Multi-Organ Distant Metastases Confer Worse Disease-Specific Survival in Differentiated Thyroid Cancer

Abstract

Background:

Differentiated thyroid cancer (DTC) is usually associated with an excellent prognosis. With appropriate management of disease in the neck, death from thyroid cancer is more commonly related to the impact of distant metastases rather than locoregional recurrence. However, many patients with distant metastases can have very long periods of progression-free survival. The aims of this study were to determine the impact of single and multi-organ distant metastases (SODM and MODM) on survival, and identify factors that predict SODM progressing to MODM.

Methods:

An institutional database of 3664 previously untreated patients with DTC who had surgery between 1986 and 2010 was reviewed. One hundred and twenty-five (3.4%) patients developed distant metastases, of whom 93 developed SODM and 32 MODM. Overall survival was determined for each group by the Kaplan–Meier method. Factors predictive of MODM were identified by univariate and multivariate analysis. Multi-organ recurrence-free survival (MORFS) is a measure of SODM progressing to MODM disease. MORFS was calculated from the time of first distant metastasis to the time of second organ involvement by distant metastases

Results:

The median age was 56 years (range 5–86 years). The median follow-up was 77 and 79 months (range 2–318 months) for the SODM and MODM groups respectively. SODM patients had five-year survival of 77.6% from the time of first distant metastasis, whereas MODM patients had a significantly poorer survival of just 15.3% from the time of second organ distant metastasis to death (p<0.001). The median time from first to second distant metastasis was 14.7 months (range 1–121 months). Seventy-one (57%) patients had M1 disease at presentation. Being aged ≥45 years (p=0.05) and having an unstimulated serum thyroglobulin (Tg) level of ≥30 ng/mL at the time of diagnosis of initial distant metastasis (p<0.001) were univariate predictors of developing MODM. Controlling for age, an unstimulated serum Tg level of ≥30 ng/mL conferred a hazard ratio of 5.77 ([CI 2.13–15.64]; p=0.001) for diagnosis of MODM.

Conclusions:

MODM are associated with a poorer survival compared to patients with SODM. A serum Tg level >30 ng/mL at the time of first distant metastases confers more than a fivefold risk of having MODM identified during follow-up.

Introduction

Differentiated thyroid cancer (DTC) is generally associated with an excellent prognosis. In the majority of patients with DTC, the cause of death is due to distant metastases rather than locoregional recurrence. Between 1% and 4% of DTC patients present with distant disease at the time of initial diagnosis (1 –3), and metastatic disease develops in an additional 7–23% of patients (4 –7). The 10-year survival of patients with metastatic DTC ranges between 25% and 70% (1,2,6,8 –14). Although some patients achieve complete remission, many patients with isolated distant metastases live with stable disease for many years. Others, however, succumb to rapidly progressive disease and die. This variability in clinical outcomes suggests significant heterogeneity in thyroid metastatic disease.

The current study is a retrospective review of 125 patients with DTC with distant metastases at presentation (M1), or distant recurrence(s) during follow-up. Our aim was to determine the outcome of patients with DTC who had single organ distant metastases (SODM) and multi-organ distant metastases (MODM) and identify factors predictive of multi-organ distant recurrence.

Materials and Methods

Following institutional review board approval, the records of 3664 consecutive patients treated with surgery for DTC at the Memorial Sloan Kettering Cancer Center (MSKCC) between 1986 and 2010 were reviewed. One hundred and twenty-five patients with evidence of distant metastases, at any point, were identified for analysis.

Patient demographics and surgical and histopathologic details were recorded. Treatment details for distant metastases, including the use of therapeutic radioactive iodine (RAI), radiotherapy, and subsequent surgeries, were recorded. All curative resection of distant metastases were carried out to treat a patient's first site of involvement and were limited to the treatment of lung and bone metastases only.

Distant recurrence was identified on imaging studies such as RAI whole-body imaging, computed tomography (CT), and positron emission tomography (PET) scans, and confirmed with cytology or histopathology when available. The date, location, and treatment of all distant recurrences were recorded. Death was recorded from the social security death index and verified by hospital record, when available. RAI avidity of the first distant metastasis was recorded using ¹³¹I dosimetry. Patients who did not receive dosimetry because their tumors were believed to be RAI nonavid (n=26) at the time of distant metastases diagnosis were included in the RAI nonavid group.

Prior to 2000, serum thyroglobulin (Tg) was not routinely used to detect recurrence. However, since 2000, this has become routine practice. All serum Tg tests performed at our institution were captured, and the levels corresponding to the first distant metastases diagnosis were analyzed. Serum Tg was considered to be nonstimulated if the concurrent serum thyrotropin (TSH) level was <5.5 mU/L. Only nonstimulated serum Tg levels without interfering Tg antibodies were included for analysis.

The primary outcome of interest was overall survival (OS). Patients were analyzed in two groups: those with SODM and those with MODM. Survival time was measured from the time of the diagnosis of distant metastases, rather than the time of the primary DTC diagnosis, in order to remove the effect of lead-time bias. Overall survival for SODM patients was calculated from the time of the diagnosis of first distant metastases to death, while survival for MODM patients was calculated from the time of second organ involvement by distant metastases to time of death. The secondary outcome of interest was multi-organ distant recurrence free survival (MORFS). Multi-organ recurrence was defined as the development of second organ involvement by distant metastases, for example initial involvement of metastatic lung disease followed by involvement of metastatic bone disease. The development of a secondary lung lesion in the opposite lung was not considered to be multi-organ disease. Hence, MORFS is a measure of SODM progressing to MODM disease. MORFS was calculated from the time of first distant metastasis to the time of second organ involvement by distant metastases. Patients who did not progress from SODM to MODM disease were censored at their last disease management team follow-up. The median follow-up for the entire patient cohort was 77 months (range 2–318 months), and both OS and MORFS were calculated at five years.

Statistical analysis was carried out using SPSS v21 (IBM Corp., Armonk, NY). A receiver operating characteristic (ROC) curve was used to determine the serum Tg level that maximized sensitivity and specificity for prediction of OS. Patient, tumor, and treatment characteristics were compared using Pearson's chi-square test. Survival outcomes were analyzed using the Kaplan–Meier method. Univariate analysis was carried out by the log-rank test, and multivariate analysis was carried out using the Cox proportional hazard model. A p-value of<0.05 was considered significant.

Results

Clinical, tumor, and treatment characteristics

The median age of patients with SODM and MODM was 62 years (range 9–89 years) and 65 years (range 30–83 years) respectively. Clinical, tumor, and treatment details are shown in Table 1. Sixty-one (48.8%) patients were female. Ten patients under the age of 18 years at the time of initial surgery were included in the study. All pediatric patients had RAI avid, SODM involving the lung, and were alive at the last follow-up.

Table 1.

Clinical, Tumor and Treatment Characteristics

	Single organ metastases		Multi-organ metastases
	n	%	n	%	p-Value
Sex
Female	48	51.6	13	40.6	0.28
Male	45	38.4	19	59.4
Age at diagnosis
<45 years	34	36.6	8	25	0.23
≥45 years	59	63.4	24	75
Pathological T stage (initial)
T1	12	13	5	13.7	0.22
T2	12	13	2	11.3
T3	32	34.8	17	39.5
T4	36	39.1	8	35.5
Pathological N stage (initial)
N0	26	28	11	34.4	0.47
N1a	13	14	2	6.3
N1b	54	58.1	19	59.4
M stage at presentation
M0/X	38	40.9	16	50	0.37
M1	55	59.1	16	50
RAI treatment
No	9	9.7	6	18.8	0.17
Yes	84	90.3	26	81.3
Initial RAI dose (mCi)
Continuous	82	M=187, SD= 93	26	M=210, SD= 111	0.33
Total no. of RAI therapies
Continuous	177	M=1.9, SD= 1.2	77	M=2.4, SD= 1.5	0.10
Systemic therapy
No	76	90.5	18	69.2	0.01
Yes	8	9.5	8	30.8
Radiotherapy to DM
No	88	94.6	25	78.1	0.01
Yes	5	5.4	7	21.9

Bold values indicate significance.

RAI, radioactive iodine; DM, distant metastases.

Ninety-three (74.4%) patients presented with T3/T4 primary tumors, and 88 (70.4%) had pathologically positive neck node metastases. The median follow-up for the SODM and MODM groups were 77 and 79 months (range 2–318 months) respectively. SODM occurred most commonly in the lung (84%), bone (12%), and distant lymph nodes (4%). The most common site of MODM, besides the lung (94%) and bone (81%), was the brain (41%). Seventy-one (57%) patients had M1 disease at or within six months of presentation, and the remaining 54 patients (43%) developed distant disease during follow-up. A total of 177 RAI therapies were given for the SODM patients (M=1.9 per patient), and 77 were given for MODM patients (M=2.4 per patient). At our institution, all patients are treated with TSH suppression. ATA low-risk patients receive suppression to a TSH level of 0.1–0.4 mU/L. Intermediate- and high-risk patients are all suppressed to a TSH level of ≤0.1 mU/L. The majority of patients in our study were intermediate- and high-risk, and had suppression to TSH ≤0.1 mU/L. There was no correlation between the level of TSH suppression and the development of distant metastases. Systemic therapy was given to eight (8.6%) SODM patients and eight (25%) MODM patients. Curative resection of distant metastases was carried out in nine patients (7.2%); four patients (44.4%) subsequently developed secondary sites of distant metastases. Resection of metastases did not correlate with improved multi-organ distant recurrence-free survival.

Overall survival

SODM patients had a five-year OS of 77.6% from the time of first distant metastases (Fig. 1A). For MODM patients, the five-year survival from time of second organ involvement to death was just 15.3% (Fig. 1B; p<0.001). The median time from first to second distant metastases was 14.7 months (range 1–121 months). MODM patients had more than a threefold increased risk of death compared to SODM patients (hazard ratio (HR) 3.077 [CI 1.822–5.194]; p<0.001) when adjusting for age at first distant metastases and RAI avidity.

FIG. 1.

Overall survival from time of (A) first distant metastases and (B) second distant metastases.

Predictors of MORFS

Progression from SODM to MODM was 76.1% at five years. MORFS was calculated from the time of the diagnosis of first distant metastasis to the time of second organ involvement by distant metastases. Factors predictive of MORFS were determined by univariate analysis and are shown in Table 2. Being aged ≥45 years (p=0.05), RAI nonavidity (p=0.03), and having an unstimulated serum Tg level of ≥30 ng/mL (by ROC curve; p<0.001) were univariate predictors of MORFS. Sex, histology, adjuvant RAI therapy, T stage, N stage, and M stage at presentation were not predictive of developing MODM. When controlling for age, an unstimulated serum Tg level of ≥30 ng/mL at the time of initial SODM diagnosis conferred a HR of 5.77 ([CI 2.13–15.64]; p=0.001) for developing MODM (Table 3a). Similarly, controlling for RAI nonavidity, serum Tg was the strongest predictor, conferring a HR of 7.540 ([CI 2.696–21.089]; p<0.001; Table 3b).

Table 2.

Univariate Predictors of Multi-Organ Recurrence-Free Survival

	n	No. events	5-year MORFS	p-Value
Sex
Female	61	13	79.6%	0.42
Male	64	19	73.0%
Age
<45 years	33	5	88.6%	0.05
≥45 years	92	27	71.4%
Histology
Follicular	8	3	62.5%	0.52
Hürthle	7	3	53.6%
Papillary non-TCV	110	26	80.5%
Papillary TCV	21	5	71.6%
T stage
T1	17	5	76.5%	0.31
T2	14	2	80.8%
T3	49	17	72.3%
T4	44	8	77.4%
N stage
N0/X	37	11	73.1%	0.67
N1a	15	2	85.6%
N1b	73	19	76.4%
M stage at presentation
M0/X	54	16	71.0%	0.22
M1	71	16	79.7%
Postop RAI treatment
No	15	6	54.5%	0.10
Yes	110	26	79.1%
Distant metastases RAI avidity
Avid	72	15	82.6%	0.03
Nonavid	53	17	66.1%
Metastectomy
Yes	9	4	64.8%	0.34
No	116	28	76.9%
Tg
<30 ng/mL	47	6	85.7%	<0.001
≥30 ng/mL	19	11	42.1%

Bold values indicate significance.

MORFS, multi-organ recurrence-free survival; TCV, tall-cell variant; Tg, thyroglobulin.

Table 3.

Multivariate Predictors of MORFS

		CI
	HR	Lower	Upper	p-Value
(a) Tg controlling for age
Age
<45 years
≥45 years	—	—	—	NS
Tg
<30 ng/mL
≥30 ng/mL	5.877	2.165	15.950	0.001
(b) Tg controlling for RAI avidity
RAI avidity
Avid
Nonavid	3.051	1.136	8.192	0.03
Tg
<30 ng/mL
≥30 ng/mL	7.540	2.696	21.089	<0.001

Discussion

Patients with distant metastases in DTC have varied survival outcomes ranging from 25% to 70% at 10 years from cancer diagnosis or detection of distant metastases (1,2,6,8 –14). The aims of this study were to determine the outcome of patients with single and multi-organ metastases, and identify factors predictive of developing multi-organ distant metastases. The main findings of the study are that patients with distant metastases limited to one organ system have a five-year survival of 77.6% (p<0.001), whilst those who develop second organ involvement by distant metastases have a very poor prognosis with a five-year survival of just 15.3%. Similar findings have also been reported by Ruegemer et al. (1).

Due to the difference in prognosis between SODM and MODM that we have reported, it would be valuable to predict the likelihood of a patient's distant metastases progressing to a second organ. We were therefore interested in determining factors predictive of multi-organ distant recurrence. From our study, we found that at the time of first distant metastases, being aged ≥45 years and having an unstimulated serum Tg level of ≥30 ng/mL and RAI nonavidity of first distant metastases were predictive of developing MODM disease (p=0.05, p=0.03, and p<0.001 respectively). We will discuss each of these factors individually.

With regards to age, the present study suggests that patients aged ≥45 years have a poorer five-year MORFS compared to patients aged <45 years (71.4% vs. 88.6%, p=0.05). This finding is in agreement with previous publication from our institution (9) as well as other publications (1,10,11,15,16).

With regards to RAI avidity, our study shows that RAI nonavidity was an independent predictor of poorer survival when controlling for serum Tg, conferring a HR of 3.051 ([CI 1.136–8.192]; p=0.03; Table 3b). This is also in agreement with studies in the literature (10 –12,17). For example, Durante et al. observed that patients with ¹³¹I uptake in their metastases, among other patient and tumor factors, had a lower risk of death on multivariate analysis (11). Lee et al. suggested that metastatic lesion iodine avidity had a significant impact on both overall survival as well as DSS in patients developing DM after initial treatment, but only significantly influenced DSS in patients with M1 disease at presentation (12). Cho et al. observed that RAI nonavidity was the only independent factor predicting poor prognosis in patients with lung metastases (17). Mihailovic et al. also found that RAI uptake in distant metastases was an independent predictor of better survival. However, in contrast to our study, it was observed that age had significantly greater influence on survival (10). In contrast to these studies, other older publications have suggested that RAI uptake is not prognostically important on multivariate analysis (1).

We hypothesized that serum Tg could also prognosticate survival outcome, since both RAI avidity and serum Tg are a measure of tumor differentiation. Using ROC analysis, we found that an unstimulated serum Tg level of ≥30 ng/mL at the time of first distant metastases was most predictive of MODM. Ronga et al. used a similar ROC method and observed that a stimulated Tg level of >70 ng/mL at presentation was a significant predictor of developing nodal or distant disease (18). Similar to our finding, Huang et al. showed that a low stimulated Tg level (<400 ng/mL) at the discovery of metastases was an independent predictor of improved prognosis (15). Our data suggest that an unstimulated serum Tg level of ≥30 ng/mL is an independent predictor of MODM, conferring a HR of 5.77 ([CI 2.13–15.64]; p=0.001) when compared to patients with an unstimulated serum Tg level of <30 ng/mL. Although we cannot directly compare an unstimulated Tg level of 30 ng/mL to a stimulated Tg level of 400 ng/mL, we can conclude that Tg levels at the time of first distant metastases are likely to be predictors of disease progression and outcome.

Before we discuss the implications of our findings, it is important to highlight the possible limitations of the study. First, of 3664 patients, only 125 (3.5%) developed distant metastases. This is lower than what one might expect from a tertiary referral center such as MSKCC. The usual rate of M1 disease in papillary thyroid cancer (PTC) is about 5%, and in follicular thyroid cancer (FTC) about 10–15%. Over the course of 20 years, about 10% of patients with PTC and about 20% of patients with FTC will develop distant metastases. The reasons for this low rate of distant metastases are multifactorial. Many of our patients have low-risk cancer and therefore will never develop distant disease. It could be argued that treatment at a high-volume center may yield improved prognosis. Our follow-up is short, and we also have some patients who are lost to follow-up. Therefore, it is likely that there are additional patients in our cohort who will develop distant metastases with ongoing follow-up. However, despite this low rate of distant metastases in our cohort, these numbers are still larger than other reports in the literature.

Second, not all data points were available for collection. For example, when clinicians thought that disease was unlikely to concentrate iodine, RAI scans were sometimes not performed. In addition, patients evaluated prior to 2000 did not have routine Tg and TSH testing, and thus have missing data. However, of the 26 patients diagnosed with their first distant metastases prior to 2000 who did not have a Tg measurement, there was an equal distribution between the SODM and MODM groups: 19 (20%) SODM patients and 7 (22%) MODM patients.

Lastly, due the retrospective nature of our study, we can never fully account for the selection bias that takes place at both the physician and patient level. However, one can also argue that we have a large cohort of patients treated by a multidisciplinary team with decades of experience in the management of thyroid cancer. Although the approach to therapy within our institution has evolved over time, our approach to risk stratification and subsequent treatment decisions has remained stable. This will help to limit the selection biases involved in patient management.

Despite these limitations, our study has important clinical implications for physicians and patients. First, patients with SODM have a relatively good five-year overall survival rate of 77.6% from time of first distant metastasis. This prognostic information is extremely useful during patient counseling. Second, it appears that it is the development of distant metastases rather than the timing at presentation or during follow-up that is prognostically important in DTC. Third, older patients with nonavid metastases and serum Tg levels of ≥30 ng/mL are at high risk of MODM. These findings may alert the physician to potential multi-organ distant recurrences and lower the threshold for follow-up imaging studies. Early identification of MODM may also allow for more timely initiation of systemic or local therapy. These high-risk patients may also be suitable for novel targeted drug therapies in future clinical trials.

Footnotes

Author Disclosure Statement

The authors have no commercial associations or potential conflicts of interests to declare.

References

Ruegemer

, Hay

, Bergstralh

, Ryan

, Offord

, Gorman

. 1988. Distant metastases in differentiated thyroid carcinoma: a multivariate analysis of prognostic variables. J Clin Endocrinol Metab, 67:501–508.

Shaha

, Shah

, Loree

. 1997. Differentiated thyroid cancer presenting initially with distant metastasis. Am J Surg, 174:474–476.

Hoie

, Stenwig

, Kullmann

, Lindegaard

. 1988. Distant metastases in papillary thyroid cancer. A review of 91 patients. Cancer, 61:1–6.

Massin

, Savoie

, Garnier

, Guiraudon

, Leger

, Bacourt

. 1984. Pulmonary metastases in differentiated thyroid carcinoma. Study of 58 cases with implications for the primary tumor treatment. Cancer, 53:982–992.

Casara

, Rubello

, Saladini

, Gallo

, Masarotto

, Busnardo

. 1991. Distant metastases in differentiated thyroid cancer: long-term results of radioiodine treatment and statistical analysis of prognostic factors in 214 patients. Tumori, 77:432–436.

Jonklaas

, Sarlis

, Litofsky

, Ain

, Bigos

, Brierley

, Cooper

, Haugen

, Ladenson

, Magner

, Robbins

, Ross

, Skarulis

, Maxon

, Sherman

. 2006. Outcomes of patients with differentiated thyroid carcinoma following initial therapy. Thyroid, 16:1229–1242.

Samaan

, Schultz

, Haynie

, Ordonez

. 1985. Pulmonary metastasis of differentiated thyroid carcinoma: treatment results in 101 patients. J Clin Endocrinol Metab, 60:376–380.

Schlumberger

, Challeton

, De Vathaire

, Travagli

, Gardet

, Lumbroso

, Francese

, Fontaine

, Ricard

, Parmentier

. 1996. Radioactive iodine treatment and external radiotherapy for lung and bone metastases from thyroid carcinoma. J Nucl Med, 37:598–605.

Shoup

, Stojadinovic

, Nissan

, Ghossein

, Freedman

, Brennan

, Shah

, Shaha

. 2003. Prognostic indicators of outcomes in patients with distant metastases from differentiated thyroid carcinoma. J Am Coll Surg, 197:191–197.

10.

Mihailovic

, Stefanovic

, Malesevic

, Markoski

. 2009. The importance of age over radioiodine avidity as a prognostic factor in differentiated thyroid carcinoma with distant metastases. Thyroid, 19:227–232.

11.

Durante

, Haddy

, Baudin

, Leboulleux

, Hartl

, Travagli

, Caillou

, Ricard

, Lumbroso

, De Vathaire

, Schlumberger

. 2006. Long-term outcome of 444 patients with distant metastases from papillary and follicular thyroid carcinoma: benefits and limits of radioiodine therapy. J Clin Endocrinol Metab, 91:2892–2899.

12.

Lee

, Soh

. 2010. Differentiated thyroid carcinoma presenting with distant metastasis at initial diagnosis clinical outcomes and prognostic factors. Ann Surg, 251:114–119.

13.

Benbassat

, Mechlis-Frish

, Hirsch

. 2006. Clinicopathological characteristics and long-term outcome in patients with distant metastases from differentiated thyroid cancer. World J Surg, 30:1088–1095.

14.

Schlumberger

, Tubiana

, De Vathaire

, Hill

, Gardet

, Travagli

, Fragu

, Lumbroso

, Caillou

, Parmentier

. 1986. Long-term results of treatment of 283 patients with lung and bone metastases from differentiated thyroid carcinoma. J Clin Endocrinol Metab, 63:960–967.

15.

Huang

, Chou

, Liu

, Tung

, Chen

, Kuo

, Hsieh

, Wang

. 2012. Long-term outcomes of distant metastasis from differentiated thyroid carcinoma. Clin Endocrinol (Oxf), 76:439–447.

16.

Sugitani

, Fujimoto

, Yamamoto

. 2008. Papillary thyroid carcinoma with distant metastases: survival predictors and the importance of local control. Surgery, 143:35–42.

17.

Cho

, Choi

, Yeom

, Lim

, Moon

, Park

, Chung

, Cho

, Yi

, Park

. 2013. Long-Term Prognosis of Differentiated Thyroid Cancer with Lung Metastasis in Korea and Its Prognostic Factors. Thyroid, 24:277–286.

18.

Ronga

, Filesi

, Ventroni

, Vestri

, Signore

. 1999. Value of the first serum thyroglobulin level after total thyroidectomy for the diagnosis of metastases from differentiated thyroid carcinoma. Eur J Nucl Med, 26:1448–1452.