The Use of Serum Albumin for Further Classification of Stage I11 Non-Oat Cell Lung Cancer and Its Therapeutic Implications Phyllis Fatzinger, R.D., Tom R. DeMeester, M.D., Hrair Darakjian, M.D., Clemente Iascone, M.D., Harvey M. Golomb, M.D., and Alex G. Little, M.D.

ABSTRACT The use of the preadmission serum albumin level for predicting survival was evaluated in 81 patients with Stage I11 disease, 59 with unresectable and 22 with resectable primary tumors. A serum albumin of less than 3.4 gm/dl in a patient with unresectable Stage I11 disease indicates a poor prognosis with an accuracy that super- sedes that obtained from a clinical assessment of the ana- tomical extent of disease.

Seventeen of the 22 patients with resectable Stage I11 disease had a preadmission level of albumin of 3.4 g d d l or greater. The median survival was 20.5 months, which was statistically longer than 9.9 months for 12 patients with unresectable Stage I11 MO disease and an albumin level of 3.4 g d d l or greater (p < 0.05). Five of the 22 patients who underwent resection had a preadmission al- bumin level of less than 3.4 g d d l . The median survival for these patients was 9.7 months compared with 20.5 months for those with a level of 3.4 gm/dl or greater. These findings suggest that resection of Stage I11 disease in pa- tients with an albumin level of 3.4 gm/dl or greater pro- longs survival.

Another group of 5 patients with resectable Stage I11 disease and an albumin level of less than 3.4 g d d l were force-fed an elemental diet while undergoing a regimen of preoperative radiation therapy. There was no im- provement in survival, and 3 died of the disease prior to resection. It is concluded that the determination of the serum albumin level in patients with Stage I11 lung cancer has prognostic and therefore decision-making value in selecting patients for aggressive medical or surgical ther- apy.

At diagnosis, 80% of patients with lung cancer have Stage III disease [l]. Based on the anatomical extent of the tumor, these patients can be further stratified into those who have disease limited to the chest (MO) and those who have distant organ metastasis (Ml). Those with MO disease can also be stratified on the basis of the resectability of the primary tumor, but additional divi-

From The University of Chicago Pritzker School of Medicine, Chicago, IL. Presented at the Nineteenth Annual Meeting of The society of Thoracic Surgeons, San Francisco, CA, Jan 17-19, 1963.

Address reprint requests to Dr. DeMeester, The University of Chicago, Department of Surgery, 950 E 59th St, Box 440, Chicago, IL 60637.

sions lead to a myriad of anatomical groupings with so few patients in each that it becomes impractical and of questionable clinical value.

A need exists to identify patients with Stage 111 disease by indicators related to survival other than the anatomi- cal extent of the neoplasm. Preferably, such an indicator should signal when a tumor has altered the metabolic function of the host beyond the point of benefit from further therapy. It has been suggested that indicators of nutritional disturbances, such as serum albumin, might serve such a purpose [2].* In this study we have pro- spectively analyzed the use of the preadmission serum albumin level to predict the survival of patients with Stage 111 non-oat cell lung cancer.

Material and Method Between January, 1978, and July, 1980, a serum albumin level was obtained prior to or on the day of admission in all patients referred to the Chest Oncology Clinic of the University of Chicago Hospital with the suspected diag- nosis of lung cancer. The standard clinical laboratory coloimetric method was used [3]. A serum albumin level of 3.4 @dl or greater was considered normal and indic- ative of an adequate nutritional state.* All patients were staged as to the extent of their tumor

based on the results of clinical examination, chest roent- genogram, gallium scan, and other organ-specific scans or roentgenograms when indicated. Each underwent bronchoscopy and, when indicated, mediastinoscopy or distant organ biopsy. A non-oat cell carcinoma was confirmed histologically in all patients, and only those who were fully ambulatory and previously untreated were admitted to the study. Once in the study, the 81 patients were grouped according to the anatomical ex- tent of their disease as follows: (1) Stage III M1 (36 pa- tients); (2) Stage III MO with an unresectable primary tumor (23 patients); and (3) Stage III MO with a resectable primary tumor (22 patients).

Patients with M1 disease were treated with combina- tion chemotherapy consisting of Cytoxan (cyclophos- phamide), Adriamycin (doxorubicin hydrochloride), methotrexate, and procarbazine (CAMP) [4]. Radiation therapy was used only for palliation. Patients with unre- sectable MO disease were treated with 3,000 rads to the tumor and mediastinum and CAMP chemotherapy [5].

*Fatzinger P, Darakjian H, Cox D, et al: Nutritional status as related to survival in Stage 111 cardnoma of the lung. (in preparation, 1984)

115

116 The Annals of Thoracic Surgery Vol 37 No 2 February 1984

PERCENT SURVIVING

N.36 I

1 0 0 , - p 2 3

80 B -- Stage III MI unresectabk

-+ Stage XI Mo unresectoble

N S 60

so

40

30

20

10

6 12 18 24 30 36

MONTHS

Fig I . Actuarial survival curves for two groups of patients with unre sectable Stage Ill non-oat cell lung cancer classified by anatomical extent of disease. Twenty-three patients had MO disease and 36 pa- tients, MI disease. (NS = not significant.)

,N* 25

- Stagem Mo or MI Alb23.4 gm/IOOml - Stagem Mo or MI Alb<3.4 gm/100ml

PERCENT SU RV IV I NG

6 12 18 24 30 36

MONTHS

Fig 2 . Actuarial survival curves for the same 59 patients in Figure 1 divided on the basis of preadmission serum albumin (Alb) lmels. For 34 the level was greater than or equal to 3.4 gmldl and for 25, less than 3.4 gmldl (p < 0.0007).

Patients with resectable primary tumor underwent ir- radiation of the mediastinum and tumor, or tumor bed, with 3,000 rads before or after operation depending on whether they were known to be Stage 111 before or after the surgical resection. Postoperatively they all received 6 to 12 months of CAMP chemotherapy.

An attempt was made to obtain for each patient the amount of recent weight loss, and whether there had been a loss of appetite, a change in taste, or both. To assess the patient’s habitus objectively, the admission weight was compared with the weight for height stan- dards of Jelliffee [ 6 ] . Within the limits of a general hos- pital admission, the patient’s average daily caloric intake was measured. The patients were provided with, and

encouraged to drink, an oral nutritional supplement (Ensure, Ross Laboratories), and the amount consumed was recorded.

A separate group of 5 patients had MO disease, a re- sectable primary tumor, and a preadmission serum albu- min level of less than 3.4 g d d l . They were continuously force-fed an elemental diet (Vital HN, Ross Laboratories) by the enteral route in an attempt to increase their re- sponse to therapy.

All patients were followed monthly at the clinic, and actuarial survival curves were constructed for the vari- ous studied groups. The differences between the curves were evaluated using the modified Lee-Desu statistical method [7].

117 Fatzinger et al: Serum Albumin for Classification of Stage Ill Lung Cancer

N.22 I,N= 12

100

90

80 - Stage IU MI Alb 23.4 gm/100ml - Stage IU Mo Alb 13.4 gm/IOOml 70

PERCENT 6o

, N = 9

SURVIVING

40

30

2 0

10

6 12 18

MONTHS

Fig 3. Actuarial survival curvesfor two groups of patients with Stage Ill non-oat cell lung cancer and preadmission serum albumin (Alb) levels greater than or equal to 3.4 gmldl. Twelve patients had MO disease with unresectable primary tumors and 22, MI disease (p < 0.01).

Nsll

- Stoge III M, Alb <3.4 qm/IOOml

p-0 Stoqe XI M,, Alb <3.4 qm/ 100ml

NS PERCENT SURVIVING 60

50i \I

6

MONTHS

Fig 4. Actuarial survival curves for two groups of patients with Stage Ill non-oat cell lung cancer and preadmission serum albumin (Alb) levels of less than 3.4 gmldl. Eleven patients had MO disease with unresectable primary tumors and 14, MI disease. ( N S = not significant.)

Results Among the 59 patients with unresectable Stage 111 dis- ease there was no difference in survival between the 23 classified as MO and the 36 classified as M1 (Fig 1). Thirty-four of these 59 patients had a serum albumin level of 3.4 g d d l or greater. They had a longer survival than the remaining 25 patients with a serum albumin of less than 3.4 g d d l (p < 0.0007) (Fig 2).

Of the 34 patients with unresectable disease and a serum albumin level of 3.4 g d d l or greater, the 12 with MO disease survived longer than the 22 with M1 disease

24 30 36

(p < 0.01) (Fig 3). Of the 25 patients with unresectable disease who had a serum albumin level of less than 3.4 gdd l , there was no difference in survival between the 11 with MO and the 14 with M1 disease (Fig 4). The 22 patients with M1 disease and a serum albumin level of 3.4 g d d l or greater survived longer than the 11 with MO disease and a serum albumin level of less than 3.4 g d d l (p < 0.04) (Fig 5). As expected, the 12 patients with MO disease and a serum albumin level of 3.4 g d d l or greater survived longer than the 14 patients with MI disease and a serum albumin level of less than 3.4 g d d l (p < 0.008).

In summary, these findings show that a serum albu- min level of less than 3.4 g d d l in a patient with unre- sectable Stage 111 disease indicates a poor prognosis with an accuracy superseding that obtained from a clinical assessment of the anatomical extent of disease present (i.e., MO or Ml) .

Seventeen of the 22 patients with Stage I11 MO disease whose primary tumor could be resected had a preadmis- sion serum albumin level of 3.4 g d d l or greater. After resection they had a statistically longer survival than the 12 patients whose lesion was unresectable and who had MO disease and an albumin level of 3.4 g d d l or greater (p < 0.05) (Fig 6). The median survival of the 17 patients with a normal albumin level and with a resectable tumor was 20.5 months. The remaining 5 patients undergoing resection had a serum albumin level of less than 3.4 gm/ dl. Their median survival was 9.7 months (Fig 7). Al- though strikingly different, the two survival curves are not statistically different because of the small number of patients with a low preadmission serum albumin level and a resectable lesion. Table 1 itemizes the patients with resectable and unresectable MO disease as to their TN classification, survival in months, and preadmission serum albumin levels.

The level of the serum albumin was not related to a marasmic appearance of the patient when objectively measured by comparing the patient’s admission weight

118 The Annals of Thoracic Surgery Vol 37 No 2 February 1984

- Stage XU M, Alb 13.4 gm/100ml - Stage m M, Alb C3.4 gml100 ml

1 - 1 I I I T 6 12 18 24 30 36

MONTHS

Fig 5. Actuarial survival curves for two groups of patients with Stage Ill non-oat cell lung cancer. Tumty- two patients had MI disease and preadmission serum albumin (Alb) levels greater than or equal to 3.4 gmldl, and 11 patients had MO disease, an unresectable prima ry tumor and a preadmission serum albumin level of less than 3.4 gmldl ( p < 0.04).

N.17

100

90

80

70

- Stage IU M, Resected Alb 13.4 gm /100ml - Stage IIt M,, Unresected Alb 234 gm/IOOml

PERCENT 60 SURVIVING ~o

40

30

20

10

6 12 18 24 30 36

MONTHS

Fig 6. Actuarial survival curves for two groups of patients with non- oat cell lung cancer, Stage 111 MO disease, and preadmission albumin (Alb) levels greater than or equal to 3.4 gmldl. In 17 patients, the primary tumor was resected and in 12, the primary tumor could not be resected. Both groups received radiation therapy and chemotherapy ( p < 0.05).

to his ideal body weight (Fig 8). Weight loss and taste aversion occurred in patients with normal as well as de- pressed preadmission serum albumin levels and was more common, but not statistically significant in the lat- ter. Among the patients with an unresectable lesion and an albumin level of greater than or equal to 3.4 gm/dl, weight loss of greater than 5% of usual body weight occurred in 13 out of 27 (48%); taste aversion, in 13 out of

31 (42%); and both weight loss and taste aversion, in 9 out of 27 (33%). Among the patients with an unresect- able lesion and a serum albumin level of less than 3.4 gdd l , weight loss of greater than 5% of usual body weight occurred in 14 out of 21 (67%); taste aversion, in 14 out of 22 (64%); and both weight loss and taste aver- sion, in 12 out of 18 (67%).

The preadmission serum albumin level did reflect the subsequent mean caloric intake. Patients whose level was 3.4 gm/dl or greater, whether the tumor was re- sected or not, consumed on average a normal intake of 25.08 and 27.20 kcal/kg/day, respectively, compared with an average abnormal intake of 21 kcal/kg/day for the group with a level of less than 3.4 g d d l (p < 0.05) (normal intake based on the 1980 recommendations of

119 Fatzinger et al: Serum Albumin for Classification of Stage Ill Lung Cancer

N=5

100 - Alb 2 3.4g/IOOml

90 - Alb < 3.4g/100ml

80

70

60 PERCENT

SURVIVING

40 N = 3

30

20

10

6 12 18 24 30 36

MONTHS

Fig 7. Actuarial survival curves for two groups of patients with non- oat cell lung cancer. Seventeen patients with Stage Ill MO disease and preadmission serum albumin (Alb) levels greater than or equal to 3.4 gmldl had the primary tumor resected in addition to receiving irradia- tion and chemotherapy; 5 patients with Stage 111 MO disease and preadmission serum albumin levels of less than 3.4 gmldl were treated similarly. (NS = not statistically significant.)

Table I . TN Classification and Survival of Patients with Stage 111 MO Non-Oat Cell Lung Cancer

Albumin 3.4 gmldl Albumin < 3.4 gmldl

Resection No Resection Resection No Resection

Patient Survival Patient Survival Patient Survival Patient Survival No. Classification (mo) No. Classification (mo) No. Classification (mo) No. Classification (mo)

1 T3N2 2 1 T3N1 2 1 T3NO 4 1 T3N1 1 2 T3N2 3 2 T3N2 4 2 T3N2 9 2 T2N2 1 3 T3NO 8 3 T3N2 5 3 T3N2 9 3 T3N1 1 4 T3N2 8 4 T2N2 6 4 T2N2 11, alive 4 T3 N2 1 5 T2N2 9 5 T3N2 7 5 T1N2 18, alive 5 T3 NO 2 6 T3N2 9 6 T3N1 9 6 T3N1 2 7 T3N2 12 7 T3N2 10 7 T3NO 2 8 T3N1 19 8 T3NO 10 8 T3N1 3 9 T2N2 20 9 T3N2 14

10 T2N2 22 10 T3NO 16 11 T3 NO 28 11 T3 N2 17 12 T2N2 30, alive 12 T3 N2 18 13 T2N2 30, alive 14 T3NO 31, alive 15 T2N2 36

9 T3N2 4 10 T3NO 6 11 T3NO 6

16 T1 N2 45, alive 17 T3NO 48, alive

120 The Annals of Thoracic Surgery Vol 37 No 2 February 1984

Oideal weight current weight

I-

(3 70 W

60

I 8o

250 40

30

20 10

A

I 2 3 4 5 6 7 8 9 10 I I 12 I3 I4 15 16 17 18 19 2021 2223 2425

PATIENT

0 ideal weight l c u r r e n t weight

100

I Oideal weight I current weight

90 Y - 80 I-

2 60 I- W

3 50 I 70

60 > 3 50 40 ' 30 P

40

20 ' 30 20

10 I0

I 70 90

Y - 80

I 2 3 4 5 6 7 8 9 I0 I I 12 13 14 15 16 17 18 1920

PATIENT B Fig 8. Comparison of ideal and current body weight in patients with Stage 111 non-oat cell lung cancer at the time of diagnosis. Only those for whom an accurate weight was obtained are included. (A) Patients with preadmission albumin levels of less than 3.4 grnldl. (B) Patients with preadmission levels greater than or equal to 3.4 gmldl and unre- sectable primay tumors. (C) Patients with preadmission levels greater than or equal to 3.4 grnldl and resectable prima y tumors.

C

I 2 3 4 5 6 7 8 9 1 0 1 1

PATIENT

121 Fatzinger et al: Serum Albumin for Classification of Stage 111 Lung Cancer

4 1 1 P " 60m0

r - i 21 1-

2 + 14 21 28 DAYS OF J FEEDING

Fig 9 . Preadmission and subsequent serum albumin levels in 5 pa- tients with Stage 111 MO non-oat cell cancer and hypoalbuminemia who were force-fed. Of these 5 patients with potentially resectable tumors, only 2 undenuent resection (R), I of whom was a long-term survivor. Review of his surgical pathology specimen suggested that his tumor may not have been correctly classified. See text for details. (U = no resection; J = jejunostorny.)

the Food and Drug Administration, 25 to 40 kcal/kg/ day). The consumption of the individual patient, how- ever, could not be predicted by the preadmission albu- min level. As expected, patients with a serum albumin level of 3.4 g d d l or greater drank substantially more diet supplement than those with a lower level.

Figure 9 shows the result of intensive nutritional sup- port of 5 Stage 111 MO patients with a potentially resect- able primary tumor and a preadmission albumin level of less than 3.4 g d d l . They were force-fed, once the full- strength formula was tolerated, with 38 to 81 kcaYkg/day of tube feedings through the period of clinical evalua- tion, preoperative radiation therapy, and after surgical resection if performed. The time to reach full-strength toleration was from 1 to 52 days (Table 2).

Of the 5 patients, 2 eventually underwent resection, and in both the albumin levels rose above 3.4 g d d l after resection. One of them remains alive beyond 60 months. His admission albumin level was the highest of the group, and a review of his surgical specimen suggested that the primary tumor might have been a T2 instead of a T3, which would make him a Stage I patient. The second

patient died at 9 months. The remaining 3 patients showed progressive systemic weakness and died with- out resection.

Comment A highly significant correlation has been found between a patient's serum albumin level, nutritional state as as- sessed with sophisticated body composition measure- ments, and survival from a malignant disease [2, 8, 9].* The confidence limits about the regressions are large, mainly as a result of different states of hydration and fluid redistribution that occur in hospitalized patients [lo]. Consequently, the use of an in-hospital serum albu- min level to assess sequentially a patient's nutritional state or his expected survival from a malignancy are not dependable.

On the other hand, the present study has shown that a normal preadmission serum albumin level is related to an improved survival of patients with Stage 111 cancer and an unresectable primary tumor regardless of the an- atomical extent of disease present (i.e., either MO or Ml). In this regard, the serum albumin level provides a means of staging patients with disease that is beyond the limits of surgical therapy. It also helps to identify which Stage 111 MO patients with resectable tumors are most likely to benefit from resection and adjuvant ther- apy. The reason for this discriminatory ability of the serum albumin appears to be the tumor's effect on host metabolism and appetite, since when the level is less than 3.4 g d d l , caloric intake falls below normal levels and appetite abates; force-feeding does little to reverse these conditions.

Theoretically, hypoalbuminemia can result from a change in the rate of production of albumin or its de- struction, dilution by the expansion of body fluid com- partments, or abnormal losses. Attempts to explain the causes of hypoalbuminemia in patients with cancer have generated an abundance of scientific literature, the anal- ysis of which suggests that the rate of albumin synthesis

*Fatringer P, Darakjian H, Cox D, et al: Nutritional status as related to survival in Stage 111 carcinoma of the lung. (in preparation, 1984)

Tablz 2 . Results of Enteral Force-Feeding in Patients with Preadmission Serum Albumin Level of Less Than 3.4 gmldl

Patient TN Days to Reach Kilocaloriesd Intake per Survival Average No. of Average Caloric

No. Classification Full-Strength FeedingsaSb (tube + orally) Tube on Full Strength Treatment (mo)

1 T3? NO 1 2 T3 N2 52' 3 T3 N1 13 4 T3 N2 4 5 T2 N2 4

40 42 81 63 38

600 1,532 3,342 1,812 2,043

Surgical resection 60+ Surgical resection 9 No resection 6 No resection 4 No resection <1

"Intramural jejunostomy feeding tube. bFull strength = 1 cal per milliliter. 'Nasogashic feeding tube. dKilocalories per kilogram of body weight

122 The Annals of Thoracic Surgery Vol 37 No 2 February 1984

is decreased in the presence of a malignancy [ l l ] . The data in support of this, however, are based on either indirect measurements or incomplete assessment of all the involved albumin pools. As a consequence, a malig- nant cause for decreased albumin synthesis is not uni- versally accepted. Yet, when the bulk of clinical and labo- ratory data is considered, the inference suggests that hypoalbuminemia is due to some extent to decreased production brought about by metabolic changes induced by the presence of the tumor.

The observations of Torporek [12] have shown that the presence of a tumor can induce an otherwise ”nor- mal” liver to produce serum proteins with an abnormal pattern. Similarly, the work of Goodlad and Raymnond [13] and Clark and Goodlad [14] has demonstrated that in tumor-bearing rats there is a marked inhibition of amino acid incorporation by muscle polyribosomes. These findings are interpreted to suggest that host meta- bolic alterations can be initiated and mediated through a serum factor released by the tumor similar to the mecha- nism postulated for the paraneoplastic syndromes asso- ciated with lung cancer. The application of these studies specifically to albumin metabolism in patients with can- cer needs to be investigated further but the current clini- cal evidence suggests that a low level of serum albumin results from a diminished synthesis of a greater mag- nitude than can be explained by a decreased intake of calories and substrate that so often occurs in the pres- ence of cancer. Consequently, the detection of pread- mission hypoalbuminemia in a patient with cancer may signal the onset of an end-stage metabolic effect of the tumor on the host’s protein metabolism.

The current study also suggests that when hypoal- buminemia occurs, its effect on survival is usually irre- versible despite the forced delivery of adequate nutri- tional needs and the removal or destruction of the tumor by a combination of irradiation, operation, and chemo- therapy. Other serum proteins with a shorter half-life than albumin, such as transferrin and prealbumin, may indicate this metabolic effect earlier, perhaps when it is still reversible.

Clinical experience has taught that hypoalbuminemia, in addition to indicating a shorter survival, can con- tribute to several other conditions. These include the inability of the patient to tolerate chemotherapy, radio- therapy, or operation [15], the loss of immunological competence [15], and the altering of respiratory homeo- stasis [16]. Thus, the cumulative effects of hypoalbu- minemia on the viability of a patient with cancer can be overwhelmingly suppressive.

On the basis of our experience, we conclude that a preadmission serum albumin level of less than 3.4 gm/ dl, in the absence of other existing known causes such as cardiac or renal failure, indicates that the effect of the malignancy on the host is far advanced regardless of the anatomical extent of the tumor and that existing thera-

peutic regimens are likely to fail. These patients should receive enteral nutritional support during their evalua- tion and therapy, if the latter is undertaken. On the other hand, Stage 111 patients with a preadmission serum albumin level of 3.4 g d d l or greater are more likely to receive positive benefits from existing therapy, and if their primary tumor is resectable, the combination of irradiation, operation, and chemotherapy can give an expectative median survival of 20.5 months.

Supported by the Partisan Lung Cancer Research Fund and a grant from Ross Laboratories.

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

Mintz U, DeMeester TR, Golomb HM, et al: Sequential staging in bronchogenic carcinoma. Chest 76653, 1979 Bozzetti F, Migleavacca S, Scolli A, et al: Impact of cancer, type, site, stage and treatment on the nutritional status of patients. Ann Surg 196170, 1982 Dow D, Pinto PV: Determination of serum albumin on the SMA-12/30 (hospital model using Bromocresol green). Clin Chem 15:1006, 1969 Bitran ID, Desser RK, DeMeester TR, et al: Treatment of metastatic (Stage 111 M1) non-oat cell bronchogenic carci- noma with cyclophosphamide, Adriamycin, methotrexate and procarbazine (CAMP). JAMA 240:2743, 1978 Bitran ID, Desser RK, DeMeester TR, et al: Combined mo- dality therapy for the treatment of Stage 111 MO non-oat cell bronchogenic carcinoma. Cancer Treat Rep 62:327, 1978 Jelliffee DB: The assessment of the nutritional status of the community. Monograph no. 53. Geneva, Switzerland, World Health Organization, 1966 Hadlai-Hill C, Wie NH (eds): The SPS Manual. Second edi- tion. New York, McGraw-Hill, 1981 Forse RA, Shiggal HM: Serum albumin and nutritional status. J Parent Ent Nutr 4:450, 1980 Levin L, Gevers W: Metabolic alterations in cancer: Part 11. Protein and fat metabolism. S Afr Med J 59553, 1981 Courtney ME, Greenc HI, Folk CC, et al: Rapidly declining serum albumin values in newly hospitalized patients: prev- alence, severity, and contributory factors. J Parent Ent Nutr 6:143, 1982 Costa G: Cachexia, the metabolic component of neoplastic diseases. Cancer Res 372327, 1977 Torporek M: Effects of whole blood on albumin fraction from tumor-bearing rats on liver protein synthesis. Cancer Res 33:2579, 1973 Goodlad GAJ, Raymnond MJ: The action of the Walker 256 carcinoma and toxohormone on amino acid incorporation into diaphragm protein. Eur J Cancer 9:139, 1973 Clark CM, Goodlad GAJ: Activity of gastrocnemius and soleus polyribosome in rats bearing the Walker 256 carci- noma. Eur J Cancer Ei:647, 1972 Costa G: Cachexia: the metabolic component of neoplastic disease. Prog Exp Tumor Res 3321, 1963 Doekel RC, Zwillick CW, Scoggin CH, et al: Clinical semi- starvation: depression of hypoxic ventilatory response. N Engl J Med 295:358, 1976