Amplification after 40 cycles is really late and may not even be from your template. Either your gene of interest primers have poor efficiency, your gene-of-interest has a VERY low expression (or isn't expressed at all) or a combination of both. If you can show that your primers have 95-105% efficiency, AND you sequence the resulting PCR product to show it's from your gene of interest, then it's just a really low abundance transcript. You should start by running out some of your reaction on a regular agarose gel to look at the product sizes and make sure you aren't seeing a primer dimer or some other artifact. Do you have a good positive control for your gene of interest? The next step is to verify your primer efficiency with a standard curve. Then try sequencing the PCR product from your gene of interest. Good luck!
As Katie said: the transcript is of really low abundance or it is not expressed at all.
Katie also mentioned a positie control. This is really important. Evaluate your primers using a positive control. If you don't have a cell line that has a known measurable expression of that lncRNA than you may consider ordering a synthetic DNA and use this as a positive control to evaluate your primers. It makes no sense to optimize a protocol if you don't know that the samples used in the optimization do express the target sequence (in measurable amounts).
Given you do know that your PCR works on positive controls:
Use more biological sample and use sequence-specific priming for the RT. This may help to get quantifiably amounts (given the target gene is expressed at all!). If this does not give you Ct values below 30, then quantification is impossible as there are too few molecules in your cells.