Prognostic influence of TNM staging and LDH levels in small cell carcinoma of the lung (SCCL)

Im J. Radiation Oncology Bid. Phys.. Vol. 12. pp. 77 1-771 Pmtcd in the U.S.A. All r@hU reserved.

??Original Contribution

PROGNOSTIC INFLUENCE OF TNM STAGING AND LDH LEVELS IN SMALL CELL CARCINOMA OF THE LUNG (SCCL) ROGER W. BYHARDT, M.D.,* ARTHUR HARTZ, M.D., PH.D.,? JOSEPH A. LIBNOCH,M.D.,* RICHARD HANSEN, M.D.* AND JAMES D. Cox, M.D.* Medical College of Wisconsin, Milwaukee, WI 53226 To better define the prognostic factors influencing the response to therapy and survival in small cell carcinoma of the lung @CCL), an expanded “TNM” type se system was developed and investigated in a series of 73 protocol treated patients. Because serum LDH levels at disease presentation have been correlatedto disease extent, response to therapy, and treatment outcome in a number of maliguancies, iacluding SCCL, these interrelationships were also analyzed in the protocolpatients. The TNM system was found to be a more descriptive and specific “shorthand” for denoting sites of involvement and for indicating the body burden of tumor than the traditional limited-extensive disease (LD-ED) system. A clear statistical advantage could not be shown over the LD-ED system for predicting chemotherapy response or survival, although there were trends suggesting the TNM system could divide patients into three prognostic subgroups. Serum LDH proved to be a useful index of disease extent and therapy outcome. LDH levels at presentation were proportionately higher with more extensive tumor, measured by either the LD-ED or TNM staging. High LDH predicted poorer responses to chemotherapy and lower survival within similar stage subgroups compared to patients with aormal LDH levels. The negativeeffect of elevated LDH was independent of hepatic involvement and did not predict subsequent hepatic failure in any consistent way. The SCCL TNM staging system proposed needs further refinement and should be tested with larger patient numbers. LDH, along with other tumor markers recently identified, need to be integrated into the staging system to form an overall prognostic index. Small cell carcinoma of lung, Staging, LDH.

INTRODUCIION

precise staging system, which assists in therapy planning, is desirable. Recently, elevated levels of serum lactic dehydrogcnase (LDH) have been correlated with the total body burden of a variety of tumor types, as well as with a less favorable treatment oufcome.3~4~8~‘1~12,‘6-‘ Others 9 have described a similar relationship of LDH to prognosis for SCCL.6*‘5 Elevated levels of LDH have been reported to be predictive of poor response to therapy and diminished likelihood of survival for testicular carcinoma,” Hodgkin’s disease,‘* non-Hodgkin’s lymphoma,3*‘2*‘7*‘9leukemia,’ Ewing’s sarcoma,4 and neuroblastoma.‘6 Recently, a similar relationship has been observed for SCCL, Ohnoshi et al. performed multivariate analysis of performance status, disease extent, and LDH levels and found a negative influence of elevated LDH on survi~al.‘~ This observation was also made by Cohen et ~1.~for both elevated LDH and low levels of serum globulin. After including these

As multi-modality therapy for small cell carcinoma of the lung (SCCL) has improved over the past decade, and as the disease has been investigated on prospective protocols, important predictors of disease outcome have become more clearly defined. Kamofsky performance status (KPS) and presenting stage have known prognostic significance for SCCL. Yet, a high KPS does not always predict a favorable outcome, and a more reliable prognostic index would be useful. The traditionally used limited disease-extensive disease (LD-ED) staging system is useful as a broad indicator of prognosis, but fails to clearly define regional extent of tumor involvement or indicate the total tumor burden. There can be vast differences in total tumor bulk between patients classified with the same LD-ED stage. As therapies improve and are interdigitated with irradiation, a more

* Dept. of Radiation Oncology. t Dept. of Medicine. $ Dept. of Medical Oncology. Reprint requests to: R. W. Byhard&M.D., Dept. of Radiation Oncology, Medical College of Wisconsin, Ave., Milwaukee, WI 53226.

authors acknowledge the physicians and staff at the Medical College of Wisconsin who aided in the care and management of these patients, Jannette Steinpas for typing this manuscript, and Mary Fischer for data analysis.

Acknowledgements-The

8700 W. Wisconsin

Accepted for publication

771

I 1 December 1985.

772

1. J. Radiation Oncology 0 Biology 0 Physics

two laboratory parameters into a prognostic index, performance status lost its significance as a predictor of survival.6 They recommended that a prognostic index could be developed for SCCL based on pretreatment laboratory results, including LDH levels. Aroney et al. noted significant correlations of disease status of SCCL after treatment with levels of serial biomarkers, including LDH, carcinoembryonic antigen (CEA), neuron-specific enolase, and adrenocorticotrophic hormone (ACTH).2 Sculier et al. also found a negative correlation of survival and elevated pretreatment levels of CEA for SCCL.20 In addition to the correlations with response to therapy and survival, serum levels of LDH have also been found to relate to disease extent.3*4.6 To develop a system of initial evaluation that is not only prognostic, but descriptive of the total body burden of SCCL, an expanded staging system was developed as a modification of the TNM system for lung cancer used by the American Joint Committee for Cancer Staging and End Results Reporting (AJC).’ This staging system was applied to a series of patients treated according to a systematic multimodal therapy protocol. LDH levels were serially recorded as part of the initial patient assessment prior to protocol entry. This report is an analysis of the ability of the modified TNM system to estimate the total body burden of tumor as well as to predict response to treatment and treatment outcome. The relationship of LDH levels at presentation to response and survival was examined by a retrospective multifactorial analysis. METHODS AND MATERIALS A total of 73 patients, treated at the Medical College of Wisconsin affiliated institutions between 198 1 and 1983, were treated according to a systematic protocol of multimodality therapy and were available for analysis. The protocol specified therapy according to stage and did not involve a randomization to different treatment op tions. Biopsies of two cytologic specimens provided an initial diagnosis. Staging evaluations at presentation included routine history and physical examination, chest radiography, complete blood count, routine serum chemistries (including LDH), bone marrow aspiration and biopsy, radionuclide bone scans, and computed tomography (CT) of the brain, chest, and abdomen to the level of the pelvis. Only patients between the ages of 20 and 73, with a minimum Kamofsky performance status of 30, were entered onto the study. All clinical, laboratory, and radiographic data were included in the clinical stage, which was first classified according to the traditional limited disease (LD) and extensive disease (ED) system. LD included tumor restricted to the lung parenchyma, mediastinum, ipsilateral hemithorax, and ipsilateral supravicular nodes and ED was defined as tumor involvement beyond the ipsilateral hemithorax and supraclavicular nodes. Patients, who on

May 1986, Volume 12, Number 5

CT scan, were found to have involvement beyond the ipsilateral hemithorax, in sites not always detectable without CT scan (retroperitoneal nodes, adrenal, liver, etc.), were included in the ED classification. Hepatic metastases were determined by dynamic computed tomography of the upper abdomen. Percutaneous liver biopsy was done in two patients with elevated hepatic enzymes and a normal upper abdominal CT. For the purpose of this analysis, all patients who had been entered onto protocol were retrospectively restaged according to a TNM system. The TNM system used was modified from the American Joint Committee (AJC) staging for lung cancer’ to better accommodate the pathophysiology of SCCL (Table 1). The aim of this modification was to arrive at a stage grouping system that would more clearly divide tumor extents according to total body burden of disease, and which would provide a “shorthand” descriptor outlining sites of major tumor involvement. Thus, the stage grouping shown in Table 1 does not break down strictly according to anatomic extent, but by both location and amount of tumor, namely, Stage I (limited), Stage II (intermediate), and Stage III (extensive). The TNM groupings within each of the three stages are according to increasing tumor burden. The protocol treatment plan involved a fixed sequence of prophylactic and consolidative irradiation with multiagent chemotherapy, which has been reported in detail elsewhere.‘,’ In brief summary, all patients received induction with Cyclophosphamide, 900 mg/M2, Doxorubicin, 60 mg/M2, and Vincristine, 2 mg/M2 (CAV), which was started concurrently with prophylactic crania-cervical irradiation (25 Gy/lO fx). Complete responders to chemotherapy received consolidative thoracic irradiation (37.5 Gy/ 15 fx) to the residual primary tumor. Some patients also received prophylactic irradiation of the lumbar axis to 25 Gy/lO fx as part of a recent protocol modification. Partial and nonresponders to chemotherapy received therapeutic irradiation to the chest primary if indicated. Response to therapy was assessed by repeating all positive tests after chemotherapy and again after consolidative irradiation. Follow-up ranged from 14 to 38 months, with a median follow-up time of 29 months. There were 56 men and 17 women ranging in age from 47 to 73, with a mean age of 6 1. Karnofsky performance status was 80 to 100 in 4 1 (56%), 50 to 70 in 29 (40%) and 30 to 40 in 3 (4%). There were 3 1 patients with limited disease and 42 with extensive disease. Stage at presentation (both traditional LD-ED and “TNM”), as well as presenting LDH values, were evaluated in relation to response to therapy and survival. LDH levels were considered elevated above 190 mu/ml. To test the significance of the difference between survival curves, we used the SPSS computer program for comparing survival curves.‘4 The statistical testing in this program is based on the Lee-Desu Statistical method.”

TNM staging and LDH levels in SCCL 0 R. W. BYHARDT

773

efal.

Table 1. “TNM” staging system for SCCL (modified from American Joint Committee System for Lung Cancer) Stage

TNM classification

I (limited) II (intermediate)

T,_3 and/or Nlaza; M0 T4 and/or Nib.zb,zcand/or M,

T (Primary, local and ipsihemithoracic extent) Stages: T-Primary tumor T-O-No evidence of primary tumor T-x-Malignant cells in sputum or washings, negative chest roentgenograph and bronchoscopy. T,--lntrapulmonary. localized; a solitary tumor 3.0 cm or less, within lung, not involving visceral pleura, not proximal to lobar bronchus. T2-Intrapulmonary, extensive; more than 3.0 cm, or any size which extends to hilar region (including associated atelectasis or pneumonitis). No pleural effusion; any atelectasis or pneumonitis involves less than whole lung. T3-Extrapulmonary intrathoracic, localized: includes lesions invading visceral pleura to pleural space or adherent to parietal pleura: lesions with associated pleural effusion, but negative cytology. T4-Extrapulmonary, intrathoracic, extensive; an extensive lesion, invading nerves for example, phrenic (Phr) or recurrent nerves (RL), major vessels for example, superior vena cava (SVC),

heart (Hrt), pericardium (Pe), esophagus (Eso), vertebrae (V), chest wall (CW) or diaphragm (Dia.) Lesions beyond parietal pleura and into chest wall, viscera and/or deep mediastinal structures. (Specify by code) N (Nodes, intrathoracic nodal extent) Stages: No-No demonstrable spread to nodes. Nx-Nodes not evaluated. N,-Ipsilateral hiiar or peribronchial nodes. N,,-Node or nodal cluster 5 3 cm. N,,-Node or nodal cluster > 3 cm.

RESULTS Staging analysis A comparison according to response to chemotherapy is made between the LD-ED staging system and the TNM system in Table 2. Complete responses were observed in 42% of LD patients and 24% of ED patients using the traditional system. For the modified TNM system, complete response rates were 36% for Stage I, 35% for Stage II, and 26% for Stage III. Some of the patients classified as Stage II would have been designated LD according to the traditional staging system (T4 primaries; Nlb,Zb.2c nodes). Some, classified Stage II on the basis of M, disease, would be considered ED in the LD-ED system, Thus, Stage II overlaps partially with both LD and ED. The responses to chemotherapy of the Stage II patients reflect an intermediate response level. Some patients, usually called ED, may have a prospect for response to drugs closer to LD than ED. Also, some LD patients with “large” primary or nodal involvement may have a poorer chance

Stage

TNM classification Any T N: Mz

III (extensive)

N2-Mediastinal nodes. N2,-Single or multiple nodes, none larger than 3 cm. N2,-Single or multiple nodes, or confluent mediastinal mass > 3 cm.: or NZn + Nib, N&-Large hilar and/or mediastinal mass with invasion of major mediastinal structures (pericardium, esophagus, etc.), or with superior vena caval obstruction, or contralateral hilar nodes.

M (Metastatic extent beyond ipsilateral hemithorax) M,,-No extra-hemithorax spread. M ,-Extrathoracic regional nodes; retroperitoneal (RP), cervical (Ce), axillary (Ax), or supraclavicular (SC) nodes. M2-Extrathoracic organ involvement M2,-Nondiffuse distant spread (i.e., involving focal area or limited volume suitable for consolidation irradiation). Adrenal (ADR), solitary pulmonary nodule contra-lateral lung (PUL-S); brain, solitary or multiple (Br); liver, focal or multiple-diffuse (Hep): bone, solitary or focal (Bo); meninges (Me); spinal . . cord (SC); eye (E); or other. [Specify stte with code i.e., M2. (SC, Br).] M2,-Diffuse distant spread (i.e., involving distant sites in diffuse fashion, not encompassable in irradiation portal for consolidation): bone marrow (Mar); bone, multiple sites (BoM); lung, contra-or ipsilateral, multiple nodules (PUL-M), skin nodules, multiple (Ski). [Specify site with code, i.e., M2b (Mar).]

of response than LD patients with a minimal to moderate

amount of disease. To evaluate the influence of bulky intrathoracic tumor on subsequent response to chemotherapy, responses were examined in LD patients classified as T4 or Nlb,Zk in the TNM system. Of 46 patients with bulky primary disease, 14 (30%) had non-response to chemotherapy, while 4/27 ( 15%) had non-response with non-bulky primary disease. Although this indicated a trend towards poorer responses

Table 2. Response to chemotherapy by stage and system Stage

No. pts

%CR

% PR

%NR

LD-ED Stage

LD ED

31 42

42 24

32 33

26 43

Modified AJC Stage

I II III

11 23 39

36 35 26

45 35 46

18 30 28

174

I.

J. Radiation Oncology 0 Biology0 Physics

with advanced local disease, the difference was not statistically significant by Chi Square analysis (p = .135). In comparison to LD-ED, the TNM system further stratified levels of survival (Table 3). Both Stage III and ED groupings had similar median survivals of 7.8 and 7.0 months, respectively. However, Stage I and II, with medians of 18 and 12 months, seemed to define two different survival levels within the LD stage, which by itself had a median of 14 months. LD, as well as Stage I and II, had about 25% of the patients in each stage grouping alive with no evidence of disease at a median follow-up time of 29 months, while ED and Stage III had none. The patient numbers in each TNM stage category are currently too small to make a statistically valid statement, but are suggestive of a trend. Further follow-up is needed to determine the relationship of TNM Stages I and II to long term survival beyond the median. LDH as a prognosticator Elevated LDH was found in 29 patients overall (330/o), 2131 (6%) with LD and 27142 (64%) with ED. Survival was significantly lower (p = .002) for patients with LDH levels above 190 mu/ml at presentation versus normal LDH (Fig. 1). Patients with liver metastases were found to have higher levels of LDH than other ED patients, therefore, it was ,necessary to determine whether elevated LDH predicted a poor outcome only because it reflected liver metastases. To rule out this possibility, the effect of liver metastases was evaluated separately and the relationship of LDH to levels of alkaline phosphatase and SGOT was examined. ED patients with and without hepatic metastases had equivalent survival. At presentation, 3/44 (7%) with normal LDH had liver involvement, whereas 13/29 (45%) with elevated LDH had liver involvement. To clarify the interaction of LDH levels and hepatic involvement in relation to survival, a pairwise analysis was performed on the following four groups: (a) low LDH, (-) liver; (b) low LDH, (+) liver; (c) high LDH, (-) liver; and (d) high LDH, (+) liver. The survival of patients without hepatic metastases is signifi-

Table 3. Relationship of staging system to survival Traditional system

ANED* Median survival (months)

TNM system

LD

ED

I

II

III

913 1 (29%)

o/42 (0%)

3/11 (27%)

6123 (26%)

o/39 (0%)

18

12

14

7

8

ANED = Alive, no evidence of disease; median follow-up interval of 29 months.

May 1986. Volume 12, Number 5

o LDH

NORMAL

.

ELEVATED

LDH

(44) (29)

O-J I

0

5

I

lo

I

I

I

1

I

I

15

20

25

30

35

40

SURVIVAL

1

45

,

50

(months)

Fig. 1. Survival in Small Cell Lung Cancer Patients (LD-ED) according to initial LDH levels.

cantly lower with an elevated LDH than with normal LDH (p = .003). Patients with liver metastases had nearly identical survival with or without elevated LDH (p = .77). Patients with low LDH had equivalent survival with or without liver metastases ( p = .18), although there was a trend toward worse survival in the liver metastases group. To rule out the possibility that elevated LDH levels merely presaged occult hepatic metastases, failure patterns were examined. Elevated LDH did not predict subsequent liver progression. Overall, 9/44 (20%) patients with a normal LDH and 6/29 (20.6%) patients with an elevated LDH demonstrated first evidence of failure in the liver. Among the subgroup of patients completing protocol therapy, with initial elevations of LDH, O/14 had first progression in the liver, and l/ 10 had second progression in the liver. LDH levels at pretherapy staging bore a closer relationship to subsequent primary site failure that bordered on statistical significance (Table 4); 15/44 (34%) of patients with normal LDH and 16/29 (55%) of patients with elevated LDH levels ultimately had initial progression at the primary site of disease and/or regional nodes (p = .075). This finding was independent of stage of disease, as well as presenting alkaline phosphatase and SGOT levels. When initial LDH was elevated and alkaline phosphatase and SGOT were normal, the initial primary failure rate was 13/19 (68%). Relationship of LDH to chemotherapy response Combining both LD and ED stages, 5123 (22%) complete responders, 5/24 (2 1%) partial responders, and 19/ 26 (73%) nonresponders had elevated LDH levels. Of 44 patients with normal LDH levels, there were 18 (4 1%) complete responses and 19 (43%) partial responses; in

TNM staging and LDH levels in SCCL 0 R:W. BYHARDT et 01.

775

Table 4. Relationship of alkaline phosphatase (AP), SGOT, and LDH levels at presentation, initial hepatic staging, and initial failure in the liver and at the primary site and/or regional nodes

No.

Staging CT Liver +

Initial liver failure

Initial primary failure

NL AP, SGOT, LDH

38

2 (5%)

ABNL AP, SGOT ABNL LDH ABNL AP, SGOT, LDH

6 19 10

1 (16%) 4 (21%) 9 (90%)

7 (18%) 2 (33%) 1(5%) 5 (50%)

14 (36%) 2 (33%) 13 (68%) 3 (30%)

Presenting enzyme levels

NL = Normal; ABNL = abnormal elevated levels;SGOT = Serum glutamic oxaloacetic transaminase; LDH = Lactic dehydrogenase.

contrast, of 29 patients were 5 ( 17%) complete sponses. The differences different by Chi square

with elevated LDH levels, there responses and 5 ( 17%) partial rein response rates are significantly analysis (p = .O18).

Relationship of LDH to stage The relationship of LDH to both LD-ED and TNM stage was examined. The survival curves for LD and ED patients with elevated and normal LDH are shown in Figure 2. Of 31 patients with LD, 2 (6%) had elevated LDH at initial staging; one of these had T2 No M,-,disease, progressed at the primary site after the fourth cycle of chemotherapy, received chest irradiation with partial response, then failed in the lung and brain at 14 months. The other patient, presented with T4 Nza Mc, disease (superior vena caval syndrome), had a non-response to chemotherapy after two cycles of CAV, and died at 3 months of progressive primary disease and pneumonia. Of the 29 LD patients with normal LDH, there is a 12.6% actuarial survival at 42 months; the median survival is 15.2 months. Pairwise comparison of the survival of

this group with that of LD patients with elevated LDH did not reveal a statistically significant difference (p = .2 1). With only 2 patients in the high LDH-LD group, more patients would be required to show a difference. Of the 42 ED patients, 20 (48%) had normal LDH, while 22 (52%) had elevated LDH. The median survivals were 10.6 months for normal LDH and 7.2 months for elevated LDH; pairwise comparison of these two groups did not show a statistically significant difference (p = .17). although a trend towards poorer survival is suggested for the high LDH group. LDH levels were examined in the 3 TNM stage groupings. For Stage I, 9/l 1 (82%) had normal LDH and 2/l 1 (18%) had slightly elevated LDH levels (average = 195 mu/ml). For Stage II, 3/23 ( 13%) were elevated to an average level of 227 mu/ml, whereas the rest had normal levels. For Stage III, 24/39 (62%) were elevated to an average level of 226 mu/ml, with 15/39 (38%) at normal levels. Patients with elevated LDH levels accounted for 2/2 (100%) 2/3 (66%), and 7/l 1 (63%) of NR’s in Stages I. II. and III, respectively. The numbers of patients were too few for meaningful statistical analysis of survival when broken down to TNM stage-LDH level categories.

DISCUSSION -

NORMAL

LDH -

LIMITED

-

NORMAL

LDH -

EXTENSIVE

DISEASE

c-a

HIGH

LDH -

LIMITED

bd

HIGH

LDH -

EXTENSIVE

(29)

DISEASE

DISEASE

(20)

(2)

DISEASE

(22)

0.70.6os0.40.30.2O.l02

r 0

5

10

15

20

25

SURVIVAL

30

35

40

45

50

(months)

Fig. 2. Survival in SCCL patients according to initial LDH level and LD-ED stage.

The traditional LD-ED staging system has served us well over the past decade. Despite improvements in therapy and overall disease control, a more detailed staging system has not yet been used or tested. It has been said that the LD-ED system is sufficient because of the frequent systemic nature of SCCL, at least until vastly improved therapy is developed, and that the key to predicting outcome is whether the disease is beyond the hemithorax of origin. Nevertheless, experience with other malignant processes would suggest that the success of therapy is inversely proportional to the total body burden of tumor. Numerous authors have noted the propensity for SCCL to fail in sites of original bulky involvement, even after apparent complete response to therapy. l3 The LD-ED system frequently does not provide a clear description of sites of involvement or their extent. For example, a 7 cm primary with asso-

776

I. J. Radiation Oncology 0 Biology 0 Physics

ciated distal atelectasis is classified as LD, as is a 2 cm peripheral nodule. A patient with a 3 cm primary with retroperitoneal nodes found only on computed tomography and a patient with diffuse bone, liver, and intracranial metastases, are both classified as ED. It is hard to accept that the prognosis of these disparate, but equally staged, patients is being accurately predicted by the LD-ED system. The modified TNM staging system described herein is a first attempt to provide a clearer means to describe the sites of initial involvement and estimate the body burden of tumor. It is intended as a “ready-reference” for identifying patients and specific sites of involvement that might benefit from consolidative treatment. Whereas this analysis of the proposed TNM system has not shown a clear statistical advantage over the LD-ED system for either the prediction of response or outcome, definite trends were observed. Patients classified as Stage I seem to have a somewhat higher total chemotherapy response rate and median survival than LD. There is an intermediate level of response and survival in the Stage II grouping. Stage III and ED seem rather similar in response and outcome. Extensive local primary disease (T4), otherwise classified as LD, shows a less favorable initial response to drug, early local progression, and poorer overall survival. Since the expected differences in response rate or survival based on the TNM system vs. LD-ED are probably small (lo15%), it will require a much larger patient population to show statistical significance. The TNM system proposed is a shorthand reference that identifies specific sites of involvement and their extent. It may be ueful as an adjunct to the LD-ED system in future protocol studies. Serum LDH proved to be a useful index of disease extent and therapy outcome in this review. With elevated

May 1996, Volume 12. Number 5

LDH, in both LD-ED, there was a trend towards decreased survival. This effect was independent of hepatic involvement at presentation and did not predict subsequent hepatic failure. In fact, this review failed to confirm a less favorable prognosis for patients with ED and hepatic metastases compared to ED without hepatic metastases, as reported by others.’ High LDH was associated with an increased risk of failure at the primary site. A higher proportion of poor responses to chemotherapy was also associated with elevated LDH. LDH levels were proportionately higher with more extensive tumor involvement as measured both by the LD-ED and TNM systems. A trend toward less favorable survival was noted in each stage grouping when LDH was elevated. The relationship of LDH level to stage in SCCL, demonstrated in this review, confirms the observations of others; in fact, it is very difficult to separate the influence of elevated LDH and advanced stage. It should be emphasized that elevated LDH levels do not necessarily signify occult hepatic metastases, but reflect the amount of tumor. As such, elevated LDH may identify high risk patients within apparent early clinical stages. Since elevated LDH seems to have a high predictive value for therapy response and disease extent, it is our current recommendation to use it as part of the initial staging evaluation for SCCL. LDH probably reflects a combination of “tumor burden” and cell kinetics, namely spontaneous cell death in a rapidly growing tumor. Used together with the TNM system proposed, a prognostic index can be formed. The usefulness of this index requires further investigation, along with other factors now being examined by other clinical investigators. These studies may sharpen our ability to design treatment strategies and predict outcome.

REFERENCES 1. American Joint Committee For Cancer Staging and End Results Reporting: Manual For Staging of Cancer, 1984. 2. Aroney, R.S., Dermody, W.C., Aldenderfer, P., Parsons, P., McNitt, K., Marangos, P.J., Whitacre, M.Y., Ruddon, R.W., Wiemik, P.H., Aisner, J.: Multiple sequential biomarkers in monitoring patients with carcinoma of the lung. Cancer Treat. Rep. 68: 859-866, 1984. 3. Arseneau, J.C., Camellos, G.P., Banks, P.M.: American

Burkitt’s Lymphoma:

Clinicopathologic study of 30 cases. 1975. 4. Brereton, H.D., Simon, R., Pomeroy, T.C.: Pretreatment lactate dehydrogenase predicting metastatic spread in Ewing’s sarcoma. Ann. Intern. Med. 83: 352-354, 1975. 5. Byhardt, R.W., Cox, J.D., Holoye, P.Y., Libnoch, J.A.: The role of consolidation irradiation in combined modality therapy of small cell carcinoma of the lung. Int. J. Radial. Am. J. Med. 58: 314-321,

Oncol. Biol. Phys. 8: 1271-1276, 1982. 6. Cohen, M.H., Makuch, R., Johnston-Early,

A., Ihde, DC., Bunn, P.A., Foss&k, B.E., Minna, J.D.: Laboratory parameters as an alternative to performance status in prognostic stratification of patients with small cell lung cancer. Cancer Treat. Rep. 65: 187-195,

1981.

7. Cox, J.D., Holoye, P.Y., Byhardt, R.W., Libnoch, J.A., Komaki, R., Hanson, R.M., Kun, L.E., Anderson, T.: The role of thoracic and cranial irradiation for small cell carcinoma. Int. J. Radial. Oncol. Biol. Phys. 8: 191-196, 1982. 8. Gomez, G.A., Sokal, J.E.. Walsh, D.: Prognostic features at diagnosis of chronic m yelocytic leukemia. Cancer 47( 10): 2470-2477, 1981. 9. Ihde, DC.. Hansen, H.H.: Staging procedures and prognostic factors in small cell carcinoma of the lung. In Small Cell Lung Cancer, Greco, F.A., Oldham, R.K., Bunn, P.A. (Eds.). NY, Grune & Stratton. 198 I, pp. 26 I-283. 10. Lee, E., Desu, M.: A computer program for comparing K samples with right-censored data. Comp. Prog. Biomed. 2: 315-321,

1972.

1 I. Lippert, M.C., Javadpour, N.: Lactic dehydrogenase in the monitoring and prognosis of testicular cancer. Cancer 45( 10): 2274-2278, 198 1. 12. Magrath, I., Lee, Y.J., Anderson, T., Henle, W., Ziegler, J., Simon, R., Schein, P.: Prognostic factors in Burkitt’s lymphoma: Importance of total tumor burden. Cancer 45(6): 1507-1515, 1980. 13. Morstyn, G., Ihde, D.C., Lichter, A.S., Bunn, P.A., Camey.

TNM staging and LDH levels in SCCL 0 R. W. BYHARDT

D.N., Glatstein, E., Minna, J.D.: Small cell lung cancer 1973- 1983: Early progress and recent obstacles. Int. J. Radial. Oncol. Biol. Phys. 10: 5 15-39, 1984. 14. Nie, N. (Ed.): Stalistical Package for the Social Sciences, SPSS User’s Guide. New York, McGraw-Hill, 1983. 15. Ohnoshi, T., Hiraki, S., Nakata, Y., Ozawa, S., Tamura, T., Seto, T., Miyai, M., Kawahara, S., Numata, T., Kimura, I.: Pretreatment serum albumin concentration and lactic dehydrogenase activity as prognostic factors in patients with small cell lung cancer. Acta Med. Okayama 36(6): 487490, 1982.

16. Quinn, J.J., Altman, A.J., Fran& C.N.: Serum lactic dehydrogenase, an indicator of tumor activity in neuroblastoma. J. Pediat. 97(I): 89-91, 1980.

efal.

777

17. Ridsway, D., Smiley, S.. Neerhout, R.C.: The prognostic value of presenting serum LDH in non-Hodgkin’s Lymphoma. J. Pediat. 99(4): 6 1 l-6 13, 198 1. 18. Schilling, R.F., M&night, B.. Crowley, J.J.: Prognostic value of serum lactic dehydrogenase level in Hodgkin’s disease. J. Lab. Clin. Med. 99(3): 382-387,

1982.

19. Schneider, R.J., Seibett, K., Passe, S.. Little, C.. Gee, T.. Lee. B.J. 3rd, Mikle, V., Young, C.W.: Prognostic significance of serum lactate dehydrogenase in malignant lymphoma. Cancer 46( 1): 139- 143. 1980. 20. Sculier, J.P., Feld, R., Evan, W.K., Shepherd. F.A., DeBoer. G., Malkin, D.G., Malkin, A.: CEA: A Useful Prognostic Marker for Small Cell Lung Cancer (SCLC). Proc Ann. Meet, Am. Assoc. Cancer Res. 25: 158. 1984.