Ectopic Adrenocorticotrophic (ACTH) Syndrome and Small Cell Carcinoma of the Lung-Assessment of Clinical Implications in Patients on Combination Chemotherapy
MARTIN D. ABELOFF, MD, DONALD L. TRUMP, MD, AND STEPHEN B. BAYLIN, MD
Small cell carcinoma of the lung is the tumor most commonly associated with ectopic ACTH production and hypercortisolism. The relationship between this paraneoplastic endocrine syndrome and the clinical course of the tumor is examined in this review of patients with the ectopic ACTH syndrome and small cell carcinoma seen at The Johns Hopkins Oncology Cen- ter between 1973 and 1979. Five of 157 (3.2%) patients with small cell carcinoma were clinically diagnosed as having the ectopic ACTH syndrome. The onset of this endocrine symdrome ap- peared to coincide with a more aggressive phase of the course of small cell carcinoma. Fur- ther analysis of these cases suggests that the development of the ectopic ACTH syndrome may reflect changes in cell populations within the tumor and/or alterations in tumor behavior with time and perhaps with the effects of drug therapy.
Cancer 48:1082-1087, 1981.
T HE PRODUCTION OF HORMONES by nonendocrine tumors with resultant clinical syndromes second- ary to hormone excess is a well-recognized concomi- tant of neoplasia. Approximately 15 hormones have been reported to be ectopically secreted by more than 20 different tumor types.1 Ectopic adrenocorticotrophic hormone (ACTH) production was the first ectopic hor- mone syndrome studied in depth and continues to be of intense interest to clinical and laboratory researchers.
Small cell carcinoma of the lung is the tumor most commonly associated with ectopic ACTH production and hypercortisolism.2 In this report, we review our experience over the past six years with small cell car- cinoma and the ectopic ACTH syndrome with em-
From The Johns Hopkins Oncology Center and Department of Medicine, The Johns Hopkins Hospital, Baltimore, Maryland. Sup- ported in part by Grant ACS PDT-108 from the American Cancer Society. Dr. Trump is the recipient of an American Cancer Society junior faculty clinical fellowship. Dr. Baylin is the recipient of a research cancer development award K04-CA-00027 from the Na- tional Institutes of Health.
Address for reprints: Dr. Martin D. Abeloff, Room 128, Oncology Center, The Johns Hopkins Hospital, 600 N. Wolfe Street, Balti- more, MD 21205.
The authors thank the laboratories of Dr. David N. Orth (Grant 5-R01-CA11685 from National Cancer Institute) and Dr. Donald S. Gann for performing the ACTH radioimmunoassays. They also thank Nancy B. Kramer for excellent secretarial assistance.
Accepted for publication March 16, 1981.
phasis on the relationship between this endocrine syn- drome and the clinical course of this tumor during treatment with modern combination chemotherapy. We describe in detail a case in which clinical and bio- chemical evidence of hypercortisolism developed after chemotherapy had resulted in a marked reduction in tumor burden and resolution of an associated syn- drome of inappropriate antidiuretic hormone secretion (SIADH).
Patients and Methods
Between July 1, 1973, and July 1, 1979, 157 patients with small cell carcinoma were entered on therapeutic research studies in The Johns Hopkins Oncology Center.3-5 These patients were carefully staged for extent of disease prior to initiation of therapy and were serially monitored in order to assess therapeutic re- sponse. Five of these patients (3.2%) were clinically diagnosed as having hypercortisolism due to ACTH production. The most recently diagnosed case which stimulated interest in the relationship between the ectopic ACTH syndrome and the course of small cell carcinoma during treatment with chemotherapy is de- scribed in detail in the case report. The clinical and laboratory features of the 5 cases are outlined in Ta- bles 1 and 2.
Case Report
Patient 1, a 55-year-old female, was found in October 1978 to have right paratracheal adenopathy and an anterior mediastinal mass on a routine chest x-ray examination. Physical examination and the remainder of the laboratory tests were normal with the exception of a serum sodium level of 121 mEq/liter. Serum potassium was 5.0 mEq/level.
The patient underwent rigid bronchoscopy with trans- tracheal biopsy. The biopsy specimen did not reveal any tumor. Mediastinoscopy was recommended but the patient postponed the procedure. In February 1979, mediastinos- copy was performed and revealed small cell carcinoma of the lung in a right paratracheal lymph node.
The patient was transferred to The Johns Hopkins Oncol- ogy Center where she was noted to be a depressed woman who had no physical stigmata of Cushing’s syndrome. An enlarged right supraclavicular node was detected but no other adenopathy or organomegaly were noted. Bone mar- row aspirate and biopsy, radionuclide scans of liver, bone and brain were normal. Endocrine evaluation revealed hyponatremia (serum sodium 123 mEq/liter), plasma hypoosmolality (275 mOsm/kg H2O), inappropriately con- centrated urine (396 mOsm/kg H2O), and increased urine sodium content (74 mEq/liter). Serum potassium was 4.6 mEq/liter. Measurements of plasma cortisol concentrations revealed a slight abnormality in diurnal variation: the 8 AM plasma cortisol level was 15.9 ug/dl and 4 PM level was 9.8 µg/dl. A low-dose dexamethasone suppression test was normal (8 AM plasma cortisol was 3.9 ug/dl after 1 mgm dexamethasone administered at 12 PM) and 24-hour urinary cortisol was also normal (29.3 ng). An 8 AM plasma ACTH level was 40 pg/ml which was within normal limits.
The patient was started on experimental chemotherapy5 with cyclophosphamide 2400 mgm/m2 (3450 mg) and doxorubicin 60 mg/m2 (85 mg) intravenously on day 1 of therapy and VP-16-213 125 mg/m2 (180 mg) on days 1, 3, and 5. Disappearance of the supraclavicular node and marked decrease in the chest masses were noted and the patient re- ceived a second course of induction chemotherapy with the same cytotoxic drugs on April 2, 1979. At that time, the hy- ponatremia had resolved without water restriction. On April 19, repeat 8AM and 4 PM serum cortisol levels were elevated at 33.7 and 20.7 ug/dl, respectively and hypokalemia (serum potassium 2.9 mEq/liter) was first noted.
In May the serum potassium dropped to 2.3 mEq/liter and hyperglycemia (glucose 170 mg/dl) was first noted. The pa- tient was treated with supplemental oral potassium and started on outpatient chemotherapy.
The patient was readmitted to The Johns Hopkins Oncol- ogy Center on June 14, 1979, for management of leukopenia. The patient reported marked proximal muscle weakness, in- creasing appetite, weight gain and headaches. Physical examination revealed a chronically ill appearing woman with Cushingoid facies. Blood pressure was 160/120 mm Hg. Generalized hyperpigmentation, a buffalo hump, nonpig- mented abdominal striae, hirsutism and trace peripheral edema were observed. Serum sodium was 145 mEq/liter, potassium 2.2 mEq/liter, and glucose 213 mg/dl. There was
no evidence of tumor other than residual right paratracheal adenopathy. Shortly after admission, the patient became fe- brile and Staphylococcus aureus sepsis was diagnosed.
Despite the successful treatment of the infection with systemic antibiotics and return of the white blood cell count to normal, the patient continued to do poorly. The clinical and laboratory data indicated hypercortisolism and the bio- chemical studies strongly suggested ectopic production of ACTH by the tumor (Table 2). Further search for clinically unidentified tumor included a negative bone scan, normal computed tomography (CT scan) of the head, and an ab- dominal CT scan which showed enlarged subcrural lymph nodes. Serum potassium level remained low despite supplementation with as much as 200 mEq potassium per day in conjunction with spironolactone. Insulin administra- tion was required to control the hyperglycemia.
In an attempt to eradicate the tumor which was the pre- sumed source of the ACTH, radiotherapy was administered to the right paratracheal area and further radiation therapy was planned to encompass the subcrural nodes. Treatment with mitotane (op-DDD) was planned if improvement in the hypercortisolism was not seen. However, on July 24 she developed fever, pancytopenia, increasing weakness and dyspnea. Localized herpes zoster and Candida pharyngitis were diagnosed. A chest x-ray film revealed pulmonary in- filtrates in right upper lobe and in the left lower lobe. Hemat- ocrit was 23, white blood cell count 175 cells/mm3, platelets 6000/mm3. Bone marrow aspirate and biopsy revealed hypoplasia of all cell lines but no identifiable tumor. Serum sodium was 144 mEq/liter and potassium was 2.4. Treatment was begun with systemic antibiotics but the patient’s respi- ratory function declined. Despite supportive care, the pa- tient died on July 30, 1979. Permission for autopsy was not granted.
Results
The clinical characteristics and the relationship of the ectopic ACTH syndrome to the course of’ the tumor in our five cases are outlined in Table 1. There was no unusual distribution regarding age at diagnosis, sex, or race. All five patients were heavy cigarette smokers. All patients presented with tumor that had extended beyond the hemithorax. The ectopic ACTH syndrome was clinically detectable in only one of the five cases (Case 3) at the time of diagnosis of small cell carcinoma. In the other four cases, the ectopic ACTH syndrome was diagnosed 3-6 1/2 months following the diagnosis of small cell carcinoma and after a partial regression of tumor had occurred as a result of cytotoxic chemotherapy. The median survival from the time of diagnosis of small cell carcinoma of the five cases with ectopic ACTH syndrome was seven months and none of the patients survived one year after the diagnosis of cancer. The median survival of 152 pa- tients with small cell carcinoma who were entered onto
| Case no. | Age (yr), race and sex | Interval between diagnosis of SCC and ectopic ACTH syndrome (mos) | Therapy* | Responset of tumor to therapy | Responset of ectopic ACTH syndrome to therapy | Survival from diagnosis of SCC (mos)/ survival from diagnosis of ectopic ACTH syndrome (mos) | Cause of death |
|---|---|---|---|---|---|---|---|
| 1 | 55,W,F | 6 | CAV | PR | NR | 91/2/31/2 | Pulmonary infection and pancytopenia |
| 2 | 50,W,M | 61/2 | CAV | PR | NR | 81/2/11/2 | Progressive tumor |
| 3 | 62,B,F | 0 | High-dose cyclophosphamide + COMB | PR | NR | 51/2 | Progressive tumor |
| 4 | 65,W,M | 6 | BCOP | PR | NR | 7/1 | Sepsis and pancytopenia |
| 5 | 76,B,M | 3 | BCOP | PR | NR | 4/1 | Progressive tumor |
* CAV = cyclophosphamide, Adriamycin (doxorubicin), VP-16-213.
COMB = cyclophosphamide, Oncovin, methotrexate, BCNU.
BCOP = BCNU, cyclophosphamide, Oncovin, procarbazine.
+ PR = partial response NR = no response
therapeutic protocols between 1973 and 1979 was 10 months and 39% (59/152) of these patients survived more than one year. At the time of diagnosis of clinical hypercortisolism, the tumor was progressing in Pa-
tients 2, 4, and 5 and appeared to be stable in Patient 1. The diagnosis of hypercortisolism occurred 1-31/2 months prior to death in the four patients who devel- oped this syndrome during their antitumor therapy.
| Endocrine assessment | Normal values | Case no. 1 | Case no. 2 | Case no. 3 | Case no. 4 | Case no. 5 |
|---|---|---|---|---|---|---|
| 24 hr urinary 17-ketosteroids | 6-15 ng/24 hr | 29 | - | 60 | - | - |
| 24 hr urinary 17-hydroxycorti- costeroids | 2-6 ng/24 hr | 73 | - | 27 | 171 | - |
| 24 hr urinary free cortisol | 14-74 ng/24 hr | >1000 | 606 | - | >100 | - |
| 8 AM serum cortisol | 7-21 µg/dl | 38 | 83 | 58 | >50 | - |
| 4 PM serum cortisol | 3-10 µg/dl | 36 | - | - | - | 15 |
| Cortisol suppression after dexamethasone | No | No | - | No | - | |
| 8 AM plasma ACTH | 10-120 pg/ml | 202(275 at 8PM) | 614 | - | - | 368 |
| Physical findings of hypercortisolism | Cushingoid facies, buffalo hump, hirsutism, hyper- pigmentation, ab- dominal striae, edema, proximal muscle weakness | Cushingoid facies, buffalo hump, hyperpigmentation, ecchymoses, edema, proximal muscle weakness | Proximal muscle weakness, edema | Hyperpig- mentation, edema | Edema | |
| Hypokalemia | Yes-required oral potassium + spironolactone | Yes-required oral potassium + spironolactone | Yes- required oral potassium + spironolactone | Yes-required oral potassium | Yes- required oral potassium | |
| Hyperglycemia | Yes-required insulin | Yes-insulin not required | Yes- required insulin | No | No | |
| Other paraneoplastic syndromes | SIADH at diagnosis of SCC | SIADH preceded ectopic ACTH, cerebellar ataxia | No | No | No |
There was no definite improvement of the ectopic ACTH syndrome as a result of antitumor therapy in any of our cases. Three of the patients (Cases 2, 3, and 5) died as a result of progressive growth of tumor while the cause of death in Patients 1 and 4 was infection secondary to leukopenia. There was little evidence of bulk deposits of tumor at the time of death in Patients 1 and 4.
Autopsy was performed in only one of these five cases. In Case 2, postmortem examination revealed small cell carcinoma in the right lung with metastases to the right hilar nodes, pleural surface of the right diaphragm, pericardium, abdominal nodes, bone mar- row, vertebrae, and liver. Extensive lymphangitic metastases were seen in both lungs. Marked bilateral adrenal cortical hyperplasia with no nodularity or adenomata were noted at autopsy. Both adrenals were also infiltrated by carcinoma. In Case 2, im- munoperoxidase studies using a specific antibody against ACTH confirmed the presence of ACTH within tumor cells from autopsy tissue.
Table 2 summarizes the biochemical and endocrine data from these five cases. The biochemical data were collected as part of the routine medical evaluation over a period of six years, and, therefore, there was varia- tion in the endocrine evaluation. Patients 1 and 2 had clinically overt hypercortisolism despite the short du- ration of their disease. The initial manifestations of hypercortisolism included hypokalemia and peripheral edema in all patients. The hypokalemia presented se- vere management problems in Patients 1, 2, and 3. One of the clues to the diagnosis of hypercortisolism in Patient 3 was exacerbation of adult onset diabetes mellitus. Plasma ACTH was measured by radioim- munoassay in Patients 1, 2, and 5 and was markedly elevated in all three. The two patients (Cases 3 and 4) in whom plasma ACTH levels were not obtained had marked elevations of serum cortisol and 24-hour uri- nary 17-ketosteroids or 17-hydroxycorticosteroids. Cortisol production was found to be nonsuppressible following dexamethasone suppression in Patient 4. The excess production of ACTH resulted in serious clinical morbidity for Patients 1-3 although the endo- crine problem could not be directly identified as the cause of death in any of these patients. SIADH was identified prior to the diagnosis of the ACTH syn- drome in Patients 1 and 2. In Patient 1, clinical evi- dence of SIADH had completely resolved at the time hypercortisolism was detected but SIADH persisted at the time of diagnosis of the ectopic ACTH syndrome in Patient 2. Cerebellar ataxia not related to a mass lesion was diagnosed in Case 2. No other paraneoplastic syn- dromes were recognized in these 5 Patients.
Discussion
Although immunoreactive ACTH has been iden- tified in the tumor tissue and circulation of a high per- centage of patients with bronchogenic carcinoma, hypercortisolism due to the ectopic production of ACTH remains an infrequent complication of lung cancer.6 Prior to the treatment of small cell carcinoma with combination chemotherapy, clinically evident hypercortisolism had been reported in 2.5 to 5% of patients with small cell carcinoma of the lung.7 This frequency is similar to our experience in patients who have been treated with modern cytotoxic chemo- therapy.
The recognition of an ectopic hormone syndrome is important to the clinician because the excess hormone production may result in significant morbidity and may also be an important clue to the diagnosis of cancer or an indicator of tumor activity. There has been little comment in the literature regarding correlations be- tween the temporal relationship of the ectopic ACTH syndrome and the course of the cancer and response to combination chemotherapy. In four of our five cases, the ectopic ACTH syndrome developed following the attainment of partial regression of tumor with chemotherapy. The recognition of this ectopic hor- mone production appeared to coincide with a more aggressive phase of the patient’s illness and the overall survival of the patients with hypercortisolism was shorter than our patients with small cell carcinoma studied during the same period. Other investigators have previously noted the very poor survival in pa- tients with small cell cancer and hypercortisolism,8 but this had not been evaluated in patients undergoing modern intensive chemotherapy.
Although an ectopic hormone syndrome may pro- vide useful clinical and biologic information regarding the behavior of a tumor, clinically evident ACTH ex- cess occurs too infrequently to be of significant value in monitoring the course of patients with small cell carcinoma. However, it has been suggested that as many as 50% of patients with small cell carcinoma have subclinical evidence of hypercortisolism as man- ifested by abnormalities in glucocorticoid produc- tion.1.9 Our patient described in the case report (Case 1) as well as 38 other patients with small cell carcinoma and 22 patients with a variety of other tumors had pretreatment measurements of 8 AM and 4 PM serum cortisol levels and dexamethasone suppression tests performed as part of a separate study on water and cortisol metabolism in cancer patients.1º Seventy-nine percent of patients with small cell carcinoma and 65% of patients with other cancers had abnormal diurnal
patterns of cortisol secretion but cortisol levels were readily suppressed by 1 mg dexamethasone in greater than 70% of all patients in this study. In fact, Patient 1 had suppressible cortisol secretion documented only two months prior to the appearance of overt signs of hypercortisolism and autonomous adrenal function. Thus, there does not appear to be a significant differ- ence in the incidence of subclinical abnormalities of adrenal function in patients with small cell carcinoma as compared with patients with other cancers. It does not appear likely that biochemical measurements of adrenal function will be useful biomarkers of disease activity in patients with small cell carcinoma who do not have clinical evidence of hypercortisolism. However, sequential measurements of cortisol sup- pressibility by dexamethasone would be required to answer this question more definitively.
As suggested by the findings in our cases, ectopic hormone syndromes may also reflect changes in cell population within the tumor and/or alterations in tumor behavior with time and perhaps with the effects of drug therapy. Two of our patients (Cases 1 and 2) had SIADH in addition to the ectopic ACTH syn- drome. There have been a number of reports regarding secretion of multiple hormones by a single lung cancer.11-13 Coscia et al. documented simultaneous production of ACTH, ADH, and ß-MSH (beta- melanocyte stimulating hormone) by small cell car- cinoma tissue (obtained at autopsy) in a patient who had clinical manifestations of ectopic ACTH syndrome and SIADH.11 However, the cell populations respon- sible for the production of multiple hormones or other tumor proteins have not been well characterized in small cell carcinoma or in most other tumors. In tes- ticular tumors, separate cell populations appear to produce human chorionic gonadotropin and alpha feto-protein.14 There is ample evidence from a variety of biologic studies for heterogeneity of cell populations in solid tumors including small cell carcinoma. 15-17 Re- cently, evidence that two hormones, calcitonin and ACTH, can appear in the same tumor cells in medul- lary carcinoma of the thyroid has also appeared.18
Patient 1 in this report developed the ACTH syn- drome at a time when the tumor had significantly re- gressed and SIADH had clinically disappeared on combination chemotherapy. Given the above cited evidence for heterogeneity of tumor cell populations, it is possible that the combination chemotherapy was successful in eradicating a population of cells that pro- duced ADH but that another group of cells, which had the capacity to produce biologically active ACTH, persisted or emerged. Marked variations in the con-
centrations of peptide hormones and enzymes in the primary tumor and metastatic deposits in patients with small cell carcinoma have, in fact, been demonstrated in several studies including our own.11,13,16
Although heterogeneity of tumor cell populations has been well documented and clinical hypercor- tisolism is an uncommon occurrence in patients with lung cancer, elevated levels of plasma ACTH have been noted in the majority of patients with all types of lung cancer. However, it appears that the im- munoreactive hormone which is being measured in these patients is not biologically active ACTH but rather “big ACTH” or “pro ACTH,” a precursor molecule of ACTH.19 The usefulness of “big ACTH” as a tumor marker in lung cancer is currently under investigation. In a recent study of patients with squamous carcinoma, large cell carcinoma, and adenocarcinoma of the lung who had undergone surgi- cal resection of their primary tumor, Yalow and col- leagues concluded that the preoperative and post- operative plasma immunoreactive ACTH did not have prognostic value for remission duration or long-term survival.20 These investigators suggested that the tumor per se was not the sole source for the elevated plasma ACTH levels but that the lung tissue might be a major contributor to the circulating ACTH. Whether or not plasma “big ACTH” measured serially in pa- tients with small cell carcinoma who are treated with modern chemotherapeutic and radiotherapeutic pro- grams will be a useful biomarker of disease activity remains to be determined.
REFERENCES
1. Gilby ED, Rees LH, Bondy PK. Ectopic hormones as markers of response to therapy in cancer. In: Proceedings of the Sixth In- ternational Symposium of Biologic Characterization of Human Tu- mors. Amsterdam: American Elsevier Publishing Co. 1976:132-139.
2. Richardson RL, Greco FA, Oldham RK, et al. Tumor products and potential markers in small cell lung cancer. Semin Oncol 1978; 5:253-262.
3. Abeloff MD, Ettinger DS, Baylin SB, et al. Management of small cell carcinoma of the lung: therapy, staging and biochemical markers. Cancer 1976; 38:1394-1401.
4. Ettinger DS, Karp JE, Abeloff MD, et al. Intermittent high- dose cyclophosphamide chemotherapy in small cell carcinoma of the lung. Cancer Treat Rep 1978; 62:413-424.
5. Abeloff MD, Ettinger DS, Khouri N, et al. Intensive induction therapy for small cell carcinoma of the lung. Cancer Treat Rep 1979; 63:519-524.
6. Gewirtz G, Yalow RS. Ectopic ACTH production in car- cinoma of the lung. J Clin Invest 1974; 53:1022-1031.
7. Azzopardi JG, Freeman E, Poole G. Endocrine and metabolic disorders in bronchial carcinoma. Br Med J 1970; 4:528-529.
8. Singer W, Kovacs K, Ryan N, et al. Ectopic ACTH syndrome: clinicopathological correlations. J Clin Pathol 1978; 31:591-598.
9. Amatruda TT, Upton GV: Hyperadrenocorticism and ACTH- releasing factor. Ann NY Acad Sci 1972; 230:168-180.
10. Trump DL, Abeloff MD: Studies of water and cortisol metab- olism in patients with small cell bronchogenic carcinoma and other neoplasms. Proc Am Soc Clin Oncol May, 1980; 21:320 (Abstract).
11. Coscia M, Brown RD, Miller M, et al. Ectopic production of antidiuretic hormone (ADH), adrenocorticotrophic hormone (ACTH) and beta-melanocyte stimulating hormone (3-MSH) by an oat cell carcinoma of the lung. Am J Med 1977; 62:303-307.
12. Horai T, Nishihara H, Tateishi R, et al. Oat-cell carcinoma of the lung simultaneously producing ACTH and serotonin. J Clin En- docrinol Metabol 1973; 37:212-219.
13. Rees LH, Bloomfield GA, Rees GM, et al. Multiple hormones in a bronchial tumor. J Clin Endocrinol Metabol 1974; 38:1090-1097.
14. Kurman RJ, Scardino PT, McIntire KR, et al. Cellular lo- calization of alpha-fetoprotein and human chorionic gonadotropin in germ cell tumors of the testis using an indirect immunoperoxidase technique. Cancer 1977; 40:2136-2151.
15. Fidler IJ. Tumor heterogeneity and the biology of cancer in- vasion and metastases. Cancer Res 1978; 38:2651-2660.
16. Baylin SB, Weisburger WR, Eggleston JC, et al. Variable content of histaminase, L-dopa decarboxylase and calcitonin in small-cell carcinoma of the lung-biologic and clinical implications. N Engl J Med 1978; 299:105-110.
17. Trump DL, Mendelsohn G, Baylin SB: Discordance between plasma calcitonin and tumor-cell mass in medullary thyroid car- cinoma. N Engl J Med 1979; 301:253-255.
18. Goltzman D, Huang SN, Broune C, et al. Adrenocorticotro- pin and calcitonin in medullary thyroid carcinoma: frequency of occurrence and localization in the same cell type by im- munocytochemistry. J Clin Endocrinol Metabol 1979; 49:364-369.
19. Wolfsen AR, Odell WD: ProACTH: Use for early detection of lung cancer. Am J Med 1979; 66:765-772.
20. Yalow RS, Eastridge MD, Higgins G, et al. Plasma and tumor ACTH in carcinoma of the lung. Cancer 1979; 44:1789-1792.
Erratum
In a recently published article (Manard AM, Chasle J, Marnay J, et al. Autopsy Findings in 11 Cases of Esophageal Cancer. Cancer 1981; 48:329-335), one of the authors, D. Bonnet, was omitted from the list of authors and mistakenly included in the acknowledg- ments footnote. We sincerely regret this error.