In Differentiated Thyroid Cancer,an Incomplete Structural Response to Therapy Is Associated with Significantly Worse Clinical Outcomes Than Only an Incomplete Thyroglobulin Response

Abstract

Background:

We previously demonstrated the clinical utility of using response to therapy variables obtained during the first 2 years of follow-up to actively modify initial risk estimates which were obtained using standard clinic-pathologic staging systems. While our proposed dynamic risk stratification system accurately reclassified patients who demonstrated an excellent response to therapy as low-risk patients, it grouped patients with either biochemical or structural evidence of disease into a single incomplete response to therapy cohort. This cohort included a wide variety of patients ranging from very minor thyroglobulin (Tg) elevations in the absence of structurally identifiable disease to widespread, progressive structural disease. Here we determined whether subdivision of the incomplete response to therapy category more precisely predicted clinical outcomes. We hypothesized that patients with an incomplete response to therapy based on persistently abnormal Tg values alone would have better clinical outcomes than patients having structurally identifiable disease.

Methods:

Following total thyroidectomy and radioactive iodine (RAI) ablation, 192 adult thyroid cancer patients were retrospectively identified as having either a biochemical incomplete response (abnormal Tg without structural evidence of disease) or structural incomplete response (structurally identifiable disease with or without abnormal Tg) as the best response to initial therapy within the first 24 months after RAI ablation. Clinical outcomes evaluated included structural disease progression, biochemical disease progression, and overall survival.

Results:

Sixty-three patients (33%) had a biochemical incomplete response while 129 (67%) had a structural incomplete response. Eleven to 156 months after evaluation of their responses (mean=70 months), patients with structural incomplete response were significantly more likely to have structural evidence of disease at final follow-up (37% vs. 17%, p=0.0004), structural progression (52% vs. 5%, p<0.001), biochemical progression (45% vs. 11%, p<0.001), and death from disease (38% vs. 0%, p<0.0001) than patients demonstrating a biochemical incomplete response. Overall survival was significantly better in patients with either a biochemical incomplete response or a loco-regional structural incomplete response than patients demonstrating a structural incomplete response with distant metastasis (Kaplan-Meier analysis, p<0.0001).

Conclusions:

A structural incomplete response to initial therapy is associated with significantly worse clinical outcome than a biochemical incomplete response to therapy.

Introduction

T he last several years have seen a renewed interest in a risk-stratified approach to the management of differentiated thyroid cancer. Initial risk estimates are calculated using one of many previously published staging systems which are primarily based on age at diagnosis, specific histopathologic features of the primary tumor, the extent of disease, and the adequacy of surgical resection (1 –13). While these initial staging systems provide important prognostic information that can guide early management decisions, our group has recently demonstrated the importance of ongoing risk stratification based on a response to therapy assessment that allows for a more accurate prediction of eventual clinical outcomes than any of the initial staging systems (14).

In a cohort of 588 well differentiated thyroid cancer patients treated with total thyroidectomy and radioactive iodine remnant ablation (RRA), the risk of having persistent or recurrent disease was significantly higher in patients having an incomplete response to therapy (96%) than in patients having an excellent (4%) or acceptable response to therapy (13%) (See Table 1 for definitions of response to therapy) (14). While the incomplete response to therapy category adequately captures patients with persistent/recurrent disease, it is a very heterogeneous cohort which includes patients with a wide variety of clinical conditions that can range from minimal thyroglobulin (Tg) elevations without evidence of structurally identifiable disease to wide spread, macroscopic structural distant metastases. So while an incomplete response to therapy accurately captures patients that had not been cured after initial therapy, the clinical outcomes of the patients in this group can widely vary.

Table 1.

Previously Proposed Response to Initial Therapy Definitions (6–24 Months After Radioiodine Ablation)

Excellent response	Acceptable response	Incomplete response
All of the following:	Any of the following:	Any of the following:
• Suppressed and stimulated Tg <1 ng/dL	• Suppressed Tg <1 ng/mL with stimulated Tg ≥1 and <10 ng/mL	• Suppressed Tg ≥1 ng/mL or stimulated Tg ≥10 ng/mL
• Neck US with NED	• Neck US with nonspecific changes or stable subcentimeter lymph nodes	• Rising Tg values
• Negative cross-sectional and/or nuclear medicine imaging (if performed)	• Cross-sectional and/or nuclear medicine imaging with nonspecific changes, although not completely normal	• Persistent or newly identified structural disease in cross-sectional imaging

Response to therapy definitions in Tuttle et al. (14).

Tg, thyroglobulin; US, ultrasound; NED, no evidence of disease.

Therefore, the goal of this study was to reanalyze the incomplete response to therapy category in an effort to subdivide this group into more homogeneous cohorts that may better correlate with final clinical outcomes. We hypothesized that patients classified as incomplete response on the basis of serum Tg elevations alone without structurally identifiable disease (biochemical incomplete response) would have significantly better clinical outcomes than patients classified as incomplete response based on persistent or newly identified structural disease (structural incomplete response).

Methods

After obtaining Institutional Review Board approval, we retrospectively reviewed the electronic medical records of 217 adult differentiated thyroid cancer patients evaluated at Memorial Sloan Kettering Cancer Center between January 1994 and December 2004 who demonstrated an incomplete response to total thyroidectomy and RRA (14). Of the 217 potentially eligible patients, 25 were excluded from this analysis for the following reasons: inadequate follow-up information (n=6), Tg determinations performed in different assays overtime (n=16), and interfering antibodies anti-Tg (TgAb) (n=3). A minimum of 3 years of follow-up was required for entry into the study unless one of the clinical end points (recurrence or death) was reached before that time point.

Each patient was risk-stratified using the 7th edition of the American Joint Committee on Cancer/Union Internationale Contre le Cancer staging system (Stage I, II, III, or IV) and the American Thyroid Association risk of recurrence stratification system (low, intermediate, or high-risk of recurrence) (5,11).

With regard to response to therapy assessment, each patient was classified as having either a biochemical incomplete response (suppressed Tg ≥1 ng/mL or stimulated Tg ≥10 ng/mL in the absence of structurally identifiable disease) or structural incomplete response (persistent/recurrent structurally identifiable disease with or without Tg elevation) based on follow-up data obtained within the first 2 years of follow-up after total thyroidectomy and RRA. Persistent/recurrent structural disease was defined as (i) positive cytology/histology; or (ii) highly suspicious lymph nodes or thyroid bed nodules on the neck ultrasound (US) (hypervascularity, cystic areas, heterogeneous content, rounded shape, enlargement on follow-up); or (iii) cross-sectional imaging highly suspicious for metastatic disease. The first fluoro-deoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT) done during the first 2 years of follow-up in 133/192 patients was used to classify the metastatic disease as either FDG avid (PET positive) or non-FDG avid (PET negative).

Patients were considered to have non iodine avid disease if at the time of the ablation they had evidence of structural disease in the single-photon emission computed tomography (SPECT)/CT or other cross-sectional images and did not have iodine uptake in the post-therapy scan or if they had a negative post-therapy scan anytime within the first 2 years with biochemical or structural evidence of disease.

During follow-up, structural disease progression was defined as: (i) enlargement of previously known metastases; (ii) development of new metastases during the follow-up, defined as findings on radioactive iodine (RAI) scans, 18-FDG-PET scans, or other cross-sectional imaging highly suspicious for metastatic disease or biopsy proven; or (iii) thyroid-cancer related death. Biochemical disease progression was defined as a progressive rise in serum Tg over time (>10% over baseline) measured in the same assay with similar levels of thyroid-stimulating hormone (TSH) suppression.

Patients were considered to have no evidence of disease (NED) at final follow-up if they had no evidence of structural disease confirmed by biopsy (cytology or histology), cross-sectional imaging (US, CT or nuclear magnetic resonance) or functional imaging (RAI scan or 18-FDG-PET scan), and a suppressed Tg <1 ng/mL.

Laboratory studies

Between 1994 and 1997, a variety of Tg assays were used with functional sensitivities of approximately 1 ng/mL. Starting in 1998, all Tg values were measured using the Dynotest-TgS immunoradiometric assay (Brahms Inc., Berlin, Germany, functional sensitivity 0.6 ng/mL normalized to CRM 457). All patients had Tg and TgAb levels while on TSH suppressive therapy at the time of final follow-up. Patients with interfering TgAb were excluded (Dynotest-TgS immunoradiometric assay; Brahms, Inc.).

Statistical analysis

Continuous data are presented as mean and standard deviations with median values. For comparing medians nonparametric Mann–Whitney test was used and for categories we used χ ² and Fisher's exact tests. Analysis was performed using SPSS software (Version 18.0.1; SPSS, Inc., Chicago, IL).

Results

Of the 192 patients demonstrating an incomplete response to total thyroidectomy and RRA included in this study, 63 (33%) had persistently abnormal serum Tg values without structurally identifiable disease (biochemical incomplete response) while 129 (67%) had structural evidence of persistent disease (structural incomplete response: 54 with only loco-regional disease, 75 with distant metastases with or without loco-regional disease).

Compared with patients with a biochemical incomplete response to therapy, patients with a structural incomplete response were older (50±16 vs. 43±14 years, p=0.002), more likely to be male (47% vs. 22%, p=0.001), more likely to have FDG PET positive disease (45% vs. 0%, p<0.0001), less likely to have RAI avid disease (60% vs. 78%, p=0.015), had a higher median suppressed Tg (20.3 vs. 3.6 ng/mL, p<0.0001) during the first 2 years of follow-up, and received a higher administered activity of RAI for ablation (150 vs. 100 mCi, p<0.0001). Patients with structural incomplete response were also more likely to have distant metastases at diagnosis (AJCC stage IVc disease in 43% vs. 3%, p<0.0001), to have been classified as ATA high risk (67% vs. 21%, p<0.0001), and to have received additional surgery (57% vs. 6%, p<0.0001), external beam radiation therapy (EBRT; 35% vs. 0%, p<0.0001), or systemic therapy after RRA (13% vs. 2%, p<0.0001) (see Table 2).

Table 2.

Description of the Cohort

	Biochemical incomplete response (n=63)	Structural incomplete response (n=129)	p-Value
Age (years)
Mean±SD	43±14	50±16	0.002
Median	39	52
Range	21–79	18–82
Gender
Female	49 (77.8%)	69 (53.5%)	0.001
Male	14 (22.2%)	60 (46.5%)
Histology
Papillary	54 (85.7%)	94 (72.9%)	0.22
Poorly differentiated	5 (7.9%)	20 (15.5%)
Follicular	2 (3.2%)	9 (7%)
HCC	2 (3.2%)	6 (4.6%)
Preparation for ablation
rhTSH	28 (44.4%)	53 (41.1%)	0.59
THW	35 (55.6%)	76 (58.9%)
PET/CT
Positive	0	58 (45%)	<0.0001
Negative	31 (49.2%)	44 (34.1%)
Nonavailable	32 (50.8%)	27 (20.9%)
RAI avid disease
Yes	49 (77.8%)	77 (59.7%)	0.015
No	14 (22.2%)	52 (40.3%)
AJCC/UICC stage
I	37 (58.7%)	27 (20.9%)	<0.0001
II	4 (6.3%)	15 (11.6%)
III	14 (22.2%)	6 (4.7%)
IVa	6 (9.5%)	24 (18.6%)
IVb	0	2 (1.6%)
IVc	2 (3.2%)	55 (42.6%)
ATA initial risk
Low	10 (15.9%)	3 (2.3%)	<0.0001
Intermediate	40 (63.5%)	40 (31%)
High	13 (20.6%)	86 (66.7%)
Follow-up duration (months)
Mean±SD	98.4±30	90.1±34.8	0.115
Median	98	42
Range	40–160	9–180
RAI ablation activity (mCi)	100 (29–150)	150 (28–590)	<0.0001
Time (months) to additional therapy (after initial surgery and RRA)	13 (5–160)	12 (4–180)	<0.0001
Additional therapy during follow-up
Surgery	4 (6.3%)	74 (57.3%)	<0.0001
RAI	53 (84.1%)	106 (82.3%)
EBRT	0	45 (34.9%)
Systemic therapy	1 (1.6%)	17 (13.2%)
Lowest suppressed Tg in the first 2 years
Mean	3.6	726.8	<0.0001
Median	3.6 (<0.3–36)	20.3 (<0.3–43500)
Structural Progression
Yes	3 (4.8%)	67 (51.9%)	<0.0001
No	60 (95.2%)	62 (48.1%)
Biochemical Progression
Yes	7 (11.1%)	58 (45%)	<0.0001
No	56 (88.9%)	71 (55%)

SD, standard deviation; HCC, Hurthle cell carcinoma; rhTSH, recombinant human thyroid-stimulating hormone; THW, thyroid hormone withdrawal; PET, positron emission tomography; CT, computerized tomography; RAI, radioactive iodine; AJCC, American Joint Committee on Cancer; UICC, Union Internationale Contre le Cancer; ATA, American Thyroid Association; RRA, RAI remnant ablation; EBRT, external beam radiation therapy.

Further, patients with a structural incomplete response to therapy were significantly more likely to have either structural disease progression (52% vs. 5%, p<0.001) or biochemical disease progression (45% vs. 11%, p<0.001) than patients with biochemical incomplete response.

The clinical status at final follow-up was significantly worse in patients having a structural incomplete response to initial therapy than those demonstrating only a biochemical incomplete response (See Table 3). A structural incomplete response was more likely to be associated with structural evidence of disease at final follow-up (37% vs. 17%, p<0.0004) and death from disease (38% vs. 0%, p≤0.0001) than a biochemical incomplete response. Conversely, patients with a biochemical incomplete response to therapy were significantly more likely to have NED at final follow-up than patients with a structural incomplete response (68% vs. 15%, p≤0.0001).

Table 3.

Clinical Status at Final Follow-Up

	Biochemical incomplete (n=63)	Structural incomplete (n=129)	p-Values
NED	68% (43)	15% (19)	<0.0001
Biochemical evidence of disease (without structural disease)	19% (12)	10% (13)	0.1
Structural evidence of disease	17% (8)	37% (48)	0.0004
Death	0% (0)	38% (49)	<0.0001

Further, patients classified as having structural incomplete response to therapy because of macroscopic distant metastasis were more likely to have progression of disease or death than patients classified as structurally incomplete on the basis of loco-regional evidence of disease without distant metastases (73.3% vs. 22.2%, p<0.0001). Both biochemical incomplete response to therapy and loco-regional structural incomplete responses were associated with a significantly better overall survival than structural incomplete response to therapy with distant metastases (see Fig. 1, Kaplan–Meier analysis, p<0.0001). Over the course of the study, no deaths were seen in the patients with biochemical incomplete response, while death was the final outcome in 11% of the patients with loco- regional disease and 57% of the patients with distant metastases (p<0.001).

FIG. 1.

Overall survival is significantly better in patients demonstrating a biochemical (B) incomplete response to therapy or a loco-regional (L-R) structural incomplete response than in patients demonstrating a structural incomplete response to therapy with distant metastases (DM) (Kaplan–Meier analysis, p<0.0001).

As can be seen from Table 3, 62 patients (43 with a biochemical incomplete response and 19 in the structural incomplete response cohorts) were reclassified as NED at the time of their final follow-up. Five of the biochemical incomplete patients had mildly elevated suppressed Tg levels that became undetectable over time without any additional therapy. All of the remaining patients with incomplete response to therapy received additional treatments during follow-up (a median of one additional treatment performed a median of 13 months after initial RRA). These additional treatments included: one additional RAI therapy in 53% (33/62), one additional neck surgery in 10% (6/62), additional neck surgery and RAI therapy in 18% (11/62), and multiple additional RAI treatments in 11% (7/62).

Discussion

Our data clearly demonstrate that two clinically distinct cohorts can be identified within patients classified as having an incomplete response to therapy using our previously proposed restaging system (14). Patients with only biochemical evidence of disease have significantly better clinical outcomes in terms of having a higher likelihood of being NED at final follow-up, a lower likelihood of having biochemical or structural disease progression, and a much lower likelihood of dying from thyroid cancer than patients exhibiting a structural incomplete response to therapy. Interestingly, after a median follow-up of 8 years, structural disease was present at the time of final follow-up in only 17% (8/63 patients) of patients initially classified as having a biochemical incomplete response. Further, patients classified as having a structural incomplete response on the basis of macroscopic distant metastases had much poorer outcomes than either patients with biochemical incomplete response or structural incomplete response based on loco-regional disease without identifiable distant metastases.

Based on these findings, we are proposing a modification of our initial response to therapy classification system to allow a distinction to be made between biochemical incomplete response and structural incomplete response (See Table 4). This modification will allow more accurate prediction of outcomes in patients with an incomplete response to therapy based on the presence or absence of structurally identifiable disease at the time of restaging.

Table 4.

Modified Response to Initial Therapy Definitions (6–24 Months After Radioactive Iodine Ablation)

Excellent response	Acceptable response	Biochemical incomplete response	Structural incomplete response
All of the following:• Suppressed and stimulated Tg <1 ng/mL• NED on neck US• Additional cross-sectional and/or nuclear medicine imaging negative (if performed)	Any of the following:• Suppressed Tg <1 ng/mL with stimulated Tg ≥1 and <10 ng/mL• Nonspecific imaging findings without definite evidence of disease	Any of the following in the absence of structurally identifiable disease:• Suppressed Tg ≥1 ng/mL• Stimulated Tg ≥10 ng/mL• Rising Tg values	Any of the following regardless of Tg values:• Persistent or newly identified disease on cross-sectional imaging and/or nuclear medicine imaging

The impact of this subclassification of the incomplete response to therapy cohort is best appreciated when data from our previous study (14) is combined with the results from this study. In Table 5, it is readily apparent that our response to therapy restaging system can identify four separate cohorts of patients that have significantly different clinical outcomes. The best clinical outcomes are seen in those patients demonstrating an excellent response to therapy with undetectable suppressed and stimulated Tg without evidence of structurally identifiable disease. Patients with an acceptable response to therapy also do very well with the majority of patients being NED at final follow-up and a smaller number (13%) having persistent biochemical evidence of disease without identifiable structural disease.

Table 5.

Composite Data Combining Results of Our Previous Study (14) with the Findings in This Study

	Excellent response	Acceptable response	Biochemical incomplete	Structural incomplete
	(n=159)	(n=95)	(n=63)	(n=129)
NED	96%	87%	68%	15%
Persistent disease, biochemical evidence	0%	13%	19%	10%
Persistent disease, structural evidence	0%	0%	17%	37%
Recurrence	4%	—^a	—^a	—^a
Death	0%	0%	0%	38%

Because the definition of recurrent disease requires a period of NED, patients with persistent disease cannot be classified as having a recurrence.

Patients with a biochemical incomplete response to therapy have higher rates of persistent biochemical (19%) or structural disease (17%), but 68% eventually become NED with no disease specific mortality over a median 8 years of follow-up. Clearly, the worst clinical outcomes are seen in those patients classified as having a structural incomplete response with only 15% being NED at final follow-up, 37% with persistent structural disease, 10% with persistent biochemical evidence of disease, and a 38% disease specific death rate over a median follow-up period of 3.5 years.

In conclusion, our findings demonstrate that restaging within the first 2 years of initial therapy provides important prognostic information in patients with differentiated thyroid cancer. While all patients with an incomplete response to initial therapy are less likely to be classified as NED at final follow-up than patients with either an excellent or acceptable response to therapy, all of the disease specific deaths were restricted to the patients with a structural incomplete response. Therefore, restaging can be used to modify initial risk estimates to provide more accurate and dynamic estimates of risk of recurrence, persistent disease, and disease specific death than static staging systems that are not usually modified over time.

Footnotes

Disclosure Statement

F.V., R.G., and H.T. have nothing to declare. R.M.T. is a consultant to and has received honoraria from the Genzyme Corporation.

References

Hay

, Grant

, Taylor

, McConahey

. 1987. Ipsilateral lobectomy versus bilateral lobar resection in papillary thyroid carcinoma: a retrospective analysis of surgical outcome using a novel prognostic scoring system. Surgery, 102:1088–1095.

Akslen

. 1993. Prognostic importance of histologic grading in papillary thyroid carcinoma. Cancer, 72:2680–2685.

DeGroot

, Kaplan

, McCormick

, Straus

. 1990. Natural history, treatment, and course of papillary thyroid carcinoma. J Clin Endocrinol Metab, 71:414–424.

Noguchi

, Murakami

, Kawamoto

. 1994. Classification of papillary cancer of thyroid based on prognosis. World J Surg, 18:552–558.

Sobin

, Gospodarowicz

, Wittekind

. 2010. TNM Classification of Malignant Tumours, 7th. Wiley: Blackwell.

Sugitani

, Kasai

, Fujimoto

, Yanagisawa

. 2004. A novel classification system for patients with PTC: addition of the new variables of large (3 cm or greater) nodal metastases and reclassification during the follow-up period. Surgery, 135:139–148.

Yildirim

. 2005. A model for predicting outcomes in patients with differentiated thyroid cancer and model performance in comparison with other classification systems. J Am Coll Surg, 200:378–392.

Byar

, Green

, Dor

, Williams

, Colon

, van Gilse

, Mayer

, Sylvester

, van Glabbeke

. 1979. A prognostic index for thyroid carcinoma: a study of the EORTC thyroid cancer cooperative group. Eur J Cancer, 15:1033–1041.

Cady

, Rossi

. 1988. An expanded view of risk-group definition in differentiated thyroid carcinoma. Surgery, 104:947–953.

10.

Shaha

, Loree

, Shah

. 1994. Intermediate-risk group for differentiated carcinoma of the thyroid. Surgery, 116:1036–1041.

11.

Cooper

, Doherty

, Haugen

, Kloos

, Lee

, Mandel

, Mazzaferri

, McIver

, Pacini

, Schlumberger

, Sherman

, Steward

, Tuttle

. The American Thyroid Association (ATA) Guidelines Taskforce on Thyroid Nodules and Differentiated Thyroid Cancer 2009 Revised American Thyroid Association Management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19:1167–1214.

12.

Pacini

, Schlumberger

, Dralle

, Elisei

, Smit

, Wiersinga

. European Thyroid Cancer Taskforce 2006 European consensus for the management of patients with differentiated thyroid carcinoma of the follicular epithelium. Eur J Endocrinol, 154:787–803.

13.

Maia

, Ward

, Carvalho

, Graf

, Maciel

, Rosário

, Vaisman

. 2007. Thyroid nodules and differentiated thyroid cancer: Brazilian consensus. Arq Bras Endocrinol Metab, 51:867–893.

14.

Tuttle

, Tala

, Shah

, Leboeuf

, Ghossein

, Gonen

, Brokhin

, Omry

, Fagin

, Shaha

. 2010. Estimating risk of recurrence in differentiated thyroid cancer after total thyroidectomy and radioactive iodine remnant ablation: using response to therapy variables to modify the initial risk estimates predicted by the New American Thyroid Association Staging System. Thyroid, 20:1341–1349.