Anybody Fill Out the Cancermath.net Form?

From past responses on this site, this calculator has never been updated so it's old stuff in it. Not sure if that means it really should be better outcomes or worse but as much as things change too bad that no one bothers to go back and make the adjustments to reflect current time.

Exactly icietla!  It increases your chance of living for 15 years by 32%.  Just saying.

Oh Beezie I wish I could supply an answer like that! Would you be willing to assess my situation as well? None of my doctors have bothered, and the calculator where I go is "broken". If you will, here are my stats. ER 90+,PR 40+, HER2 neg. 2.1 cm lymph node with 3mm cancer cells, occult primary tumor. Small DCIS area removed in lumpectomy, .5cm.. On Letrozole from APR 9 2015 until July 27 2017. Surgery with two margin clearings - 3 in all. Radiation 30 + 5 boosts, whole breast, includes clavicle and ALND. No chemo. Stage IIA, grade 2..

Thank you Beesie! Large armpit lymph node found by myself - had pain in my shoulder, like it popped up overnight, but pain/funny feeling persisted for two months previous. No Ki67 (because no primary?). No Oncotype because no chemo. Age at dx 65, age now - 67 1/2.

I'm off to another eye dr. appt. Be back in a few hours.

Thank you so much Beesie. I will accept a guestimate with a disclaimer. You're the best.

KB, you're welcome!

I know nothing about chemo - I was fortunate that I didn't need it - so I can't answer your question about Taxol Herceptin. Since Herceptin is relatively new, I would think that it might be 2nd or 3rd generation, but that's a complete guess. Hopefully someone who actually knows will come by and answer.

Beesie, can you do mine and explain to me in plain english? I'm on Tamox v AI which MO says gives (if it works) 3-4% additional benefit.

Age at dx- 50

Mode of detection: both symptoms and screening

size 7 cm

grade 3

1 SN with micromets

ER positive (95%) so

Her 2- negative

Ki67- positive (75%)

chemo- I think the ACTaxotere is third

And not just with what the calc says but if it were you, would you think if it were you, worse than what it says? Thanks!

If it is appropriate to use PREDICT, then consideration should be given to using the most current version (currently PREDICT Version 2.0).

I think Marijen received neoadjuvant endocrine therapy. Is CancerMath or PREDICT designed and validated for use in patients who received neoadjuvant treatment, and neoadjuvant endocrine therapy in particular? If not, it should not be used in such patients.

[EDIT (09/11/2018): As of this date, the latest version of PREDICT is now Version 2.1. The overview page states (emphasis added): "It is not designed for women who have had neo-adjuvant treatments (chemotherapy given before surgery) or have already been treated for cancer, for women whose breast cancer is in both breasts or has already spread to distant parts of the body at the time it is diagnosed or for women with non-invasive breast cancer, such as as Ductal Carcinoma In Situ or Lobular Carcinoma In Situ. It is also not designed for men with breast cancer.]

Here is what I found for PREDICT. The tool was described in this 2010 paper, which discusses the dataset used for development and the independent dataset used for validation. Reference is made to adjuvant chemotherapy and adjuvant endocrine therapy.

See also, this 2017 paper re Version 2.0 and noting that: "The predicted benefit of adjuvant chemotherapy classified as first-, second- or third-generation and adjuvant hormone therapy was taken from the meta-analyses of the Early Breast Cancer Trialists Collaborative Group [4]." Reference 4 corresponds to this 2005 Lancet paper which appears to concern adjuvant treatments (i.e., in the adjuvant setting (surgery-first)).

One recent Review article states:

"The PREDICT tool (http://www.predict.nhs.uk/) was developed using cancer registry data from 5,694 patients in the UK.[5] Validation of the model was made on 5,000 other patients from the U.K. and 3140 patients from Canada.[6] An estimation of therapy and prognosis of HER 2 tumors was later incorporated.[7] The PREDICT tool utilizes data on patient age and tumor characteristics (the mode of detection (i.e., screening versus discovery of a palpable mass), size, grade, ER status, and KI67 status) to provide a choice for estimating the value of endocrine therapy alone, or endocrine therapy and second-generation chemotherapy (anthracycline-containing, >4 cycles or equivalent) versus third-generation (taxane-containing chemotherapy regimens).[8] The [PREDICT] model allows one to estimate the effects of adjuvant endocrine and chemotherapy treatment on survival at 5 and 10 years, but there is no estimate of relapse and it does not account for non-breast cancer causes of mortality in the overall survival estimate. However, unlike Adjuvant!, the PREDICT model can estimate the benefits of anti-HER2 therapy in patients with HER-2 positive tumors.[7] In addition, a recent study has validated this tool's ability to provide accurate estimates of the potential benefits of treatment at 5 years for older patients.[9] Table 2 provides several scenarios showing the effects of treatment selection on survival using the PREDICT model."

Reference [9] was a study of the performance of the test conducted in older women in the adjuvant setting only: "For this study, all patients with unilateral, unicentric, invasive, local adenocarcinoma were included, provided that they received adequate locoregional treatment (either mastectomy or lumpectomy with radiotherapy and axillary lymph node staging) and they did not receive neoadjuvant treatment."

Has anyone found any documentation indicating that CancerMath and PREDICT can be used for patients who received neoadjuvant treatment?

Some on-line tools do not perform well in certain patient sub-groups. Patients with invasive breast cancer with pending treatment decisions (to initiate or continue treatment) should always discuss such on-line tools with their Medical Oncologist to ensure they are considered applicable and adequately validated in the individual case, that the parameters are entered correctly, and that the implications of the outputs are explained, along with any caveats.

BarredOwl

______________

Note: The passage quoted above re PREDICT notes "there is no estimate of relapse". Version 2.0 of the PREDICT tool appears to provide information regarding 5-year and 10-year Overall Survival, which is typically a mortality assessment. Overall survival benefit is important; however, it typically measures whether the study participants are alive or not at a specific time-point, not whether they have remained disease- or recurrence-free. Those who are alive may be living with metastatic disease, living with a loco-regional recurrence or new disease (i.e., a new primary in the same or contralateral breast), or may be disease-free. Endocrine therapy can reduce the risk of these events. Thus, for patients with invasive breast cancer and any significant risk of distant recurrence (relapse), consideration is typically given to the potential benefit of endocrine therapy in reducing the risk of suffering a distant (metastatic) recurrence; yet to my layperson knowledge, this information is not provided by PREDICT.

BarredOwl, as an FYI, I have been using Predict V. 2.0.

I'm glad that you posted your caution, because I was going to also post a caution before providing the info Artista requested. I posted initially in this thread to simply explain her CancerMath results to Hippie Girl, because she asked, and because those who replied before me seemed to be confused by the results, which in fact do seem to be quite reasonable for Hippie Girl's type of diagnosis.

These models are handy, but they don't replace your MOs and the information that they provide. Nor do they replace the Oncotype test, for those of you who've had that. Everyone should understand that both CancerMath and Predict are generalized models, and at best provide ballpark information about the survival and mortality rates for any particular diagnosis. If the models suggest that someone of your age with your type of diagnosis and treatment will have an 80% 10-year survival rate, well, at least you know that it's not 40%. And of course you have to understand that if the model suggests that the survival rate is 80%, the model is also therefore saying that 20% of women with this type of diagnosis will die... and unfortunately there is no one on earth who can tell you if you will be one of the 80, or one of the 20.

With those cautions, Artista, here is what Predict and CancerMath come up with for your diagnosis. Unlike Hippie Girl's situation where the absolute benefit of treatment appears to be quite low (in terms of mortality risk reduction), with your type of diagnosis and at your age, the addition of chemo and Tamoxifen appears to save many lives.

It is interesting to note that the CancerMath projection appears to be more positive than the Predict projection, which again highlights that these figures are generalized ballpark estimates.

• For someone diagnosed at 50, Predict suggests that after 10 years, 41.5% will be alive without any adjuvant treatment, an additional 12.3% will be alive because of chemo, and an additional 12.8% will be alive because of Tamoxifen.

.

• For a similar patient, CancerMath suggests that after 15 years, 41% will be alive without any adjuvant treatment, and an additional 32% will be alive because of either chemo or Tamoxifen (or a combination of both). Therefore CancerMath is actually showing a higher survival rate at 15 years than Predict is showing at 10 years. CancerMath also indicates that of the 27% of women who will have died within 10 years, 7% will have died from non-cancer causes, and 20% will have died as a result of this breast cancer diagnosis.

.

Artista, I'm not in a position to make an assessment as to whether the Predict projection might be more accurate than the CancerMath projection, or vice versa. Nor can I assess whether the outlook might be worse than what either model says, or better than what they say. I have no way of knowing. If it were me, I would look at these results and net them down to a simple "Okay, with the type of diagnosis I had, after 10 - 15 years, about 65% - 70% of women who have chemo and hormone therapy are still living." And with that I would know that unfortunately it's not 90%, but at least it's not 40%.

And with that, I'm closing up shop!

marijen, intraductal is DCIS. 'Ductal" would have been IDC, however because since your IDC was never found and your DCIS was, I put you down as having DCIS.

Thank you Beesie! I've tried to understand these calculators and never could. You made it clear for me so thanks so much! xo

BarredOwl, good catch! I was getting confused with all of the inputs I was doing yesterday.

I do default to using Predict V.2.0 and did that for Hippie Girl and Artista, and started off with V. 2.0 for marijen. The problem is that V 2.0 version requires an input for the grade of the tumor, and since marijen's tumor was occult, we don't have a grade. So I couldn't get results from V 2.0 and moved to V. 1.2, which allows an input of "unknown" for the grade. I forgot about that little hitch when I posted my reply to you.

You are correct that there is nothing to suggest that these tools can be used for those who received neoadjuvant therapy. However if the purpose of the tool is to just get a ballpark estimate of risk, I think they still provide generalized information that can be helpful. Both CancerMath and Predict present the risk with and without adjuvant therapy; conservatively, I'd say that someone who had neoadjuvant therapy but who did not continue with adjuvant therapy for 5 years after surgery, as in marijen's case, should probably look at the 'without adjuvant therapy' results to get her estimates of risk.

There are a lot of "ifs, ands and buts" related to using these models. It would be ideal if every patient received personalized information from her oncologist and/or had access to genetic testing such as the Oncotype test. Unfortunately we know that this doesn't happen and that genetic tests are not yet available for everyone. Given that, and given how difficult it is to not know where we stand with our diagnoses and our future risk ("do I have a 5% chance of dying from this, or a 50% chance of dying from this?"), I think that these models can be helpful in providing an approximate range of risk. I would hope that nobody uses these models for anything more than that.

Another concern relates to applying either CancerMath or PREDICT in the setting of occult disease (T0), which is a relatively rare diagnosis. For example, the FAQ for CancerMath notes:

"How accurate are the [CancerMath] projections?

When verifying the accuracy of the calculators against independent data sets (using different patients than the patients that the model parameters were dervied from) the mean and median error across patient stratifications are typically ~2% Thus, if a calculator outputs a projected 15-year mortality of 16%, the actual value may in fact vary between 14-18%. The accuracy is generally higher for the most common types and sizes of tumors, and lower for the rarer and the most advanced tumors. For details, see our technical reports"

The absence of any indication of applicability or validation in the neoadjuvant setting remains a concern for me. Perhaps I am obsessed with design parameters and validation.

Using CancerMath or PREDICT in Marijen's case (after neoadjuvant endocrine therapy) seems to reflect her post-treatment pathologic staging only? Her clinical staging (N status) was higher than her pathologic staging. I note that efforts to develop prognostic tools for use in the neoadjuvant chemotherapy setting account for such additional factors. For example, tools developed in 2008 ("CPS" and "CPS + EG") used datasets from patients who received neoadjuvant chemotherapy, and a scoring system that accounted for both pretreatment clinical staging and post-treatment pathologic staging ("CPS") and biologic markers ("CPS + EG") (See also, later validation in HR+HER2- pts receiving neoadjuvant chemotherapy). This approach accounts for the possibility that pretreatment clinical staging might have an additional impact on prognosis. In 2016, a modification of "CPS + EG", named "Neo-Bioscore" was described here here (see also, ScienceDaily).

A more recent effort to assess prognosis following neoadjuvant chemotherapy incorporates assessment of Residual Cancer Burden. The closing remark of the commentary is: "For now, pCR remains the clinical trial standard to assess the effect of a given neoadjuvant chemotherapy-based treatment."

These newer tools may also be less reliable in those with more rare diagnoses. That said, prognostic tools have been specifically developed for use in the neoadjuvant setting. Patients who received either neoadjuvant chemotherapy or neoadjuvant endocrine therapy should seek expert professional advice from their Medical Oncologist regarding the best available tool and/or prognostic approach in their specific case.

BarredOwl

Hi Marijen:

I am not sure, but the neoadjuvant calculators may also not be well-suited for use with occult disease either, because an element is pathological response in the breast (which seemingly cannot be assessed with occult disease).

I am really not very clear on CancerMath or PREDICT calculators. For example, I do not understand why CancerMath includes the option of "Intraductal + LCIS" as histological type. The CancerMath calculator doesn't work properly for in situ disease, so why is that even an option? I wrote CancerMath this morning to inquire. The substance of my inquiry was:

>> "In my layperson understanding the CancerMath Breast Cancer tools are for use in patients with invasive breast cancer. I noticed that "Histological Type" in the various Breast Cancer calculators include the option of "Intraductal + LCIS" or "Intraductal + Lobular in situ." These are forms of "in situ" disease. However, Technical Report #2 (Equation Parameters) does not refer to either "Intraductal" or "LCIS" or "in situ". Instead, it refers to "Ductal and lobular", which would seem to be a "Mixed" invasive histology.

Is it possible that the the "Histological Type" options in the calculators are incorrect and should actually be "Ductal and lobular" or "Mixed ductal and lobular"?" <<

BarredOwl