in some studies. ACKNOWLEDGMENT We wish to thank Ms. B. Ottema, R.N., and Mrs. L. Millband, R.N., for their assistance in data collection, and Mrs. J. Russell for her careful preparation of the manuscript.

REFERENCES 1.

2. 3. 4.

5.

6.

7.

6.

9. 10.

800

ShanbromE, Miller S, Haar H: Herpes zoster in hematologic neoplasms: some unusual manifestations. Ann Intern Med 1960; 53: 523-533. Wright ET, Winer LH: Herpes zoster and malignancy. Arch Dermatol 1961; 84: 242-244. Sokal JE, Firat D: Varicella-zoster infection in Hodgkin’s disease. Am J Med 1965; 39: 452-463. Schimpff S, Serpick A, Stoler B, et al.: Varicella-zoster infection in patients with cancer. Ann Intern Med 1972; 76: 241-254. Goffinet DR. Giatstein EJ, Merigan TC: Herpes zoster-varicella infections and Ivmohoma. Ann Intern Med 1972: 76: 235-240. . Wilson JF, Mama GW, Johnson RE: Herpes zoster in Hodgkin’s disease. Clinical, histologic and immunologic correlations. Cancer 1972; 29: 461-465. Dolin R, Reichman RC, Mazur MH, Whitley RJ: Herpes zoster-varicella infections in immunosuppressed patients. Ann Intern Med 1978; 89: 375-388. Rosenberg SA, Diamond HD, Saslowitz B, Craver LF: Lymphosarcoma: a review of 1269 cases. Medicine (Baltimore) 1961; 40: 31-84. Mazur MH, Dolin R: Herpes zoster at the NIH: a 20-year experience. Am J Med 1978; 63: 738-744. Williams HM. Diamond HD, Craver LF, Parsons H: Herpes zoster. In: Karnofsky DA, ed. Neurological complications of lymphoma and leukemias. Springfield, IL: Charles C

December

1992

The American Journal of Medicine

11. 12.

13.

14.

15.

16.

17.

18.

Volume 73

Thomas, 1959: 101. Feldman S, Hughes WT, Kim HY: Herpes zoster in children with cancer. Am J Dis Child 1973; 126: 178-184. Luby JP, Ramirez-Ronda C, Rinner S, Hull A, Vergne-Marini P: A longitudinal study of varicella-zoster virus infections in renal transplant recipients. J Infect Dis 1977; 135: 659-663. Livingston RB, Moore TN, Heilbrun L, et al.: Small cell carcinoma of the lung: combined chemotherapy and radiation: a South-West Oncology Group study. Ann Intern Med 1978; 88: 194-199. Johnson RE. Brereton HD, Kent HC: “Total” therapy for small cell carcinoma of the. lung. Ann Thorac Surg -1978; 25: 510-515. Einhorn LH, Bond WH, Hornback N, Beng-Tek J: Long term results in combined modality treatment of small cell carcinoma of the lung. Semin Oncol 1978; 5: 309-313. Feld R, Pringle J, Evans WK. et al.: Combined modality treatment of small cell carcinoma of the lung. Arch Intern Med 1981; 141: 469-473. Feld R, Evans WK, DeBoer G: Herpes zoster in patients with small cell carcinoma of the lung receiving combined modality treatment. Ann Intern Med 1980; 93: 282-283. Feld R, Evans WK, Yeoh JL, et al.: Combined modality induction therapy without maintenance therapy for small cell carcinoma of the lung (SCCL) (abstr). Proc Am Assoc Cancer Res and Am Sot Clin Oncol 198 1; 22:

HERPES ZOSTER IN LUNG CANCER PATIENTS-FELD

19. 20. 21. 22.

23.

Cutler SJ, Ederer F: Maximum utilization of the life table method in analyzing survival. J Chron Dfs 1958; 8: 899. Kaplan EL, Meier P: Nonparametric estimation from incomplete dxlervations. Am Stat Assoc J 1958; 53: 457-481. Gehan EA: A generalized Wilcoxon test for comparing arbitrarily singly censored samples. Biometrika 1985; 52: 203-223. Peto R, Pike MC, Armitage P, et al.: Design and analysis of randomized clinical trials requiring prolonged observation of each patient. II. Analysis and examples. Br J Cancer 1977; 35: l-39. Wittes RE, Hopfan S, Wittes JT, Colby RB, Martini N: Oat cell carcinoma of the lung: combination chemotherapy with cyclophosphamkte, vincristine and procarbazine. Clin Bull

24.

25. 26. 27.

December 1082

ET AL.

1977; 7: 21-24. Sorenson S, Hansen HH, Dombernowsky P: Phase II study of vinblastine, Adriamycin and procarbazine in small cell carcinoma of the lung. Cancer Treat Rep 1978; 80: 1283-1286. Huberman M, Fossieck BEJr, Bunn PA Jr, Cohen MH, lhde DC, Minna JD: Herpes roster and small cell bronchogenic carcinoma. Am J bled 1980; 08: 214-218. Cohen MH, Chretien PB, lhde DC, et al.: Intensive combination chemotherapy proionging the survival of patients with small cell lung cancer. JAMA 1979; 241: 1813-1815. Feld R, Bodey GP: Infections in patients with malignant lymphoma treated with combination chemotherapy. Cancer 1977; 39: 1018-1025.

The American Journal of Medlclne

Volume 73

801